ASPECT

AND

ARGUMENT STRUCTURE

IN JAPANESE

A Thesis submitted to the University of Manchester

for the degree of Doctor of Philosophy

in the Faculty of Arts

September 2000

Chiaki Taoka

Department of Linguistics

1
TABLE OF CONTENTS

TITLE ….…………………………………………………………………… 1
TABLE OF CONTENTS ………………………………………………… 2
ABSTRACT …. ……………………………………………………………. 10
DECLARATION…………………………………………………………… 11
COPYRIGHT…..……………………………………………………………. 12
ACKNOWLEDGEMENTS ………………………………………………… 13
ABBREVIATIONS …………………………………………………………. 14

1. Introduction .…………………………………………………………….. 16
1.1. Verbal semantic structure ……………………………………………. 16
1.2. Argument linking ……………………………………………………. 16
1.3. Aims of the thesis …………………………………………………… 20
2. Theories of argument linking ……………………………………………... 24
2.1. Introduction …………………………………………………………. 24
2.2. Literature on Japanese verbs of putting and removing ……………… 27
2.3. Croft’s causal analysis (1991, 1993, 1994ab, 1998a) ………………. 35
2.3.1. Event structure ………………………………………………… 36
2.3.2. The verbal profile………………………………………………. 41
2.3.2.1. Spray/load alternation……………………………………. 41
2.3.2.2. The caused-motion construction …………………………. 43
2.3.3. Summary ……………………………………………………… 43
3. Aspect in Japanese ……………………………………………………… 45
3.1.Representational model of aspect and aspectual types ………………. 45
3.1.1. Aspectual classes of Vendler (1967) ………………………….. 45
3.1.2. Croft’s (2000) basic aspectual representation model…………… 47
3.1.2.1. Basic notations and concepts …………………………… 47
3.1.2.2. Aspectual types of events ………………………………. 51
3.1.2.2.1.States ………………………………………………. 51
3.1.2.2.1.1. Point states …………………………………… 51
3.1.2.2.1.2. Inherent states ………………………………. 52
3.1.2.2.1.3. Transitory states ………………………………. 54
3.1.2.2.2. Activities (unbounded processes) ………………….. 55
3.1.2.2.3. Achievements (punctual bounded processes) ……… 57
3.1.2.2.3.1.Cyclic achievements …………………………… 57
3.1.2.2.3.2. Reversible directed achievements………………58
3.1.2.2.3.3. Irreversible directed achievements ……………. 59
3.1.2.2.4. Accomplishments (extended bounded processes) ….. 59
3.1.2.2.5. Runup achievements ………………………………. 63
3.1.3. Summary of the aspectual types ……………………………….. 65
3.2. Analysis of Japanese TA constructions and aspectual classes ………..66
3.2.1. Basic background for analysis …………………………. 67
3.2.2. Criteria …………………………………………………. 69
3.2.3. Analysis ………………………………………………… 71
3.2.3.1. The Present construction …………………………. 71
3.2.3.2. The Te-iru Construction ………………………….. 75
3.2.3.2.1. On te-iru ……………………………………. 75
3.2.3.2.2. Senses of the Te-iru construction ………….. 80
3.2.3.3. The Past construction …………………………….. 90

2
 3.2.3.4. Summary …………………………………………. 98
3.3. Summary ……………………………………………………………. 104
4. Analysis and method ………………………………………………………. 105
4.1. Incremental theme/verbal scale and the argument linking ………….. 105
4.2. The Integration of aspect and causal structure—
Croft’s (2000) new version …………………………………… 111
4.2.1. Notation ………………………………………………………… 111
4.2.2. Verbal scale and related principles …………………………….. 115
4.2.2.1. Verbal scale and holistic theme ……………………………116
4.2.2.2. Governing principles …………………………………….. 116
4.2.3. More examples with path ………………………………………. 117
4.2.3.1. Verbs of motion ………………………………………….. 117
4.2.3.2. Spray/load verbs …………………………………………. 121
4.2.4. Summary ……………………………………………………….. 124
4.3. Methodology—data collection and analysis …………………………. 125
4.3.1. Selection ………………………………………………………... 125
4.3.2. Searching for translation ……………………………………….. 126
4.3.3. Confirming and screening ……………………………………… 128
4.3.4. Looking for more members ……………………………………. 129
4.3.5. Investigating syntactic behaviours and patterns ……………….. 129
4.3.6. Subcategorising into smaller semantic classes ………………… 130
4.4. Basic information on Japanese grammar ……………………………. 130
4.4.1. Particles and antecedent/subsequent obliques ………………….. 130
4.4.1.1. Subject and object markers ………………………………. 131
4.4.1.2. Oblique case markers …………………………………… 132
4.4.1.2.1. Ni ………………………………………………….. 132
4.4.1.2.2. E …………………………………………………... 137
4.4.1.2.3. Made ………………………………………………. 138
4.4.1.2.4 Kara …………………………………………………. 138
4.4.1.2.5. Yori …………………………………………………..139
4.4.1.2.6. De ………………………………………………….. 139
4.4.1.2.7. To ………………………………………………….. 140
4.4.1.3. Summary …………………………………………………. 142
4.4.2. Optional arguments—DNI vs FNI ………………………………143
4.4.3. On path argument ………………………………………………. 145
4.4.4. Type of verbs ………………………………………………….. 147
4.4.4.1. Compound verbs …………………………………………. 147
4.4.4.1.1. Pair compounds ……………………………………. 148
4.4.4.1.2. Means compounds ………………………………….. 149
4.4.4.1.3. Compounds with semantically deverbalized V2 …… 150
4.4.4.1.4. Compounds with semantically deverbalized V1 …….151
4.4.4.1.5. Frozen compounds …………………………………. 151
4.4.4.2. VN-suru verbs ………………………………………….. 152
4.5. Summary …………………………………………………………….. 153
5. Verbs of putting ………………………………………………………….. 154
5.1. Introduction …………………………………………………………. 154
5.2. Basic linking patterns …………………………………………………154
5.3. Analysis …………………………………………………………….. 156
5.4. Classification ………………………………………………………….157
5.4.1. Type 1 Oku ‘put’ verbs ………………………………………….157

3
 5.4.1.1. Members ………………………………………………….. 157
5.4.1.2. Semantics …………………………………………………. 157
5.4.1.3. Syntactic patterns ………………………………………….158
5.4.1.4. Causal and aspectual patterns …………………………….. 159
5.4.1.5. Semantic representations …………………………………. 161
5.4.2. Type 2 Ireru ‘put into’ verbs:
spatial caused-transfer verbs with a specified direction……. 162
5.4.2.1. Members ………………………………………………….. 162
5.4.2.2. Semantics …………………………………………………. 162
5.4.2.3. Syntactic patterns ………………………………………….162
5.4.2.4. Causal and aspectual patterns ……………………………. 165
5.4.2.5. Semantic representations…………………………………. 168
5.4.3. Type 3 Hitasu ‘soak’ verbs …………………………………….. 170
5.4.3.1. Members …………………………………………………. 170
5.4.3.2. Semantics …………………………………………………. 170
5.4.3.3. Syntactic patterns ………………………………………….171
5.4.3.4. Causal and aspectual patterns ……………………………. 171
5.4.3.5. Semantic representations ………………………………… 173
5.4.4. Type 4 Sosogu ‘pour into’ verbs ………………………………. 173
5.4.4.1. Members ………………………………………………… 173
5.4.4.2. Semantics …………………………………………………. 173
5.4.4.3. Syntactic patterns ………………………………………… 174
5.4.4.4. Causal and aspectual patterns ……………………………. 175
5.4.4.5. Semantic representations………………………………….. 176
5.4.5. Type 5 Kabuseru ‘put on’ verbs ……………………………….. 177
5.4.5.1. Members …………………………………………………. 177
5.4.5.2. Semantics …………………………………………………. 177
5.4.5.3. Syntactic patterns ……………………………………….. 177
5.4.5.4. Causal and aspectual patterns …………………………… 178
5.4.5.5. Semantic representations ………………………………… 179
5.4.6. Type 6 Maku ‘scatter’ verbs: verbs of applying and scattering …180
5.4.6.1. Members …………………………………………………. 180
5.4.6.2. Semantics …………………………………………………. 180
5.4.6.3. Syntactic patterns ………………………………………….180
5.4.6.4. Causal and aspectual patterns ……………………………. 181
5.4.6.5. Semantic representations …………………………………. 182
5.4.6.6. Digression on pair compounds …………………………….185
5.4.7. Type 7 Tsurusu ‘hang’ verbs:
verbs of putting in a spatial configuration ……………………. 189
5.4.7.1. Members …………………………………………………. 189
5.4.7.2. Semantics …………………………………………………. 189
5.4.7.3. Syntactic patterns ………………………………………….190
5.4.7.4. Causal and aspectual patterns ……………………………. 191
5.4.7.5. Semantic representations …………………………………. 192
5.4.8. Type 8 Tsumu ‘load’ verbs ………………………………………193
5.4.8.1. Members …………………………………………………. 193
5.4.8.2. Semantics …………………………………………………. 193
5.4.8.3. Syntactic patterns ………………………………………….194
5.4.8.4. Causal and aspectual patterns ……………………………. 194
5.4.8.5. Semantic representations ………………………………… 195

4
5.4.9. Type 9 Oou 'cover' verbs ……………………………………….. 197
5.4.9.1. Members ………………………………………………….. 197
5.4.9.2. Semantics …………………………………………………. 197
5.4.9.3. Syntactic patterns ………………………………………….197
5.4.9.4. Causal and aspectual patterns ……………………………. 199
5.4.9.5. Semantic representations …………………………………. 200
5.4.10. Type 10 Yogosu ‘dirty’ verbs …………………………………..202
5.4.10.1. Members ………………………………………………….202
5.4.10.2. Semantics ……………………………………………….. 202
5.4.10.3. Syntactic patterns ………………………………………. 203
5.4.10.4. Causal and aspectual patterns …………………………… 203
5.4.10.5. Semantic representations ………………………………. 205
5.4.11. Type 11 Kazaru ‘decorate’ verbs: verbs of locative alternation..206
5.4.11.1. Members ………………………………………………… 206
5.4.11.2. Semantics ……………………………………………….. 206
5.4.11.3. Syntactic patterns ……………………………………….. 206
5.4.11.4. Causal and aspectual patterns ………………………….. 207
5.4.11.4.1. Kazaru ‘decorate’ …………………………. 207
5.4.11.4.2. Nuru ‘smear/paint’ …………………………. 210
5.4.11.4.3. Tsumeru ‘pack/cram’ ………………………..212
5.4.11.4.4. Mitasu ‘fill’ ………………………………….214
5.4.11.5. Semantic representations ……………………………….. 215
5.4.12. Type 12 Tsutsumu ‘wrap’ verb ………………………………..220
5.4.12.1. Members ……………………………………………… 220
5.4.12.2. Semantics ……………………………………………….. 220
5.4.12.3. Syntactic patterns ………………………………………. 220
5.4.12.4. Causal and aspectual patterns …………………………… 223
5.4.12.5. Semantic representations ……………………………….. 224
5.4.13. Type 13 Hou.boku-suru ‘graze’ verbs: ground incorporation …226
5.4.13.1. Members ………………………………………………… 226
5.4.13.2. Semantics ……………………………………………….. 226
5.4.13.3. Syntactic patterns ………………………………………. 226
5.4.13.4. Causal and aspectual patterns …………………………… 227
5.4.13.5. Semantic representations ………………………………...227
5.4.14. Type 14 San.sui-suru ‘sprinkle.water-do’ verbs:
figure incorporation 1 ………………………………228
5.4.14.1. Members ………………………………………………… 228
5.4.14.2. Semantics ……………………………………………… 229
5.4.14.3. Syntactic patterns ……………………………………… 229
5.4.14.4. Causal and aspectual patterns …………………………… 230
5.4.14.5. Semantic representations ……………………………….. 231
5.4.15. Type 15 Shoku.rin-suru ‘afforest’ verbs:
figure incorporation 2 ……………………………… 232
5.4.15.1. Members ………………………………………………… 232
5.4.15.2. Semantics ………………………………………………. 232
5.4.15.3. Syntactic patterns ………………………………………. 232
5.4.15.4. Causal and aspectual patterns ………………………….. 233
5.4.15.5. Semantic representations ……………………………….. 234
5.4.16. Type 16 Hame-komu ‘fit.into-put.into’ verbs:
means compound 1 with specified direction ………. 236

5
 5.4.16.1. Members ………………………………………………… 236
5.4.16.2. Semantics ……………………………………………….. 236
5.4.16.3. Syntactic patterns ………………………………………. 237
5.4.16.4. Causal and aspectual patterns …………………………… 238
5.4.16.5. Semantic representations ………………………………. 238
5.4.16.5.1. Middle compounds:
Hame-komu ‘fit.into-put.into’ ……………….239
5.4.16.5.2. Pure means compounds …………………………244
5.4.16.5.2.1. Oshi-komu ‘push-put.into’ …………… 244
5.4.16.5.2.2. Tataki-komu ‘hit-put.into’ ……………. 247
5.4.17. Type 17 Nuri-tsukeru ‘smear-apply’ verbs:
means compound 2 ……………………………. 250
5.4.17.1. Members ………………………………………………… 250
5.4.17.2. Semantics ………………………………………………. 251
5.4.17.3. Syntactic patterns ……………………………………….. 251
5.4.17.4. Causal and aspectual patterns …………………………… 252
5.4.17.5. Semantic representations ……………………………….. 252
5.5.17.5.1. Middle compounds:
Nuri-tsukeru ‘smear-apply’ .…………………253
5.4.17.5.2. Pure means compounds:
Nui-tsukeru ‘sew-attach’ .……………………255
5.4.18. Type 18 Ooi-kakusu ‘cover-hide’ verbs: means compound 3 …258
5.4.18.1. Members ………………………………………………… 258
5.4.18.2. Semantics ……………………………………………….. 258
5.4.18.3. Syntactic patterns ……………………………………….. 259
5.4.18.4. Causal and aspectual patterns …………………………… 259
5.4.18.5. Semantic representations ……………………………….. 260
5.5. Digression on compounds ……………………………………………. 262
5.6. Summary ……………………………………………………………... 266
5.6.1. Locative alternation …………………………………………….. 267
5.6.2. Semantic structures of each argument linking construction …….268
5.6.2.1. L-linking construction …………………………………… 269
5.6.2.2. I-linking construction …………………………………….. 270
5.6.2.3. FA-linking construction ……………………………………271
5.6.2.4. GA-linking construction …………………………………. 272
5.6.2.5. GO-linking construction ………………………………….. 273
5.6.3. Comparison with English ………………………………………. 273
6. Verbs of removing …………………………………………………………276
6.1. Introduction ………………………………………………………….. 276
6.2. Basic linking patterns …………………………………………………276
6.3. Japanese Genitive no and the semantic relation
between figure and ground ………………………………………. 278
6.4. Classification ………………………………………………………….281
6.4.1. Type 1 Tori-nozoku ‘remove’ verbs ……………………………. 281
6.4.1.1. Members …………………………………………………. 281
6.4.1.2. Semantics ………………………………………………… 281
6.4.1.3. Syntactic patterns ………………………………………… 282
6.4.1.4. Causal and aspectual patterns …………………………… 283
6.4.1.5. Semantic representations …………………………………. 284
6.4.2. Type 2 Dasu ‘take out’ verbs:

6
 spatial caused-transfer verbs with a specified direction ……… 287
6.4.2.1. Members …………………………………………………. 287
6.4.2.2. Semantics ………………………………………………… 287
6.4.2.3. Syntactic patterns ………………………………………….287
6.4.2.4. Causal and aspectual patterns …………………………… 289
6.4.2.5. Semantic representations …………………………………. 290
6.4.3. Type 3 Katazukeru ‘clear’ verb …………………………………292
6.4.3.1. Member ………………………………………………….. 292
6.4.3.2. Semantics …………………………………………………. 292
6.4.3.3. Syntactic patterns ………………………………………….292
6.4.3.4. Causal and aspectual patterns ……………………………. 293
6.4.3.5. Semantic representations …………………………………. 294
6.4.4. Type 4 Haku ‘sweep’ verbs ……………………………………. 295
6.4.4.1. Members ………………………………………………….. 295
6.4.4.2. Semantics …………………………………………………. 295
6.4.4.3. Syntactic patterns ………………………………………….296
6.4.4.4. Causal and aspectual patterns …………………………….. 296
6.4.4.5. Semantic representations …………………………………. 298
6.4.5. Type 5 Muku ‘peel’ verbs ………………………………………300
6.4.5.1. Members ………………………………………………….. 300
6.4.5.2. Semantics …………………………………………………. 300
6.4.5.3. Syntactic patterns ………………………………………….300
6.3.5.4. Causal and aspectual patterns ……………………………. 302
6.4.5.5. Semantic representations …………………………………..303
6.4.6. Type 6 Hagasu ‘peel’ verbs …………………………………….305
6.4.6.1. Members ………………………………………………….. 305
6.4.6.2. Semantics …………………………………………………. 305
6.4.6.3. Syntactic patterns ………………………………………….305
6.4.6.4. Causal and aspectual patterns …………………………….. 306
6.4.6.5. Semantic representations …………………………………..307
6.4.7. Type 7 Arau ‘wash’ verbs ……………………………………… 308
6.4.7.1. Members ………………………………………………….. 308
6.4.7.2. Semantics …………………………………………………. 308
6.4.7.3. Syntactic patterns ………………………………………….309
6.4.7.4. Causal and aspectual patterns …………………………….. 310
6.4.7.5. Syntactic representations …………………………………. 312
6.4.8. Type 8 Nusumu ‘steal’ verbs …………………………………… 315
6.4.8.1. Members ………………………………………………….. 315
6.4.8.2. Semantics …………………………………………………. 315
6.4.8.3. Syntactic patterns ………………………………………….316
6.4.8.4. Causal and aspectual patterns ……………………………. 317
6.4.8.5. Semantic representations …………………………………. 318
6.4.9. Type 9 Tsui.hou-suru ‘expel’ verbs ……………………………..321
6.4.9.1. Members ………………………………………………….. 321
6.4.9.2. Semantics …………………………………………………. 321
6.4.9.3. Syntactic patterns ………………………………………….321
6.4.9.4. Causal and aspectual patterns …………………………….. 323
6.4.9.5. Semantic representations …………………………………. 324
6.4.10. Type 10 Ha.mon-suru ‘excommunicate’ verbs ……………….. 324
6.4.10.1. Members ………………………………………………… 324

7
 6.4.10.2. Semantics ……………………………………………….. 325
6.4.10.3. Syntactic patterns ……………………………………….. 325
6.4.10.4. Causal and aspectual patterns …………………………….326
6.4.10.5. Semantic representations ……………………………….. 327
6.4.11. Type 11 Kai.hou-suru ‘liberate’ verbs ……………………….. 327
6.4.11.1. Members ………………………………………………… 327
6.4.11.2. Semantics ………………………………………………. 327
6.4.11.3. Syntactic patterns ……………………………………….. 328
6.4.11.4. Causal and aspectual patterns …………………………… 328
6.4.11.5. Semantic representations …………………………………331
6.4.12. Type 12 Toku ‘relieve’ verbs …………………………………..331
6.4.12.1. Members ………………………………………………… 331
6.4.12.2. Semantics ……………………………………………….. 332
6.4.12.3. Syntactic patterns ……………………………………….. 332
6.4.12.4. Causal and aspectual patterns ………………………….. 333
6.4.12.5. Semantic representations ……………………………….. 338
6.4.13. Type 13 Kai.nin-suru ‘dismiss’ verbs ………………………… 339
6.4.13.1. Members ………………………………………………… 339
6.4.13.2. Semantics ……………………………………………….. 339
6.4.13.3. Syntactic patterns ……………………………………….. 339
6.4.13.4. Causal and aspectual patterns …………………………… 340
6.4.13.5. Semantic representations………………………………… 342
6.4.14. Type 14 Sen.patsu-suru ‘wash hair’ verbs:
ground incorporation ……………………………… 343
6.4.14.1. Members ………………………………………………… 343
6.4.14.2. Semantics ……………………………………………….. 343
6.4.14.3. Syntactic patterns ………………………………………. 344
6.4.14.4. Causal and aspectual patterns …………………………… 344
6.4.14.5. Semantic representations ………………………………. 345
6.4.15. Type 15 Hai.sui-suru ‘drain water’ verbs
figure incorporation 1 ……………………………. 346
6.4.15.1. Members ………………………………………………… 346
6.4.15.2. Semantics ………………………………………………. 346
6.4.15.3. Syntactic patterns ……………………………………….. 347
6.4.15.4. Causal and aspectual patterns …………………………… 348
6.4.15.5. Semantic representations ………………………………. 349
6.4.16. Type 16 Jo.setsu-suru ‘clear of snow’ verbs:
figure incorporation 2 …………………………….. 349
6.4.16.1. Members ………………………………………………… 349
6.4.16.2. Semantics ……………………………………………….. 349
6.4.16.3. Syntactic patterns ……………………………………….. 350
6.4.16.4. Causal and aspectual patterns …………………………… 351
6.4.16.5. Semantic representations ……………………………… 352
6.4.17. Type 17 Nuki-dasu ‘pull.out-take.out’ verbs:
means compound 1 with specified direction ………. 354
6.4.17.1. Members ………………………………………………… 354
6.4.17.2. Semantics ……………………………………………….. 354
6.4.17.3. Syntactic patterns ……………………………………….. 355
6.4.17.4. Causal and aspectual patterns …………………………… 356
6.4.17.5. Semantic representations ……………………………….. 357

8
 6.4.17.5.1. Middle compounds:
Nuki-dasu ‘pull.out-take.out’ ………………..357
6.3.17.5.2. Pure means compounds ……………………….. 359
6.3.17.5.2.1. Mochi-dasu ‘take-take.out’………………..359
6.4.17.5.2.2. Shime-dasu ‘shut-take.out’ ………………. 362
6.4.18. Type 18 Arai-otosu ‘wash-remove’ verbs:
means compound 2 ……………………………….. 365
6.4.18.1. Members ………………………………………………… 365
6.4.18.2. Semantics ………………………………………………. 365
6.4.18.3. Syntactic patterns ……………………………………….. 366
6.4.18.4. Cauasl and aspectual patterns …………………………… 366
6.4.18.5. Semantic representations ……………………………….. 366
6.4.18.5.1. middle compounds: Nusumi-toru ‘steal-take’ …. 367
6.4.18.5.2. Pure means compounds …………………………369
6.4.18.5.2.1. Sori-otosu ‘shave-remove’……………….. 369
6.4.18.5.2.2. Arai-otoru ‘wash-remove’ and
damashi-toru ‘deceive-take’ …………… 371
6.4.19. Type 19 Haki-kiyomeru ‘sweep-cleanse’: means compound 3 .. 376
6.4.19.1. Members ………………………………………………… 376
6.4.19.2. Semantics ………………………………………………. 376
6.4.19.3. Syntactic patterns ……………………………………….. 376
6.4.19.4. Causal and aspectual patterns ………………………….. 377
6.4.19.5. Semantic representations ………………………………. 377
6.4.20. Other verb type ………………………………………………. 380
6.5. Summary …………………………………………………………….. 381
6.5.1. Semantic structures of each argument linking construction …… 382
6.5.1.1. A-linking construction …………………………………… 382
6.5.2.2. G-linking construction …………………………………… 383
6.5.1.3. RA-linking construction ………………………………….. 385
6.5.1.4. FA-linking construction ………………………………….. 386
6.5.1.5. GA-linking construction …………………………………. 387
6.5.2. Differences between English and Japanese ……………………….. 388
6.5.3. Asymmetries between verbs of putting and removing
found in Japanese ………………………………………….. 388
7. Conclusion ………………………………………………………………….398

REFERENCES ………………………………………………………………. 403

APPENDIX A Senses of the constructions and situation types ………….. 410

APPENDIX B Levin’s classification of verbs of putting ………………….. 444
APPENDIX C List of classes of verbs of putting in Japanese ……………. 447
APPENDIX D List of verbs of putting in Japanese ………………………. 450
APPENDIX E Levin’s classification of verbs of removing ………………. 464
APPENDIX F List of classes of verbs of removing in Japanese ………….. 468
APPENDIX G List of verbs of removing in Japanese …………………….. 472
APPENDIX H List of classes of verbs of putting and
removing in Japanese…………………………… 484

9
ABSTRACT

Aspect, described by Comrie (1976: 3) as the ‘temporal structure of events’, and

argument structure are two important facets of verbal semantics. Individual verbs, in
linguistic expressions, always occur with a certain tense-aspect (TA) construction
such as the Present and the Present Progressive and with a certain argument linking
construction such as the Transitive construction and the Resultative construction.
Verbal lexical semantics combined with these constructions determine the
grammaticality and acceptability of, and the interpretative sense of, a predicate phrase
as a whole. Therefore, aspect and the argument structure of verbs are fundamental
information every speaker has to know in using a certain language.
Croft (2000) represents aspect as a two dimensional model, which has a time scale
and a qualitative scale. Argument structure is derived directly from the causal
structure where the force-dynamic relationship between participants in event
determines the order of participants according to Croft (1990, 1991, 1993, 1994ab,
1995ab, 1998a, 1999a). These are ranked in the causal order and mapped into
syntactic arguments via the linking rules.
These two dimensions of verbal semantics, which are independent but related, are
represented in the causal-aspectual model (Croft 2000), which combines the two
dimensional representation of aspect and of the force dynamic causal structure of
events.
The main purpose of this thesis is to apply the causal-aspectual representation of
verbal semantics proposed by Croft (2000) to Japanese predicates. First of all, the
aspectual dimension of Japanese predicates is focused on. I analyze forty-eight
situation types of Japanese predicates in terms of their behaviour in relation to three
constructions: the Present, the Te-iru, and the Past constructions. Through an
examination of the situation types that occur in these constructions, the Present is
revealed to have four senses, the Te-iru to have eight senses, and the Past to have eight
senses.
Secondly, I focus on both the causal and aspectual structures and analyse verbs of
putting and removing in terms of the causal-aspectual model for two reasons. Firstly,
these two classes of verbs are important because they refer to situation of motion and
location which are within the essential experience of human beings. Secondly, since
causal structures with these two classes of verbs have three arguments (agent, figure,
and ground), they are more complicated than the structures involved in verbs that
denote non-causal relations or that involve only two participants. The verbs are
subcategorised mainly according to the linking constructions. Various occurrences of
verbs with the constructions are examined and their semantic structures are
represented in the causal-aspectual model. A semantic structure for each construction
is also proposed.
Finally, systematic differences between English and Japanese verbs of putting and
removing are observed and syntactic asymmetries between the two verb classes are
explained in terms of the differences between the semantic natures of the events that
they denote.

10
DECLARATION

No portion of the work referred to in the thesis has been submitted in support of an

application for another degree or qualification of this or any other university or other

institute of learning.

11
COPYRIGHT

(1) Copyright in text of this thesis rests with the Author. Copies (by any process)

either in full, or of extracts, may be made only in accordance with instructions

given by the Author and lodged in the John Rylands University Library of

Manchester. Details may be obtained from the Librarian. This page must form

part of any such copies made. Further copies (by any process) of copies made

in accordance with such instructions may not be made without the permission

(in writing) of the Author.

(2) The ownership of any intellectual property rights which may be described in

this thesis is vested in the University of Manchester, subject to any prior

agreement to the contrary, and may not be made available for use by third

parties without the written permission of the University, which will prescribe

the terms and conditions of any such agreement.

Further information on the conditions under which disclosures and exploitation may

take place is available from the Head of the Department of Linguistics.

12
ACKNOWLEDGEMENTS

I wish to express my gratitude to Professor William Croft for his patience and the
encouragement he has given me during my preparation of this thesis. A lot of his
helpful suggestions and comments, as well as his emotional support, helped me a great
deal in finishing the dissertation.
I am also indebted to the University of Manchester for granting me a University
Research Studentship and to the Committee of Vice-Chancellors and Principals of the
Universities of the United Kingdom for granting me an Overseas Research
Studentship. It was due to their generous financial support that it was possible for me
to do a PhD programme in the University of Manchester.
I am also grateful to Professor Noriko Ue for helping me academically giving me
and sending me books and articles and encouraging me. I should like to thank Kobe
College Alumnae Association for the scholarship given to me, Kazuhiro Taoka for his
consistent financial support, and the Rotary International Foundation for giving me a
chance to study in Manchester in the first place.
I am very grateful to Melvyn Cole, who carefully proofread the thesis and made
necessary stylistic amendments even though he was busy finishing his own thesis.
I also wish to express my gratitude for all my friends, dance teachers and dance
partners, who made my stay in Manchester enjoyable, and all the people who
encouraged me from outside the UK: Chiranut Sa-ngiamsak and her friends, Fuji
Kawata and her friends, Joel Daou, Ayumi Tsukiashi, Joan and Bill Woodall, Mitsue
Kishida, Steve Crowther, Yayoi Nagahama, Bob Dale, Wayne Newhouse, Mo-Ji,
Jacqui Murrey, Katerina Frantzi and her family, Ruben Romero, Veena Yamada,
Markus Baldus and his family, Graham Heron, Felix Kou, Kris Lai, Tatsuru Uchida,
Noriko Hiki, Nao Taniguchi, Yusuke, J.P. and Kirstie, Satoshi, Amy, Darius,
Christina, Matthias, Patrick, Julie, Olive, Sandy, Steve, John and Brigitte, Richard,
Yumiko, Sachie, Fumi, Soko, Sayuri, Shinako, Bilal, Ayumi and others in Manchester
Students’ Dance Society, Cadmans, Cuba Café, Copacabana, and the Kobe College
Aikido club.
Last, but not at least, I wish to express my sheer appreciation and gratitude to my
mother, Emi Taoka, who always cares for me, helps and encourages me. Without her,
I would not have completed the thesis.

13
ABBREVIATIONS

ABL Ablative
ACC Accusative
AGT agent
ALL Allative
A.OBL antecedent oblique marker
CAS cause
CHI Chinese VN suru verb
CLSS classifier
COM Comitative
DA da in Japanese (copula)
dev1 deverbalized V1 compound
dev2 deverbalized V2 compound
ENG English VN suru verb
GEN Genitive
frozen idiomatic (frozen) compound
INST Instrument
JAP Japanese VN suru verb
LINK te-linkage
LOC Locative
means means compound
NOM Nominative
OBJ Object
OBL Oblique
OJCT ni which serves like an accusative marker in English
ONMP1 non-reduplicated mimetic/onomatopoeic phrase
ONMP2 reduplicated mimetic/onomatopoeic phrase
PA passive agent
pair pair compound
PASS Passive
PAST Past
PURP purposive
QSS quotation and sound symbolism
REC Recipient
RST result marker
SBJ Subject
SCP scope marker
simple simple verb
S.OBL subsequent oblique marker
SRC Source (non-spatial)
TE-IRU te-iru in Japanese
TLOC temporal locative marker
TOP Topic

14
References:

GJD the Great Japanese Dictionary

JSD Dictionary of Synonyms in Japanese
Kouji-en Kouji-en [Japanese Extensive Dictionary]
Sakabikijiten Nihongo Sakabiki Jiten [Reverse Dictionary of Japanese]

15
1. INTRODUCTION

1.1. Verbal semantic structure

Aspect, described by (Comrie 1976: 3) as ‘temporal structure of events’, and argument

structure are well discussed topics in the semantics of verbs. In linguistic expressions,

individual verbs always occur with a certain tense-aspect (TA) construction (such as

the Present, the Past, and the Progressive) and with a certain argument linking

construction (such as the Transitive construction and the Resultative construction).

Verbal lexical semantics combined with the meaning of constructions determine the

grammaticality and acceptability of a predicate phrase and bear a certain interpretative

sense. The lexical aspect of a verb is correlated with its distributional patterns, with

the TA constructions determining what kind of the tense-aspect constructions it occurs

with and what sense it gets, and argument structure of the verb is correlated with its

argument linking patterns determining what kind of argument linking construction it

occurs in. That is, the two are important facets of verbal lexical semantics every

speaker has to know in using a certain language.

1.2. Argument linking

Linguists have made numerous attempts to analyse argument linking over the past

thirty years since Fillmore’s (1968) case grammar appeared. Argument linking is an

important component of linguistic theory which is expected to illuminate the interface

between syntax and semantics. Verbs require arguments, that is,

16
 Chapter 1 Introduction

participants of the situations1 they denote. These arguments are realised syntactically

by way of grammatical relations such as subject, object, or oblique. Are there rules as

to how semantics interacts with syntactic realisation? In what way do semantic

properties of the arguments determine the grammatical relation in which they are

expressed? These issues have been discussed in order to establish argument linking

theories that properly specify which arguments are associated with which grammatical

relations. Nevertheless, the patterns of argument linking have not been fully

explained.

There has been a near-consensus that a verb lexically carries information about

which arguments are realised in which grammatical relations. This information is the

argument structure of verbs. It is referred to argument-taking properties or

grammatically-relevant elements of meaning by Rappaport Hovav et al. (1988, 1999).

In the past, it was also argued that a verb’s lexical semantic representation should

contain a list of thematic/semantic roles which specifies the number of arguments the

verb requires and identifies the role each of the verb’s arguments plays in the event in

question. What is called the thematic role hierarchy, varying according to linguists,

then, is seen as functioning as a linking construct which determines the syntactic

realisation of these roles.

However, recent studies on argument linking (Croft (1991, 1998a), Rappaport

Hovav et al. (1988), Levin et al. (to appear: 11-17), Dowty (1991) to name a few)

criticise and admit some theoretical and empirical setbacks to the concept that the

verbal lexical representation contains an inventory of thematic roles. One of the

empirical problems of the thematic role lists is that one-to-one relationship between

1
The term “situation” refers to an incident or chain of incidents which are encoded by a single verb
plus its arguments. It includes states and processes following Croft (2000) based on Comrie (1976).

17
 Chapter 1 Introduction

semantic roles and syntactic arguments is not always maintained; one semantic role

can be associated with multiple syntactic arguments and vice versa. My specific

concern in this thesis is the former phenomenon.

This phenomenon is exemplified by the facts that many verbs realise their

arguments in various ways (verbs’ alternations) (Levin 1993) or that the order of

syntactic arguments is interchangeable in some verbs that encode symmetrical

relationship (Langaker (1991a, b), Levin et al. (to appear), Croft (1991)). Verbs’

alternations are illustrated using the locative alternation (which subsumes the

spray/load alternation) as in (1)2 and the dative alternation as in (2):

(1a) She loaded the hay onto the truck.

(1b) She loaded the truck with the hay. (Croft 1998: 34)

(2a) He threw the ball to Fiona.

(2b) He threw Fiona the ball.

In terms of thematic roles, which are notional, the three participants in the above

sentences should be identical in (a) and (b). Leaving aside the consistent realisation of

the agents (She and He), the other two participants are realised differently. The theme

(hay) is assigned to the direct object position and the location3 (truck) is realised as

oblique in (1a), while the latter is realised as the direct object and the former is

assigned an oblique role in (1b). (2a) expresses the theme (ball) in the direct object

position and the goal/recipient (Fiona) as oblique while (2b) expresses both of them as

direct object. Thus, the semantically identical sets of participants appear in two

distinct alternations, that is, two distinct syntactic realisations. The mere list of

semantic roles does not explain why they can have these two distinct alternations.

2
Examples, figures, and lists are all numbered from one in each chapter. An example (1) refers to (1) in
the present chapter. Example (1. 10) means example (10) in Chapter 1 and this approach is followed in
the other chapters. Figures and lists also follow this rule.
3
The terms, theme and location, are originally from Anderson (1971).

18
 Chapter 1 Introduction

In most recent approaches to argument linking, linguists still use thematic roles for

convenience’ sake to refer to certain participants of situations but do not regard a list

of thematic roles as a proper representation of verbal argument-taking properties.

Instead, a verbal semantic representation is supposed to have an event structure4 over

which thematic roles are defined and which contains the grammatically relevant

information relating to each verb. It is assumed that once the event structure is

properly articulated it can serve as an important part of verbs’ lexical representation as

verbs denote the situations. Therefore, argument-linking theories are concerned with

how to represent situations properly, first of all. Argument linking theories also need

mapping rules which specify how these semantic roles which are defined in event

structure are expressed in syntax. Some linguists use a thematic role hierarchy while

others use other constructs. Different linking theories have different semantic models

of event structure, notation, and mapping rules, and also account for the alternations

above, for example, in different ways. Sometimes an attempt is made to analyse and

explain multiple syntactic realisation of semantic participants of a single verb.

Croft (1990, 1991, 1993, 1994ab, 1995ab, 1998a, 1999a) has already attempted to

represent the event structure of verbs based on force-dynamic relationships instead of

thematic role hierarchies (causal analysis). His earlier representational model reflects

the behaviour of verbs concerning argument linking. Croft’s newly proposed semantic

representation of event structure, however, illustrates even more subtle semantic

differences between several alternations in which a verb occurs. The new model

combines his past causal analysis with aspectual analysis. Some linguists (Dowty

1991; Tenny 1992; Jackendoff 1996b, Rappaport Hovav et al. (1999), for example)

4
The term ‘event’ is used to refer to bounded processes in this dissertation. (This will be discussed in
Chapter 3). However, I ask readers to understand that ‘event’ in ‘event structure’ is equivalent to
‘situation’.

19
 Chapter 1 Introduction

have focused their attention on aspect and treated it as another important factor that is

related to argument linking. Croft incorporates the aspectual scene of verbal semantics,

as an independent but related dimension, into his earlier representation of event

structure by adopting the idea of “incremental theme” proposed by Dowty (1991) and

formalising the notion in his geometric representation. Croft (1999a, 2000), with his

new model, analyses English verbs extensively including problematic and

often-discussed verbs with alternations such as the locative alternation and the dative

alternation (examples (1) and (2)) as well as prototypical transitive verbs and verbs of

motion and location.

1.3. Aims of the thesis

The main purpose of the thesis is to explore the verbal semantic structure of aspect

and the argument structure in terms of a semantic representation using Croft’s (1999a,

2000) causal-aspectual representational analysis. I shall look at two different though

related dimension of verb meaning by introducing his model and demonstrating its

applicability to the Japanese language.

His approach, which is the one pursued in the dissertation, is cognitively-based. I

introduce two of the important theoretical assumptions of cognitive approaches here.

First of all, the construal of the speaker plays a major role in semantic

representation. That is, meanings involve speakers’ construals/conceptualisation of

situations. It is a subjective operation which each speaker exercises in encoding any

situation in linguistic expression. To explain it more detail, I shall first introduce

Langacker (1976)’s three levels5 of relationship between cognition and language. He

assumes that there is ‘uncoerced’ or ‘raw’ conceptual structure at the lowest level.

5
At the upper level, there is a linguistic expression. However, I only mention the two structures below

20
 Chapter 1 Introduction

This is a structure which is captured in a speaker’s cognition and can be very

complicated due to the complexity of our experience in the world. As Langacker

(1976: 320) explains, conceptual structure is universal; “cognition or conceptual

structure is essentially the same for speaker of all languages.” In the middle level,

there is coerced/construed semantic structure, whose purpose is for linguistic

expressions. This is the level which is represented by ‘linguistic’ semantic

representation or event structure which the linguists have attempted to represent. To

put a conceived situation (‘raw’ event) into a linguistic expression, one must select

“pertinent aspects of his current conceptual structures and cast them in a form

appropriate for linguistic operation” (Langacker 1976:322) or even “alter” conceptual

structures (Croft 1998a: 24)6. A language consists of a finite set of constructs and

because of this limitation, the speaker has to submit his conceptualisations to the

exigencies of the linguistic system, which is language specific. The

construals/conceptualisation is this operational process of coercion7. When we talk of

semantic representation of event structure, it refers to this semantic/coerced semantic

structure, at least, in cognitive approaches. In the cognitive accounts of semantic

structure, thus, there is a distinction between ‘raw’ conceptual structure and ‘coerced’

semantic structure, which is made by the process of construals/conceptualisation; the

speakers’ construal/conceptualisation of any ‘raw’ event is done before the semantic

structure, whose representation cognitive linguistics regard as related to aspectual

behaviours and the argument linking of a verb.

this.
6
Specifically, altering is occasionally done for the argument linking of non-canonical events. See Croft
(1991, 1993, 1998a).
7
The construal operation is open to the conventions of each language. Depending on what is available
in the language in question, speakers are accustomed to view an event in a certain way or are allowed to
have an alternative image. Langacker (1987) calls this “conventional imagery.”

21
 Chapter 1 Introduction

The second important theoretical background is that constructions as well as verbs

have meanings, therefore, the constructions as a “conventional symbolic unit”

(Langacker 1987) have semantic import and should have semantic representation.

Constructions include various linking patterns such as the Transitive construction [SBJ

VERB OBJ] and the Ditransitive construction [SBJ VERB OBJ1 OBJ2], or tense-aspect

(TA) constructions such as the Progressive or the Present tense. Verbs are always used

in these constructions, so, we need to see empirically what sense each verb gets with a

particular construction and how this is represented in the semantic model. Through

analysing verbs’ occurrences with the construction, we could also propose a rather

abstract meaning of the construction which integrates all the semantic senses it shows

with verbs and this could be possibly represented as a schematic causal-aspectual

model.

In applying Croft’s causal-aspectual representation of verbal semantics to the

Japanese language, I shall focus on verbs of putting and verbs of removing, which are

conventionally three-argument verbs that require an agent, figure and ground.

Situations with three participants are more complicated than verbs with one or two

participants in terms of force-dynamic relationships. Three-argument verbs in English,

as we already know, posit some difficult issues on various alternations in argument

linking. Moreover, the spatial situation such as location and motion is an essential

experience to human beings. Therefore, verbs of putting and removing, which are

thought of as caused-motion situations, are be of particular interest in verbal semantic

representation. It is also challenging to examine some three-argument verbs in

Japanese to see how Croft’s new model of verbal semantic representation can be

applied to Japanese and to investigate the interrelations between the verbal semantics

and argument structure of Japanese verbs.

22
 Chapter 1 Introduction

The thesis contains seven chapters. Chapter 2 briefly reviews Fukui et al.’s (1985)

argument linking study relating to Japanese verbs of putting and removing and also

Croft’s causal analysis (the old version of his argument linking theory). Chapter 3

focuses on aspect and its representation. I introduce Croft’s two-dimensional

representation of aspect and universal aspectual types. Then, I examine the

distributional behaviour of Japanese predicates for three constructions: the Present, the

Te-iru8, and the Past construction. Looking at various situation types in Japanese in

terms of what sense each type bears in the three TA constructions, I show the distinct

senses of each construction, namely its aspectual types. Situation types are

subcategorised into aspectual classes based on their distributional aspectual behaviour.

Chapter 4 describes Croft’s new semantic representation of event structure, which

combines the causal analysis (discussed in Chapter 2) with the aspectual analysis

(discussed in Chapter 3). The chapter also demonstrates how his causal-aspectual

model works in English giving some notable examples. It further describes the

methodology I have used to collect, read, and analyse the data of verbs of putting and

removing as well as giving basic Japanese grammatical information necessary to look

at the data. Chapter 5 and Chapter 6 classify verbs of putting and removing,

respectively, in terms of argument linking constructions presenting semantic

representations of each type, and also give schematic semantic representations for

each construction. Chapter 7 concludes the discussion of the thesis and proposes some

points of the causal-aspectual analysis that can be pursued in future study.

8
Te-iru is the aspectual construction as opposed to the unmarked form. Sometimes it behaves like the
English progressive, and at other times describes the resultative state of something. Detailed
explanations are given in Chapter 3.

23
2. THEORIES OF ARGUMENT LINKING

2.1. Introduction

The chapter gives a brief summary of some theories of argument linking. In the later

sections, I discuss the literature on verbs of putting and removing in Japanese and

introduce Croft’s causal analysis for argument structure, which this study is based on.

Argument linking of English has been discussed extensively in the literature by

Jackendoff (1990, 1996b, 1997), Croft (1990, 1991, 1993, 1994ab, 1995ab, 1998a,

1999a, 2000), Dowty (1991), Fillmore and Kay (1993), Goldberg (1995), Langacker

(1991ab), Levin et al. (1991, 1995), and Rappaport Hovav et al. (1988, 1993, 1998,

1999). These are recent argument linking theories which do without an inventory of

semantic roles as the event structure model. An ideal articulated argument linking

theory has to cover and explain many facets of the syntactic realisation of languages.

The multiple syntactic realisation of a single thematic role is one of them. This

includes various alternations such as the locative alternation, the dative alternation and

the causative alternation, for example. Each of these has its own verbal lexical

representation model.

Though their representational models vary, there are basically two levels of

representations in verbal meanings. One is a purely semantic and idiosyncratic

representation and the other is a representation which relates to the argument taking

properties of verbs. The lexical conceptual structure (LCS) and the

predicate-argument structure (PAS) in Rappaport Hovav et al. (1988, 1993) apparently

correspond respectively to the two levels. Even in other theories, where only one

representation is available, still two different components in the representation can be

identified. In Jackendoff (1990), the pure semantic representation is illustrated by the

24
 Chapter 2 Theories of argument linking

decompositional style of event structure while the argument structure is encoded by

stipulated A-marking. In Rappaport Hovav et al. (1995, 1998, 1999), a constant serves

as an idiosyncratic part of verbal meanings and an event structure template serves as

the structural part, which is regulated by argument-linking. Goldberg (1995), however,

includes in a verbal meaning only an idiosyncratic part and attributes properties of

argument structure to a constructional meaning. Langacker (1991a, b) and Croft (1991,

1993, 1994ab, 1998a)1 represent purely the semantic part of verbal meaning in event

structure based on force-dynamic relations and derive argument structure directly

from the event structure. However, it is possible to say that the profile or the verb

profile, which is closely associated with the syntactic realisation of verbs' arguments,

serves as another component in verb meaning.

None of the above theories can do without stipulations when linking the

representation of argument-taking properties or the structural part of verbal meaning

to the syntactic representation. These stipulations include Jackendoff's A-marking and

correspondence rules such as adjunct fusion, Rappaport Hovav et al.’s (1988) linking

rules, Goldberg's profiling and argument structure in each construction, and

Langacker's and Croft's profile. However, the last two are preferable since the

different stipulations are semantically motivated. In fact, Croft’s notion of a verbal

profile will be discussed shortly.

In the theories of argument linking, there are basically two main approaches to

analysing verbs that allow more than one alternation: the lexical rule approach and the

non-lexical approach. (See Croft 1999a, 2000 for detail.) The former posits separate

verbal representations for distinct alternations that verbs manifest while the latter

accounts for the multiple representations by distinct extra-lexical correspondence rules.

1
I shall describe Croft’s theory in more detail later in the chapter.

25
 Chapter 2 Theories of argument linking

The former can also be said to be a polysemous approach in that multiple

representations of verbs are related to each other. This approach is taken by Rappaport

Hovav and Levin, Croft, and Langacker. The last two presuppose that the profile is

also a part of verbal meanings and also their polysemous approach is not as systematic

as that of Rappaport Hovav and Levin in that they do not claim monotonicity2. The

extra-lexical approach is taken by Jackendoff and Goldberg. Jackendoff reduces the

information in verbal lexical representation and makes different correspondence rules

(including various adjunct rules) to yield different syntactic realisations. Goldberg,

who advocates the concept of constructional meanings, is a convinced defender of the

constructional approach. Her account is still extra-lexical in that she attributes the

existence of different alternants to a single verb's being fused with different

constructions. Whichever approach is taken, the two approaches are not so distinct in

terms of descriptive adequacy; there is just a difference in the components they are

applied to.

Concerning the lexical and constructional approach, as Croft (1999a, 2000) points

out, we cannot clearly attribute a whole semantic meaning of a verbal phrase either

only to the verb’s lexical entries or only to constructions, as verbs and constructions

are not separable when they appear in syntax. Moreover, Rappaport Hovav et al.

(1999) and Levin et al. (to appear) note that use of a constructional approach merely

results in a theory being articulated from a different aspect that would have been in the

case with a lexical approach.

Compared to English, argument linking in Japanese has been relatively little

discussed. The rest of this chapter deals with the discussion about verbs of putting and

2
Monotonicity in verbs’ meanings claims that verbs’ meaning can only be built up in a one way fashion
from the basic one to more complex one, not the basic one is reduced to other ones.

26
 Chapter 2 Theories of argument linking

removing in Japanese in Fukui et al. (1985). Then, I introduce the Croft’s old model of

argument linking, leaving discussion of his new model to Chapter 4.

2.2. Literature on Japanese verbs of putting and removing

Argument linking in Japanese has been discussed in general terms. Such discussions

have been concerned with what kinds of verbs take how many arguments and how

they are realised in syntax, and sometimes there has been a comparative study of

English and Japanese (Teramura 1982, Kageyama 1996, Tanaka et al. 1997,

Yoshikawa 1995). However, there has been no detailed discussion concerning

three-argument verbs, especially the locative alternation, although the subject is

generally discussed in Kageyama 1980 and Fukui et al. 1985. The work of Fukui et al.

(1985) is closer to what I am trying to do in this dissertation in that they have

proposed the conceptual structures for verbs that allow the locative alternation, so I

shall briefly discuss their arguments as a preliminary to the analysis of verbs of

putting and removing3.

Fukui et al. (1985) explain the locative alternation in the Lexical Conceptual

Structure (LCS), over which semantic roles are defined. They propose that verbs (in

English and Japanese) that appear in the locative alternation illustrated in examples

((1)-(4)) have the two shared semantic properties of (5), which will be restated in the

LCS-like term later (Fukui et al. (1985: 22-24)).

nuru/smear type (‘material-adding’ verbs)

(1a) smear paint on the wall

(1b) smear the wall with paint

3
Kageyama's (1980) approach stipulates semantic roles (semantic role approach) according to Fukui et
al. (1985). Therefore, I do not discuss it here. (See Chapter 1, where I have briefly discussed why a
semantic role approach is not adequate.)

27
 Chapter 2 Theories of argument linking

kabe-ni penki-o nuru

(2a) wall-on paint-ACC smear
‘smear paint on the wall’

kabe-o penki-de nuru

(2b) wall-ACC paint-with smear

katazukeru/clear type (‘removing’ verbs)

(3a) clear the plates from the table

(3b) clear the table of plates

teeburu-kara sara-o katazukeru

(4a) table-from plates-ACC clear

teeburu-o katazukeru4
(4b) table-ACC clear

(5) semantic conditions for the alternation

(i) it must have two arguments (other than the subject)
(ii) it must include the semantic notion of ‘Affect Y’

In order to support their account5, they compare the nuru/smear type, which allows

the alternation, with other types that do not allow it. First of all, they argue for the

condition (5i), by comparing the nuru/smear type with the maku/spray type, which

does not allow the alternation in Japanese.

mizu-o hodoo-ni maku

(6a) water-ACC sidewalk-on spray
spray water on the sidewalk

*hodoo-o mizu-de maku

(6b) sidewalk-ACC water-with spray
spray the sidewalk with water
(Fukui et al. (1985: 38))

They conclude that the nuru/smear type requires two arguments (other than the

subject) while the maku/spray type has only one. They present two types of syntactic

evidence: deletability of arguments and compounding.

4
Japanese does not have the equivalent linking to (3b).
5
I only present the discussion of nuru/smear type because katazukeru/clear type is also argued to have
the same semantic condition of (5) above and is shown to have same syntactic behaviour to verify this
as the alternating type of putting have, which will be discussed below.
28
 Chapter 2 Theories of argument linking

The first test, deletability of arguments, supposes that if a required argument of the

verb is not represented, there is a clear intuition that something is missing. They test

haru (‘hang’ a non-alternating verb in Japanese) and hari-tsukusu (a compound

(‘hang’-‘exhaust’ that means ‘hang completely’ which allows the with-variant) to

support their idea.

haru (‘hang’--one-argument)

ano posutaa-o haru

(7a) that poster-ACC hang (no sense of missing)
‘hang that poster’

kabe-ni haru
(7b) wall-on hang (sense of missing)
‘hang on the wall’
(Fukui et al. (1985: 26))

It is concluded that the only argument is Material (fulfilled by ‘poster’)6.

hari-tsukusu (‘hang-completely’--two-argument)

kabe-o hari-tsukusu
(8a) wall-ACC hang-exhaust (sense of missing)
‘completely hang the wall’

posutaa-de hari-tsukusu
(8b) poster-with hang-exhaust (sense of missing)
‘completely hang with posters’

Thus, it is concluded that the two arguments are the wall (Entity/location) and the

poster (Material).

The second test, compounding, is to see the grammaticality when verbs are

compounded with their argument. The compounding is a process of combining an

infinitive verb with an argument required by the verb in order to make a noun phrase

such as (kitte-atsume ‘stamp-collecting’ (from ‘collect stamps’) and shoku-sagashi

6
I think (7a) is actually a case of Definite Null Instantiation (this will be discussed in Chapter 4).
Speakers should know where the poster is hung. I would say it is difficult to say this without contexts.
However, it is at least true that wall in the example is less necessary than poster.

29
 Chapter 2 Theories of argument linking

‘job hunting’ (from ‘hunt job’). The nuru/smear type (which is supposed to have two

arguments) should allow compounds for each of the argument and the non-alternating

type (which is supposed to have one argument) should allow only one compound for

the argument. This is proved to be true.

nuru/smear type (alternating):

(9a) penki-nuri (paint-smearing)
(9b) kabe-nuri (wall-smearing)

maku/spray type (non-alternating):

(10a) mizu-maki (water-spraying)
(10b) *hodoo-maki (sidewalk-spraying)
(Fukui et al. (1985: 29))

Thus, condition (5i) is shown to be complied with in the two syntactic tests above.

However, condition (5i) is not sufficient since it does not preclude the oku/put type

verbs, which require two arguments other than the subject, from the alternation. They

argue that the nuru/smear (alternating) type has the additional meaning of

‘Affectedness.’

nuru/smear example:
Taroo-ga kabe-ni akapenki-o nutta.
(11a) Taroo-NOM wall-on red.paint-ACC smeared
‘Taro smeared red paint on the wall.’

Taroo-ga akapenki-de kabe-o nutta.

(11b) Taroo-NOM red.paint-with wall-ACC smeared
‘Taro smeared the wall with red paint.’
(Fukui et al. (1985: 41))

They say that the two sentences above mean that Taro carried out the action of

smearing using the Material ‘red paint’; and, as the result of that action, the nature of

the Entity/Location, kabe (‘wall’), was ‘affected’ (i.e., it changed colour). However,

they say the maku/spray type and the oku/put type do not imply any clear sense of

“affectedness.” For example, putting an entity in a certain location does not change

30
 Chapter 2 Theories of argument linking

the nature of the location7.

Following these arguments, they propose the LCS for the three type of verbs:

(12) a. LCS of oku ‘put’: OKU x at some place y.

(it has two arguments; x is simply an object of the action OKU)

b. LCS of maku ‘spray’: Realize the action MAKU by using the Material x.
(material is ‘inherently necessary’ for the realization of the action)

c. LCS of nuru ‘smear’: Realize the action NURU by using the Material x &
Affect y

(Fukui et al. (1985: 42))

(it has two arguments, material is ‘inherently necessary8’ for the realization of
action; and the meaning also implies some effect on y, which is an important
part of the meaning of nuru)

Then, the conditions (5i) and (5ii) can be restated as follows.

(13) Conditions for the alternation

(i) The verb takes two arguments x, y in its LCS; and

(ii) One of its arguments (y) is affected by the action represented by the meaning of
the verb (‘Affect y’).
(Fukui et al. (1985: 43))

The oku/put type satisfies (i) but not (ii). The maku/spray type fails to satisfy (i), so

condition (ii) is irrelevant.

Fukui et al. (1985) also refer to the fact that English has many more alternating

verbs than Japanese. They say, however, that Japanese is rich in morphology and can

make the alternation possible with non-alternating verbs by attaching, for example,

tsukusu ‘exhaust’.

7
This idea is different from Dowty 1991: the location is the incremental theme (the argument that
measures out the event) when it is the direct object. Actually, if we spray something onto a location
such that it is covered with it, we can say it is affected. The notion of affectedness should be clarified in
their argument.
8
“Material is ‘inherently necessary’ for the realisation of action” means that the verbs have special
semantic restriction on Material, which the oku/put type lacks.

31
 Chapter 2 Theories of argument linking

haru ‘hang’ vs. hari-tsukusu ‘hang-exhaust’

haru
kabe-ni posutaa-o haru
(14a) wall-on poster-ACC hang
‘hang posters on the wall’

*kabe-o posutaa-de haru

(14b) wall-ACC poster-with hang
‘hang the wall with posters’

hari-tsukusu
kabe-ni posutaa-o hari-tsukusu
(15a) wall-on poster-ACC hang-exhaust
‘hang all the posters on the wall’

kabe-o posutaa-de hari-tsukusu

(15b) wall-ACC poster-with hang-exhaust
‘completely hang the wall with posters’

maku ‘spray’ vs. maku-tsukusu ‘spray-exhaust’ (Fukui et al. (1985: 44))

maku
mizu-o hodoo-ni maku
(16a) water-ACC sidewalk-on spray
‘spray water on the sidewalk’

*hodoo-o mizu-de maku

(16b) sidewalk-ACC water-with spray
‘spray the sidewalk with water’

maki-tsukusu
mizu-o hodoo-ni maki-tsukusu.
(17a)9 water-ACC sidewalk-on spray-exhaust
‘spray all the water on the sidewalk.’

hodoo-o mizu-de maki-tsukusu.

(17b) sidewalk-ACC water-with spray-exhaust
‘spray the sidewalk completely with water.’

It is argued that -tsukusu ‘exhaust’ in the above examples adds the semantic

component of ‘Affect y’ to the LCS of non-alternating verbs. As a result, the verbs

combined with tsukusu can have the alternation. Returning to the difference between

English and Japanese, the semantic element ‘Affect y’occurs in the LCS of English

verbs more often than in Japanese verbs, which often acquire that semantic element by

9
The example is mine as Fukui et al. do not mention clearly the locative version of maki-tsukusu.

32
 Chapter 2 Theories of argument linking

morphology such as -tsukusu.

Thus, Fukui et al. (1985) explain the locative alternation in terms of conceptual

structure. However, the crucial problem is that they propose the semantic properties of

the verbs that manifest the alternation, but do not discuss the subtle semantic

differences between the locative variant and the with-variant. In the discussion of

nuru/smear ((11)), they say that Location is affected in both of the variants. According

to their explanation, verbs that alternate have the meaning of affecting the Location

and both of the two variants imply this “affected” meaning. This is against the

standard view that the with-variant has a holistic interpretation, which the locative

variant lacks (at least, in English).

I can show counterexamples even in Japanese. According to them, hari-tsukusu

(‘hang’-‘exhaust’) and maki-tsukusu (‘spray’-‘exhaust’) should have an “affect y”

interpretation. However, they do not imply that the location is “affected” in the

locative variant since the notion of “affectedness” would have to be changed or

respecified in order for them to do so. We can say the following in Japanese. The

structure of the bold part is the same as (15a) above.

Kabe ni posutaa o hari-tsukushi-ta ga

(18a) wall ALL poster ACC hang-exhaust-PAST but

mada supeisu ga at-ta node soko ni tapesutorii o kake-ta.

still space NOM is-PAST so there ALL tapestry ACC hang-PAST

I hung all the posters on the wall, but there was still space there.
So, I hung a tapestry there.

If the wall is not fully covered, or has not changed its nature, I do not think we can

interpret that the wall changes its nature by being partly covered with posters. The

same expression is not possible with the with-variant. The bold part has the same

argument linking as (15b) above.

33
 Chapter 2 Theories of argument linking

*Kabe o posutaa de hari-tsukushi-ta ga

(18b) wall ACC poster INST hang-exhaust but

mada supeisu ga at-ta node soko ni tapesutorii o kake-ta.

still space NOM is-PAST so there ALL tapestry ACC hang-PAST

I completely hung the wall with posters, but there was still space there.
So, I hung a tapestry there.

The other example is maki-tsukusu ‘spray-exhaust’. The bold parts are the same as

(17a) and (17b), respectively.

Mizu o hodoo ni maki-tsukushi-ta ga

(19a) water ACC sidewalk ALL spray-exhaust-PAST but

mada sukoshi kawaita tokoro ga aru.

still little dry part NOM is

I sprayed all the water on the sidewalk,

but there is still some dry part there.

*Hodoo o mizu de maki-tsukushi-ta ga

(19b) sidewalk ACC water INST spray-exhaust-PAST but

mada sukoshi kawaita tokoro ga aru.

still little dry part NOM is

I sprayed the sidewalk completely with water,

but there is still some dry part there.

The sidewalk partly wet with water cannot be interpreted as changing its nature in

(19a). Actually, Fukui et al. even say that maku ‘spray’ does not imply any clear sense

of ‘affecteness.’ According to them, in example (16a), “hodoo (‘sidewalk’) simply

indicates the location at which the action of spraying took place by using the Material

mizu (‘water’).” Is the sidewalk less affected by spraying water on it (encoded in the

locative variant of maku ‘spray’) than by spraying all the water on it (encoded in the

locative variant of maki-tsukusu (‘spray-exhaust’)? I do not think this is plausible. I

would say it is better to attribute the affecteness of the location to the with-variant, not

the alternation, following the standard view. Also I would not think –tsukusu ‘exhaust’

adds the “affected y” component to the LCS of the verbs. Rather, it assigns the

34
 Chapter 2 Theories of argument linking

“affectedness” to the direct object. So, “affectedness” is assigned to the Material in the

locative variant and to the Location in the with-variant.

The article focuses on the locative alternation in Japanese, which is rare in the

literature, and also explains the alternation in the conceptual structure. What the

present study, on the other hand, will reveal are the subtle semantic differences

between the two variants of the locative alternation, which the article fails to do. Fukui

et al. (1985) also have a list of some related Japanese simple verbs and compound

verbs with their syntactic configurations, which is similar to the present study.

However, the study in this thesis concentrates on the two verb classes only and has a

much larger number of verbs. It will also look more closely at syntactic configurations

other than the locative and with-variants. Now, I introduce the framework adopted in

this thesis, which is Croft’s causal analysis.

2.3. Croft’s causal analysis (1991, 1993, 1994ab, 1998a)

I shall discuss Croft’s older version of his argument linking theory here. Though his

basic idea has been maintained, some notation of his semantic representation has

changed since 1991 and I shall mostly cite his recent version (1998a).

His argument linking theory, which is based on cognitive linguistics, also uses

some theoretical constructs such as event structure. What is distinctive in his analysis

is that he adopts two semantic notions, which are (i) force-dynamic relations and (ii)

the verbal profile (cf. Langacker 1991a, b). Croft (1991, 1993, 1994ab, 1998a)

dispenses with thematic role hierarchy by employing force dynamic relations, and

another layer of representation such as syntactic arguments or super-roles (cf. Dowty

(1991)’s proto-roles) by employing the verbal profile, and makes linking rules much

simpler.

35
 Chapter 2 Theories of argument linking

2.3.1. Event structure

Croft’s original semantic representation for event structure which is considered as

the foundation for argument linking, goes beyond the other theories such as

Jackendoff (1990), Rappaport Hovav et al. (1988), who regard event structure as a

representation from which thematic roles are derived, by proposing a semantic

representation from which linking properties of arguments are directly derived (cf.

Langacker 1991); he does not use a separate argument structure such as Jackendoff's

A-marking or Rappaport Hovav et al.’s PASs. Focusing on relational nature between

participants, he argues for an event structure which can encode the ranking of

participants. His event structure is based on the causal chain model, which is not

dissimilar to Langacker's action chain. In the model, it is force-dynamic relationships

that count; the direction of transmission of force determines the order of participants.

If one participant acts on another force-dynamically, then the acting participant is

thought to outrank the other (is antecedent to the other). For example, in the event of

John’s breaking the vase as in John broke the vase, John causes the vase to break,

therefore, John is antecedent to the vase in the causal chain. In this example, the order

of participants is straightforward.

(20) John broke the vase. (cf. Croft 1998a: 48 causative verb type)

x y (y) (y)
* * * * * * * * * * * * * * break * * * * * * * * * * * *
Sbj Obj
John vase

The force-dynamic relation between the two participants is indicated by the first

row. They are distinguished with labels such as x and y. The causal chain not only

orders the participants in terms of the transmission of force but also individuates the

event as consisting of subevents or segments at the proper level of granularity (cf.

Croft 1991: 163-166). In the above sentence, for example, the process of the vase

36
 Chapter 2 Theories of argument linking

undergoing the change has its own subevent (CHANGE segment), the state of the vase

broken also has a distinct subevent (STATE segment) as well as the force-dynamic

relation between John and the vase (CAUSE segment)10. Every segment has two

participants on each end, which are called the initiator (for antecedent participant) and

the endpoint (for subsequent participant) of the segment. In cases of

non-force-dynamic subevents such as in the change segment and the state segment,

the initiator and the endpoint refer to the same participant and do not need to be

realised as distinct arguments in syntax. The repeated participant is represented by

parenthesised constants such as "(y)" in the notation.

As Langacker (1991ab) argues, events do not always denote a prototypical

force-dynamic relationship. Some events seem to exhibit no force-dynamic

relationship between participants or the relationship is less clear than in prototypical

cases. Non-force-dynamic events are those denoted by statives of symmetrical

relations (such as Marsha resembles Hilda as discussed in Langacker (1991a: 222),

stative mental verbs11 and those of spatial relationship in which two participants are

labeled as figure (theme) and ground (location). Croft also argues that the construal of

the event on the part of the speaker explains these non-force-dynamic relations. The

"raw" event structure (conceptual structure) of non-causal relations is possibly coerced

to asymmetrical causal relations to encode the event in question in linguistic

expressions which are available in each language through a construal operation. If a

language allows, they may be realised equally in syntax as either double subjects or

double objects. These ambiguities of the syntactic realisation of non-causal events are

motivated by the ambiguous semantic properties of events; because events themselves

10
In his new notation (1998a), two parallel lines indicate process (CHANGE), a unbroken line indicates
state (STATE), and a round dotted line indicates force-dynamic relationship (CAUSE).
11
For details, see Croft (1991, 1993) and for mental verbs in Japanese see Taoka (1995)).

37
 Chapter 2 Theories of argument linking

are indeterminate in terms of force-dynamic relations between participants, the

argument linking of events which is based on their semantic structure turns out to be

ambiguous. Thus, the force-dynamic relation representing the event structure, as Croft

argues, is not invalid.

Concerning the order of figure and ground, however, Croft (1991, 1998a) claims

that there is a universal Figure-first coercion pattern. That is, the figure (theme) is

coerced as antecedent to the ground (location). He (1998a: 39) argues that this

construal is semantically motivated by the fact that it is the figure that is directly acted

upon by the initiator. For example, in John put the blanket over the sofa, the blanket

(figure) and the sofa (ground) are in a non-causative spatial relationship, which is

brought about by an actor. However, they are coerced into asymmetrical relations so

that the blanket comes first being realised as direct object12. It is easily interpretable

that John acts on the blanket (such as taking hold of it) and changes its location to

over the sofa.

(21) John put the blanket over the sofa. (cf. Croft 1991: 200)

x y (y) z
* * * * * put * * * * * * * * * * * * * • • • over• • • •
Sbj Obj S.Obl
John blanket sofa

Following Langacker (1987, 1991ab), Croft's semantic representation of verbs

involves a distinction between profiled parts and unprofiled parts in event structure. A

whole causal chain serves as the base or frame (cf. Fillmore) and verbs are assumed to

denote some parts or all of it. The segment parts which are denoted lexically by the

verb are called the verb profile and indicated by "* * *" in the representation (see (20)

12
For other various examples of Figure-first coercion, see Croft (1991: 198-206).

38
 Chapter 2 Theories of argument linking

and (21) above13). These are the part/parts which are asserted by the verb semantically.

Adpositions can additionally profile other segment/segments of the causal chain,

which are indicated by • • • (Croft 1998a: 59-60). In (21), the cause segment and the

change segment are profiled by the verb and the state segment is profiled by the

preposition.

The lexical representation of verbs is posited as above and the next discussion is

how the linking of arguments is reflected from the event structure representation.

Croft presents four universal linking rules as follows14 (Croft 1998a: 24):

(22) Linking rules:

1. The verbal profile is delimited by Subject and Object (if any)

2. Subject <15 Object
3. Antecedent Oblique < Object < Subsequent Oblique
4. Subject < Incorporated Noun < Object (if any)

The verb profile which Croft regards as a part of semantic representation is actually

stipulated based on the first linking rule. However, the stipulation itself is not a

problem as long as it has an explanatory power which can provide consistent and

proper predictions concerning argument linking. We cannot help to a certain extent

stipulating argument linking properties from the way they are actually realised in

syntax; other linguists also use the stipulation (Jackendoff's A-marking, Rappaport

Hovav et al.’s PASs, and Goldberg's argument structure in constructional meanings).

The second rule is not inconsistent with Langacker's (1991) claim that the head of

the profiled action chain will be the subject and the tail of the profiled action chain

will be the object when there is an asymmetrical interaction between two participants.

That is, the subject always precedes the object in the force-dynamic chain. The

13
The second row specifies the part which is profiled or denoted by the verb.
14
As the fourth rule is not related to the English language nor the Japanese, I shall not discuss it further.
15
X<Y indicates that "X antecedes Y in the causal chain".

39
 Chapter 2 Theories of argument linking

initiator of the action is unmarkedly encoded as the subject rather than the receiver of

the action.

The third rule involves Croft's original idea about oblique. He distinguishes two

types of oblique, the antecedent oblique and the subsequent oblique, in his causal

order hypothesis (Croft 1991: 186), according to its positioning relative to the object

in the causal chain; the antecedent oblique precedes the object, which in turn precedes

the subsequent oblique. It is agreed among linguists that the grammatical relation

hierarchy is SBJ < OBJ < OBL, which corresponds to the order of participants in the

causal chain in Croft. However, only subsequent obliques comply with these

hierarchical relations. Antecedent obliques which come before the object are relatively

marked forms. A good example of antecedent obliques is with, the instrument case

marker in English, as in I broke the coconut with a hammer. The order of the causal

chain is "I > hammer > coconut". With (antecedent marker) case-marks the hammer

which precedes the coconut, which is realised as direct object. Croft (1995b, 1998a:

40 etc) summarises that antecedent obliques in English are with (comitative, manner,

instrument), by (passive agent, means), of (theme as in Bears stripped trees of bark)

and nonspatial from (cause) and out of (cause) while the subsequent obliques are to

(result, recipient), for (benefactive) and the spatial Path prepositions16. However, some

prepositions do not mark participants in an event but only refer to a circumstantial

setting (time, place) in which an event takes place; He danced at five o’clock or She

baked a cake in the kitchen. Setting is like a stage where an event (which is

16
Croft defends the antecedent/subsequent oblique distinction giving cross-linguistic evidence (Croft
1991: 187-190) and developmental evidence (Croft 1998a: 40) for syncretisms among oblique markers
of the same type. In short, a single oblique marker is likely to have several functions to mark thematic
roles which belong to the same category (by is used for passive agent and means, which belong to
antecedent thematic roles) or children wrongly use one oblique marker in the place of another which
belongs to the same category.

40
 Chapter 2 Theories of argument linking

represented as the causal chain) unfolds. Therefore these prepositional phrases are

irrelevant to the causal analysis.

The grammatical relations are indicated in the third row of the representation.

"Sbj" stands for Subject, "Obj" for Object, "S.Obl" for subsequent oblique, and

"A.Obl" for antecedent oblique. The final fourth row represents the participants

referred to.

2.3.2. The verbal profile

To illustrate the operation of the verbal profile, the spray/load alternation and the

caused-motion construction will now be discussed.

2.3.2.1. Spray/load alternation

The verb profile as part of the semantic representation treats the locative alternation

without positing another layer of representation for argument-predicate structure.

Croft (1998a: 39) gives the following two geometric representations for the two

variants of spray:

(23) Jane sprayed paint on the wall. (the locative variant)

x y (y) z
* * * * * spray * * * * * * * * * * * • • • on • • • • • •
Sbj Obj S.Obl
Jane paint wall

(24) Jane sprayed the wall with paint. (the with-variant)

x y (y) z
* * * * * spray * * * * * * * * * * * * * * * * * * * * *
Sbj A.Obl Obj
Jane paint wall

According to the linking rules, spray of the locative variant is represented as having

the cause segment and the change segment as its verb profile, while it covers all the

41
 Chapter 2 Theories of argument linking

segments of the event as the verb profile in the with-variant. Wall is marked as a

subsequent oblique in the former variant and the paint is marked as an antecedent

oblique in the latter variant. Moreover, the assignment of direct object is different,

which is the reason why their coverage in the verb profile is different. Croft argues

that the different degree of affectedness of the end participant (wall) is attributed to

the difference of profiling. As we can see from the semantic representations of the two

sentences, that for (24) is profiled wholly by the verb including the final segment

which denotes the achievement state while this is not the case in (23). The final

subevent (achievement) in the causal chain is profiled complementarily by the

prepositional phrase and not by the verb. That is, “a lesser degree of affectedness

indicates a less complete achievement of the event, which is expressed by less of the

event being profiled by the verb” (Croft 1998a: 44)17.

Croft also defends the validity of the verbal profile by demonstrating that it is "not

idiosyncratically stipulated." As above, he shows that differences in verbal profile on

the same event base/frame, which cause different argument patterns, are motivated by

semantic differences (Croft 1991, 1994ab, 1998a). As above, the stipulated verb

profile can predict the semantic difference18.

Thus, the two variants with the semantic difference are attributed to the two

17
Croft (1998a: 43) also points out that some of the spray/load verbs such as spray, spatter, and scatter
occur in simple transitive constructions with the figure being assigned Accusative as in The broken fire
hydrant sprayed water all afternoon. This case leaves a path (trajectory) unspecified. This also supports
the idea that the final segment is profiled not by the verb, but by the prepositional phrase. The
prepositional phrase (not the verb) profiles the final segment in the semantic representation. Another
example to support the relation between a difference in profile and a different degree of affectedness is
the conative alternation. For details, see Croft (1998a: 45).
18
Croft's different stipulations for the verb profile correctly predict different semantic differences.
Unlike Croft's verb profile, Jackendoff’s (1990) A-marking theory has no correlation with semantic
interpretation; different A-marking stipulation occur in different conceptual structures and there is no
correlation between the two.

42
 Chapter 2 Theories of argument linking

different verb profiles. On the other hand, the identical frame/base confirms that they

denote the “same” force-dynamic relation of event structure. His theory does not refer

to another layer of predicate-argument representation nor posit two absolutely distinct

meanings of the verb.

2.3.2.2. The caused-motion construction

Croft’s theory explains the caused-motion construction by taking the view that the

prepositional phrase profiles the final state segment. Actually, the caused-motion

construction is not dissimilar to that with a verb of putting in that the actor causes the

figure to move to the ground (or from the ground). Let us take the following example:

(25) Bill pushed the piano into the orchestra pit.

x y (y) z
* * * * * push * * * * * * * * * * * • • • into• • • • • •
Sbj Obj S.Obl
Bill piano pit

The two segments of the verb profile of push are marked with “* * * * * *”. The third

segment marked with “• • • • •” is the prepositional profile, which is occupied by into.

The representation makes clear which part is the meaning contributed by the verb19.

2.3.3. Summary

Croft's semantic representation of verbs has two parts; the causal chain and the verb

profile. The causal chain is based on the force-dynamic relations between participants,

which is of a relational nature. This enables us to derive the argument linking

19
In this sense, the possible analysis of the caused-motion construction can be said to be similar to that
of Jackendoff (1990) or Goldberg (1995), who do not attribute the change-of- location meaning of the
whole phrase to the verbs' lexical representation of verbs.

43
 Chapter 2 Theories of argument linking

properties of the verbs directly from the event structure representation, without

reference to an additional layer of argument structure. The verb profile, which is

independent from the force-dynamic relations and which is another construal

operation, makes alternative argument linkings of verbs possible without positing

totally distinct event structure representations. Croft’s approach to argument

alternation is polysemous in that he posits two different representations for verbs

regarding the verb profile as a part of the semantic representation. However, unlike

Rappaport Hovav et al. (1998, 1999), there is no distinction between basic and derived

meaning, and also Croft's analysis is not so rigid as theirs in that he does not claim the

monotonicity of verbal meanings.

Another remarkable product of his theory is the antecedent/subsequent oblique

distinction. Subject and object linking is extensively discussed in argument linking

theories but not so many theories deal seriously with obliques. By distinguishing the

two types of oblique markers, the theory can predict the encoding of obliques in a

much better and more systematic way. In particular, the locative alternation, which

involves oblique markings, is nicely analysed in Croft's representation.

In the next chapter, I shall turn to the aspectual part of verbal semantic structure

introducing Croft’s (2000) aspectual representation of verbs. Then, using Croft's new

model, the aspectual behaviour of predicates in Japanese will be examined.

44
3. ASPECT IN JAPANESE

This chapter focuses on the dimension of aspect of predicates. The contents of the

chapter are as follows. Section 1 introduces the basic idea about lexical aspect and the

aspectual representation model proposed by Croft (2000). Section 2 analyses some

Japanese predicates in terms of Japanese tense/aspect (TA) constructions, namely the

Present, the Past, and the Te-iru form, following Croft’s (2000) analysis of English

TA constructions.

3.1. Representational model of aspect and aspectual types

3.1.1. Aspectual classes of Vendler (1967)

Vendler (1967) proposes four aspectual classes for English predicates. These are states,

activities, achievements, and accomplishments. There follows a summary of his

classification.

Table 1. Vendler’s classification of aspect

States (unbounded and durative): X loves Y

Processes Unbounded (atelic); activities: X is running
Bounded (telic) Punctual; achievements: X reaches the top
Durative; accomplishments: X builds a house

The distinction between the four types has been widely accepted and is motivated by

grammatical behaviour. In brief, states and achievements are distinguished from

activities and accomplishments in that the former do not have continuous senses (this

is shown by the ungrammaticality with the Present Progressive) and do not indicate

processes going on in time while the latter do (e.g., John is running and Jack is

building a house).

45
 Chapter 3 Aspect in Japanese

States denote situations which last for a period, while achievements

denote situations which are predicated for only single moments of time. The former

are used with the temporal adverbial indicating “interval” while the latter are used

with temporal adverbials which refer to a point in time, as the following examples

show:

(1) At what time did you reach the top? At noon sharp. (Vendler 1967: 102)
(2) For how long did you love her? For three years. (Vendler 1967: 103)

Activities are atelic and achievements and accomplishments are telic. The

difference is manifested by the use of adverbials. In-adverbials (Croft’s (2000)

container adverbials) occur with telic predicates and for-adverbials (Croft’s (2000)

durative adverbials) occur with atelic predicates.

(3) I ran for an hour/*in an hour. (atelic)

(4) I reached the top in an hour/*for an hour. (telic)
(5) I built a house in one year/*for a year. (telic)

Croft (2000) adds to Vendler’s classification of aspectual distinctions by adopting

a finer-grained analysis. The other important point Croft (2000) makes is that verbs

have varying aspectual behaviour, that is, even a verb can have more than one of the

aspectual types above, depending on different Tense/Aspect constructions and context

as well. For example, the event of pushing a cart is an activity but that of pushing the

cart to New York is an accomplishment. Therefore, we cannot classify “push” as

belonging to one aspectual type. To give a more felicitous account of the aspectual

behaviour of predicates, Croft (2000) distinguishes aspectual types and aspectual

classes. Aspectual types are universal and describe the semantic structures which are

found in languages. Accomplishments or activities are aspectual types. Aspectual

classes are language-specific and are defined by the range of construals that the class

of verbs allows with various TA constructions. Situation types are divided into

aspectual classes by examining their distributional behaviour in relation to the

46
 Chapter 3 Aspect in Japanese

different TA constructions which are available in a certain language, namely what

interpretation verbs get in a certain TA construction (i.e, TA constructions are also

language-specific). Predicates which have the same distributional behaviour in TA

constructions are grouped together as in the same aspectual class. These points will be

further clarified in what follows.

First of all, I introduce Croft’s basic model for aspectual representations with

notations, and illustrate them with his finer-grained universal aspectual types that are

possible in the real world. Croft (2000) starts from Vendler’s four aspectual types and

further classifies them into smaller categories.

3.1.2. Croft’s (2000) basic aspectual representation model

The tense and aspect system has often been discussed with representations using a

linear diagram that refers to a temporal transition from left (past) to right (future) with

the present moment in the middle. (Comrie (1976), Dowty (1979), Machida (1989)).

Croft (1999b, 2000), however, uses a two-dimensional diagram approach to aspectual

representation.

3.1.2.1. Basic notations and concepts

Croft’s aspectual representation is a two-dimensional diagram consisting of a

qualitative state/change dimension (notated as Δ) and a time dimension (notated as t)

as follows (Croft 2000: 2, Figure 1):

47
 Chapter 3 Aspect in Japanese

Figure 1. The basic representation for verbal aspect

The addition of the qualitative dimension to a common time scale dimension is a

unique property of his representation. The qualitative dimension encodes the

“relevant” qualitative states and changes in an event. When we say The door opened,

the state of the door concerning its openness is encoded in the qualitative dimension

and not its colour, for example. The qualitative dimension is represented on the

ordinate and the time dimension is represented on the abscissa. (We shall see the

motivation for the two-dimensional representations later).

The course of an event from its beginning to the end is described as the aspectual

contour in the diagram. Each participant in the event is supposed to have its own

aspectual contour that represents its state and processes1. The following is an example

for The door opened (Croft 2000: 3, Figure 3).

Figure 2. The aspectual representation for the door opened.

open
door
not open
Δ

t
<P

1
However, in this chapter, I limit myself to representing a single aspectual contour per event to show
introductory and fundamental ideas of the aspectual representation.

48
 Chapter 3 Aspect in Japanese

The participant, the door, undergoes a transition (represented by the vertical line) from

being not open (a rest state that is represented by a lower horizontal line) to being

open (a result state that is represented by a higher horizontal line). All these

geometrical representations make up the aspectual contour2 (in the above case, the

aspectual contour for “the door”). Each composite piece of the aspectual contour is

called a phase.

Following one of the main innovations in semantic representation in cognitive

linguistics (cf. Croft’s causal analysis in the previous chapter), we maintain a

distinction between profiled phases and unprofiled ones in the aspectual contour.

Profiled phases are those that are asserted by a sentence. The other phases which are

not profiled function as necessary/background information which is required to

understand the meaning of the profiled part encoded by a sentence. They are

considered as the frame or base.

As in Figure 2, a bold black line means that the phase is profiled and a broken line

means that the phase is not profiled. The phase which is asserted in the event is, thus,

the transitional change of the door from being closed to being open. The phase prior to

the transition is presupposed: the door is presupposed to have been closed before the

change happened. The phase following the transition is implied as part of the natural

course of events: it can be implied that the state of the door being open holds, at least,

for a while.

Finally, the past tense is used with the example. That is, the transition is

interpreted to occur in the past. The transition is placed in the time scale with the

notation “<P” in the time prior to the present in Figure 2.

A difference in profiling can denote another related but distinct event. For example,

2
In some cases, only one phase makes an aspectual contour. We shall observe examples later.

49
 Chapter 3 Aspect in Japanese

here is the diagram for the door was open:

Figure 3. Aspectual representation for The door was open.

open

door not open

Δ

The be-verb plus the adjective open asserts the state after the transition had occurred.

If we compare Figure 3 with Figure 2, the two sentences use the same stem open, but

they are different in profile. The profiled part should precede the present moment. The

past (<P) can denote an interval referring to the state that was true for a certain period

which precedes the present.

The geometric semantic representation for sentences is, to be more precise, a

representation of a verb stem combined with the tense/aspect (TA) constructions (and

sometimes with adverbial phrases), as no verb appears without one of the TA

constructions. Both a verb and the TA construction contribute to a meaning of a

sentence. As Croft (2000: 4) points out, in the case of the first example (The door

opened), the verb stem open contributes to the profile in the aspectual contour and the

simple past specifies that the profiled part is mapped onto a temporal point which is

prior to the present.

The above example is a simple one. However, there are many different classes of

verbs that interact with many different TA constructions, so we need to examine the

wide varieties from both sides. What is more complex is that the combination of a

verb stem and the TA construction can involve a construal operation. I shall return to

this issue in §3.2.1.

50
 Chapter 3 Aspect in Japanese

3.1.2.2. Aspectual types of events

Now that we have a basic representational tool for the aspectual analysis, we can

examine the different kinds of event aspectual types and how they can be shown in the

geometric representations. Each aspectual type is accompanied by English examples.

3.1.2.2.1.States

There are three subcategories of states: point states, inherent states, and transitory

states.

3.1.2.2.1.1. Point states

Point states are states that occur only at a point in time, such that “the state does not

hold before the point in time, and ends immediately after that point in time” (Croft:

2000: 6). Only the point state is profiled. The inceptive phase immediately precedes

the point state and a terminative phase immediately follows it. (These all occur in the

same point on the t scale).

Figure 4. The aspectual type of point states (Croft 2000: 6):

examples: be 5 o'clock, be on time

There are two states on the Δ dimension. One is the state which is denoted by the

predicates and the other one is the state of not having that state. In the example of be 5

51
 Chapter 3 Aspect in Japanese

o’clock, there is a state which is five o’clock and the other which is not five o’clock.

Ideally, these two states should not be connected by a line but understood as two

separate points as there is a quantum leap from one to the other. Only as a convention,

we use a continuous line to indicate a transition (quantum leap) between the two states.

The first (left) transition phase is called the directed3 transition (d-transition) and

denotes a change from the rest state to another state. The second (right) transition

phrase is called the r-transition (r can be that of a return, reverse, or rest state) and

denotes a change back to the rest state. A point between the two transitions only holds

for a single point in time, so that point is called the point state (p-state). Only this

phase is profiled. It is worthy of note that, in the representation of aspectual types, it is

important to show which parts are profiled in their aspectual contour.

Technically, the two transition lines and a point state have to be located at the

same point on the time scale. Because of the difficulty in representing them, I

illustrate the two transitions as two different lines.

Finally, as a convention, an original state is represented lower in the Δ scale and a

result state is represented higher.

3.1.2.2.1.2. Inherent states

Inherent states are states that are conceptualised as an inherent property of the

individual. There are two types of inherent state; original inherent states and acquired

inherent states. An original inherent state is a property which is held by the individual

from its origination. For example, the inherent state of being a lizard is fixed from the

birth of the referent and cannot change. An acquired inherent state is an inherent

property that can be acquired and which stays forever. Being tall as a human’s

3
The word, directed, is used to indicate that the individual ends up with a new result state, which is

52
 Chapter 3 Aspect in Japanese

physical property, for example, will stay, once an individual has acquired that state.

The following are the aspectual types of the two types of inherent state.

Figure 5. The aspectual type of original inherent states (Croft 2000: 7)

example: be a lizard

The phrase above is called an inherent state (i-state) and represents a state without a

beginning point or end. The state is construed to last till the end of the history

(existence) of the individual.

Unlike original inherent states, an acquired inherent state presupposes the

beginning point of that property being held by the individual.

Figure 6. The aspectual type of acquired inherent states (Croft 2000: 8)

example: be tall

The rest state and the d-transition which are represented by broken lines in the above

different from the rest state.

53
 Chapter 3 Aspect in Japanese

Figure indicate that there is a time where a non-i-state holds and there is a change of

acquiring an i-state.

3.1.2.2.1.3. Transitory states

Transitory states are states that occur in a finite period of time (such as the door being

open). According to Croft (2000: 8), a transitory state may be a point in time, or at

least profile just a point in time, because of its compatibility with temporal locative

adverbials (as in She was ill at 5: 00.(She was in the state of being ill at 5: 00)), or can

be an interval (such as She was ill all week). The following is the representation for a

transitory state.

Figure 7. The aspectual type of transitory states (Croft 2000: 8)

examples: (door) be open, (Jack) be ill

A transitory state presupposes the inception of the event as an acquired inherent state,

but does not include its termination. Croft (2000: 8), who points out that whether a

terminative part should be represented as a base or not is an interesting and difficult

problem for the representation, argues for the asymmetric treatment of inception and

termination. That is, the inceptive phase but not the terminative phase should be

included in the aspectual contour. He attributes the asymmetrical treatment of

inception and termination to the asymmetry of time. It is clear that there is an

54
 Chapter 3 Aspect in Japanese

inception phase leading to the transitory state. However, it is not certain whether the

transitory state will come to terminate or persist in the future, though it is expected to

end at a certain time in the future. Therefore, Croft leaves the termination phase

unexpressed in the representation, noting that only those parts of the aspectual contour

which can be safely presupposed should be included in the representation.

Finally, the profile part of state above is called the transitory-state (t-state).

3.1.2.2.2. Activities (unbounded processes)

Unlike states, activities involve continuous change on Δ, as they refer to a kind of

process. However, the continuous change does not reach any new resulting state, and

is thus unbounded. Activities are also extended over the t scale as they do not occur

with a punctual locative adverbial 4 . They are not inherent because they have a

beginning and an expected end just as transitory states do. There are two types of

activities: undirected activities and directed activities. The former involve a change

which is not construed as a directed change, such as dancing which only denotes the

activity of moving one’s body. The latter involves a directed change such as the

balloon expanded slowly (Croft 2000: 10) where the size of the balloon becomes

spatially bigger as time goes by.

4
She danced at 11pm is not allowed under the interpretation of “she was dancing at the time of 11: 00”
(Croft 2000: 9).

55
 Chapter 3 Aspect in Japanese

The aspectual types of undirected activities are represented as follows.

Figure 8. The aspectual type of undirected activities (cf. Croft 2000: 9):

examples: run, swim, dance

The undirected activity is represented as a zigzag line which is extended over the Δ

dimension and the t dimension to show it involves a kind of cyclic process such as

taking a series of steps in the undirected activity of walking. The profiled part is called

an undirected process (u-process). The inception to the u-process is not profiled and

the u-process is open-ended to show the transitory nature of the activity (as in the

representation of the transitory state).

Directed activities can be represented as follows:

Figure 9. The aspectual type of directed activities (Croft 2000: 10):

examples: widen, expand

56
 Chapter 3 Aspect in Japanese

The profiled part is called the directed process (d-process), which is a vector that

indicates that the change involved in the directed activity has a direction and is

gradual. It has no specific endpoint. As in transitory states and undirected activities,

the inception is not profiled and the termination is not represented, so as to show the

nature of transitoriness.

3.1.2.2.3. Achievements (punctual bounded processes)

Achievements are punctual bounded processes. In bounded processes an event is

bounded by two distinct states, that is the rest state and the result state, and the process

traverses the Δ dimension from the rest state to the result state. ‘Punctual,’ means that

the process traverses from the rest state to the result state instantaneously. That is, the

event is construed as taking place instantaneously.

Only the inception (d-transition above) is profiled and not the result state. Croft

argues that such a representation is felicitous in reflecting the entailment that the

participant has achieved the result, since the first point of the result state is profiled.

Achievements are classified into three subtypes depending on the type of result

state they take.

3.1.2.2.3.1.Cyclic achievements

Achievements are classified as cyclic achievements if the result state of achievements

is a point state where the participant immediately returns to the rest state after

momentarily achieving the point state.

57
 Chapter 3 Aspect in Japanese

Figure 10. The aspectual type of cyclic achievements5

examples: flash, squeak, tap (in semelfactive reading)

3.1.2.2.3.2. Reversible directed achievements

Achievements are classified as reversible directed achievements if their result state is

a transitory one. Such achievements are called reversible, because the transitory state

is expected to end in the future.

Figure 11. The aspectual type of reversible directed achievements

example: X open, X fall ill

5
To make the representation clear, I put a point state in the representation separately from the inception
phase, but technically it should be understood as the upper point of the inceptive transition (and also of
the terminative transition).

58
 Chapter 3 Aspect in Japanese

3.1.2.2.3.3. Irreversible directed achievements

Achievements are classified as irreversible directed achievements if the result state of

achievements is an inherent state.

Figure 12. The aspectual type of irreversible directed achievements

example: break, shatter, die

3.1.2.2.4. Accomplishments (extended bounded processes)

In accomplishments, the process does not only extend over the Δ dimension

(involving a change from one state to another) but also over the t dimension. They

include directed processes which “incrementally” lead to the result state. A good

example of an accomplishment is eating a pizza. Every bite of eating the pizza

gradually brings about the final result state where all the pizza is consumed. Dowty

(1991: 567-571) introduces the notion of incremental theme, by which he refers to an

entity that undergoes incremental change. This determines the aspect of a situation

denoted by a predicate, with its parts corresponding to those of the event itself. In the

above example, the pizza is the incremental theme as its status reflects the aspectual

status of the situation of somebody’s eating the pizza. If the pizza is half gone, then

the situation of somebody’s eating the pizza is half-done. If the pizza is completely

gone, then the situation of eating the pizza is also completely finished. That is, the

59
 Chapter 3 Aspect in Japanese

incremental theme “measures out” (Tenny 1992) the event.

An incremental theme is not necessarily an argument of a verb. It can be the path,

for example, with verbs of directed motion such as Jack went to New York. In this case,

it is the path to New York that undergoes an incremental change by being traversed. A

participant which travels the path is a holistic theme (Dowty 1991: 569). This term

refers to a participant who has a close relationship to the incremental theme. In the

above example, Jack does not undergo the incremental change, but his relative

positioning with respect to the path undergoes the change.

Hay et al. (1999) argue that an incremental theme is “properly construed as a

measure of some property of an argument of a verb, not an argument”, pointing out

that the nature of a gradable property depends on the lexical meaning of the verb.

They argue that “the lawn” in mowing the lawn, which is the original example of an

incremental theme given by Dowty, is actually not an incremental theme but that a

property of the lawn (its area) is an incremental theme. The lawn is the argument

which possesses the measurable property. That is, it is identical to the holistic theme

of Dowty in that it is closely associated with a kind of scalar property. Croft (2000:

12) calls the scalar property/incremental theme the verbal scale and also uses the term

holistic theme for an argument which is associated with the verbal scale.

Croft (2000: 12) further argues that the directed activity actually has a verbal scale,

since it involves gradual directed change. The example he gives is the act of eating

pizzas. As “pizzas” here is a bare plural noun, it does not specify the endpoint.

However, every bite of eating represents an incremental change on the verbal scale, i.e.

the consumption of pizzas. The difference between accomplishment and directed

activity is that the former has a definite measurement on the verbal scale and the latter

does not. The aspectual type of an accomplishment, therefore, involves a d-process as

60
 Chapter 3 Aspect in Japanese

follows:

Figure 13. The aspectual type of accomplishment (Croft 2000: 12)

Croft argues that three phases of accomplishments: the inception of the directed

process, the directed process itself, and the completion of the directed process, are all

profiled, at least in the simple past tense. Since accomplishments are bounded

processes, they are bounded by the two d-transition phases, which are the inception

and the completion.

According to Croft (2000: 13), justification for profiling the inception and

completion phases is provided by the difference in behaviour of accomplishments

compared to that of undirected activities and transitory states with interval locative

adverbials such as in the afternoon.

(6) She danced in the afternoon. (activity)

(7) She was miserable in the afternoon. (transitory state)

In the above examples, locative adverbials only indicate what state of affairs held

during the afternoon; they do not indicate whether or not the state is true before or

after that interval. However, in an accomplishment as in the (8), the entire event is

construed as having happened in that interval.

(8) She wrote her term paper in the afternoon. (accomplishment)

The entire event of writing her term paper including inception and completion is

asserted to have taken place in the time period denoted by the adverbial; the sentence

61
 Chapter 3 Aspect in Japanese

entails that she began and completed the term paper in the afternoon.

Every bounded event is seen to involve a verbal scale. Directed achievements (i.e.

reversible achievements and irreversible achievements) can actually be seen as trivial

versions of accomplishments (Dowty 1991: 568 and Croft 2000: 13). For example, in

John broke the glass, the glass undergoes definite change instantaneously. That is, the

change is so trivial that the event is construed as having no intervening process

(d-process), and the inception and the completion are represented as coinciding in

directed achievements. Croft (2000: 13) describes the directed achievements possess a

trivial verbal scale.

Another important point to note is that a determinately quantified argument

functions as a verbal scale (Croft 2000: 14 following Dowty 1991: 570). Typical

examples are plural subjects or direct objects: John visited 25 cities or 2,500 tourists

visited Atlanta (Dowty 1991: 570). Because of the quantification, the events

themselves are delimited and also have intermediate stages (John visited 25 cities one

by one and 2,500 tourists visited Atlanta one after another). In these cases, the verbal

scale is entailed not by a meaning of the verb but by the semantics of the argument.

This is called a derived verbal scale by Croft (2000: 14) following Dowty’s derived

incremental theme.

Accomplishments are further classified into three subtypes according to what type

of result state they entail after the completion phase. An accomplishment whose result

state is a point state is called a cyclic accomplishment. An example is Judith danced

the kopanica. It takes time to dance the kopanica from its beginning to its end.

However, when the dance finishes, the performance will exist no more.

Accomplishments are classified as reversible (directed) accomplishments, if the result

state of an accomplishments is a transitory state. An example is I pruned the hedge. It

62
 Chapter 3 Aspect in Japanese

takes time to prune the hedge, but after a while it will grow again. The representation

of Figure 13 above is that of reversible (directed) accomplishments. Finally,

irreversible (directed) accomplishments are accomplishments whose result state is an

inherent state as in I burned down the shed. Once the shed has been burned down, it

will stay burnt down. The representation for cyclic accomplishments and irreversible

directed accomplishments is the same as in Figure 13 except for the result state.

3.1.2.2.5. Runup achievements

The last aspectual type is runup achievements. In runup achievements, a participant

undergoes the process “which culminates in the achievement” (Croft 2000: 15); the

process has a specified resulting state, which occurs instantaneously, as in the case of

achievements. However, before the completive phase, runup achievements require a

process that is extended over the t scale but that does not involve an incremental

change to achieve that state.

Prototypical runup achievements are manifested as basic achievement verbs with

the progressive form in English such as die in He’s dying. The sentence refers to the

process that leads to his dying state. However, that dying process does not occur

incrementally. The transition to dying should be captured as happening

instantaneously as we do not have the state of being half-dead or two-thirds dead.

According to Croft, most English achievement verbs can have the progressive form

under the runup achievement reading. The representation for runup achievements is as

follows:

63
 Chapter 3 Aspect in Japanese

Figure 14. The aspectual type of run-up achievements (Croft 2000: 15)

example: He's dying.

Technically, we should interpret that the instantaneous transition of achievement

events is reconstrued and extended as consisting of the three profiled parts (as in

accomplishments): inception, activity, and completion. The non-incrementality of the

activity, however, is represented as a u-process unlike in accomplishments. Croft

(2000: 15) justifies profiling the inception and the completion in runup achievements

by their co-occurrence with container adverbials such as in a week. An example is She

died in a week. A week specifies the temporal interval from a certain reference point

when she is construed as starting “dying” to the completion of the death. Runup

achievements are bounded by an inception and a completion. To summarise, they are

“bounded processes which lack a verbal scale and yet are extended in time” (Croft

2000: 15).

Theoretically, there should be three variations of runup achievements as with other

aspectual types according to the type of result state that follows. Cyclic runup

achievements have a point state following the runup achievement. An example is flash

in The lighthouse is flashing, with which we can imagine a situation where the

lighthouse light gradually becomes visible as it slowly rotates and an observer sees the

light as coming and going instantaneously. Reversible (directed) runup achievements

have a transitory state as their result state as in falling asleep. Irreversible (directed)

64
 Chapter 3 Aspect in Japanese

runup achievements have an inherent state as in She’s dying as in Figure 14.

3.1.3. Summary of the aspectual types

Basic aspectual types are summarised as follows:

Table 2. Basic aspectual types (cf. Croft: 2000: 16):

profiled contour type of (result) state:

preceding (result)
state Point interval history
none point state transitory state inherent state
achievement cyclic reversible irreversible
achievement (directed) (directed)
achievement achievement
accomplishment cyclic reversible irreversible
accomplish-me (directed) (directed)
nt accomplish-me accomplishment
nt
runup cyclic runup reversible irreversible
achieve-ment achievement (directed) (directed) runup
runup achievement
achievement
profiled contour type of process:
preceding process: undirected directed
none (process undirected directed
profiled) activity activity

There are fourteen possible aspectual types, each of which can be represented as an

aspectual contour which comprises one phase or a combination of phases, and profile.

There are seven types of phases, namely, the p-state, the i-state, the t-state, the

d-transition, the r-transition, the d-process, and the u-process. Thus, it is possible to

denote aspectual types in terms of a sequence of phases. For example, irreversible

(directed) accomplishments are described as having the contour of t-state –

d-transition – d-process – d-transition – i-state. (The bold parts are profiled parts.)

The above table is just a brief summary, to serve as a guide to what kinds of aspectual

event types are possible in a language. A full two-dimensional representation for

aspect will be employed later in the thesis in analysing verbs of putting and removing.
65
 Chapter 3 Aspect in Japanese

In the rest of this chapter, we shall examine how Japanese predicates are classified

into smaller aspectual classes in terms of the behaviour and interpretation they get in

the three TA constructions, which are the Present, the Te-iru, and the Past

constructions. We also see how these three constructions can be polysemous, having

different aspectual construals.

3.2. Analysis of Japanese TA constructions and aspectual classes

Now I shall analyse Japanese predicates in three constructions, the Present

construction, Te-iru construction, and the Past construction. Japanese predicates

include “noun + da (desu) form”, “nominal adjectives 6 + da (desu) form”, and

adjectives as well as verbs. The copula da can be combined with either a noun or a

nominal adjective to produce the simple Present assertive sentences as in examples

(9)-(10) below. Desu is a polite version of da. However, I only deal with the da case

in the present study because it is not our concern whether the sentences are in a polite

or bare form.

(9) Noun plus da

Kare wa isha da.
he TOP doctor DA7
He is a doctor.

6
Nominal adjectives are a unique syntactic category in Japanese which have the characteristics of both
adjectives and nouns.
7
In examples of this chapter, I use “DA” as indicating this unmarked copula. Another thing is to
mention is that Japanese uses quite a few postpositional markers. In the following examples in this
chapter, TOP means topic marker, NOM means nominative marker, ACC means accusative marker,
GEN is genitive marker, and CLSS is classifier. OJCT is assigned to the use of ni which is similar to
English accusative in its meaning, but which is not an accusative marker. I gloss other postpositions as
English prepositions whose meaning is closer to each postposition in Japanese. The detailed description
of Japanese postpositions will be given in the next chapter.

66
 Chapter 3 Aspect in Japanese

(10) Nominal adjective plus da

Watashi wa amaimono ga suki da.
I TOP sweets NOM be.fond.of DA
I like sweets.

Adjectives end with i in the simple Present tense.

(11) Adjective
Kanojo wa utsukushii.
she TOP beautiful
She is beautiful.

These predicates have the tense distinction (non-past vs. Past) but do not take an

aspectual distinction (which is manifested as te-iru, which will be discussed later). For

the past tense, the Past tense marker ta is used at the end of the predicate.

Kare wa isha dat-ta.

(12) he TOP doctor DA-PAST
He was a doctor.

Watashi wa amaimono ga suki dat-ta.

(13) I TOP sweets NOM be.fond.of DA-PAST
I used to like sweets.

For adjectives, ta cannot be attached to the adjectival root directly and kat must be

inserted between the stem and ta.

Kanojo wa utsukushi-kat-ta.
(14) she TOP beautiful-inflection-PAST
She was beautiful.

3.2.1. Basic background for analysis

I shall now review Croft (2000)’s analysis for English predicates since it is the model

used in the analysis of Japanese in this chapter. He classifies English verbs into

different aspectual classses according to the aspectual types they get in various TA

constructions. As he points out, ‘many classes of verbs display schizophrenic

aspectual behaviour, in that they sometimes behave like one of the basic aspectual

types and sometimes like another.’ What happens is that many verbs belong to one

67
 Chapter 3 Aspect in Japanese

basic aspectual type in one TA construction and to another one in another TA

construction or the same TA construction under another interpretation. The

complexity here resides in the nature of the encyclopaedic knowledge of event types

and in what is called a construal operation by the speaker. Speakers of different

languages or even of the same language possibly construe the same event that has

happened in the world in distinct ways.

Verbs appear in a variety of TA constructions such as the Simple Present tense, the

Past tense, the Present Progressive, profiling changing constructions, such as the

inceptive construction (start to Verb) or the completive construction (finish Verb-ing),

adverbials such as iterative adverbials, durative adverbials, and container adverbials

etc. (For details, see Croft (2000)). The combination of a verb stem and a certain TA

construction, in most cases, involves a construal operation. There are two types of

construal operation: (i) only shift in profile or (ii) reconstrual of the aspect contour

itself. One example is an achievement type verb break. We interpret the verb itself as

having a profiled inception phase in the “t-state – d-transition – i-state” contour.

When it is used in the resultative adjective construction as in The glass is broken, there

is a shift in profile, i.e., “t-state – d-transition – i-state”, as the construction profiles

the result state alone. The verb appearing in the simple present tense such as The glass

breaks easily construes an event as a state “i-state”. The process is reconstrued as a

generic/inherent property of the glass in that it is capable of undergoing the process

easily.

Since verbs always appear with (a) certain TA construction(s) in a sentence, we

need to examine how a single verb actually behaves with various TA constructions in

order to distinguish which part of the meaning of a whole sentence is contributed by a

verb. Croft (2000), through investigating complex behaviours of verbs stems and the

68
 Chapter 3 Aspect in Japanese

T/A construction in English, has attempted to find invariant meanings for them, has

tentatively identified definitions for each TA construction and has classified 44

situation types into 27 aspectual classes.

Limitations of space prevent me investigating every TA construction in Japanese,

so the rest of the chapter focuses on the three TA constructions: the Present, the

Te-iru, and the Past constructions. In order to produce my analysis I looked at a

certain range of predicates and saw the range of construals they allow with the three

constructions above and separated them into Japanese aspectual classes. I deal with

forty-eight situation types, the details of which are presented in Table 4 in the

summary at the end of this chapter. Situation types are types of “the qualitative

semantic classification in terms of types of experiences described by predicates”

(Croft 2000: 1). Examples are Social Role, Dispositions, Manner of Motion, and so

on.

3.2.2. Criteria

I shall give criteria for identifying construction senses for the Japanese Present, Te-iru,

and Past constructions based on Croft’s (2000) analysis of English. First of all, the

following senses are excluded from the present discussion.

Present—Future Present (future time reference), Historical Present

Past—Past of realisation of expectation8 (Teramura 1984: 105-106)

Secondly, the following features of the TA constructions are either fixed or left to

8
This is the Past which does not refer to past events. An example is the expression of Hon ga at-ta
(book NOM exist-PAST = Here was the book.) This can be used for the situation where the book exists
in the present moment. One can say the phrase when one finds the book after looking for it for a while.
What this shows is that, in reality, the book has been there but that the speaker did not know about this
till the moment he found it. This use should be treated in irrealis mental space and is excluded from the
discussion.

69
 Chapter 3 Aspect in Japanese

vary, conforming to Croft’s analysis:

Referent Specificity (fixed)—The argument referent is fixed as a single and

specific referent. That is, accomplishment construals with a derived incremental

theme/verbal scale are excluded. Japanese expressions can be ambiguous in terms

of number of arguments. For example, mushi ‘insect’ can be interpreted as singular

or plural. In this case, I disambiguate the referent as singular by using the

demonstrative adjectives such as sono (‘that, the’).

Argument Structure (varies)—The argument structure varies while the situation

type for a predicate in question remains the same. For example, oou ‘cover’ can

appear in a two-argument transitive construction (as in The fog covered the city) or

three-argument transitive construction (as in My father covered his car with a

sheet). The situation type in both constructions is the same in that it refers to an

event of somebody/something covering another entity. Therefore, the verb in the

two constructions is treated as one situation type.

Referent Type (varies)—The referent type can also vary. For example, miru ‘watch,

look at’ can be used to describe an event of looking at “the sky” and an event of

watching “the video”. (Interestingly, English uses different verbs for the two

events.) The referent types are different, but the situation type of miru is still the

same one of paying one’s perceptual attention to something. Therefore, the

referent types can vary in a single situation type.

Adverbial Support (varies)—Adverbial support can be used to get a certain

construal. For example, some achievement verbs can get the runup achievement

reading with te-iru only with the support of a container adverbial. In this case, I

adopt the analysis that a predicate can occur with that construal in a certain

construction, following Croft (2000: pers.comm) who does “not distinguish

70
 Chapter 3 Aspect in Japanese

between predicates requiring adverbial support for a particular construal in a

particular TA construction and predicates not requiring adverbial support.”

Finally, because of limitation of space, I do not list all the examples with each

distinct sense the constructions have. For this full data, the reader is referred to

Appendix A.

3.2.3. Analysis

3.2.3.1. The Present construction

Formally, in the Present construction, verbs end with the unmarked form “u/ru”,

nouns and nominal adjectives are combined with the copula “da” (e.g. isha da ‘doctor

DA’ = be a doctor or kirei da ‘beautiful DA’ = be beautiful), and adjectives end with

“i” form.

The Present construction in Japanese has four different construals: inherent state

(coded as 1I), transitory state (coded as 1T), habitual (coded as 1H), and

uninterpretable (coded as 1*). I regard “uninterpretability” as an independent

construal following Croft (2000).

The first construal is of an inherent state. This construal is quite common with the

following predicates.

[1-1Natural Kinds]
Sono ishi wa daiamondo da.
(15) that stone TOP diamond DA
That stone is a diamond.

[1-2 Ethnicity]
Kanojo wa igirisujin da.
(16) she TOP English DA
She is English.

[1-3 Biological Kinds]

Are wa tokage da.
(17)
that TOP lizard DA
That is a lizard.

71
 Chapter 3 Aspect in Japanese

[1-4 Physical Properties]

Kare wa se ga takai.
(18) he TOP height NOM tall
He is tall.

The inherent state construal includes the generic interpretation that describes a

property of the individual as an inherent state. The following are examples:

[18 Blooming]
Sono hana wa umibe ni saku.
(19) the flower TOP seaside at bloom
The flower blooms (grows) at the seaside.

[21-1 Light Emission]

Sono dentou wa sukoshi no enerugi de hikaru.
(20) the light TOP little GEN energy with shine
The light shines with little energy.

[26 Motion/Movement]
Sono omocha wa denchi de ugoku.
(21) the toy TOP battery with move
The toy moves with a battery.

Some other predicates may get adverbial support, such as jouzuni “well”, hayaku

“fast”, or kantanni “easily”, to refer to the inherent property of the subject, namely,

the ability of the subject.

[20 Open/Close]

Sono mesu wa kantanni kanbu o hiraku.

(22) the surgical.knife TOP easily diseased.part ACC open
The surgical knife opens diseased parts easily.

[24 Manner of Motion 1]

Taro wa totemo hayaku hashiru.

(23) Taro TOP very fast run
Taro runs very fast.

[27 Performance]
Markus wa warutsu o jouzuni odoru.
(24) Markus TOP waltz ACC well dance
Markus dances the waltz very well.

72
 Chapter 3 Aspect in Japanese

The second construal of the Present construction is a transitory state. This is a

common construal with the following situation types:

[3 Social Role]
John wa shachou da.
(25) John TOP president DA
John is president.

[5 Bodily States]
Kare wa byouki da.
(26) he TOP ill DA.
He is ill.

Some situation types which can take an 1I construal are likely to get an 1T

construal when they are modified with adverbials which specify the temporary nature

of events, such as kyou “today”, ima “now”, or mezurashiku “unusually”.

[2-2 Necessity]
Watashi ni-wa ima kane ga iru2
(27) I to-TOP now money NOM need/necessary
I need money now.

[2-3 Dispositions]
Jack wa kyou wa shinsetsu da.
(28) Jack TOP today TOP kind DA
Jack is kind today.

[9 Perception]
Kyou wa mezurashiku Huji-san ga mieru.
(29) today TOP unusually Mt.Fuji NOM be.visible
Mt.Fuji is visible today, which rarely happens.

The third construal of the Present construction is associated with the habitual

interpretation. It describes a series of repeated events being reconstrued as an inherent

property of an individual under a “coarse-grained scalar adjustment” on the qualitative

dimension and the time dimension. This is a good motivation for the two dimensional

representation for the aspect since the reconstrual of a situation operates on the two

dimensions.

The habitual interpretation is found with most of the situation types. In particular,

the regularity of events is specified with adverbials such as itsumo “always”, yoku

73
 Chapter 3 Aspect in Japanese

“often”, or translations equivalent to English “every” or “whenever”.

[2-3 Dispositions]

Jack wa tanomigoto ga aru toki dake shinsetsu da.

(30) Jack TOP favor NOM be when only kind DA
Jack is kind only when he receives a favour.

[13 Sleeping]
Chichi wa maiban 11ji ni nemuru.
(31) father TOP every.night 11o'clock at sleep
My father goes to bed at 11 o’clock every night.

[17 Change of Physical Properties]

Kono keito no uwagi wa sentaku-suru tabi ni chijimu.
(32) this wool GEN jacket TOP wash time at shrink
This woolen jacket shrinks every time I wash it.

[28 Create Mark/Defect]

Watashi wa yoku juuyouna shorui o ayamatte yaburu.
(33) I TOP often important document ACC mistaking tear.up
I often tear up important documents by mistake.

[30 Covering]
Hazukashigariya no kanojo wa itsumo te de kao o oou.
(34) shy.person GEN she TOP always hand with face ACC cover
She, as a shy person, always covers her face with her hands.

The last construal is uninterpretability. The following two situation types are

ungrammatical with the Present construction. They are “Type 4 verbs” which only

occur with the Te-iru construction. These will be dealt with shortly in the discussion

of Te-iru.

[7-1 Relation 2]
*Kare no ronbun wa kono ronbun yori sugureru.
(35) he GEN article TOP this article than excel
His dissertation is superior to this dissertation.

[7-2 Posture 1]
*Huji-san ga me no mae ni sobieru.
(36) Mt.Fuji NOM eye GEN front at tower
Mt. Fuji towers high in front of us.

The situation type of dying and discovering are also uninterpretable with the Present

unless the subject is interpreted as plural, i.e., a derived verbal scale. This is because

74
 Chapter 3 Aspect in Japanese

the event of dying and discovering are prototypical irreversible (directed)

achievements. The following is an example for the situation type of dying.

[35 Dying]
*John wa maishuu shinu.
(37) John TOP every.week die
*John dies every week.

To summarise, the Present construction in Japanese reconstrues the situations

denoted by verbs as a state. The internal structure of the situation is focused and it

does not specify where a starting point or endpoint of the situation is. The state does

not include the endpoint, thus allowing for the possibility that the situation is never

completed (inherent state) as well as the alternative of it being completed (transitory

state). The Present tense profiles only a point (which is the present moment) in that

state phase.

3.2.3.2. The Te-iru Construction

Okuda (1977), Suzuki (1979) cited by Kudou (1995: 36) state that this construction is

one with an aspectual distinction from the Present. It is quite unique to Japanese in

that some of its uses are similar to the Present Progressive of English and others are

similar to the Present Perfect in that language. Therefore, before analysing the

construction, I shall give some basic information about it, citing other scholars’ work.

3.2.3.2.1. On te-iru

Kindaichi (1976), cited by Tsujimura (1996) and Machida (1989) among others,

classifies Japanese verbs into four classes, which have been widely adopted by many

scholars. They are stative, continuative, instantaneous, and “Type 4”. This

classification is based on the aspectual properties, or to be more concrete, on the

75
 Chapter 3 Aspect in Japanese

interpretation that arises from the Te-iru construction9.

Stative verbs refer to a state being true in the present moment in its unmarked form,

that is, the u/ru form. In other words, the unmarked form of stative verbs denotes that

an event is true at the moment of utterance. They are characterised as not being able to

occur with the Te-iru construction.

(38) the Present construction

Hon ga tsukue no ue ni aru.
book NOM desk GEN surface at be
The book is on the table.

(39) the Te-iru construction

*Hon ga tsukue no ue ni at-te-iru.
book NOM desk GEN surface at be-TE-IRU
The book is on the table.

Continuative verbs refer to an event or process which normally has a perceptible

beginning and ending and which extends over a certain amount of time. They cover

accomplishment and activity type events in English. In order to indicate that an event

is true at the moment of utterance, the Te-iru construction is used. The unmarked form

(u/ru form) cannot indicate that the event is in progress. The construction with

continuative verbs functions as the Present Progressive in English.

Taro wa hashit-te-iru.
(40) Taro TOP run-TE-IRU
Taro is running.

Instantaneous verbs, as the term indicates, describe an event that occurs and is

done instantaneously. A good example is an event such as shinu ‘die’. Instantaneous

verbs refer to instantaneous change from the point where the denoted event is not true

to the one where it is true. Though it needs some time for an event to occur in the real

world (for example, normally it takes some time before a person dies. i.e. he could be

ill for some time or suffer some time), what these verbs indicate in their lexical entry

9
The construction has another phonological variant, de-iru, which is in complementary distribution

76
 Chapter 3 Aspect in Japanese

is the immediate transition, just as in achievement type events in English. In the case

of dying, it is a transition from being alive to being dead.

Because they are instantaneous, these verbs cannot indicate that a denoted event is

true at the moment of utterance. When the utterance is made, the event has already

finished. It is only possible to indicate that a result of an event is true at the moment of

utterance. The Te-iru construction is used for this purpose:

Inu ga shin-de-iru.
(41) dog NOM die-TE-IRU
The dog is dead. (The dog has been dead).

According to Teramura (1984:124), cited by Machida (1989: 24), “Type 4” verbs

do not have any temporal structure but indicate that something takes on a property of a

state/stativity. They are always used in the Te-iru construction and never in the u/ru

form (the Present construction). Type 4 verbs with the Te-iru construction can be

regarded as frozen idiomatic expressions used as adjectives. They are actually similar

to “be boring/surprising/interesting” in English though semantically different.

Watashi no ronbun wa kare no ronbun yori sugure-te-iru.

(42) I GEN dissertation TOP he GEN dissertation than excel-TE-IRU
My dissertation excels his dissertation.
My dissertation is more outstanding than his dissertation.

*Watashi no ronbun wa kare no ronbun yori sugureru.

(43) I GEN dissertation TOP he GEN dissertation than excel
My dissertation is more outstanding than his dissertation.

We can summarise the types of Japanese verbs with their occurrence and meaning

in the Te-iru construction as follows:

Table 3. Meanings of te-iru

Type of verbs: Meaning of the te-iru form:
Stative verbs N/A
Continuative verbs ACTIVITY IN PROGRESS
Instantaneous verbs RESULTATIVE STATE
Type 4 verbs Almost Frozen Expression (they could
be always used with the Te-iru form.)

with te-iru. The term, te-iru form/construction, is used to include both these phonological patterns.

77
 Chapter 3 Aspect in Japanese

The Te-iru construction also has another meaning of retrospective ((Machida

1989: 160-161), (Nitta 1997: 237), (Washio 1997: 115), which is applied to all types

of verbs that take the Te-iru form (except “Type 4” verbs). In this sense it functions as

the existential use of the English Present Perfect. The following examples are cited

from Machida (1989: 161):

stative verb:

Huji-san wa izenni kono basho kara nido mie-te-iru.

(44) Mt.Fuji TOP before this place from twice be.seen-TE-IRU
Mt. Fuji has been seen from here twice before.

continuative verb:

Taro wa sannenmaeni kono yamamichi o arui-te-iru.

(45) Taro TOP three.years.ago this mountain.trail ACC walk-TE-IRU
Taro (has) walked this mountain trail three years ago.

instantaneous verb:

Koronbusu wa 1492nen ni amerika o hakkenshi-te-iru.

(46) Columbus TOP 1492year in America ACC discover-TE-IRU
Columbus (has) discovered America in 1492.

According to Machida (1989: 161), these verbs get retrospective interpretation when

used with past time adverbials such as izenni 'before' or sannenmaeni 'three years ago'

or in contexts where activity in progress and resultative state meanings are excluded.

However, some instantaneous verbs such as hakkensuru ‘discover’ or mokugekisuru

‘ witness’) that refer to an event that leaves no perceptible result automatically bear a

retrospective interpretation in the Te-iru form, even though there are no appropriate

adverbials or contexts. (46) will be interpreted to have retrospective meaning even

without the phrase ’in 1492.’ (For details, see later discussion on the Te-iru form.)

Finally, I shall discuss the distinction between perceptible and non-perceptible

states. The resultative state reading with te-iru is only possible where there is a

perceptible state of the action such as in example (41). Here, we see the state of the

78
 Chapter 3 Aspect in Japanese

dog, and know that it is dead. That is the dog’s being dead is the result of its dying

(change of state from being alive to being dead). In this sense, intransitive verbs that

can be construed as achievement verbs are likely to get the resultative state reading.

However, the transitive achievement types (prototypically korosu ‘kill’) can be argued

not to have the resultative state reading as it is difficult to see any perceptible state in

the subject. The following example illustrates:

Kare wa jibun no inu o koroshi-te-iru.

(47) he TOP own GEN dog ACC kill-TE-IRU

When we see the state of the dog, we can perceive that it is dead and may see some

directed evidence that the dog was killed (such the knife stuck in its stomach).

However, merely looking at the dog cannot indicate who has carried out the action.

Also we have no direct evidence whether the agent has killed the dog or not, UNLESS

he is standing there, with a knife in front of the dog. In this case, we can infer what

has happened from the scene. However, this interpretation largely depends on contexts.

The distinction between the perceptible and non-perceptible, and thus, that of between

the resultative and retrospective readings, is not so clear with transitive achievement

verbs (korosu ‘kill’) as with intransitive achievement verbs (shinu ‘die’). In most of

the cases, the direct object of the transitive achievement does not show any perceptible

indication of who has done the act, nor does the subject manifest any perceptible

result of his action. Therefore, these verbs with te-iru (such as (47)) are better

analysed as having the retrospective rather than the resultative state reading10. Two

exceptions are (i) when the agent is realised as dareka “somebody” (cf. Machida

1989: 39) since this expression can obscure who has carried the action and the dead

state of the dog, for example, can indicate that it is the result of somebody’s killing the

10
Please note that some transitive achievement verbs can possibly get the activity in progress reading of
the runup achievement construal with te-iru, which will be discussed shortly.

79
 Chapter 3 Aspect in Japanese

dog and (ii) when sentences have reflexive construction as in she dyed her hair since

the state of her hair (a body part of the subject) can indicate that it is the result of her

carrying out the action of dying. In these two cases, verbs of transitive achievement

can get the resultative reading.

Though we have seen several readings of the Te-iru construction, the classification

of its distinct uses is coarse-grained. They can be subdivided into small categories in

terms of the aspectual types of Croft (2000), reflecting the complexity of the real

world situation and of the speaker’s construal operation when a certain predicate is

used with a certain TA construction.

3.2.3.2.2. Senses of the Te-iru construction

In a more subtle classification of aspectual types, the Te-iru construction has actually

eight senses: the undirected activity construal (coded as 2U), the directed activity

construal (coded as 2D), the transitory state construal (coded as 2T), the inherent state

construal (coded as 2I), the runup achievement construal (coded as 2R), the habitual

interpretation (coded as 2H), the perfect construal (coded as 2F), and the

uninterpretability construal (coded as 2*). The resultative state reading I have

discussed in relation to te-iru can be split into 2T and 2I depending on the nature of

the result state. The activity in progress reading is divided into 2U, 2D, and 2R. The

habitual interpretation is a subtype of activity in progress since each event is construed

as a cyclic achievement and a series of each event is reconstrued as undirected activity.

A retrospective reading is equivalent to 2F in the analysis.

The first construal is undirected activity. This is found as prototypical with the

following situation types:

80
 Chapter 3 Aspect in Japanese

[24 Manner of Motion 1]

Taro wa ima hashit-te-iru.
(48) Taro TOP now run-TE-IRU
Taro is running now.

[25 Manner of Motion 2]

Hanako wa ima oyoi-de-iru.
(49) Hanako TOP now swim-TE-IRU
Hanako is swimming now.

[26 Motion/Movement]
Jishin de tsukue ga ugoi-te-iru!!
(50) earthquake with desk NOM move-TE-IRU
The desk is moving because of the earthquake!!

[27 Performance]
Jacqui wa ima utat-te-iru.
(51) Jacqui TOP now sing-TE-IRU
Jacqui is singing now.

The first three types are verbs of motion without any goal phrases. Performance,

without a direct object, also refers to an undirected activity without reaching any result

point but merely activity as with singing in example (51).

Some situation types get the undirected activity construal of iterated cyclic

achievements.

[22-1 Contact]
Jack ga mado o tatai-te-iru.
(52) Jack NOM window ACC knock-TE-IRU
Jack is knocking at the window.

[22-2 Bodily Motion]

Mary wa John ni te o hut-te-iru.
(53) Mary TOP John to hand ACC wave-TE-IRU
Mary is waving her hand to John.

What the above examples denote is that the cyclic achievement of one knocking at the

door or one waving one’s hand is iterated without achieving any specific result.

This construal is reinforced in other situation types by using support of

reduplicated mimetic/onomatopoeic phrases. Japanese is abundant in onomatopoeic

and mimetic expressions that imitate sounds or describe manners. For example, chu

81
 Chapter 3 Aspect in Japanese

imitates the sound of a mouse’s cry and chika describes the flashing of a light.

Reduplicated means that the words are doubled up with a resultant meaning that an

event they describe is also repeated more than once. As a result, chuchu or pikapika

indicate or reinforce the fact that the events are repeated. I gloss these phrases as

“ONMP2” in the examples below:

[21-1 Light Emission]

Sono dentou wa chikachika hikat-te-iru.
(54) the light TOP ONMP2 shine-TE-IRU
The light is flashing.

[21-2 Sound Emission]

Nezumi ga chuchu nai-te-iru.
(55)
mouse NOM ONMP2 cry-TE-IRU
The mouse is squeaking.

Other situation types whose force-dynamic relation is similar to inactive action

have also the undirected activity construal. A typical example of inactive action is

holding something, where one participant is engaged in an activity which is not

overtly explicit. In the following situation types, the activities the subject is engaged

in are not so overt.

[12 Emotional Activity]

Haha wa watashi no seikou o yorokon-de-iru.
(56) mother TOP I GEN success ACC rejoice-TE-IRU
My mother is happy with my success.

[13 Sleeping]
Chichi wa ima shizukani nemut-te-iru.
(57) father TOP now quietly sleep-TE-IRU
My father is sleeping now.

[14 Perceptual Activity]

Chichi wa ima terebi no nyuusu o mi-te-iru.
(58) father TOP now TV GEN news ACC watch-TE-IRU
My father is watching news on TV now.

In the situations denoted above, the subjects are engaged in non-overt activity. The

sentences can be ambiguous between an undirected activity construal and a transitory

82
 Chapter 3 Aspect in Japanese

state construal, especially in Japanese, since te-iru can have the activity in progress

reading and the resultative state reading. These above verbs get the achievement

reading with the Past construction (see section §3.2.3.3). That is, the situations above

may be construed as a resultative state of the achievements. However, I argue that

they are undirected activities in that people other than the subject can still see what

kind of activities are going on. For example, all the sentences above (from example 56

to 58) can be used to answer the question What is she/he doing?. In the case of the

first one (emotional activity), this question may sound strange. However, one can

show one’s emotion in an explicit way. For example, one can ask the daughter the

above question, watching her mother crying and wondering what is going on, and she

can answer the question with saying the example (56) and explain that the mother is

crying with joy. Another good reason to treat them as activities is that they can be

modified with a manner adverbial such as shizukani ‘quietly’, as in example (57).

The second sense of the Te-iru construction is the directed activity construal.

Unlike with the undirected activity construal, the situations are interpreted as

increasing the degree on a certain scale (the verbal scale). The following are typical

examples.

[27 Performance]
Jacqui wa ima Corrs no shinkyoku o utat-te-iru.
(59) Jacqui TOP now Corrs GEN new.song ACC sing-TE-IRU
Jacqui is singing Corrs’ new song now.

[29 Consumption]
Joel wa yuushoku no pasuta o tabe-te-iru.
(60) Joel TOP dinner GEN pasta ACC eat-TE-IRU
Joel is eating the pasta for his dinner.

[31 Creation]
Haha wa watashi no doresu o tsukut-te-iru.
(61) mother TOP I GEN dress ACC make-TE-IRU
My mother is making a dress for me.

A characteristic of this construal is that it can occur with adverbial phrases which

83
 Chapter 3 Aspect in Japanese

illustrate that the situation goes into directed change incrementally such as

sukoshizutsu ‘little by little’ or masumasu ‘more and more’. The following situation

types even require this adverbial support to have this interpretation with te-iru:

[15 Attitudes]
Kare wa masumasu hinekure-te-iru.
(62) he TOP more.and.more get.warped-TE-IRU
He is getting more and more warped.

[16 Change of Weight]

Mary wa sukoshizutsu hutot-te-iru.
(63) Mary TOP little.by.little get.fat-TE-IRU
Mary is getting fatter little by little.

[17 Change of Physical Properties]

Suupu ga sukoshizutsu atatamat-te-iru.
(64)
soup NOM little.by.little get.warm-TE-IRU
The soup is getting warm.

The third construal associated with the Te-iru construction is transitory state. The

following are examples:

[17 Change of Physical Properties]

Suupu ga atatamat-te-iru.
(65) soup NOM get.warm-TE-IRU
The soup is warm.

[19 Directed Motion]

John wa jibun no koya ni hait-te-iru.
(66) John TOP own GEN kennel ALL go.into-PAST
John is in his kennel (as a result of going there).

[26 Motion/Movement]
Kinou no jishin de tsukue ga sukoshi ugoi-te-iru.
(67) yesterday GEN earthquake with desk NOM a.little move-TE-IRU
The desk has moved a little because of yesterday’s earthquake.

[33 Dyeing and Shaving]

Kanojo wa kami o {kiiroku} some-te-iru.
(68) she TOP hair ACC {yellow} dye-TE-IRU
She has dyed her hair yellow. (Now she has yellow hair.)

Actually, the nature of the result state, especially whether it is transitory or

inherent, is not always easy to judge, since it depends on various kinds of possible

context. Some situation types may have both transitory state and inherent state

84
 Chapter 3 Aspect in Japanese

construals, depending on the context or adverbial support. The following types also

have an inherent state construal (to be discussed shortly) with te-iru:

[16 Change of Weight]

Mary wa mezurashiku ima hutot-te-iru.
(69) Mary TOP unusually now become.fat-TE-IRU
Mary is fat now, which is unusual for her.

[30 Covering]
Kiri ga machi o sukkari oot-te-iru.
(70) fog NOM city ACC completely cover-TE-IRU
The fog completely covers the city.

In example (69), ima ‘now’ forces the transitory state construal and in (70) the types

of the arguments give its interpretation as the transitory state.

Situation types which denote situations which can be thought of as inactive action

are also thought to have a transitory state reading as well. Unlike the situation types of

Emotional Activity, Sleeping, Perceptual Activity, which I referred to as having the

undirected activity reading, the following types have the transitory state construal with

te-iru:

[8 Posture 2]
Steve wa asoko ni tat-te-iru.
(71) Steve TOP there at stand-TE-IRU
Steve is standing over there.

[9 Perception]
Koko kara Huji-san ga mie-te-iru!!
(72) here from Mt.Fuji NOM be.visible-TE-IRU
I can see Mt. Fuji from here!!

[11 Cognition 2]
Watashi wa kami o shinji-te-iru.
(73) I TOP God ACC believe-TE-IRU
I believe in God.

One motivation for assigning the transitory state reading to these situation types is that

the above sentences cannot be used as answers to what are you/is she/he doing?.

Another motivation is that the manner adverbial shizukani ‘quietly’ cannot modify

them either. That is, situations denoted by these predicates are closer to ‘state’ than to

85
 Chapter 3 Aspect in Japanese

‘activity’11.

The fourth construal of te-iru is the inherent state construal. “Type 4” verbs get

this construal.

[7-1 Relation 2]
Kare no ronbun wa kono ronbun yori sugure-te-iru.
(74) he GEN article TOP this article than excel-TE-IRU
His dissertation is superior to this dissertation.

[7-2 Posture 1]
Huji-san ga me no mae ni sobie-te-iru.
(75) Mt.Fuji NOM eye GEN front at tower-TE-IRU
Mt. Fuji towers high in front of us.

Intransitive achievement type of situations also appear with te-iru with an inherent

state interpretation:

[35 Dying]
Neko ga michibata de shin-de-iru.
(76) cat NOM roadside on die-TE-IRU
The cat is dead on the roadside.

[36 Fracture/Breaking (intra)]

Heya no mado ga ware-te-iru!
(77) room GEN window NOM break-TE-IRU
The window in the room is broken!

Some situation types that have the 2T interpretation appear in this construal as

well.

[16 Change of Weight]

Mary wa hutot-te-iru.
(78) Mary TOP become.fat-TE-IRU
Mary is fat.

[30 Covering]
Midori no kigi ga sono kuni o oot-te-iru.
(79) green GEN trees NOM the country ACC cover-TE-IRU
Green trees cover the country.

Example (78), without the adverbial ima ‘now’, can denote an inherent property of a

person (compared with (69)). In the situation type of Covering (79), the fact that the

11
Actually, [8 Posture 2] accepts the adverbial. This shows the ambiguous status of this situation type

86
 Chapter 3 Aspect in Japanese

arguments are different from those of (70) can make the predicate get the inherent

state reading.

The fifth construal is the runup achievements construal. This is identical to runup

achievements in English (She’s dying!), though the number of situation types which

allow this construal with te-iru is small. They involve a directed transition to a result

state (bounded process), and this transition is construed as extended on the t scale.

However, this transition is not an incremental process, which is shown by the

unacceptability of adverbials such as sukoshizutsu ‘little by little’. The following

sentences are all grammatical but would become ungrammatical with the adverbial.

[34 Dressing]
Jacqui wa ima tonari no heya de kimono o ki-te-iru.
(80) Jacqui TOP now next GEN room at Jap.dress ACC put.on-TE-IRU
Jacqui is putting on her Japanese dress in the next room.

[38 Killing]
Haha wa kanshasai ni shichimencho o kososhi-te-iru.
(81) mother TOP thanksgiving.day for turkey ACC kill-TE-IRU
Mother is killing a turkey for Thanksgiving Day.

[39 Winning]
Nihon chiimu wa kankoku chiimu ni genzainotokoro kat-te-iru.
(82) Japan team TOP Korea team against at.this.moment win-TE-IRU
The Japanese team is leading the Korean team at this moment.

The sixth construal of the construction is the habitual interpretation. Situation

types that have this 2H construal also have the 1H construal but not vice versa. Unlike

the 1H construal, 2H construal especially implies the agent’s will to have control over

the repeated events of habitual interpretation, such as the agent making it a rule to go

to bed at eleven at night or to eat rice as a breakfast. For this interpretation, adverbials

that manifest regularity such as maitsuki ‘every month’ or mainichi ‘every day’ are

required. A lot of situation types allow this construal, but the following examples

adequately illustrate the matter:

between the result state and the undirected activity of inactive action.
87
 Chapter 3 Aspect in Japanese

[13 Sleeping]
Chichi wa maiban 11ji ni nemut-te-iru.
(83) father TOP every.night 11o'clock at sleep-TE-IRU
My father goes to bed at 11 o’clock every night.

[29 Consumption]
Watashi wa maiasa gohan o tabe-te-iru.
(84) I TOP every.morning rice ACC eat-TE-IRU
I eat rice every morning.

The seventh construal associated with the Te-iru construction is the perfect

construal. With the adverbial support of past time adverbials such as kako ni ‘in the

past’ or iterative adverbials such as ikkai ‘once’ or nikai ‘twice’, most of the situation

types have this construal. However, the following three examples adequately

illustrate:

[14 Perceptual Activity]

Chichi wa kako ni ichido yurei o mi-te-iru.
(85) father TOP past in once ghost ACC see-TE-IRU
My father has seen a ghost once in the past.

[24 Manner of Motion 1]

Chichi wa ichido Honolulu marason de hashit-te-iru.
(86) father TOP once Honolulu marathon in run-TE-IRU
Father has run once in the Honolulu marathon.

[40 Discovery]
Sono kenkyuu chiimu wa {kyonen} shinsei o hakkenshi-te-iru.
(87) the research team TOP {last.year} new.star ACC discover-TE-IRU
The research team has discovered a new star (*last year).

The Discovery type is the only situation type that has only an 2F construal with te-iru.

This is because the irreversible directed achievement that this type denotes cannot be

construed as taking time like a runup achievement nor can it leave any perceptible

result on its direct object.

The last construal with te-iru is uninterpretability. Situation types which are

expressed by non-verb predicates are ungrammatical since such predicates cannot be

combined with the Te-iru construction, to begin with. Moreover, the situation types of

Existence and Necessity are uninterpretable with the construction.

88
 Chapter 3 Aspect in Japanese

[2-1 Existence]
*Hon ga soko ni at-te-iru.
(88) book NOM there at exist-TE-IRU
The book is there.

[2-2 Necessity]
*Watashi ni-wa kane ga it2-te-iru.
(89) I to-TOP money NOM need-TE-IRU
I need money.

To summarise all the above uses, I propose a single unified meaning of the

construction, which is that it takes the temporally extended phase after an

achievement (inception or completion) and construes it as an extended state. One

exception is when it is applied to an inherent state. In this case, since the inherent state

has no inception phase, the construction takes the temporally extended phase of

inherent state and construes it as a transitory state. This case is examplified by [6

Relation1] which has both 1I and 2I construals.

[6 Relation 1]

Gengogaku wa bunkengaku to-wa kotonaru.

(90) Linguistics TOP philology with-TOP differ
Linguistics differs from philology.

[6 Relation 1]
Gengogaku wa bunkengaku to-wa kotonat-te-iru.
(91) Linguistics TOP philology with-TOP differ-TE-IRU
Linguistics differs from philology.

Actually, the subtle difference between the above sentences should be attributed to a

difference in style. Basically, when the same predicate has the same construals in the

Present and with the Te-iru constructions as above, the former sounds more assertive

and the latter sounds more colloquial and soft12.

Other construals of the Te-iru construction are explained by the above hypothesis.

The 2U, 2D, 2I, and 2T construals all have an extended phase after the inception phase

(achievement). 2H is thought to be a derived undirected activity type whose

89
 Chapter 3 Aspect in Japanese

fine-grained construal is iterated cyclic achievements. 2F would be treated like the

English present perfect, which Croft (2000) defines as “taking a bounded event

preceding the present moment, construing it as an achievement”. For example, in the

case of somebody’s having divorced twice, the d-transition is thought of as that from

the event of getting divorced twice having not occurred to that of the event of getting

divorced twice having occurred. The aspectual contour of 2F takes the form of

irreversible directed achievements as what has happened is unchangeable. Thus, even

2F has a temporally extended phase after the inception.

Finally, it has to be mentioned that the present moment should be included in the

state that results from the reconstrual of situations due to the use of the Te-iru

construction.

Now let us compare our analysis of te-iru and that of Jacobsen (1992), who also

gives a unified account. He proposes that a unified meaning of te-iru is that “it

presents a state of affairs as existing in homogeneous, unchanging fashion over a

given interval of time” (Jacobsen 1992: 200). The two analyses are not contradictory

but harmonious; both of them explain that te-iru refers to a state which does not

undergo any changes. However, our analysis is more precise than Jacobsen’s in

specifying what kinds of event phase can be construed as states.

3.2.3.3. The Past construction

The Past construction in Japanese uses the ending “ta” with predicates. It has eight

construals: the accomplishment construal (coded as 3P), the achievement construal

(coded as 3V), the cyclic achievement construal (coded as 3C), the runup achievement

construal (coded as 3R), the undirected activity construal (coded as 3U), the directed

12
Some of my informants have agreed.

90
 Chapter 3 Aspect in Japanese

activity construal (coded as 3D), the transitory state construal (coded as 3T), and the

uninterpretability (coded as 3*).

A first construal with the Past construction is the accomplishment. A 3P construal

is compatible with container adverbials such as gohun de ‘in five minutes’ or ichijikan

de ‘in an hour’, which measure the interval from the inception to the completion of the

events. Normally, situation types in this construal have a bounded direct object or a

resultative phrase which can serve as a delimiter of the event.

[14 Perceptual Activity]

Watashi wa sono bideo o nijikan de mi-ta.
(92) I TOP the video ACC two.hours in watch-PAST
I watched the video in two hours.

[16 Change of Weight]

Mary wa sankagestsu de {kirei ni} yase-ta.
(93) Mary TOP three.months with {beautiful RST} become.thin-PAST
Mary became beautifully slender in three months.

[29 Consumption]
Jack wa ookina sandoicchi o ippun de tabe-ta.
(94) Jack TOP big sandwich ACC one.minute in eat-PAST
Jack ate the big sandwich in one minute.

Some intransitive verbs also occur in this construal as well.

[17 Change of Physical Properties]

Suupu wa gohun de atatamat-ta.
(95) soup TOP five.minutes with get.warm-PAST
The soup got warm in five minutes.

[20 Open/Close]
Sono tsubomi wa gojikan de hirai-ta.
(96) the bud TOP five.hour in open-PAST
The bud opened in five hours.

An accomplishment construal, which involve the profiled d-process, can be

modified with sukoshizutsu ‘little by little’, which verifies the characteristics of

incremental change in the directed activity.

91
 Chapter 3 Aspect in Japanese

[14 Perceptual Activity]

Watashi wa sono bideo o sukoshizutsu mi-ta.
(97) I TOP the video ACC little.by.little watch-PAST
I watched the video little by little.

Tochuumade ‘halfway’ is also compatible with 3P13, which is bounded with a profiled

d-transition to the result state as well as having the profiled d-process.

[14 Perceptual Activity]

Watashi wa sono bideo o tochuumade mi-ta.
(98) I TOP the video ACC halfway watch-PAST
I watched the video halfway.

The second construal is an achievement. The following situation types are typical

examples of 3V construals:

[35 Dying]
3V
Chichi wa totsuzen nakunat-ta.
(99) father TOP suddenly die-PAST
My father died suddenly.

[36 Fracture/Breaking (intra)]

3V
Sono tsubo wa totsuzen ware-ta.
(100) the vase TOP suddenly break-PAST
The vase broke suddenly.

[39 Winning]
3V
John wa 5ji 5hun ni kesshou.sen ni kat-ta.
(101) John TOP 5o'clock 5minutes at final.game OJCT win-PAST
John won the final game at 5:05.

The achievement construal, as we can see above, is compatible with a pointlike

adverbial which specifies a point in time as ‘at 5: 05’ or totsuzen ‘suddenly’ which

construes events as punctual.

Other situation types which typically denote events that are extended over the

t-scale get the inception interpretation with the above adverbials. This interpretation is

13
This adverbial is not compatible with unbounded events such as in the directed activity reading, since
the meaning of the adverbial (halfway) presupposes that there is a delimiter (endpoint).

92
 Chapter 3 Aspect in Japanese

regarded as an achievement construal with the Past construction, since the inception

itself is an achievement (d-transition).

[24 Manner of Motion 1]

?Taro wa totsuzen hashit-ta.
(102) Taro TOP suddenly run-PAST
Taro suddenly ran. (Taro suddenly started to run.)

[26 Motion/Movement]
Sono kuruma wa totsuzen ugoi-ta.
(103) the car TOP suddenly move-PAST
The car suddenly moved.

[27 Performance]
?Jacqui wa totsuzen utat-ta.
(104) Jacqui TOP suddenly sing-PAST
Jacqui suddenly sang. (Jacqui suddenly started to sing.)

The achievement construal is also possible if an event as a whole can be construed as

having happened in an instant.

[17 Change of Physical Properties]

Sono kinzoku wa sono saishin gijutsu de shunkanni hie-ta.
(105) that metal TOP that latest technology with in.an.instant cool-PAST
The metal cooled down in an instant with the latest technology.

[21-1 Light Emission]

Sono dentou wa totsuzen chikachika hikat-ta.
(106) the light TOP suddenly ONMP2 shine-PAST
The light flashed suddenly.

[38 Killing]
Chichi wa shunkanni shichimenchou o kososhi-ta.
(107) father TOP in.an.instant turkey ACC kill-PAST
Father killed a turkey in an instant.

The third construal with the Past construction is the cyclic achievement reading.

This refers to the one-off reading of an event. In contrast to the reduplicated

onomatopoeic/mimetic phrase, a non-reduplicated onomatopoeic/mimetic word plus

the quotation marker ‘to’ is used to reinforce the semelfactive reading. The phrase is

assigned ONMP1 in the English gloss as follows:

93
 Chapter 3 Aspect in Japanese

[21-1 Light Emission]

Sono dentou wa chikat-to ichido hikat-ta.
(108) the light TOP ONMP1 once shine-PAST
The light flashed only once.

[21-2 Sound Emission]

Nezumi ga ichido chut-to nai-ta.
(109) mouse NOM once ONMP1 cry-PAST
The mouse squeaked once.

[28 Create Mark/Defect]

Kanojo wa birit-to sono tegami o yabui-ta.
(110) she TOP ONMP1 the letter ACC tear-PAST
She tore the letter.

The fourth construal associated with the Past construction is the runup

achievement reading. With a container adverbial, such as gohun de ‘in five minutes’,

quite a number of situation types get this construal. In some situation types which

denote a prototypically punctual event such as dying, the container adverbial measures

the interval up to the d-transition to the result state.

[10 Cognition 1]
Taro wa ikkagestsu de sono suushiki o rikaishi-ta.
(111) Taro TOP one.month in the formula ACC understand-PAST
Taro understood the formula in one month (one month later).

[32 Dismantle]
Sono kodomo wa juppun de sono omocha o kowashi-ta.
(112) the child TOP ten.minute in the toy ACC destroy-PAST
The child destroyed the toy in ten minutes (ten minutes later).

[35 Dying]
Chichi wa hatsubyou kara ichinen de nakunat-ta.
(113) father TOP falling.ill from one.year in die-PAST
My father died in one year after he fell ill.

In the above examples, the container adverbial measures the interval till Taro’s

understanding (maybe he made efforts and understood the formula suddenly at the

end), till the child’s destroying of the toy (maybe he kept on doing something wrong

with the toy, and finally destroyed it at the last moment), and till the father’s death (he

was ill for some time and died in the end).

94
 Chapter 3 Aspect in Japanese

In other cases, the container adverbial with the Past construction measures the

interval up to the inception of the activity, as the following examples illustrate:

[24 Manner of Motion 1]

?Taro wa ippun de hashit-ta.
(114) Taro TOP one.minute in run-PAST
Taro ran in one minute (one minute later).

[26 Motion/Movement]
Sono kuruma wa sanpun de ugoi-ta.
(115) the car TOP three.minute in move-PAST
The car moved in three minutes (three minutes later).

[27 Performance]
?Jacqui wa gohun de utat-ta.
(116) Jacqui TOP five.minutes in sing-PAST
Jacqui sang in five minutes (five minutes later).

In the above examples, the container adverbial measures the interval till the inception,

that is, the time till Taro started running, the time till the car started moving, and the

time till Jacqui started singing.

The fifth construal is the undirected activity reading. This construal is made clear

and reinforced with the support of a durative adverbial such as ichijikan no aida ‘for

an hour’, which specifies a length of time for which an event lasts.

[12 Emotional Activity]

Haha wa nagai aida watashi no seikou o yorokon-da.
(117) mother TOP long period I GEN success ACC rejoice-PAST
My mother was happy with my success for a long time.

[22-1 Contact]
(118)
Jack wa Brigitte no kata o ichijikan no aida tatai-ta.
Jack TOP Brigitte GEN shoulder ACC one.hour GEN period pat-PAST
Jack patted Brigitte on the shoulder for an hour.

[25 Manner of Motion 2]

Hanako wa ichijikan no aida oyoi-da.
(119) Hanako TOP one.hour GEN period swim-PAST
Hanako ran for one hour.

[27 Performance]
Jacqui wa gohun no aida utat-ta.
(120) Jacqui TOP five.minutes GEN period sing-PAST

95
 Chapter 3 Aspect in Japanese

Jacqui sang for five minutes.

The sixth construal is the directed activity reading. Like the 3P (accomplishment)

construal, it can be modified with sukoshizutsu ‘little by little’ or masumasu ‘more and

more’. However, it does not imply that a certain new state is achieved, unlike in the

case of an accomplishment.

[15 Attitudes]
Kare wa masumasu hinekure-ta.
(121) he TOP more.and.more get.warped-PAST
He got more and more warped.

[16 Change of Weight]

Mary wa sukoshizutsu hutot-ta.
(122) Mary TOP little.by.little become.fat-PAST
Mary became fat little by little.

[17 Change of Physical Properties]

Suupu wa sukoshizutsu atatamat-ta.
(123) soup TOP little.by.little get.warm-PAST
The soup got warm little by little.

In the above examples, the subjects only moves up on the scalar property denoted by

the verbs but no definite result state is specified.

Some situation types which prototypically denote accomplishment can have the

directed activity construal as well. With a durative adverbial, verbs in the examples

below can indicate that the subject is engaged in the directed activity with a

prospective result state (i.e. accomplishments). However, they do not necessarily

entail that the final state is achieved, so the event is construed as directed activity and

not as an accomplishment:

[27 Performance]
Jacqui wa gohun no aida sono shinkyoku o utat-ta.
(124) Jacqui TOP five.minute GEN period the new.song ACC sing-PAST
Jacqui sang Corrs’ new song for five minutes.
(She did not necessarily finish the song.)

[29 Consumption]
Joel wa gohunkan no aida yuushoku o tabe-ta.
(125)
Joel TOP five.minutes GEN period supper ACC eat-PAST
Joel ate his supper for five minutes.

96
 Chapter 3 Aspect in Japanese

(He did not necessarily finish his supper.)

The seventh construal is the transitory state. As the examples below show, this

construal can be used with a pointlike adverbial (126), an interval adverbial (127), and

a durative adverbial (128). Even without adverbial support, this construal is felicitous

(129). The following are examples with the transitory state construal.

[2-1 Existence]
Gohun mae wa soko ni hon ga at-ta.
(126) five.minute ago TOP there at book NOM be-PAST
The book was there five minutes ago.

[2-2 Necessity] iru2 ‘need’

Kyonen kare ni-wa kane ga it2-ta.
(127) last.year he to-TOP money NOM need/necessary-PAST
He needed money last year.

[3 Social Role]
John wa gonenkan shachou dat-ta.
(128) John TOP for.five.years president DA-PAST
John was president for five years.

[5 Bodily States]
Kare wa byouki dat-ta.
(129) he TOP ill DA-PAST.
He was ill.

The last construal of the Past construction is uninterpretability. Three classes,

Relation 1, Relation 2, and Posture 1, are ungrammatical with the Past tense

construction. Here is an example:

[7-1 Relation 2]

*Benkyou de-wa Joe wa kanojo no otouto yori sugure-ta.

(130) study in-TOP Joe TOP she GEN brother than excel-PAST
Joe excelled (did better than) her brother in study.

Other four situation types which prototypically denote inherent states (Natural

Kinds, Ethnicity, Biological Kind, and Physical Properties) cannot occur with the Past

construction unless the referent is interpreted as no longer existing or the construction

is used for “a distal perspectival construal” (Croft 2000: 43). This is similar to

English:

97
 Chapter 3 Aspect in Japanese

[1-2Ethnicity]
Haha wa igirisujin dat-ta.
(131) mother TOP English DA-PAST
My mother was English. (She is not alive now.)

[1-3 Biological Kinds]

Kinou mi-ta mono sore wa tokage dat-ta.
(132) yesterday see-PAST entity that TOP lizard DA-PAST
The thing I saw yesterday was a lizard.

To summarise, in relation to the Past construction, it is safe to say that the profiled

part of the aspectual contour precedes the present moment and the aspectual contour

includes an unprofiled state in the present (Croft (2000: 38)) as in the English Past

construction.

3.2.3.4. Summary

There follows a table summarising situation types and aspectual classes of Japanese.

There are forty-eight situation types, which belong to forty aspectual classes. Situation

types which have identical distributional behaviour in the three constructions are

grouped together as one aspectual class. The data is not exhaustive and more types

and classes will be added in future study. However, for the present purpose of

showing different senses of the three TA constructions, the number of situation types

and aspectual classes is satisfactory.

98
 Chapter 3 Aspect in Japanese

Table 4. Japanese situation types and aspectual class

situation type example Present (1) Te-iru (2) Past (3)

1 Natural Kinds daiamondo da I * *
‘be a diamond’
Ethinicity Igirisujin da I * *
‘be English’
Biological Kinds tokage da I * *
‘be a lizard’
Physical Properties se ga takai I * *
‘be tall’
2 Existence aru, iru1 ITH * T
‘be’
Necessity iru2 ‘need’ ITH * T
Dispositions shinsetsu da ITH * T
‘be kind’
3 Social Role shachou da T * T
‘be president’
4 Physical States tsumetai ITH * T
‘be cold’
5 Bodily States byouki da TH * T
‘be ill’
6 Relation 1 kotonaru I I *
‘differ’
7 Relation 2 sugureru * I *
‘excel’
Posture 1 sobieru * I *
‘tower’
8 Posture 2 tatsu ‘stand’ H THF VR
9 Perception mieru ITH TF VRT
‘see,be visible’
10 Cognition 1 rikaisuru H IRHF? VR
‘understand
11 Cognition 2 omou ‘think’ ITH THF? VR?
12 Emotional Activity yorokobu H UF VU
‘rejoice,
be happy’
13 Sleeping nemuru ‘sleep’ H UHF VRU
14 Perceptual Activity miru ‘watch’ H UDHF PVR?UD
15 Attitudes hinekureru IH DTIHF? VD
‘become warped’
16 Change of Weight hutoru ‘become H DTIF PD
fat’
17 Change of Physical chijimu ‘shrink’ IH DTF PVD
Properties
18 Blooming saku ‘bloom’ IH TF PVR?U?
19 Directed Motion hairu ‘go into’ H THF PVR
20 Open/Close hiraku ‘open’ IH DTHF PVR?D

99
 Chapter 3 Aspect in Japanese

21 Light Emission hikaru ‘shine’ IH UHF VCR?U

Sound Emission naku ‘cry’ IH UHF VCR?U
22 Contact tataku ‘hit’ H UHF VCR?U
Bodily Motion te o huru ‘wave H UHF VCR?U
one’s hand’
23 Touching hureru ‘touch’ H UTHF VCR?U
24 Manner of Motion hashiru ‘run’ IH UIHF PV?R?U
1
25 Manner of Motion aruku ‘walk’ IH UHF PV?R?U
2
26 Motion/Movement ugoku ‘move’ IH UDTF VRUD
27 Performance odoru ‘dance’ IH UDHF PV?R?UD
28 Create Mark/Defect hikkaku ‘scratch’ IH UDHF VCR?U
29 Consumption taberu ‘eat’ IH DHF PV?D
30 Covering oou ‘cover’ H DTIHF PVR?D
31 Creation tsukuru ‘make’ IH DHF PD
32 Dismantle torikowasu IH DHF PVRD
‘dismantle’
33 Dyeing and kami o someru IH DTHF PV?D
Shaving ‘dye one’s hair’
34 Dressing kiru ‘dress’ IH TRHF VR
35 Dying shinu ‘die’ * IF VR
36 Fracture/Breaking wareru ‘break’ I IF VR
(intra)
37 Attachment kuttsuku ‘stick’ IH TF VR
38 Killing korosu ‘kill’ IH RF VR
39 Winning katsu ‘win’ H RHF VR
40 Discovery hakkensuru * F VR
‘discover’

For convenience only, the left column of the table shows the number of aspectual

classes. The second column lists all the situation types discussed in this chapter.

Sometimes one situation type relates to two types of aspectual behaviour. In this

circumstance, the situation type is subdivided into 1 and 2 (as with Posture 1 and

Posture 2). The third column gives a typical example of that Japanese situation type or

subtype and its English translation. Then, the following three columns indicate the

aspectual sense each situation type gets with the constructions using alphabetical

codes I used in the previous discussion; the order of the columns is the Present (1), the

Te-iru (2), and the Past (3) constructions. Question marks as in “R?” indicate that the

construal in question is not perfectly natural but acceptable.

100
 Chapter 3 Aspect in Japanese

Though cutting across these situation types into a more general types of aspectual

classes is not easy (as there are forty), I try to put the situation types in a relatively

understandable order in terms of aspectual behaviour. At least, I put the types which

have similar aspectual properties close together in the list. I start with “inherent state”

types (prototypically type 1 of aspectual class), then proceed to “transitory state” types

(types 2 to 5). Types 6 to 7 are inherent state but I position them as the last of the

“state” types as they are peculiar in that te-iru refers to inherent state. The positioning

of type 6 illustrates its “in between” status; it is similar to type 1 in that it refers to

inherent state in the Present but also similar to type 7 in that it also refers to inherent

state with te-iru.

Then, types 8 to 14 are ambiguous in interpretation between “state” and “activity”

with te-iru. They are situation types of inactive action (posture and sleeping),

cognition, perception, and emotion. Because of the semantic nature of these verbs, it is

difficult to see if a certain situation denoted by them is a state or activity. For example,

somebody’s standing on the hill may be construed as the transitory state of the person

after his standing up or may be construed as the invisible activity of the person’s

holding standing position. What is shared by these situation types is an action that is

invisible. I actually distinguish types 8, 9, 10 and 11 as being closer to “state” with

te-iru and types 12, 13, and 14 as being closer to “activity”. I use two criteria to assign

them to activity; (i) whether the Te-iru construction can be used as an answer to the

question What is “one” doing? and (ii) whether they are compatible with adverbials

such as shizukani ‘quietly’ that modifies the manner of action. However, in this

respect, Posture 2 satisfies the condition (ii). This is a good example to show the fuzzy

boundaries that exist between larger aspectual types such as “states” or “activities”.

Only a type that satisfies the two tests has been classified as an activity.

101
 Chapter 3 Aspect in Japanese

Types 15 to 20 are “roughly” intransitive change of state verbs. (Type 20 can be

used transitively as well.) Again, the classification can not be so clear-cut. Other than

types 18 and 19, they have a directed activity construal with te-iru. That is, the change

of state can be construed as extending over time. Types 18 and 19 are closer to

“achievement” types (compare them with Types 34 to 40) in that they do not have any

activity in progress reading (specifically, 2U and 2D) with te-iru. However, they can

get the accomplishment reading with a container adverbial in the Past, so they are not

prototypically achievements.

Types 21, 22, and 23 are verbs that can denote cyclic achievement. Basically, they

all have the one-off achievement reading in the Past (3C) and the undirected activity

reading with iterated achievements with te-iru (2U).

Types 24 to 33 form the continuum between “activity” and “accomplishment”.

Type 24 and 25 (Manner of Motion), allowing the undirected activity construal but not

the directed activity construal with te-iru, are prototypical “activity” verbs. Types 26,

27, and 28 have fuzzy boundaries as they may either get an undirected or directed

activity construal depending on other factors, namely, on whether an event can be

bounded with a direct object or a goal phrase. Types 29 to 33 are basically

accomplishment types as they get the accomplishment construal in the Past and also

the directed activity reading (not the undirected one) with te-iru. It is also worthy of

note that types 26, 30 and 33 can also have “achievement” characteristics in that they

will get the resultative state reading (transitory state) with te-iru. That is, we can either

focus on the activity part or the result part after the action of the “accomplishment”

situation types denoted by these verbs.

Types 34 to 40 are “achievement” types. However, again, the boundaries are not

so clear. Type 34 is quite peculiar in that it gets both the resultative state reading

102
 Chapter 3 Aspect in Japanese

(specifically, the transitory state reading) and the activity in progress reading

(specifically, the runup achievement reading) with the Te-iru construction. Actually,

type 34 is similar to type 33 in that both have a reflexive reading; the agent causes

changes to him/herself (e.g. by dyeing his/her own hair and dressing him/herself). The

only difference is that type 34 Dressing does not denote situations that have

intermediate/incremental stages. Other achievement type verbs (types 35 to 40) either

get the resultative reading (transitory state or inherent state) or the runup achievement

reading, or only the retrospective reading with te-iru. They all take achievement (3V)

and runup achievement reading (3R) in the Past, which is characteristic of

achievements verbs.

Not only are there fuzzy boundaries between the general aspectual types, but also

some aspectual classes which are distant in the table share similar aspectual properties.

For example, type 10 Cognition 1, and type 39 Winning, are different only due to the

presence of 2I construal in the former Type. Other aspectual behaviours are identical

(1H, 2H, 2R, 2F, 3V and 3R). Also, type 13 Sleeping, and types 21, 22, and 23 are

similar in that they all have 1H, 2U, 2H, 2F, 3V, 3U,and 3R construals. Type 8

Posture 2, and type 19 Directed Motion, are also similar in allowing 1H, 2T, 2H, 2F,

3V, and 3R construals. Type 14 Perceptual Activity has prototypical characteristics of

accomplishment as in type 32 or 33 in allowing 2D, 3P and 3D construals.

Thus, there are quite a few aspectual classes and they cannot easily be categorised

into general larger types of state, achievement, activity, or accomplishment. These

aspectual classes form a continnum of general aspectual types with some of their

aspectual characteristics belonging to one general aspectual type and others belonging

to another general aspectual type.

Analysis of these complicated aspectual classes and types will continue with the

103
 Chapter 3 Aspect in Japanese

elaboration of semantic maps proposed by Croft (2000), where the meanings of

constructions and distribution classes (in this case aspectual types) are organised into a

conceptual space. In such maps, categories which have similar distributional patterns

will occupy connected regions. However, I leave this subject for future research.

3.3. Summary

This chapter introduced Croft’s (2000) two-dimensional representation for aspect,

which is the other semantic dimension of verbal semantics following the causal

structure discussed in Chapter 2. It also saw that there are more aspectual types than

those indicated by Vendler’s (1967). Then, it showed how Japanese predicates can be

categorised into aspectual classes in terms of their distributional behaviour with three

constructions: the Present, the Te-iru, and the Past constructions. At the same time, it

illustrated that the three constructions have several aspectual construals.

The next chapter will discuss how the causal structure representation and the

aspectual representation will be combined into verbal semantic representation and

then Chapter 5 and Chapter 6 will analyse verbs of putting and removing using the

new causal-aspectual representational model.

104
4. ANALYSIS AND METHOD

The purpose of this chapter is to introduce the causal-aspectual model on which the

analysis of verbs of putting and removing is based and to explain the methodology of

collecting the data of these verbs. Section 1 of this chapter gives a background to the

relation between the verbal scale/holistic theme and argument linking. Section 2

introduces and describes Croft’s (1999a) new version of argument linking theory,

which combines his causal structure analysis with the aspectual representation. I shall

examine how it works citing his geometrical representations for some English verbs.

Section 3, then, gives details on methodology, describing how the data presented in

this thesis has been collected and how it was examined. Section 4 provides some

necessary information on Japanese grammar and Section 5 briefly summarises and

leads into the subsequent chapters.

4.1. Incremental theme/verbal scale and the argument linking

The correlation between verbal scale and argument selection is discussed by Dowty

(1991) and Tenny (1992). Verbal scale is roughly equivalent to incremental theme in

Dowty and to measuring out in Tenny.

Dowty introduces the notion of incremental theme in his subject/object argument

selection theory. He proposes a set of entailments for Proto-agent and Proto-patient.

There are five properties for each (Dowty (1991: 572)), and an argument which has

more properties as Proto-agent is realised as subject and one which has more

properties as Proto-patient is realised as direct object. An incremental theme is one of

the properties of the Proto-patient. He accounts for the spray/load alternation, which

concerns direct object/oblique selection, by the very property of incremental theme. In

105
 Chapter 4 Analysis and method

brief, Dowty argues that the direct object is always the incremental theme in the two

variants. The following examples cited from Dowty (1991: 591-592) support his

argument:

(1a) John sprayed this wall with paint in an hour. (telic)

(1b) (#)John sprayed this wall with paint for an hour. ((#)atelic)
(1c) #John sprayed paint onto this wall in an hour. (#telic)
(1d) John sprayed paint onto this wall for an hour. (atelic)

(2a) #John sprayed subway cars with this can of paint in an hour. (#telic)
(2b) John sprayed subway cars with this can of paint for an hour. (atelic)
(2c) John sprayed this (whole) can of paint onto subway cars in an hour. (telic)
(2d) #John sprayed this (whole) can of paint onto subway cars for an hour. (#atelic)

As we can see, it is the direct object and not the other arguments that affects the

aspectual meaning of the whole sentence. In (1), when the direct object is bounded

(this wall), the telic reading is natural, while the atelic reading is more natural when

the direct object is a mass noun and thus nonbounded (paint). The same is true of (2);

subway cars are nonbounded, thus the whole event is interpreted as atelic, while the

bounded status of the direct object (this can of paint) also delimits the event. That is,

Dowty claims that the two variants of the spray/load alternation should be lexically

distinct but related, emphasising that there is an aspectual difference between the two,

which is reduced to the incrementality of the direct object.

Tenny (1992) argues for a stronger claim which directly relates the argument that

measures out the event to the direct object in her Aspect Interface Hypothesis. She

also gives examples of the locative alternation to support her hypothesis. One of the

supporting facts is that both the instrument role and the material role can appear in the

oblique position in the with-variant of spray, but only the latter is possible in the direct

object position in the locative variant. The following examples are from Tenny (1992:

15-16):

(3a) spray the wall with water (material as oblique)

(3b) spray the wall with a hose (instrument as oblique)

106
 Chapter 4 Analysis and method

(4a) spray water on the wall (material as direct object)

(4b) *spray a hose on the wall (*instrument as direct object)

Tenny’s explanation of the difference in grammaticality of (4a) and (4b) is that

material such as water can be moved to the wall little by little until it is gone, so the

event can be measured out while an instrument such as a hose cannot be construed as

moving it part by part. The instrument is not a measuring-out-argument and cannot

appear in the direct object position.

Dowty (1991) opposes Tenny’s hypothesis, which relates incremental theme

‘exclusively’ to the direct object position, arguing that the path but not the direct

object is an incremental theme in verbs of motion plus a goal phrase (push the cart to

New York) and that there are cases of an incremental theme subject crossing a

boundary (John entered the icy water very slowly). However, as we have seen above,

Tenny (1992) and Dowty (1991) treat the locative alternation in a similar way with the

direct object in each variant interpreted as the incremental theme.

Unlike Tenny and Dowty, Jackendoff (1996) admits little correlation between the

direct object and incremental themehood. One of his main arguments about the lack of

correlation between the incremental theme and the argument selection is that aspectual

properties are not necessarily entailed in the lexical entry of verbs, as Tenny claims

and Dowty implies. Jackendoff points out that the direct object, which is supposed to

be the incremental theme, does not always measure out the event, and that pragmatic

factors interact with the locative alternation of spray/load verbs. First of all, he claims

that the holistic interpretation, which is associated with the with-variant, is favoured

but not required (especially with content-oriented verbs like spray or splash). He gives

an example that the with variant can be atelic: Bill sprayed/smeared/dabbed/splashed

the wall with paint (for ten minutes), but it still wasn’t covered (Jackendoff 1996: 346).

Secondly, he argues that spray/load verbs have alternative aspectual construals

107
 Chapter 4 Analysis and method

(accomplishment or achievement). It is pointed out that one can load a truck with

many successive shovelfuls (accomplishment) or with one scoop of a steam shovel

(achievement) for load, and that one can spray the wall with paint successively,

covering different areas bit by bit (accomplishment) or spray one’s shoes with water

in a single burst (achievement) with spray. Thus, Jackendoff emphasises that

incrementality is not entailed as a part of the LCS of spray/load verbs, since the

incrementality or non-incrementality of the direct objects does not affect the argument

structures of these verbs. Rather, incrementality is an extraneous pragmatic factor. For

the spray/load alternation, he does not give any representation but only suggests that

two independent factors are involved in favouring the holistic interpretation; (i)

container-oriented verbs (load or pack) have a bias toward the holistic reading, while

content-oriented verbs (spray or splash) are more neutral1, and (ii) the with-variant

creates a bias toward the holistic reading, while the locative variant is more neutral.

As a result, load the truck with dirt (two factors involved) heavily favours the holistic

reading while spray paint on the wall (no factors involved) is neutral.

To summarise Jackendoff’s idea, certain cases (spray/load verbs) of telicity and

the measuring out phenomenon result from the interaction of lexical semantics and

pragmatics. In particular, these verbs should not be regarded as measuring out verbs.

Therefore, he admits that there is little correlation between measuring out and

argument selection.

I agree that verbs have different aspectual construals and they normally do not

1
A similar claim is made by Dowty (1991: 590-592). He ascribes subtle differences to spray and load
concerning the interaction of the telicity of the theme or direct object and that of the event as a whole to
pragmatic factors. First of all, spraying the wall and loading the hay are different in purpose; the former
is normally to cover the wall and the latter is normally to move the theme around. Secondly, one can
put paint on the wall indefinitely by putting paint repeatedly on the same place (atelic event), while one
cannot put hay onto the truck indefinitely (telic event) as the truck has a spatial limit.

108
 Chapter 4 Analysis and method

belong exclusively to one aspectual class. Most verbs have different aspectual

construals of situations. In Chapter 3, we saw that there can be different aspectual

construals of the same event. However, my belief is that it is still useful to a certain

extent to classify each verb according to the most common construal. It should be

noted that when there is more than one representation/construal possible for each verb

or predicate, cognitive linguistics does not differentiate which is basic and which is

derived. It only admits that one is more common and one is less common. For spray

and load verbs, the bit-by-bit spraying/loading, which is represented as an

accomplishment (with a verbal scale), is more common than the instantaneous version

that is represented as an achievement (with a trivial verbal scale). We can say that the

incremental theme in the common context is construed as a trivial verbal scale in the

less common context, in which case it is possible to maintain the idea that the

aspectual property of the incremental theme/verbal scale” has something to do with

that of the event.

We should distinguish the verbs mentioned above from typical achievement verbs.

In the former, the accomplishment construal is more common than the achievement

construal while in the latter the achievement construal is more common than the

accomplishment construal. Let us take typical achievement verbs like break in break

the glass. Even though we think of a situation where a very precise camera catches a

snapshot-like moments of the breaking the glass, we still cannot say the glass half

broke. They are real achievements and their theme cannot be construed as a

non-derived incremental theme. Most achievement verbs can only have a runup

achievement reading. At least, we can distinguish the following types of verbs; (i)

verbs that have an accomplishment reading as the common one and an achievement

reading as a more special construal, (ii) verbs that have an achievement reading as the

109
 Chapter 4 Analysis and method

common one and a runup achievement as a less common construal, (iii) and real

achievement type verbs such as shinu ‘die’ in Japanese, which does not allow the

accomplishment reading nor the runup achievement reading.

Another point is that incrementality is still valid and necessary to distinguish the

two variants of the spray/load alternation, at least. In cognitive linguistics, we assume

that different forms have different meanings (even subtly different ones). Therefore,

if there is more than one syntactic realisation of a single verb with the same

participants, we need to explore how they have differences in meanings as well.

Actually, Jackendoff (1996) himself agrees that the with-variant has a bias towards a

holistic interpretation. If the notion of incrementality can differentiate the two variants,

it is worth symbolising this in a formal representation of construals of situations, even

though we do not regard it as the determinant factor for argument selection as Tenny

argues. That is, the approach taken here is similar to Dowty’s, which claims a partial

correlation between the argument structure and aspectual properties. In short, what the

two theoretical points of view above suggest is that we seek to generalise and classify

verbs as much as possible according to their possible (common and uncommon)

construals (including aspectual and causal ones) and the argument linking patterns,

and to differentiate different argument linking patterns in a proper representational

model while admitting the influence of pragmatic factors on the construal of situations.

The next section explains the formalism of Croft’s new representational models.

110
 Chapter 4 Analysis and method

4.2. The Integration of aspect and causal structure—Croft’s (2000) new version

4.2.1. Notation

Croft (1999a, 2000) has developed a new model of event representation by

integrating the causal structure described in Chapter 2 and the aspectual structure

described in Chapter 3 as two important dimensions of verbal semantics which are

independent but related. His main hypothesis is that argument linking is derivable

from force-dynamic (causal) structure and not from aspectual structure (Croft 2000:

55). However, aspectual structure and causal structure are related to a certain extent,

as we saw in the previous section.

First of all, I shall introduce the basic notation of his representation with some

examples. (i) The new representation involves an aspectual contour for each

participant in an event. All participants in the event, which have their own aspectual

contour, are conventionally positioned from bottom upward according to the direction

of the transmission of force.2 Those which are positioned more forward in the causal

analysis come lower in the integrated representation. The following are examples for

the prototypical transitive events of Jack broke the vase (punctual event) and Jack ate

the lasagne (extended causation with holistic theme)3:

2
Croft (2000, §3.3) argues for a three-dimensional representation for causal and aspectual structure,
where each participant belongs to its own t/∆ scale, which is represented as a two-dimensional, and
with the causal relations being specified by the third dimension. However, for the sake of convenience
in order to represent it on paper, it is presented as a 2D model, where the participants are piled up
vertically along the ∆ dimension, as most of the discussion in this thesis can be well illustrated with the
2D model.
3
Figure 1 is cited from Croft (2000: 57, Figure 3) and Figure 2 is from Croft (2000: 59, Figure 4).

111
 Chapter 4 Analysis and method

Figure 1. causal-aspectual representation for Jack broke the vase

(5) Jack broke the vase.

be broken

vase

Jack impact
Δ

Figure 2. causal-aspectual representation for Jack ate the lasagne

(6) Jack ate the lasagne.

be eaten

lasagne

Jack eat

As we can see in the above Figures, (ii) the names of participants are presented to

the left of the contour and the kinds of action they undergo or the state in which

they are is named to the right of the contour. (iii) the verbal profile is represented

by solid black lines while unprofiled parts of situations are represented as broken

112
 Chapter 4 Analysis and method

lines. (iv) Vertical arrows represent force-dynamic relationships. Breaking a vase

is the punctual causation where the application of force is spontaneous and holds at

only a point in time4. A single arrow symbolises punctual causation as in Figure 1.

Eating the lasagne is the extended causation which involves continuous application of

force over time (Croft 2000: 65). Two arrows symbolise and bind extended causation

as in Figure 2.

Figure 1 illustrates the fact that Jack’s act on the vase is punctual in Jack’s

aspectual contour (the bottom contour)5. The upper aspectual contour for the vase

specifies the immediate change of state of vase from not being broken to being broken.

Figure 2 represents the fact that Jack is in the activity of eating, which causes the

lasagne to undergo the gradual change of being consumed. The two solid arrows

specify that Jack’s eating continued during a whole process from the beginning to the

end of the lasagne being consumed. The incrementality/verbal scale is illustrated by a

slanted line (directed activity) and vertical lines represent boundary (§ 3) and the

representation captures the accomplishment construal of the event. Jack is construed

as undergoing an undirected activity of eating. Its inception is profiled as it triggers

the accomplishment subevent of the lasagne. The termination is not profiled as it does

4
In the causal-aspectual analysis, we distinguish punctual causation and extended causation as above.
Talmy constrasts extended causation with onset causation. Croft (2000: 65) argues that onset causation
should be analysed as punctual causation by showing the example of ballistic motion such as kick the
ball across the field. In Talmy’s (1988) sense, it is the onset causation where the initiator transmits
force only at the beginning, and the acted on entity will move or change state before the end of the
event. However, all the three interpretations of the progressive form (Jack was kicking the ball across
the field) indicate that the event is an achievement (See Croft 2000:98-102, §4.5). Croft argues that
‘there are only two types of causation from an aspectual point of view, punctual causation and extended
causation’. Therefore, we use the term punctual causation to cover onset causation.
5
Because of the difficulty in representing it using the wordprocessor, the reversed transition cannot be
represented as a pure vertical line. Instead, the line is slightly inclined. I request readers to regard it as a
vertical line (hence punctual).

113
 Chapter 4 Analysis and method

not cause the completion of the accomplishment (for details, see Croft 2000: 60). The

termination and the rest state are the implied frame as the activity has to terminate

when the lasagne is eaten completely. We treat this contour as being reconstrued and

extended from a single change of state, like the one for vase in Figure 1. As a result,

the lasagne undergoes change of state twice (inception and completion) since it has

two vertical lines in its aspectual contour.

Now we see how the non-causal relation is represented in the new analysis. The

following is a representation of a typical state of location with be (Croft 2000: 76,

Figure 1):

Figure 3. causal-aspectual representation for The ball is in the house.

(7) The ball is in the house. (basic noncausal spatial predication)

house
exist

be located
ball

(v) Solid grey segments represent the complement (preposition) profile6. In the

example, the grey segment represents parts of the event that the prepositional phrase,

in the house, profiles while the solid black line segment is the state of the ball and is

profiled by the verb. (vi) Conventionally, a vertical line represents the

non-force-dynamic relationships between the two participants. Two vertical lines

6
However, Croft (2000) uses dotted lines for the prepositional profile.

114
 Chapter 4 Analysis and method

symbolise the extended non-force-dynamic relationships between the two. The

non-force-dynamic relation between house and ball in the example is symbolised by

the two grey vertical lines to indicate that the state is maintained for a certain period of

time. In the example, the ground is construed as a mere existing reference point

(inherent state) with which the figure has a spatial relation. In turn, the contour of the

figure is represented as a transitory state.

Another important facet of the causal-aspectual representation is the identification

of a subevent that each participant is associated with in the situation. Croft (2000: 58)

sets general rules for labelling subevents:

Extended processes are generally named by the process

Directed achievements are generally named by the resulting state

Moreover, he mentions that certain subevent types will recur. However, he does not

admit the existence of a small finite number of subevent types, but argues that

subevents are defined by the real-world situation, which is conceptually quite rich. In

some cases, a subevent label is described as identical to that of the whole situation

description. For example, Jack’s subevent in Jack ate the lasagne is represented as

“eat”, which is the same as the overall situation. Croft (2000: 58) assumes that the

subevent description is derived from the overall situation type in these cases.

Finally, concerning the causal representation, there is the following constraint on

the causal structure of situations that are expressed by verbs (Croft 2000: 58).

Nonbranching Causal Chain: the semantic structure of verbs construes the

encoded situation as a nonbranching causal chain

4.2.2. Verbal scale and related principles

Some of the more complicated examples are verbs of motion and spray/load

alternating verbs. The latter is important because the Japanese verbs discussed in this

thesis will include the equivalent of the spray/load alternation. The notion of verbal

115
 Chapter 4 Analysis and method

scale/holistic theme is important with these verbs. Therefore, before looking at the

more complicated examples in the causal-aspectual analysis, I shall move on to

discuss verbal scale and the principles of the analysis in Croft (2000).

4.2.2.1. Verbal scale and holistic theme

In Chapter 3, I introduced the notion of verbal scale, which is a measurable scale of

some property included as the lexical meaning of the verb. The holistic theme is the

argument of the verb which possesses that relevant measurable property. For example,

the measurable property is ripeness in The fruits ripened and the holistic theme is the

fruits. In the example of John ate the lasagne, the incremental eating of the lasagne is

a verbal scale and the holistic theme is lasagne. In The cart rolled to NY, the path the

cart travels is the verbal scale and the cart is the holistic theme. A verbal scale can also

be trivial as in Jack ate a handful of peanuts in one gulp. In this example, verbal scale

and therefore holistic theme are reconstrued as one without any internal/incremental

aspectual changes. The above examples are cases where situations are bounded, but

the verbal scale and holistic theme can be unbounded as in Jack ate pizzas, where

there is no endpoint in the relevant measurable scale because of the bare plural form

of the holistic theme.

4.2.2.2. Governing principles

Concerning the relations between the verbal scale, the situation being construed, and

the verbal profile, Croft (2000: 60) proposes three principles as follows.

Verbal Scale Construal: if there is a verbal scale in one of the subevents of a

complex situation, then the overall aspectual type of the situation is determined by
the verbal scale.

Verbal Scale Uniqueness: there is only one verbal scale/holistic theme per situation
encoded by a clause.

116
 Chapter 4 Analysis and method

Verbal Scale Profiling: the verbal scale/holistic theme contour is always part of the
verbal profile.

As a piece of evidence for the verbal scale construal principle, Croft (2000: 60) argues

that in Jack ate the lasagne (as in Figure 2), the contour of the holistic theme (the

lasagne) has more of the time interval of the situation than the agent’s contour. The

next two principles, Croft says, appear to hold for English in his analysis. I shall give

further examples of his English examples based on these principles.

4.2.3. More examples with path

This section examines how the path which expresses a spatial relation between the

figure and the ground is represented in the causal-aspectual representations. The

previous example of Figure 3 which illustrates the basic noncausal spatial relation

includes the path (the non-causal relation between the figure the ball and the ground

the house) which is linked by undirected vertical lines. Now we shall observe other

examples.

4.2.3.1. Verbs of motion

As is well known, English has two types of verbs of motion, those of manner of

motion and those of directed motion (cf. Croft 2000, § 4.1). The former are

exemplified by The bottle floated in the water and the latter are exemplified by She

entered the room. The following is an example of the intransitive verb of motion float

(Croft 2000: 77, Figure 2).

117
 Chapter 4 Analysis and method

Figure 4. the causal-aspectual representation for manner of motion

(8) The bottle floated in the water. (manner of motion verb)

move with floating

bottle

Without a directed preposition, the verb only specifies that the bottle was in the

activity of floating without any specified direction. The bottle can be interpreted as

floating around in the same spot.) Let us compare it with the following example,

which includes a manner of motion that describes directed motion (Croft 2000: 78,

Figure 3).

Figure 5. the causal-aspectual representation for directed manner of motion

(9) The bottle floated into the cave.

(directed motion verb from manner of motion verb)

cave exist

travel with floating

bottle

Goldberg (1995) calls this construction an intransitive motion construction (inherited

by the subpart inheritance from the caused-motion construction), Rappaport Hovav et

118
 Chapter 4 Analysis and method

al. (1999) include it in the Bare XP pattern of the resultative construction, and

Jackendoff (1990) subsumes it under the resultative construction.

The prepositional phrase added to the manner of motion verb profiles the grey

segment above; that is, it profiles an extensive non-force-dynamic relation between

the ground and the figure. What is unique to Croft’s approach is that the PP expresses

the relationship between the figure and the ground. The bottle is the holistic theme

which travels the path, which is the verbal scale, and its motion determines how far

the event in question is achieved.

Croft (2000: 80) argues for the appropriateness of the analysis of Figure 4 and

Figure 5 stating that the motion of the figure causes the figure to travel; the manner of

motion (floating in the example above caused the directed motion) is included in the

subevent of the motion.

The analysis has another descriptive advantage. Verbs of sound emission such as

rumble can also be used in the caused-motion construction (The car rumbled down the

hill/into the driveway.) It is not the rumbling sound that causes the car to move down

the hill but the motion of the car that causes the rumbling sound. Rappaport Hovav et

al. (1999: 39) argue that this should get the non-causative analysis and is best

paraphrased as The car went down the hill, rumbling. This sentence can be analysed in

Croft (2000: 150) in the same way as the above floating example (The bottle floated

into the cave). The path expression that denotes the non-causal relation between the

figure and the ground is represented by the grey lines and the profile of the holistic

theme. The new representation does not specify that the subevent of manner of motion

causes the motion or vice versa. As we can see, the descriptive label for the subject in

the example, the aspectual contour has two different descriptions of “move with

floating” in the floating example. The two actions which are conflated as one

119
 Chapter 4 Analysis and method

aspectual contour cause the change of location of the subject at the end. This

representation analyses the caused-motion construction of verbs of sound emission in

the same way as verbs of motion. In The car rumbled down the hill, the event

representation is similar to that of Figure 5 and the car will be analysed as having the

descriptive label “move with rumbling”. Likewise, these two actions are conflated and

together cause the final change of location. Thus, the new analysis represent verbs of

manner of motion and those of sound emission in the caused-motion construction in

the same way, reflecting their similarities in syntactic realisation. Moreover, it predicts

and allows the paraphrases like The bottle moved into the cave, floating and The car

moved down the hill, rumbling.

The transitive “derived” directed motion verbs such as Jack pushed the cart into

the house is represented in a similar way. Push itself does not necessarily entail the

change of position of the direct object. Because of the PP, the spatial relation between

the direct object (the figure) and the oblique referent (the ground) becomes clear and

apparent in the event structure representation. The following is the causal-aspectual

representation of the sentence:

120
 Chapter 4 Analysis and method

Figure 6. the causal-aspectual representation of push with directed motion

(10) Jack pushed the cart into the house. (Croft 2000: 91-92):

house
exist

travel

cart

push

Jack apply pushing

Δ force

Croft’s analysis is consistent between the intransitive version and the transitive

version in that the only difference in the latter is that it adds the agent contour to the

representation of the former. The cart travels the path, which is the verbal scale. The

agent’s continuing to apply force is represented as extended causation bound by the

two arrows.

4.2.3.2. Spray/load verbs

Now we shall examine how the alternation of spray/load verbs can be represented in

the new causal-aspectual analysis. Croft (2000: 82-86) calls the locative variant the

Path construction and the with variant the With construction. The revised version of

the event representation of spray/load verbs follows (Croft 2000: 82-86):

121
 Chapter 4 Analysis and method

Figure 7. Application verbs in the Path construction

(11a) Jack sprayed the paint on the wall.

wall
exist

become located

paint

spray

Jack emit
Δ

Figure 7 is similar to that of the caused-directed motion verb with a PP complement in

Jack pushed the cart into the house, where the verbal scale is the path travelled by the

cart. The paint is represented as the holistic theme as the verbal scale measures “how

much paint has ended up on the wall.” However, the labelling of the holistic theme is

slightly different from that of directed motion verbs. What is measured here is not

how far the paint has travelled as we cannot imagine that the paint is halfway between

the wall and the spray can. Rather, the measurable property is how much paint is

located on the wall, so the subevent that the paint undergoes is described as “become

located” rather than “travel”. To summarise, the situation is construed as locating

paint on the wall in the Path construction.

On the other hand, in the With cosntruction, the situation is construed as the event

of covering the ground, as the following example illustrates.

122
 Chapter 4 Analysis and method

Figure 8. Application verbs in the With construction

(11b) Jack sprayed the wall with the paint.

be covered

wall

spray

paint
cover

Jack
apply
Δ

The verbal scale measures how much of the wall is covered with the paint. The verbal

scale is the surface of the wall that is covered by the paint, but the wall itself is close

enough to be identified with its surface. Therefore, the wall is represented as the

holistic theme here. The two Figures 7-8 above are similar in terms of the order of the

participants, but with a few differences. The major difference in the verbs’ profiles

basically captures the aspectual difference between the two variants. That is, we

maintain the approach of the first version discussed in Chapter 2, which assigns a

different verbal profile to the two constructions.

It should also be noted that the labels of the events each participant undergoes

are quite different between the two variants. This, with the other differences,

results in two distinct conceptualisations of the event: one variant captures it as the

paint moving to the location (as an emission verb plus a locative complement) and the

other captures it as the location being covered (as a covering verb plus an instrument).

123
 Chapter 4 Analysis and method

The holistic interpretation of the with variant is well illustrated by the status of the

location as the holistic theme.

4.2.4. Summary

The new representation solves three problems with Croft’s earlier model. First of all,

the earlier model represents the causal and aspectual structure of situations together

without differentiating the two; it indicated the aspectual relation (processes) such as

the state or change of state as well as the force-dynamic relationship. Because of this,

we needed to represent the same participant repeatedly in parenthesis when it was

involved in more than one segment of the causal chain, which also made it difficult to

specify to what extent the parts of the causal chain should be profiled. Moreover,

events could not be given the finer-grained aspectual analysis. That is, the distinction

could not be made between a transitory state and an inherent state or an achievement

and an accomplishment. Secondly, the causal structure representation of the ealier

model did not capture the causal relation between multiple events, though it did

represent the causal or non-causal relations between participants. The new

representation can capture both the causal structure of an event causing another event

and that of a participant acting on another participant. Thirdly, the earlier

representation did not have any formal notation to represent a verbal scale/holistic

theme while the new one does.

Thus, in Croft’s new model, aspectual status is also presented formally and in a

fine-grained way. In particular, the notion of incremental theme is formalised.

However, his basic approach to argument realisation is maintained. That is, (i) a single

verb appearing in a different syntactic realisation is attributed to verbal polysemy

(though it is not necessary to be systematic). (ii) He uses a model of force-dynamic

124
 Chapter 4 Analysis and method

relationship and the convention of profiling.

4.3. Methodology—data collection and analysis

This thesis analyses two main classes of Japanese three-argument verbs, verbs of

putting and verbs of removing, following the semantic classification of equivalent

English verbs in Levin (1993). The verbs require the agent, the figure and the ground

as arguments7. The procedure of collecting and analysing the data consisted of six

stages as follows; (1) selection, (2) searching for translation, (3) confirming and

screening, (4) looking for more members, (5) investigating syntactic behaviours and

patterns, and (6) subcategorising into smaller semantic classes. These will now be

explained in turn.

4.3.1. Selection

First of all, reference was made to the English verbs presented in each of the classes in

Levin’s English verb classes and alternations (1993) and, in particular, to Section 9

for verbs of putting and Section 10 for verbs of removing.

Levin’s above-mentioned semantic classes also contain non-three-argument verbs

such as pocket verbs (verbs of putting) and debone verbs (verbs of removing). The

following are examples:

(12) Lydia pocketed the change.

(i.e., Lydia put the change in her pocket.) (Levin 1993: 121)

7
It should be recalled that the agent typically is a person who exerts force and carries out an action to
cause a change in another entity. Even though no changes happen, at least, it transmits the energy to
another entity. The figure is a part which is seen to stand out from the other parts (ground), which is a
setting, in cognitive linguistic terms where people perceive or recognise a scene (Langacker 1987:
120-122). In the discussion of verbs of putting and removing, the figure can be regarded as equivalent
to the theme or locatum, and the ground to the locational goal or source.

125
 Chapter 4 Analysis and method

(13) The cook deboned the fish.

(i.e., The cook removed the bones from the fish.) (Levin 1993: 131)

Even though three participants are actually involved in the semantic frame, these

verbs manifest only two arguments syntactically; the agent and the figure in the

former example and the agent and the ground in the latter. They are cases of noun

incorporation in English. The ground and the figure are incorporated to the meaning of

the verbs, pocket and debone, respectively.

In this initial process, every subclass of the verbs of putting and removing was

referred to as we still did not know how these verbs are syntactically realised in

Japanese. It was quite possible that pocket/debone verbs in English might be

three-argument verbs in Japanese.

4.3.2. Searching for translation

The relevant meanings of these English verbs were selected and translated into

Japanese by consulting dictionaries such as DDWin CD-ROM dictionary (for verbs of

putting), Genius English-Japanese dictionary CD-ROM version 1.11 (for verbs of

removing), Shogakukan Progressive English-Japanese dictionary, and Sanseido’s

Daily Concise English-Japanese dictionary8.

In this procedure, I only selected meanings relevant to the semantic descriptions

for each class of verbs in Levin’s book. Most English verbs have multiple senses. For

example, recall (as in recall an ambassador from London), which is classified as a

banish verb of verbs of removing, has another sense of remembering as in I recalled

having read the book. Only the “banish” sense and its translation equivalent were used

in this case.

8
The sources are paper-version unless they are specified as the CD-ROM version.

126
 Chapter 4 Analysis and method

The meanings of English verbs can be realised in various syntactic ways in

Japanese. The simplest case is that one English verb corresponds to one Japanese verb

of equivalent translation such as put which is translated into oku ‘put’ in Japanese. In

more complex cases, what an English verb refers to is realised in Japanese as a phrase

consisting of multiple words. In one case, a phrase can consist of a predicate and an

adverbial; blanket (type 8 of verbs of putting) in English can be translated as

ichimennni ‘the whole surface’ plus oou ‘cover’ in Japanese. That is, blanket is

paraphrased as ‘cover a whole surface’ in Japanese. In another case, a phrase consists

of a predicate and one of its argument followed by a postpositional particle; blanket

(type 9 of verbs of putting) in English is translated as mouhu (‘blanket’) de (INST)

kurumu (‘wrap/tuck’), that is, ‘wrap with a blanket,’ in Japanese9. However, I decided

not to deal with periphrastic predicative expressions like the above, but only picked

out the main verbs and examined if their equivalents in Japanese are verbs of putting

and removing (cf. § 4.3.3).

Another thing worthy of mention is that different types of main verbs also vary in

form. Japanese has simple verbs, compound verbs, and verbal noun (VN) plus suru

(‘do’) verbs, which express identical or at least almost equivalent meanings to those of

simple English verbs. I shall discuss these three types of predicate in this thesis.

(14) Types of verbs

(i) simple verbs

These are verbs that consist of a single word such as oku ‘put.’

(ii) compound verbs

Compound verbs are those into which two different simple verbs are
combined. For example, mochi-ageru ‘have-lift.’

9
Verbs are sentence-final in Japanese, that is, it is an SOV language.

127
 Chapter 4 Analysis and method

(iii) VN plus suru verbs

These consist of a verbal noun plus suru the ‘do’ verb. The verbal noun
is a syntactic category which is peculiar to the Japanese language.
Verbal nouns can be of Japanese origin, Chinese origin or English
origin. For example, kaimono-suru (‘shopping-do’), haijo-suru
(‘excluding-do’) or fairu-suru (‘file-do’)

The next two chapters on verbs of putting and removing are devoted to an

examination of the argument linking patterns of the three types of Japanese predicates

in order to propose detailed semantic representations for them. The reason for this is

that only three types can be regarded as equivalent to simple verbs in English.

4.3.3. Confirming and screening

Japanese translation equivalents of the English verbs were confirmed and filtered by

using Japanese-English dictionaries such as Kenkyusha’s New Collegiate

Japanese-English dictionary, Kenkyusha’s New Japanese-English Dictionary, and

Progressive Japanese-English Medium-sized Dictionary CD-ROM version 1.11, and

also the Great Japanese dictionary, which is an encyclopaedic-type dictionary.

In this process, the main verbs representing Japanese translation equivalents of the

English were checked through the dictionaries because some of the verbs might

possibly have belonged to another semantic class even though their meaning as a

whole VP (with an object or modifiers) seemed to comply with the meaning of a

certain semantic class of English verbs. For example, cage (a pocket verb of verbs of

putting) has a translation of ‘kago (cage) ni (location) kau (keep as a pet),’ which is

‘keep a pet in a cage.’ Kau (keep) itself is definitely not a verb of putting, since it is

not a three-argument verb and it does not relate to putting something into a location.

Verbs like this were eliminated from the data.

128
 Chapter 4 Analysis and method

4.3.4. Looking for more members

In this fourth stage, after reducing the number of Japanese verbs from the data by

leaving out those which were irrelevant to the semantic verb classes in question, I

added some more Japanese verbs that are synonymous to the Japanese translation

equivalents that had already been found. I consulted the dictionaries, Tsukaikata no

wakaru ruigo reikai jiten: a dictionary of synonyms in Japanese (abbreviated as JSD)

and the Great Japanese dictionary (abbreviated as GJD), in order to find more

Japanese verbs which might belong to the verb classes in question. The reason for

doing this was that dictionaries might use certain basic and simpler verbs for

translation and there may be more infrequently used or special verbs for each verb

class which had not been found in the English-Japanese dictionaries I used.

4.3.5. Investigating syntactic behaviours and patterns

The next process was to investigate the syntactic behaviour of the Japanese verbs to

the fullest extent, especially as regards of argument linking. Nihongo Goi Taikei

[Japanese Vocabulary Compendium] and Nihongo Kihon Doushi Youhou Jiten

[Dictionary of Usage of Japanese Basic Verbs] were consulted because they provide

information on argument linking patterns each verb can have as well as semantic

restrictions on its arguments. However, the opinions of the two dictionaries were seen

to vary concerning the optionality of arguments. Moreover, they do not cover all of

the verbs; there was no information on some Japanese verbs I had picked out for

translation at all. In these cases, example sentences from the Japanese-English or

English-Japanese dictionaries and Yourei Kensaku Sohuto 2 [Reference soft for

examples 2] were referred to and I also used my own intuition to determine the

argument linking pattern. Regarding the optionality of arguments, I used the

129
 Chapter 4 Analysis and method

distinction between definite null instantiation and free null instantiation as a guideline.

These notions will be discussed later.

The aspectual properties of the verbs have also been examined through an

interpretation when they are used in the te-iru form. Basically, Nihongo Kihon Doushi

Youhou Jiten [Dictionaries for Usage of Japanese Basic Verbs] was consulted.

However, the number of verbs it treats is not extensive (actually, it only contains basic

Japanese verbs). Again, I have used some examples from the dictionaries and

depended on my own intuition and that of other Japanese speakers10 based on the

discussion in the chapter on aspect in Japanese.

4.3.6. Subcategorising into smaller semantic classes

The Japanese verbs were further subcategorised into more detailed semantic classes

according to the syntactic patterns they bear and subtle semantic differences

(aspectual properties), as Levin did with English verbs. The semantic representation in

the framework of the causal-aspectual analysis was given to each class of the Japanese

verbs.

4.4. Basic information on Japanese grammar

In this subsection, some syntactic characteristics of the Japanese language which are

particularly relevant to the argument linking are discussed.

4.4.1. Particles and antecedent/subsequent obliques

One of the characteristics of the Japanese language concerning argument linking is

that postpositional particles case-mark arguments and adjunct NPs. A description is

10
I thank Ayumi Tsukiashi, Sachie Funahashi, Fuji Kawata, and Shinako Imaizumi for being

130
 Chapter 4 Analysis and method

given below of use of particles in Japanese with particular attention to those relevant

to the causal analysis. Moreover, it is observed which use of the particles functions as

antecedent and subsequent oblique markers. When the uses are not relevant to the

causal analysis because they refer to circumstantial settings, this is indicated.

4.4.1.1. Subject and object markers

The subject is marked by either ga or wa. Basically, ga is a Nominative marker11 and

wa is a Topic marker. The subject which is realised by ga is normally replaceable by

wa depending on contexts. The proper uses of the two particles are distinguished at

the pragmatic rather than syntactic level12. However, this is not the main point in the

present study and therefore, it is not relevant to discuss the functional differences

between the two particles here. They are here treated equally as particles which can

mark the subject.

The accusative case is marked by o. The following is a typical transitive sentence

using the subject (topic) marker, ga (wa), and the object marker, o:

John ga/wa mado o wat-ta.

(15) John NOM/TOP window ACC break-PAST
John broke the window.

One special use of the accusative marker in Japanese is that it can also indicate “a

space in/on/across/through/along which someone or something moves.” (Makino

1986: 349). This use is allowed only with verbs of motion such as aruku ‘walk,’ and

hashiru ‘run’ and is quite contrastive with English, which represents the location as

informants.
11
Ga is also used to mark the argument which is realised as the direct object in English (see Kuno
1973: 81). This particle is used in what is called the double subject phenomenon; both experiencer and
stimulus can be realised as subjects in stative predicates of emotion in Japanese. For details, see Taoka
(1995). This phenomenon is not relevant to the present study.
12
For details, see Kuno (1973: 37-61).

131
 Chapter 4 Analysis and method

oblique.

Watashi wa Oxford Road o arui-ta.

(16) I TOP Oxford Road ACC walk-PAST
I walked along Oxford Road.

4.4.1.2. Oblique case markers

4.4.1.2.1. Ni

The particle ni has an extensive number of uses, whose classification varies according

to different linguists. The main uses are classified and exemplified below following

previous work by Makino (1986), Morikawa (1997) based on Kuno (1973), Shirota

(1993), Kouji-en [Japanese extensive dictionary], Kitagawa et al. (1988) and McClain

(1981)13:

Uses of ni:

1 Location (not relevant in the causal analysis—circumstantial): this indicates the

location where someone or something exists (‘in,’ ‘at,’ ‘on’).

Kanojo wa pari ni sun-de-iru.

(17) she TOP Paris LOC live-TE-IRU
She lives in Paris.

Makino (1986: 303) refers to this use as locational existence to distinguish it from a

use of de, which specifies a place where an action is taking place.

2 Temporal Location (not relevant to the causal analysis—circumstantial): this

indicates a point of time at which something takes place (‘at,’ ‘in,’ ‘on’)

13
In the parenthesis below, I indicate which oblique marker (antecedent or subsequent) is assigned to
each use. This is also applied to the other particles that follow. For the abbreviation under each particle,
see the summary of the use of particles at the end of the section or see the abbreviation page before the
Introduction chapter. Where the uses of ni classified and presented in previous works are not exhaustive,
I have presented some additional uses that I, myself, have observed.

132
 Chapter 4 Analysis and method

Watashi wa roku ji ni oki-ta.

(18) I TOP six o'clock TLOC get.up-PAST
I got up at six o’clock.

3 Goal (subsequent oblique marker): this indicates a place toward which someone

or something moves (‘to,’ ‘toward’) with verbs of motion.

Kanojo wa kinou gakkou ni it-ta.

(19) she TOP yesterday school ALL go-PAST
She went to school yesterday.

I also include in this category the use of direct contact referred to by Makino (1986:

295), according to whom, the particle indicates ‘the surface of something upon which

some action direction takes place’. In this use, the translation equivalent in English is

‘on’ or ‘onto’.

uwagi o hangaa ni kakeru

(20) coat ACC hanger ALL hang
hang a coat on a hanger

The motivation for subcategorising this use under Goal is that the location where the

theme (entity) has direct contact can be interpreted as the Goal that it reaches, that

direct contact here is limited to a spatial one only, and that ni in this use is replaceable

by e, which indicates the direction in which a certain entity moves.

4 Recipient (subsequent oblique marker): this marks a recipient of something or of

an action (‘to,’ ‘for’).

Watashi wa Taro ni hon o age-ta.

(21) I TOP Taro REC book ACC give-PAST
I gave a book to Taro.

As this use is almost equivalent to the English dative case in the double object

construction, scholars like Makino (1986), Morikawa (1997), and McClain (1981)

regard it as an indirect object marker. I simply call it “Recipient” here because verbs

that occur with this use are verbs of related possessional transfer; these verbs typically

133
 Chapter 4 Analysis and method

relate to an transitive action that involves an entity being transferred from one person

to another as ageru ‘give’, miseru ‘show’, nageru ‘throw’, or oshieru ‘teach’.

5 Object 1 (subsequent oblique marker); this indicates an object towards which an

action has been taken.

Kinou kanojo ni at-ta.

(22) yesterday she OJCT meet-PAST
I met her yesterday.

Yatto saigo no shiai ni kat-ta.

(23) finally last GEN game OJCT win-PAST
Finally, we won the last game.

The term, object, can be abstract and is used extensively. What I mean by this word is

that it marks the object which is acted on. Other scholars have used this term to

describe the use of ni; Kouji-en (p. 1941) lists a use of ni as to “specify object”.

Kitagawa et al (1988: 127) refer to ni being used to “specify the object on which an

action has an effect”, and Shirota (1993: 75) to it being used to “specify the scope

where the action or state has an effect”. As the examples above show, these objects

can be realised as direct objects in English in some cases. The construction as in (22)

and (23) is called the transitive nominative-dative form in Morikawa (1997: 16). That

is, verbs require two arguments, one of which is realised as the subject and the other

marked with ni (dative). Another characteristic of the use of ni is that the argument

marked by it cannot appear as the subject in the passive sentence, (so the example

sentences are not passivised).

6 Object 2 (subsequent oblique marker); this indicates an object towards which an

action has been taken.

134
 Chapter 4 Analysis and method

John wa Alex ni koishi-ta.

(24) John TOP Alex OJCT fall.in.love-PAST
John fell in love with Alex.

Kinou kanojo wa watashi ni butsukat-ta.

(25) yesterday she TOP I OJCT bump.into-PAST
She bumped into me yesterday.

Jackie wa John ni sara o araw-ase-ta.

(26) Jackie TOP John OJCT dishes ACC wash-CAUSE-PAST
Jackie made John wash the dishes.

Some scholars do not distinguish between this use and the previous one (Morikawa

(1997), Kitagawa et al. (1988), and McClain (1981)). The objects marked with ni are

equally acted on in both uses. However, the object in this use can be realised as a

subject in the passive. (These sentences allow passivisation.) This is the main reason

that I classify this as an independent category. Morikawa (1997: 16) proposes an

independent use of ni marking the Causee, but I have included this under the current

category (example (26)) since the causee is acted on and also appears in the passive.

7 Result (subsequent oblique marker): this specifies a result or state after a change.

John wa kabe o aka ni nut-ta.

(27) John TOP wall ACC red RST paint-PAST
John painted the wall red.

John no musuko wa isha ni nat-ta.

(28) John GEN son TOP doctor RST become-PAST
John’s son became a doctor.

8 Scope (subsequent oblique marker): this specifies the scope or the object where a

certain state is valid.

Kono kabe wa hi ni tsuyoi.

(29) this wall TOP fire SCP strong
This wall is fire-resistant.

Ano ko ni-wa inu ga kowai.

(30)
that child SCP-TOP dog NOM afraid
That child is afraid of dogs. (Dogs are scary to that child.)

135
 Chapter 4 Analysis and method

Kitagawa et al. (1988: 127) points out that this use of ni can be replaced by ni-totte

(‘for’, ‘from the standpoint of’).

9 Purpose (subsequent oblique marker): this indicates a purpose of an action

(motion). ‘in order to’.

Nomi ni iku.
(31) drinking PURP go
go for drink

In this construction, only verbs of motions are used; ni indicates a purpose of an

entity’s movement to a place.

10 Source/Agent (antecedent oblique marker): this indicates an agent or a source in

the passive and other receiving constructions (‘by,’ ‘from’)

John wa Mary ni but-are-ta.

(32) John TOP Mary PA hit-PASS-PAST (passive)
John was hit by Mary.

Watashi wa John ni hon o morat-ta.

(33) I TOP John PA book ACC receive-PAST (receiving)
14

I received a book from John.

11 Cause (antecedent oblique marker): this indicates a cause of an event.

Kare no kougekitekina taido ni odoroi-ta.

(34) he GEN aggressive attitude CAS get.surprised-PAST
I was surprised with his aggressive attitude.

Ni is used as an antecedent and subsequent oblique markers in the causal analysis

as well as non-causally. Makino (1986: 289-303) gives the synchronic explanation

that the basic meaning of ni refers to contact, from which all the seven uses he

proposes for the particle are derived. On the other hand, Kouji-en (p. 1941) states that

the original meaning of ni is to specify temporally, spatially, or psychologically the

point of existence, activity, or the place which is acted on; it specifies an entity, an

136
 Chapter 4 Analysis and method

event, or a person that the effect of an action or the action itself reaches. A diachronic

approach would be more appropriate to show the explanation of the syncretism of

oblique markers that Croft (1991: 187-190) proposes. In brief, his proposal about such

syncretism states that no surface oblique markers can subsume both antecedent and

subsequent ‘thematic’ roles. Ni does not seem to comply with the hypothesis, as it is

used as both an antecedent and a subsequent marker. Since ni originally marked

non-causal circumstantial arguments, (i.e. which are neutral in position), it is no

wonder that the particle has developed to be used as both an antecedent and

subsequent marker.

4.4.1.2.2. E

E is a particle that indicates ‘the direction toward which some directional movement

or action proceeds,’ according to Makino (1986: 116). It is almost equivalent to ‘to’ or

‘towards’ in English. E is a subsequent oblique marker because it is a spatial Path

postposition.

Watashi wa kyouto e it-ta.

(35) I TOP Kyoto ALL go-PAST
I went to Kyoto.

Watashi wa tomodachi e tegami o dashi-ta.

(36) I TOP friend ALL letter ACC send-PAST
I sent a letter to my friend.

As Makino (1986) points out, e is used interchangeably for ni in the ‘direction’ sense

(equivalent to the Goal use at §4.4.1.2.1 example 3). Many speakers also use it in the

place of ni even in the ‘direct contact’ sense (I have classified this sense as a Goal use

as well), as the following example illustrates:

14
Even for source of possessional transfer, I use the same abbreviation for passive agent, PA.

137
 Chapter 4 Analysis and method

uwagi o hangaa e kakeru

(37) coat ACC hanger ALL hang
hang a coat onto a hanger

4.4.1.2.3. Made

This is a particle that indicates “a spatial, temporal or quantitative limit or an

unexpected animate/inanimate object” (Makino 1986: 225). Its equivalent English

translation is ‘as far as,’ ‘till,’ ‘up to,’ ‘until,’ ‘through,’ or ‘even.’ The substitution of

made in the spatial meaning of ‘up to’ for e or ni (in the meaning of the ‘direction’

(Goal) sense above) causes a very slight semantic change focusing on the course

rather than on direction. Made can be considered as a subsequent oblique marker

and is represented as Allative as ni and e are. (See the summary below).

Watashi wa kyouto ni/e/made it-ta.

(38) I TOP Kyoto ALL go-PAST
I went (up) to Kyoto.

4.4.1.2.4. Kara

When it is used as a postpositional particle, kara indicates a temporal or spatial

starting point or a source. A source can be a person, material, cause or reason. Its

English equivalents are ‘from,’ ‘since,’ or ‘out of.’ The meaning of the temporal

starting point is irrelevant in the causal analysis. When kara indicates a spatial starting

point, it is a subsequent oblique marker as a spatial postposition. When it is a

personal source or material, it is an antecedent oblique marker. As Croft’s (1991:

194) argues that spatial prepositions which are non-causal and subsequent obliques are

metaphorically extended to causal relations, in which case, the ablative form is

mapped to the antecedent oblique in the space-causal metaphorical extension. (The

allative is used as the subsequent oblique in the causal anlaysis. For details, see Croft

(1991: 194)).

138
 Chapter 4 Analysis and method

Spatial starting point (subsequent):

Nimotsu o Kyoto kara Kobe made hakobu.
(39) luggage ACC Kyoto ABL Kobe ALL carry
carry the luggage from Kyoto up to Kobe.

Source (antecedent):
Budou kara wain o tsukuru.
(40) grape SRC wine ACC make
Wine is made from grapes.

Agent (antecedent):
Watashi wa John kara hon o kari-ta.
(41) I TOP John SRC book ACC borrow-PAST
I borrowed a book from John.

4.4.1.2.5. Yori

This indicates a set point in terms of space. Kara can replace it when a starting point

in space is indicated, in which sense, it is a subsequent oblique marker.

Kare wa nimotsu o Kyoto yori hakon-de15-ki-ta.

(42) he TOP luggage ACC Kyoto ABL carry-LINK-come-PAST
He carried the luggage from Kyoto.

4.4.1.2.6. De

According to Makino (1986: 105-111), de has four uses, two of which are antecedent

oblique marker. The use of de are as follows:

1 Instrument or means (antecedent oblique marker): it indicates the use of

something for doing something.

Chichi wa basu de kaisha e iku.

(43) father TOP bus INST office ALL go
My father goes to the office by bus.

This use is relevant to container adverbial de discussed in the chapter on aspect in

Japanese. It indicates required time, which is subcategorised under ‘instrument’ since

the time required for doing something can be an instrument in a broader sense (‘do

15
De is an allophonic variant of the te-participial (as in te-iru), which connects two verbs. I gloss of

139
 Chapter 4 Analysis and method

something using a certain amount of time.’)

2 Place (not relevant to the causal analysis—circumstantial): in this use it indicates

location, other than location of existence (‘at,’ ‘in,’ ‘on’), which is marked by ni.

Chichi wa unsouya de hatarai-te-iru.

(44) father TOP transportation.company LOC work-TE-IRU
My father is working at the transportion company.

3 Cause and Reason (antecedent oblique marker): in this use it indicates a weak

causal relationship, which is translated into English as ‘and,’ ‘because of,’ ‘due to,’ or

‘because.’

John wa byouki de gakkou o yasun-da.

(45) John TOP illness CAS school ACC be.absent.from-PAST
John was absent from school because of illness.

4 Time (not relevant to the causal analysis): in this use it indicates the time that

something terminates16.

Harugakki wa gogatsu de owaru.

(46) spring.term TOP May in end
The spring term ends in May.

4.4.1.2.7. To

To has three uses as a postpositional particle. Its basic meaning is ‘with.’

The uses of to are as follows (Makino 1986: 473-483):

te/de as ‘LINK’ in this use.

16
De in this use is replaceable by ni in the locative (point of time) sense.

140
 Chapter 4 Analysis and method

1 Accompaniment (antecedent oblique marker17): in this use it marks a noun phrase

which maintains a reciprocal relationship with the subject of a clause. The English

translation will be ‘with,’ ‘as,’ or ‘from.’

Mary wa John to odot-ta.

(47) Mary TOP John COM dance-PAST
Mary danced with John.

2 Exhaustive Listing (not relevant to the causal analysis): in this use it lists things

exhaustively (‘and’).

Markus wa eigo to doitsugo to huransugo o hanasu.

(48) Markus TOP English and German and French ACC speak
Markus speaks English, German, and French.

3 Quotation and sound symbolism (antecedent oblique marker): in this use it marks a

quotation, mimetic or onomatopoeic expression which indicates sound or the manner

in which someone/something does something. The equivalent translation is ‘that,’

‘with the sound of,’ or ‘in the manner of.’

Kodomo wa batabata to hashiri-mawat-ta.

(49) child TOP clattering.noise QSS run-round-PAST
The child ran around with a clattering noise.

As it can be used with mimetic/onomatopoeic expressions to describe a certain

manner in which the action is carried out, I would classify it as an antecedent marker.

I am not sure if the quotation use is relevant to the causal analysis or not. It will be

recalled that the sound symbolism use was mentioned as the quotation marker in

Chapter 3.

17
The Comitative is regarded as an antecedent oblique marker (see Croft 1991: 178, 184-187). This can
be represented nicely in the three-dimensional representation. However, it is difficult to represent it in
the simplified two-dimensional representation.

141
 Chapter 4 Analysis and method

4.4.1.3. Summary

There follows a summary of the particles and such of their uses which appear in and

are related to the causal analysis:

Table 1. Summary of particles

particle uses case/ Abbreviation/ position

(mainly from Makino: meaning translation in causal
1986) analysis
ga --- Nominative NOM SBJ
wa --- Topic marker TOP SBJ
o --- Accusative ACC OBJ
ni Location (existential) Locative LOC S.OBL
Temporal Location Locative TLOC N/A
Goal Allative ALL S.OBL
Recipient Recipient REC S.OBL
Object 1 Accusative-lik OJCT S.OBL
e

Object 2 Accusative-lik OJCT S.OBL

e
Result result RST S.OBL
Scope scope SCP S.OBL
Purpose Purposive PURP S.OBL
Source/Agent Source/Passive PA A.OBL
Agent
Cause cause CAS A.OBL
e --- Allative ALL S.OBL
made --- Allative ALL ‘up to’ S.OBL
kara spatial starting point Ablative ABL ‘from’ S.OBL
source Source/Agent SRC/AGT A.OBL
yori --- spatial set ‘from’ S.OBL
point
de Instrument/means Instrument/ INST A.OBL
means
Place Locative LOC N/A
cause and reason Cause CAS A.OBL
time temporal set ‘at’ ‘in’ ‘on’ N/A
point
to accompaniment Comitative COM A.OBL
exhaustive listing listing ‘and’ N/A
quotation and sound quotation and QSS A.OBL?
symbolism sound
symbolism

In the second column (‘uses’), basically, Makino (1986)’s terminology is used when a

142
 Chapter 4 Analysis and method

particle has more than one meaning (except ni). In the third column, I use the popular

name of case or put its meaning when terminology from case-marking is not available.

The fourth column indicates the abbreviation used in the glosses of example sentences.

The final column is for the position in the causal analysis each role marked by the

particles occupies. N/A (not applicable) means that some use of the particles is

irrelevant to the position of the causal analysis. Some particles do not mark

participants in an event but only refer to a circumstantial setting (time, place) in which

an event takes place, like locative ni or de.

A final remark here is that the Japanese language is quite flexible in terms of word

order even though it is predominantly SOV. Thus, NPs that are marked by particles

appear in a relatively free order (Kuno 1973: 3-16, Morikawa 1997: 15).

4.4.2. Optional arguments—DNI vs FNI

In discussing argument linking, we have to consider another characteristic of the

Japanese language, which is the extensive ellipsis of elements in a sentence. Elliptical

elements can be words, particles, and clauses. As Morikawa (1997: 17) notes, ‘any

elements recoverable from context may be deleted without making a sentence

ungrammatical’ and ‘speakers of Japanese are more tolerant toward ellipsis than

speakers of languages such as English.’ Japanese even omits the subject. Japanese

speakers are very likely to omit an oblique argument of three-argument verbs, in

which case they look like two-argument verbs. The problem is how to distinguish

two-argument verbs and three-argument verbs with an omitted argument. Some verbs

are ambiguous in terms of their valency. Even the two dictionaries I have consulted to

check the linking pattern of each verb (Nihongo Goi Taikei [Japanese Vocabulary

Compendium] and Nihongo Kihon Doushi Youhou Jiten [Dictionary of Usage of

143
 Chapter 4 Analysis and method

Japanese Basic Verbs]) have different opinions about the optionality of arguments for

a few verbs.

In an attempt to make clear the distinction between two-argument verbs and

three-argument verbs in Japanese, I employ the distinction of FNI (Free Null

Instantiation) and DNI (Definite Null Instantiation) discussed by Fillmore and Kay

(1993). A Null Instantiation represents an absent complement whose semantic roles

are a part of the interpretation of the sentence. FNI and DNI are two cases of the Null

Instantiation.

According to Fillmore and Kay (1993, §7.2), DNI requires “an appeal to

something in the conversants’ context for its interpretation.” According to Croft

(pers.comm), it needs “a highly accessible referent which is implicit.” For example, I

already sent the invitations is acceptable only when the people who receive the

invitations are inferable from the context or from previous utterances. At least, a

hearer should have information that the speaker is going to have a party. Therefore,

send in English is a three-argument verb even though it superficially can appear as a

two-argument verb as in the example sentence above.

For FNI, on the other hand, there are no contextual constraints on a missing

element. A referent for a null complement is arbitrarily accessible. For example, we

say The prime minister was assassinated yesterday by the man called John Smith with

the specified killer in mind. We could also omit the passive agent by simply saying

that The prime minister was assassinated yesterday. In this case, we do not need any

contexts that would specify who the killer is. A speaker can utter this sentence without

giving a hearer any previous information about this. In that case, a killer is unspecified

and can be anybody, but still the sentence is acceptable. This is a case of FNI.

To determine if a verb is a two-argument or a three-argument verb in Japanese, I

144
 Chapter 4 Analysis and method

examine whether the referent of an omitted argument is specified (DNI) or unspecified

(FNI) in context. If a definite noun is required from context, it is specified and the

verb will be considered as a three-argument verb. If it is not required from context,

which means that an omitted argument can denote anybody, the verb is considered as

a two-argument verb.

4.4.3. On path argument

Since this dissertation also treats cases of caused motion (i.e, the agent causes the

figure to move from one location to another), it is necessary to speak of the path

argument. Some verbs of putting and removing allow both an ablative oblique (spatial

source phrase) and an allative oblique (spatial goal phrase), so I shall further clarify

what is meant by a path argument using the folllowing typical verb of sending:

Chichi wa sono nimotsu o ima kara naya ni hakon-da.

(50) father TOP the luggage ACC living.room ABL shed ALL carry-PAST
My father carried the luggage from the living room to the shed.

Nihongo Goi Taikei [Japanese Vocabulary Compendium] describes the argument

pattern for hakobu ‘carry’ as superficially having four arguments, which are agent,

theme (figure), spatial source, and spatial goal. From the example, it may seem that

there are really four arguments in the event denoted by (50). However, concerning this

case, Jackendoff (1990: 46-47) brings a Path-function into his semantic structures,

with a source and a goal as arguments; he treats the path, which is the whole extent of

the line where the figure (theme) is travelling, as an argument of motion predicates. In

the example, the line from the living room to the shed is regarded as one argument

rather than two components each represented by a different location.

The existence of a path argument is also further motivated in Japanese, where the

path can be marked by the Accusative:

145
 Chapter 4 Analysis and method

Kare wa Kobe kara Osaka made 50km o 10jikan de arui-ta.

(51) he TOP Kobe ABL Osaka ALL 50km ACC 10hour in walk-PAST
He walked a distance of 50 km from Kobe to Osaka in 10 hours.

In the above example, the path argument is realised as having a distance of 50 km, that

from Kobe to Osaka.

In the causal-aspectual analysis, therefore, I treat cases like hakobu ‘carry,’ which

superficially require four arguments, as a verb requiring three arguments, that is, agent,

figure (theme) and path. The path is profiled with the figure, which is the holistic

theme. The ground serves as the reference point in the non-causal relation with the

path. The type of caused motion verbs such as hakobu ‘carry’ should be distinguished

from those of real four-argument verbs. One semantic class of four-argument verbs is

represented by verbs of commercial transaction such as uru ‘sell,’ which requires an

agent, recipient, theme and instrument (or exchanged entity) as follows.

John wa kuruma o Fred ni 1000pondo de ut-ta.

(52) John TOP car ACC Fred REC 1000pound INST sell-PAST
John sold the car to Fred for 1000 pounds.

Real four-argument verbs are not discussed in this dissertation18.

Some of the verbs that are discussed in the dissertation may only have either a

spatial goal or a spatial source besides the agent and the figure. The view is still taken

that there is a path expression. This can be for a path to the spatial goal or for a path

from the spatial source. In either case, “there is a single Path semantically, which can

be expressed by one or more syntactic elements such as NPs or adverbs.” (Croft:

pers.comm.) Jackendoff (1990: 290) also mentions in the footnotes 4 of Chapter 2 that

‘The innovation of a Path-constituent corresponding to PP permits all prepositions of

Path to be handled uniformly, an important descriptive advance.’ That is, the path with

a goal or a source, with a goal only, and with a source only should be treated in the

18
For discussion of verbs of commercial transaction, see Croft et al. (2000, § 5).

146
 Chapter 4 Analysis and method

same way.

4.4.4. Type of verbs

Various type of verbs have already been mentioned briefly in ‘4.3.2 Searching for

translation.’ This section further discusses some verb types in more detail, especially,

compound verbs.

4.4.4.1. Compound verbs

The Japanese language is very rich in compound verbs, which are a combination of

two lexical verbs. In this dissertation, the initial verb of such compounds is referred to

as V1 and the second as V2. There is a variety of compound verbs differing in their

morphological characteristics. I only focus on lexical i-compound verbs19 which are

related to the dissertation, however.

Compound verbs are classified semantically in terms of the semantic relations

between V1 and V2. Matsumoto (1996: 198-219) has classified lexical compound

verbs into seven types based on semantics, as follows:

1) pair compounds
2) cause compounds
3) manner compounds
4) means compounds
5) compounds exhibiting other relations
6) compounds with semantically deverbalized V2
7) compounds with semantically deverbalized V1

In the following two chapters on verbs of putting and removing, I examine the

semantic representations of compound verbs, focusing on how those of component

19
I only focus on what are called i-compounds (as opposed to te-compounds) and lexical compounds
(as opposed to aspectual compounds) in that these two types of compounds are thought to be
“single-word” predicates by the soo suru (‘so do’ = do so ) test (cf. Kageyama 1993: 80, Matsumoto
1996: 36)

147
 Chapter 4 Analysis and method

verbs, V1 and V2, are combined to form a semantic representation as a compound

verb following the classification of Matsumoto above. However, I do not examine

cause compounds because they all have an unaccusative verb as V2 and are therefore

not relevant to a discussion of three-argument verbs. In the category of compounds

exhibiting other relations, Matsumoto lists some exceptional cases that are not

relevant to three-argument verbs, either. Manner compounds, which can be confused

with means compounds, are not discussed in verbs of putting and removing. Therefore,

the other four types of compound are focused on.

4.4.4.1.1. Pair compounds

Matsumoto (1996: 198) states that “two verbs with similar meanings are compounded

to indicate the repetitiveness or intensity of the described process” in pair compounds.

This category is equivalent to what Kageyama (1993: 99) calls parallel relation.

Component verbs in pair compounds not only have similar meanings (i.e. belong to

the same semantic class of verbs) but also identical argument patterns. Two

component verbs with identical argument structures are combined and the resulting

argument pattern for that compound verb is identical to those of the component verbs.

One example is from verbs of putting: maki-chirasu (sprinkle-scatter).

Douro ni jari o maku.

(53a) road ALL gravel ACC sprinkle (V1—locative variant)
sprinkle gravel on the road

Douro ni jari o chirasu.

(53b) road ALL gravel ACC scatter (V2—locative variant)
scatter gravel on the road

Douro ni jari o maki-chirasu.

(53c) road ALL gravel ACC sprinkle-scatter (compound-locative variant)
scatter gravel on the road

148
 Chapter 4 Analysis and method

As the examples above show, the linking pattern of V1 and V2 is L-linking, and the

resulting compounds preserve the same linking pattern. I examine how the semantic

representations of each component verb contribute to that of the compounds in the

relevant chapters later.

4.4.4.1.2. Means compounds

The V1 component specifies the means of an event denoted by V2. The V2 often

denotes a result or a change of state that has been brought about. The number of

means compounds found for the present study is quite large. The following examples

show that means compounds are translated into English either as ‘V2 by V1-ing’ or ‘a

verb plus prepositional phrase’ construction:

(54) Examples of means compounds (Matsumoto 1996: 213-226):

ii-nogareru (say-escape) ‘evade by speaking’

nage-katsu (throw-win) ‘defeat … by throwing’
oshi-akeru (push-open) ‘push open’
mushiri-toru (pluck-take) ‘pluck off’
hari-tsukeru (paste-attach) ‘paste on’
uchi-komu (hit-put.into) ‘hammer in’
arai-otosu (wash-drop) ‘wash (dirt) off’
ii-makasu (say-defeat) ‘defeat by talking’
kui-tsubusu (eat-waste) ‘use up … by eating’
sori-otosu (shave-drop) ‘shave off’
hori-dasu (dig-take.out) ‘dig out’

Means compounds vary in terms of the way the V1 and V2 are integrated into an

argument structure of a compound as a whole. The V1 and V2 are often different

either in the number of arguments they require or in their argument linking patterns

and hence the way of integration also varies. The relation between V1 and V2 is

described as a right-headed relation (Kageyama 1993: 101-103). This simply means

that the verb on the right (V2) is a head and decides the argument linking pattern. This

is illustrated in the following chapters.

149
 Chapter 4 Analysis and method

It is not always easy to differentiate means compounds from manner compounds,

where V1 specifies a manner in which an action denoted by V2 is carried out or an

action which temporally synchronises with that of V2. One way of distinguishing the

two types of compounds is to remember that means compounds can basically be

paraphrased as ‘V1 ni-yotte V2’ (‘V2 by V1-ing’) and manner compounds can be

paraphrased as ‘V1 nagara V2’ (‘V2 while V1-ing’)20.

Means compounds with extended causation can be misinterpreted as manner

compounds since the agent applies force to another participant continuously along its

movement (as in John pushed the cart to NY) and the two subevents synchronise.

However, in a strict sense, the two synchronized events in manner compounds should

be carried out by the same participant such as the cart moved, rumbling. In relation to

the verbs of putting and removing discussed in this dissertation, compounds of this

misleading type are treated as means compounds. I return to this in the next chapter.

4.4.4.1.3.Compounds with semantically deverbalized V2

As the term indicates, V2 verbs have lost their original verbal meanings in this

category and modify V1 verbs with additional adverbial meaning instead. Some

examples of these V2 verbs and compounds are listed in Matsumoto (1996: 218). The

following are compounds in this category.

(55) Examples of left-headed compounds:

V2 tateru kazari-tateru ‘decorate-actively/too much’

nuri-tateru ‘smear-actively/too much’
tsukeru okuri-tsukeru ‘send-hard/harshly’
nage-tsukeru ‘throw-hard/harshly’

20
Ni-yotte is a particle that indicates the means for doing something according to Makino (1995: 297).
Nagara is ‘a conjunction which indicates that the action expressed by the preceding verb takes place
concurrently or simultaneously with the action expressed in the main clause’ (Makino 1986: 269).

150
 Chapter 4 Analysis and method

In my data for verbs of putting, tateru in kazari-tateru ‘decorate-actively’ and

nuri-tateru ‘smear-actively’ is a deverbalised V2. Tateru as a full verb meaning ‘stand

(transitive)’, but has an adverbial meaning ‘actively’ in its deverbalised sense.

Okuri-tsukeru is a verb of sending and nage-tsukeru is a verb of throwing. Tsukeru

means ‘attach’ or ‘put on’ as a full verb. However, it is deverbalised in the above

examples and it means ‘hard’ or ‘harshly.’ Since the two V2s above are deverbalised,

they do not contribute to the semantic structure of the compound verbs. That is, the

semantic structure of V1 is carried over to that of the compound verb and the linking

pattern of V1 is preserved. Therefore, it is ‘left-headed.’ Compound verbs of this type

are treated like simple verbs in the discussion of verbs of putting and removing.

4.4.4.1.4. Compounds with semantically deverbalized V1

V1 component verbs of compound verbs of this category have lost their verbal

meaning. Two examples of such V1 are sasu (‘thrust’ as a full verb) and toru (‘take’

as a full verb). According to The Great Japanese Dictionary, they behave like prefixes

to intensify the meaning of the V2 or to simply make verbs sound better. Examples

from the data of verbs of putting in Japanese are tori-chirakaru ‘toru-scatter,’

tori-kakomu ‘toru-surround,’ and tori-maku ‘toru-wind.’ In this case, the semantic

structure and linking pattern of the V2 is preserved in that of the compound verbs so

that they are right-headed. Again, they are treated like simples verbs in the thesis.

4.4.4.1.5. Frozen compounds

The last category of compound verbs is frozen compounds (my terminology). They

are not included in the classification of Matsumoto. Kageyama (1993: 103) calls them

lexicalised compounds. They are idiomatic and fixed expressions, whose component

151
 Chapter 4 Analysis and method

verbs V1 and V2 do not seem to contribute to the meaning of the compound as a

whole. In some cases, even the argument structure of V1 or V2 are not maintained.

One apparent example of this kind of compounds is maki-ageru (‘wind-lift’) from

verbs of removing. On the one hand, it literally means ‘lift something by winding’ or

‘roll something up.’ On the other hand, it also means ‘take away,’ which has nothing

to do with the original meaning of V1 and V2. As the following examples illustrate,

the V1 and the V2 of the frozen compound cannot take the same semantic type as their

argument.

Patrick wa tomodachi kara kane o maki-age-ta.

(56a) Patrick TOP friend ABL money ACC wind-lift-PAST (compound)

Patrick took money away from his friend.

*Kane o maku.
(56b) money ACC wind (V1)
*wind money

*Kane o ageru.
(56c) money ACC lift (V2)
*lift money

They are also treated like simple verbs in this thesis.

4.4.4.2. VN-suru verbs

VN-suru consists of a verbal noun plus suru ‘do’ verb. Verbal nouns are nouns which

can be linked to suru ‘do’ directly, as a result of which, they behave as full verbs.

Verbal nouns can be of Japanese origin, Chinese origin, or English origin. Those of

Japanese origin are nouns whose origin is purely Japanese, such as irezumi-suru

(‘tattoo-do’=to get a tattoo) (Uehara 1998: 134). They can be verbal nominalisations

as well. For nominalisation, verbs appear in the renyoo-kei ‘renyoo-form’ that is one

of the conjugations of Japanese verbs. An example is uotsuri-suru (‘fish.angling-do’ =

do fishing). Verbal nouns can be English loan words, too, such as fairu-suru

152
 Chapter 4 Analysis and method

(‘filing-do’ = file).

Those of Chinese origin dominate in terms of number. Most of them are composed

of two Chinese characters, each of them representing a meaning. The semantic

relationship between the meaning of the two characters varies. It may manifest a

synonymous relation (two synonymous verbs combined together), a complement

relation (one is a complement of the other, which is a verb), or an adverbial relation

(one modifies the other). In the present study, I am not going to look at the relations of

the two Chinese characters of verbal nouns but treat them as single verbs. One

motivation for this is that, as Croft suggests, we do not use the component character

separately. For example, in hai.chi-suru (‘arrange, post’), hai signifies ‘distribute’ and

chi signifies ‘put.’ They make ‘distribute-put-do’, which translates into English as

‘arrange, post.’ However, we do not have such a word as *hai-suru or *chi-suru. I

only treat the verb as ‘arranging/posting-do’ and hence ‘arrange/post.’

4.5. Summary

This chapter has introduced the causal-aspectual representation for verbal semantics

and also explained the methodology of collecting data as well as giving basic

grammatical information about Japanese. The next two chapters will examine verbs of

putting and removing in Japanese.

153
5. VERBS OF PUTTING

5.1. Introduction

This chapter discusses and analyses the first class of Japanese three argument verbs

which are equivalent to verbs of putting in English. Chapter 9 of English Verb Classes

and Alternations by Levin (1993) has been consulted in order to collect data on

Japanese verbs of putting. According to her comments in Chapter 9 of her book, verbs

of putting refer to situations of putting an entity in some location, covering surfaces,

and putting things into containers. Levin classified verbs of putting in English into ten

subclasses from 9.1 to 9.10 based on their semantic characteristics and syntactic

behaviour. Her classification is set out in Appendix B.

5.2. Basic linking patterns

Japanese verbs of putting manifest two linking types, the FA-type (figure-Accusative)

and the GA-type (ground-Accusative). L-linking (the locative pattern which is

equivalent to the locative variant in English) belongs to the FA-type and I-Linking

(the instrumental pattern which is equivalent to the with variant in English) belongs to

the GA-type. The terms, figure and ground, are used to refer to the two internal

arguments of predicates of verbs of putting1. Examples illustrating the construction

types and associated linking types follow:

1
They are what Rappaport and Levin (1988) call Locatum and Goal, respectively, in their discussion of
spray/load verbs.

154
 Chapter 5 Verbs of putting

FA-type constructions: (2 linking patterns)

L-linking construction (locative pattern)2: figure—o (ACC), ground—ni (ALL)

Watashi wa heya ni shokubutsu o oi-ta.

(1) I TOP room ALL plant ACC put-PAST
I put some plants in my room.

In some cases, the third argument may be omitted. Then, the example manifests FA-

linking (the figure is in the Accusative).

FA-linking construction: figure—o (ACC), ground—omitted

Watashi wa mizu o mai-ta.
(2) I TOP water ACC sprinkle-PAST
I sprinkled water.

GA-type constructions (2 linking patterns):

I-linking construction (instrumental pattern): figure—de (INST), ground—o (ACC)

Watashi wa hon o kami de oot-ta.

(3) I TOP book ACC paper INST cover-PAST
I covered the book with paper.

GA-Linking construction: ground—o (ACC), figure—omitted

Watashi wa denchi o juu.den-shi-ta.

(4) I TOP battery ACC fill.electricity-do-PAST
I charged the battery.

There is another linking pattern, which does not belong to either the FA-type nor GA-

type. This is the GO-linking construction, of which an example follows:

GO-Linking construction: ground—ni (ALL), figure—omitted

Watashi wa denchi ni juu.den-shi-ta.

(5) I TOP battery ALL fill.electricity-do-PAST
I charged the battery.

L-linking entails FA-linking, and I-linking entails GA-linking.

2
Ni in the allative use is replaceable by e in most of the cases (cf. § 4).

155
 Chapter 5 Verbs of putting

5.3. Analysis

There are 18 types for verbs of putting. These types may be further subcategorised

into small subtypes. Each type is discussed in turn as below in terms of the following

points of analysis:

(a) Members
(b) Semantics
(c) Syntactic patterns
(d) Causal and aspectual patterns
(e) Semantic representations (causal-aspectual analysis)

Firstly, some members of each type are extracted from the data list. This indicates

which verbs belong to a certain event class. Secondly, the semantics of the class, that

is, which kinds of situation it denotes, is briefly introduced. Thirdly, the syntactic

patterns of the event class are discussed and the argument linking patterns that are

allowed and prohibited are indicated3 . Then, the causal pattern and an aspectual

pattern are investigated. In doing this, an examination is made of the force-dynamic

causal chain of events (the order of participants in the chain) and also the aspectual

type of causation, which involves whether or not the causation is extended over time.

The aspectual pattern also deals with the interpretation with the te-iru form4 and

verbal scale. The aspectual characteristics can from time to time be ambiguous since

pragmatic factors are also involved (cf. § 4). However, it is still worthwhile discussing

them because a certain subclass of verbs may lexically (that is, conventionally, not

pragmatically) favour or reject a certain interpretation (activity in progress etc.) of te-

iru or lexically entail a (non-trivial) verbal scale. Finally, based on the causal and

3
I focus on L-linking , I-linking, and whether verbs allow the source phrase instead of the goal phrase.
FA-linking, GA-linking, and GO-linking should be regarded as unacceptable unless otherwise stated.
4
In this chapter and the next chapter, I will use the terms, retrospective, activity in progress and
resultative state reading of te-iru, unless otherwise specified such as undirected activity or directed
activity, and so on. It is because which construal verbs get with te-iru will be made clear in detailed
causal-aspectual analysis.

156
 Chapter 5 Verbs of putting

aspectual patterns, a semantic representation based on the causal-aspectual analysis is

proposed.

Appendix C presents a table of the classes and subclasses with their characteristics.

Appendix D is a list of verbs of putting in Japanese with information about which

class category each verb belongs to.

5.4. Classification

5.4.1. Type 1 Oku ‘put’ verbs

5.4.1.1. Members

The following are some members of this type:

verbs of putting:
oku ‘put’, sueru ‘place/lay’, noseru ‘put’, sue-tsukeru ‘install’, tori-tsukeru
‘install’
verbs of displaying:
chin.retsu-suru ‘display’, hai.chi-suru ‘arrange’

5.4.1.2. Semantics

This category roughly corresponds to type 1 put verbs in English Levin’s

classification. They relate to moving an entity to another location. The entity is the

figure and the location is the ground. The figure and ground are normally still

separable even after the former is located at the ground. The spatial relationship

between the two (i.e, contact) is reversible and not permanent, as in the way a vase put

on the shelf can be movable to somewhere else. Some verbs of this type also refer to

an event of causing a change of location of an entity in an extended sense, an event of

employing a person in an office or opening a branch in a location.

157
 Chapter 5 Verbs of putting

5.4.1.3. Syntactic patterns

This verb type manifests L-linking but not I-linking as illustrated below:

L-linking:
John wa ima ni kabin o oi-ta.
(6) John TOP living.room ALL vase ACC put-PAST
John put the vase in the living room.

*I-linking:
*John wa ima o kabin de oi-ta.
(7) John TOP living.room ACC vase INST put-PAST
*John put the living room with the vase.

They do not allow a source phrase:

*John wa daidokoro kara kabin o oi-ta.

(8) John TOP kitchen ABL vase ACC put-PAST
*John put the vase from the kitchen.

Some verbs in this category allow a variety of locational phrases, which indicates

the neutralness of their meaning of changing location (they do not specify how or to

what position the entity changes its location)5.

Terebi no mae/ushiro ni kabin o oku.

(9) TV GEN front/behind ALL vase ACC put
put the vase in front of/behind the TV

Todana no ue/shita ni hon o oku.

(10) shelf GEN upper.part/lower.part ALL book ACC put
put the book on/under the shelf

5
Unlike English, which is rich in locational/path prepositions such as in, over, under, inside, beside, in
front of, behind, Japanese does not have a lot of postpositions that specify location/path; these ‘rare’
postpositions are kara (‘from’), yori (‘from’), ni (‘to’, ‘in’), made (‘upto’), e (‘to). (Matsumoto 1997:
142, also see Chapter 4). Instead Japanese uses phrases with locational nouns such as naka (‘inside’),
soto (‘outside’), ue (‘upper part’ or ‘surface’), shita (‘lowever part’ or ‘under’), mae (‘front’), and
ushiro (‘back’) and so on. These words are used in combination with ‘ground’, ‘no(GEN)’, ‘locational
noun’, and ‘locational/path postpositions’ as the above examples show.

158
 Chapter 5 Verbs of putting

5.4.1.4. Causal and aspectual patterns

The causal pattern for this type of verb is straightforward. The agent acts on the figure

and changes its location to the ground. Even though the event of ‘putting’ something

on somewhere seems to take a certain amount of time (as taking it, holding it, moving,

and putting) in the real world, it is construed as a punctual causation. The verbs

lexically select and encode only the punctual transition.

This is shown by the ‘more general/typical’ sense te-iru acquires with these verbs.

The normal interpretation of te-iru is retrospective. The verbs basically do not have an

activity in progress meaning with te-iru.

*John wa ima ni kabin o oi-te-iru.

(11) John TOP living.room ALL vase ACC put-TE-IRU
John has put the vase in the living room.
(*activity in progress sense)

The activity in progress sense is not allowed except in two special circumstances.

Since we know we need time to prepare (as for holding it) before the figure is really

put on the ground (i.e, contact) in the real world, a situation is still interpreted as

taking time with te-iru. One special circumstance is the ‘derived verbal scale’ case

and the other is the ‘runup achievement’ case.

When the figure is interpreted as a derived verbal scale, the activity in progress

meaning is possible. This means that an event is construed as having multiple actions

of putting, that is, a multiple figure is to be put. For example, when the speaker has in

mind a certain number (even it is unspecified in the language)6 of vases to be put, (12)

is possible.

John wa kabin o hitotsuzutsu ima ni oi-te-iru.

(12) John TOP vase ACC one.by.one living.room ALL put-TE-IRU
John is putting vases one by one in the living room.

6
Once the number is specified, the event becomes an accomplishment.

159
 Chapter 5 Verbs of putting

The other case is when an event is interpreted as runup achievements. It is possible

for (11) to have an activity in progress meaning if the speaker is watching John in the

activity of trying to put the vase in the living room such as lifting it to move it into the

room.

In both cases, the d-transition of the figure’s aspectual contour in punctual

causation is thought to extend on the t-scale (as inception + d-process or u-process +

completion).

The resultative state reading is possible with te-iru as well. The distinction

between a resultative interpretation and a retrospective interpretation is related to

whether evidence of a resultative situation is perceptible or imperceptible (cf. §3). If

the speaker goes into John’s living room and sees the vase, sentence (11) has the

resultative meaning; he sees the evidence because the vase is there in John’s living

room. This case is similar to reflexive sentences where the perceptable resultative

state of the direct object implies the action of the subject. However, the difference

between retrospective and resultative meanings does not make a difference in the

definition of an aspect type of verb for semantic representations; verbs which only

take the retrospective reading with te-iru and those which can take both the

retrospective and resultative reading are achievements. Therefore, these are not

discussed in the rest of the dissertation. Readers are asked to understand that the

retrospective reading with te-iru can get the resultative reading when contexts allow

(which is rarely). The important fact is whether or not verbs get the activity in

progress reading in the conventional sense. Thus, verbs of this type are punctual

(achievements) as they do not get the activity in progress reading with te-iru.

160
 Chapter 5 Verbs of putting

5.4.1.5. Semantic representations

The semantic representation is the same as in that for put verbs from Croft (2000: 83,

Figure 10, §4.2).

Figure 1. achievement construal of L-linking of oku ‘put’ in (6)

living room

exist
ni
be located

vase oku ‘put’

John cause become

Δ located

The figure (vase) is represented as undergoing a change of location. The spatial

relation with the ground (living room) after the transition is represented by the grey

line.

161
 Chapter 5 Verbs of putting

5.4.2. Type 2 Ireru ‘put into’ verbs: spatial caused-transfer verbs with a specified
direction

5.4.2.1. Members

The following are the members of this verb type:

subtype 1:
ireru ‘put into’, dasu ‘take out’
subtype 2:
ageru ‘raise’, sageru ‘lower’, orosu ‘drop’, otosu ‘drop’

5.4.2.2. Semantics

This category is roughly equivalent to English verbs of putting with a specified

direction under 9.4 in Levin’s classification. They relate to moving an entity to a

location whose direction is specified with respect to its original place. That is, they

entail the location/path. Ireru ‘put into’ and dasu ‘take out’ refer to spatial transfer

between inside and the outside of something. Ireru ‘put into’ entails that an entity

ends up being inside of something by being moved from outside of something and

dasu ‘take out’ entails that the entity ends up being outside of something by being

moved from inside of something. Ageru ‘raise’ means moving an entity up and sageru

‘lower’ and orosu ‘drop’ mean moving an entity down. Otosu ‘drop’ also entails

moving an entity down but it is done by stopping holding/keeping it (letting causation

of motion) rather than by actively exerting a force to lower it. Because of the gravity,

the entity will go down to the earth naturally.

5.4.2.3. Syntactic patterns

This class of verbs takes an agent as subject, the figure as accusative, a spatial goal as

allative, and a spatial source as ablative, as the example below shows. They take a

path argument which is specified by spatial goal/source prepositions as the third

argument.

162
 Chapter 5 Verbs of putting

path pattern: (with spatial goal and spatial source)

Sue wa neko o soto kara heya no naka ni ire-ta.

(13) Sue TOP cat ACC outside ABL room GEN inside ALL put.into-PAST
Sue let the cat into the room from outside.

Sue wa nimotsu o ikkai kara nikai ni age-ta.

(14) Sue TOP luggage ACC ground.floor ABL first.floor ALL lift-PAST
Sue lifted the luggage from the ground floor to the first floor.

However, it often happens that one of the spatial prepositions is omitted as DNI in an

extended sense (see the discussion below). The spatial source phrase is likely to be

omitted except with dasu ‘take out7’. Without the source phrase, the verbs manifest L-

linking but not I-linking.

L-linking
Sue wa neko o heya no naka ni ire-ta.
(15) Sue TOP cat ACC room GEN inside ALL put.into-PAST
Sue let the cat into the room.

Concerning the optionality of goal/source phrase, I argue that ireru ‘put into’ and

dasu ‘take out’ show DNI. If an entity ends up in a certain location, it should come

from inside/outside of that location. With the other verbs, we normally need

information from context for omitted sources (DNI) as well. Because the goal

location is always specified in terms of direction with respect to the source and vice

versa, at least a listener knows/supposes that the source is lower/higher than the goal

location if the location is specified. Actually, when we compare ireru ('put into')-type

7
With dasu ‘take out’, quite often the goal phrase is omitted instead of the source phrase. That is the
reason why this verb and compound verbs with dasu as V2 components appear a lot in verbs of
removing in the next chapter. Basically, all the verbs of this category can appear as verbs of removing
when the goal phrase is the one that is omitted. However, the frequency of appearance of these verbs as
verb of removing is much lower than that of dasu ‘take out’. I would say that the reason is that we tend
to focus on the resultative state for communicative purposes; that is, what kind of change is caused is
naturally more important than what the original state/location is. Therefore, when there is a path
argument, the resultative state (i.e. goal) is more likely to be chosen as an argument. This is what
happens for the other verbs of this type. Dasu ‘take out’, lexically probably, implies a greater focus on
the original state (source) than on the result (goal).

163
 Chapter 5 Verbs of putting

and ageru ('raise')-type verbs, it is more likely for the latter to omit the goal phrase (i.e.

they omit the path argument as a whole and end up in FA-linking construction). This

is because the latter type can be construed as directed activity without a delimiter such

as goal phrase. When we raise or lower something (without specifying a source or a

goal), the entity moves on the degree scale of lower and higher. A closer examination

shows that there are two types of aspectual construal with ageru ‘raise’. Example

(16a) shows a construal in which somebody raises his hand to the top of the shelf to

pick something up:

L-linking
John wa tana no ue ni te o age-ta.
(16a) John TOP shelf GEN over ALL hand ACC raise-PAST
John raised his hand up to the shelf.

There is a goal and it is the achievement of ‘raise his hand to the top.’ Another

construal is exemplified in (16b).

FA-linking
John wa hata o age-ta.
(16b) John TOP flag ACC raise-PAST
John raised the flag.

This lacks a goal, but we can still say that is can convey an achievement of ‘putting

the flag higher than before.’ Therefore, it does not matter how much higher the flag is

raised by John. Even ‘one metre’ is enough to say that the person raised the flag.

Alternatively, he can raise the flag indefinitely keeping on putting it up higher

specifying the higher point on the scale of “low and high” 8 . This is a crucial

difference between ireru ‘put into’ type and ageru ‘raise’ type verbs. The latter can

occur without goal phrases or source phrases as in (16b). The former should either

8
The verbs in subtype 2 are called “atelic verbs of directed motion” by Hay et al. (1999). Verbs of this
type show aspectual duality of atelic and telic situation. In the above example, (16a) is typically a telic
situation as the event is delimited by the goal phrase while (16b) is ambiguous as the situation can be
construed as directed activity without the goal phrase.
164
 Chapter 5 Verbs of putting

have goal phrases or, at least, source phrases since without these we do not set any

boundary of ‘in’ and ‘out.’

Another syntactic property of the verbs of this category is that they make

compounds very productively as V2 components (see the discussion of means

compounds later in this chapter). Moreover, they also have intransitive counterparts in

verbs of motion such as deru (go out) for dasu (put out), agaru (rise) for ageru (raise)

and sagaru (fall/drop) for sageru (lower).

5.4.2.4. Causal and aspectual patterns

The causation pattern is that the agent acts on the figure to move it (away from a

source to a specified location). These verbs show indeterminacy between punctual

causation and extended causation; depending on contexts, they can be either punctual

or extended. Taking ireru ‘put into’, for example, what the agent needs to do to move

an entity like air or a cat into the room is to open the door or the window (i.e, punctual

causation, or more specifically, letting causation of motion). In the case of putting the

car into the garage, the driver needs to continue acting on the car (i.e, extended

causation). In an extended causation construal, the motion of the figure can describe a

trajectory of motion, making the imaginary line of moving possible.

However, ireru ‘put in’ and dasu ‘take out’ basically refer to punctual causation

except in special circumstances, as there is a clear line of being inside or outside of

something, and hence there should be a clear line between putting an entity into or out

of a location. Te-iru with these two verbs normally does not have an activity in

progress reading unless an event is construed as an activity of an iterated cycle, or the

figure is thought to be a derived verbal scale, or the event is forcibly construed as a

runup achievement.

165
 Chapter 5 Verbs of putting

*Haha wa neko o heya no naka ni ire-te-iru.

(17) Mother TOP cat ACC room GEN inside ALL put.into-TE-IRU
Mother is putting the cat into the house.
(Mother is letting the cat into the house.)
(*activity in progress sense)

The special case where an event denoted by the two verbs is durational is also

related to pragmatic factors; the nature of the figure and the ground may allow the

accomplishment reading. Even though there is a clear line between being in and out of

something, it is possible that something is on the line. For example, putting a car into

the garage, we can say putting a car halfway into a garage. In this case, the imaginary

trajectory line on the car is crossing the boundary which differentiates outside and

inside the garage. Another case is that we can put one suitcase in the middle of the

front door of the house and say the suitcase is half inside and half outside the house.

The suitcase is crossing the boundary that differentiates the inside and outside of the

house. In these cases, the car and the suitcase are holistic themes which are closely

associated with the verbal scale. Moreover, te-iru also can bear the meaning of

activity in progress.

Chichi ga shako ni kuruma o ire-te-iru.

(18) Father NOM garage ALL car ACC put.into-TE-IRU
Father is putting his car in the garage.

When the figure measures out the event as in (18), the whole event is regarded as

being of an accomplishment type. To allow this interpretation, the figure must be a

concrete object and also there should be a clear distinction between inside and outside

of the ground; the whole event is measured as the figure object that has concrete form

or surface crosses the boundary. The relative positioning of the figure with respect to

the line serves as a verbal scale.

Ageru 'raise' and sageru 'lower', which entail transfer of an entity upwards or

downwards, also have a verbal scale. In this case, the figure travels from one place to

166
 Chapter 5 Verbs of putting

another place but there is not any clear boundary which distinguishes between the

two; the verbal scale is the path either from a source or to a goal (or both). The event

is measured out according to how far the figure has travelled on the path.

However, te-iru tends to have a meaning of directed activity when this path is

specified, as in (19)9.

Kureen ga piano o {ikkai kara} jukkai ni age-te-iru.

(19) crane NOM piano ACC {ground.floor ABL} ninth.floor ALL lift-TE-IRU
The crane is lifting the piano {from the ground floor} to the ninth floor.

Pragmatic factors also interact with aspectual behaviour with ageru ‘raise’ and

sageru ‘lower’. When a goal is specified (even as DNI) without the source phrase, the

sentence is more likely to be interpreted as having resultative meaning as in (20).

Seito ga te o {atama no ue ni} age-te-iru.

(20) student NOM hand ACC {head GEN over ALL} raise-TE-IRU
The student’s hand is up over his head as a result of having raised his hand.

World knowledge indicates that it does not take much of an effort for a person to raise

his hand. In this example, the path from the normal position of the hand to over the

head is very small, and it is construed as an achievement (trivial verbal scale).

9
Curly brackets mean that the phrases inside are optional.

167
 Chapter 5 Verbs of putting

5.4.2.5. Semantic representations

The semantic representations of ireru ‘put into’ and ageru ‘raise’ are as follows.

Figure 2. achievement construal of L-linking of ireru ‘put into’ in (15)

room

exist
ni
be located inside

cat ireru ‘put into’

mother cause become

Δ located

Figure 3. accomplishment construal of L-linking of ireru ‘put into’ in (18)

garage exist
ni
travel

car
ireru ‘put into’

Father cause move

Δ

168
 Chapter 5 Verbs of putting

Figure 4. accomplishment construal of L-linking of ageru ‘raise/lift’ in (19)

ninth floor exist

ni
travel

piano
ageru ‘lift’

crane cause move

Δ

Figure 5. accomplishment construal of path pattern of ageru ‘raise/lift’ in (19)

ninth floor exist

ground floor exist

travel

piano
ageru ‘raise/lift’

crane cause move

Δ

This Figure is based on the semantic structure of a multiple ground expression in

directed motion as in Croft (2000: 79, Figure 4, §4.1). The two “ground” expressions

are treated as reference points for the path.

169
 Chapter 5 Verbs of putting

Figure 6. achievement construal of FA-linking of ageru ‘raise/lift’ in (20)

be up

hand

ageru ‘raise/lift’

student cause move

5.4.3. Type 3 Hitasu ‘soak’ verbs

5.4.3.1. Members

The following are some member of this verb type:

verbs of soaking:
tsukeru ‘soak, steep in’, hitasu ‘soak/dip in’, shizumeru ‘sink into’
verbs of burying:
uzumeru ‘bury’,umeru ‘bury’
verbs of confinement:
kankin-suru ‘confine/imprison’, toji-komeru ‘shut/lock in’, kakumau
‘hide/shelter’
verbs of stocking:
takuwaeru ‘store’, chozou-suru ‘store’

5.4.3.2. Semantics

Verbs of this category refer to putting things into containers or places, or putting

people into places (especially in order to keep them or keep them in order) such that

the figure is completely inside the ground. The category includes verbs of soaking

which refer to putting things (completely) into liquid or at least making an entity

totally wet with liquid and also verbs of burying which refer to putting things

completely under ground.

170
 Chapter 5 Verbs of putting

5.4.3.3. Syntactic patterns

Verbs in this category take L-linking but not I-linking or a path pattern (including a

source phrase), as the following examples illustrate:

L-linking
Claire wa tenugui o mizu ni hitashi-ta.
(21) Claire TOP towel ACC water ALL soak-PAST
Claire soaked a towel into water.

*I-linking
*Claire wa tenugui o mizu de hitashi-ta.
(22) Claire TOP towel ACC water INST soak-PAST
*Claire soaked water with a towel.

*Source-phrase/*path pattern:
*Claire wa tenugui o temoto kara {mizu ni} hitashi-ta.
(23) Claire TOP towel ACC at.hand ABL {water ALL} soak-PAST
*Claire soaked a towel from her hands {in water}.

Another characteristic of these verbs is that they predominantly occur with no

naka ni ‘GEN inside ALL = to the inside of’ as an alternative to the very general and

neutral postposition ni. Because of their semantics, these verbs do not occur with a

variety of postpositional phrases such as Japanese translation equivalents of outside of,

over, under etc.

Claire wa tenugui o mizu no naka ni hitashi-ta.

(24) Claire TOP towel ACC water GEN inside ALL soak-PAST
Claire soaked a towel in water.

5.4.3.4. Causal and aspectual patterns

The agent acts on the figure to move it into the ground. This can be either punctual

causation or extended causation. This pattern is in some ways parallel to that for type

2 above. When the figure is measurable with respect to the boundary of containers or

places, the action can be construed as an accomplishment (extended causation), but

otherwise as an achievement (punctual causation). However, verbs of confinement and

verbs of stocking conventionally get the punctual construal.

171
 Chapter 5 Verbs of putting

(accomplishment construal)
Claire wa tenugui o mizu ni hitashi-te-iru.
(25) Claire TOP towel ACC water ALL soak-TE-IRU
Claire is soaking a towel in water. (activity in progress)

As a accomplishment construal, the above sentence can be modified by sukoshizutsu

‘little by little’:

Claire wa tenugui o mizu ni sukoshizutsu hitashi-ta.

(26) Claire TOP towel ACC water ALL little.by.little soak-PAST
Claire soaked a towel in water little by little.

(achievement construal)
Claire wa te o mizu ni hitashi-te-iru.
(27) claire TOP hand ACC water ALL soak-TE-IRU
Claire soaked her hand in water. (resultative state reading)

Verbs of soaking refer to an event of putting an entity completely into liquid10, which

(27) implies. In the accomplishment reading, the figure is thought to be associated

with a verbal scale crossing the boundary from outside the water to inside the water

since ‘little by little’ in (26) above measures how much of the towel is put into the

water.

10
For tsukeru2 'soak', this entailment is a little weaker, however.

172
 Chapter 5 Verbs of putting

5.4.3.5. Semantic representations

The semantic representation of hitasu ‘soak’ with the achievement construal is as

follows:

Figure 7. achievement construal of L-linking of hitasu ‘soak’ in (27)

water

exist
ni
be located inside

hand hitasu ‘soak’

Claire cause become

Δ located

5.4.4. Type 4 Sosogu ‘pour into’ verbs

5.4.4.1. Members

The following are some member of this verb type:

sosogu ‘pour into’, shimau ‘put away’, shiireru ‘stock’

5.4.4.2. Semantics

Verbs of this type refer to an event of putting or pouring an entity into a container.

Semantically, they are similar to type 3 hitasu ‘soak’ verbs, but syntactically they are

slightly different. Some verbs such as sukuu ‘scoop’ and suu ‘suck’ are cross-listed in

verbs of removing.

173
 Chapter 5 Verbs of putting

5.4.4.3. Syntactic patterns

Sosogu ‘pour into’ type verbs manifest an L-linking pattern but not I-linking. Unlike

type 3, they allow a path pattern where both goal and source phrases are realised.

L-linking
Haha wa chawan ni oyu o sosoi-da.
(28) mother TOP cup ALL hot.water ACC pour-PAST
My mother poured hot water into her cup.

Lindsey wa omocha o hako ni shimat-ta.

(29) Lindsey TOP toy ACC box ALL put.away-PAST
Lindsey put her toy into the box.

path-pattern
Haha wa kyuusu kara chawan ni oyu o sosoi-da.
(30) mother TOP teapot ABL cup ALL hot.water ACC pour-PAST
My mother poured hot water from the teapot into her cup.

Lindsey wa omocha o yuka kara hako ni shima-ta.

(31) Lindsey TOP toy ACC floor ABL box ALL put.away-PAST
Lindsey put her toy away from the floor to the box.

However, the grammaticality is marginal when only a source phrase appears instead

of a goal phrase. Some of them that are cross-listed as verbs of removing allow a

source phrase, though.

??source-phrase (some OK)

?Haha wa kyuusu kara oyu o sosoi-da.
(32) mother TOP teapot ABL hot.water ACC pour-PAST
My mother poured hot water from the teapot.

??Lindsey wa omocha o yuka kara shimat-ta.

(33) Lindsey TOP toy ACC floor ABL put.away-PAST
Lindsey put toy away from the floor.

Lindsey wa tsumetai mizu o oke kara sukut-ta.

(34) Lindsey TOP cold water ACC bucket ABL scoop-PAST
Lindsey scooped cold water from the bucket.

174
 Chapter 5 Verbs of putting

5.4.4.4. Causal and aspectual patterns

The agent acts on the figure and causes it to move to the ground. The verbs vary in

terms of aspectual analysis. For example, sosogu ‘pour into’ gets the activity in

progress reading with te-iru and also allows the tochuumade ‘halfway’ adverbial that

forces the event into an accomplishment construal.

Haha wa chawan ni oyu o sosoi-de-iru.

(35) mother TOP cup ALL hot.water ACC pour-TE-IRU
My mother is pouring hot water into her cup.

Haha wa chawan ni oyu o tochuumade sosoi-da.

(36) mother TOP cup ALL hot.water ACC halfway pour-PAST
My mother poured hot water into her cup up to halfway.

On the other hand, shimau ‘put away’ is conventionally punctual as it gets the

activity in progress reading “only” in special cases.

Lindsey wa omocha o hako ni shimat-te-iru.

(37) Lindsey TOP toy ACC box ALL put.into-TE-IRU
Lindsey is putting her toys into the box.

Example (37) is regarded as having an activity in progress meaning when omocha 'toy'

is treated as plural (derived verbal scale). Since Japanese does not distinguish singular

and plural as English does, by the presence/absence of a plural form, omocha 'toy/toys'

can be thought of as singular or plural. With te-iru (37), it is naturally interpreted as

plural and the whole sentence bears the activity in progress reading. If the figure is

fixed/specified as sono omocha 'that/the toy,' te-iru bears either a retrospective

meaning or forced runup achievement meaning.

Lindsey wa sono omocha o hako ni shimat-te-iru.

(38) Lindsey TOP the toy ACC box ALL put.into-TE-IRU
Lindsey is putting the toy into the box.
(activity in progress by runup achievement)
Lindsey has put the toy into the box.(retrospective)

175
 Chapter 5 Verbs of putting

5.4.4.5. Semantic representations

The following are the semantic structures for the L-linking of sosogu ‘pour’ and

shimau ‘put away’.

Figure 8. accomplishment construal of L-linking of sosogu ‘pour into’ in (28) and (35)

cup exist
ni
be located

hot water
sosogu ‘pour’

mother pour
Δ

Figure 9. achievement construal of L-linking of shimau ‘put away’ in (29)

box

exist
ni
be located inside

toy shimau ‘put away’

Lindsey cause become

Δ located

176
 Chapter 5 Verbs of putting

5.4.5. Type 5 Kabuseru ‘put on’ verbs

5.4.5.1. Members

The following are some of the members of this verb type:

verbs of covering/putting on of L-linking

kakeru ‘put over’, kabuseru ‘put on’
verbs of filling of L-linking
tataeru ‘fill’, haru ‘fill’
verbs of putting on
kiseru ‘dress, cover’

5.4.5.2. Semantics

Verbs of this category refer to events of putting an entity on another or between others

or filling a container with liquid. The category is actually quite close to type 1 oku

‘put’ verbs and type 3 hitasu ‘soak’ verbs. I distinguish this type from type 1 because

it does not allow various kinds of locational phrase as type 1 does. The difference

between type 3 and this type is very subtle; in type 3 the figure goes totally inside the

ground by moving, which this type does not entail. Rather, the figure may cover a

whole/some part of the ground (for verbs of covering and wearing) or the figure may

just located in a container or a place which is not confined/closed.

5.4.5.3. Syntactic patterns

Verbs of this type take L-linking. I-linking and source phrases (including a path

pattern) are ungrammatical with them. The following examples illustrate:

L-linking
John wa Mary no atama ni boushi o kabuse-ta.
(39) John wa Mary GEN head ALL hat ACC put.on-PAST
John put a hat on Mary’s head.

Mary wa ohuro ni mizu o hat-ta.

(40) Mary TOP bathtub ALL water ACC fill-PAST
Mary filled the bathtub with water.

177
 Chapter 5 Verbs of putting

*I-linking

*Mary wa ohuro o mizu de hat-ta.

(41) Mary TOP bathtub ACC water INST fill-PAST
Mary filled the bathtub with water.

*source-phrase
*Mary wa jaguchi kara mizu o hat-ta.
(42) Mary TOP tap ABL water ACC fill-PAST
*Mary filled the bathtub with water from the tap.

5.4.5.4. Causal and aspectual patterns

The agent causes the figure to move on/between/into the ground. The aspectual

patterns are different for individual verbs. For example, putting a hat on one’s head is

conventionally interpreted as achievement and filling the bathtub with water is

interpreted as accomplishment as the te-iru form shows.

John wa Mary no atama ni boushi o kabuse-te-iru.

(43) John wa Mary GEN head ALL hat ACC put.on-TE-IRU
John has put a hat on Mary’s head. (retrospective)

Mary wa ohuro ni mizu o hat-te-iru.

(44) Mary TOP bathtub ALL water ACC fill-TE-IRU
Mary is filling the bathtub with water. (activity in progress reading)

(43) is at best interpreted as involving runup achievement with te-iru. On the other

hand, haru ‘fill’ can be further modified with tochuumade ‘halfway’, which indicates

an accomplishment construal of the event.

Mary wa ohuro ni mizu o tochuumade hat-ta.

(45) Mary TOP bathtub ALL water ACC halfway fill-PAST
Mary half filled the bathtub with water.

178
 Chapter 5 Verbs of putting

5.4.5.5. Semantic representations

The following are the semantic structures for kabuseru ‘put on’ and haru ‘fill’ with L-

linking.

Figure 10. achievement construal of L-linking of kabuseru ‘put on’ in (39)

Mary’s head

exist
ni
be located

hat kabuseru ‘put on’

John cause become

Δ located

Figure 11. accomplishment construal of L-linking of haru ‘fill’ in (40) and (45)

bathtub exist
ni
be located

water
haru ‘fill’

Mary cause move

Δ

179
 Chapter 5 Verbs of putting

5.4.6. Type 6 Maku ‘scatter’ verbs: verbs of applying and scattering

5.4.6.1.Members

The following are some members of this verb type:

subtype 1:
verbs of applying: tsukeru 'apply on'
subtype 2:
verbs of scattering: maku 'scatter', maki-chirasu 'scatter'

5.4.6.2. Semantics

Verbs in this class denote an event of applying or scattering an entity onto a surface of

something. A special characteristic of this class is that the figure is generally

uncountable; it can be liquid, cream, powder, or a granular type of entity like sugar or

gravel.

5.4.6.3. Syntactic patterns

Verbs in this class appear with L-linking but not with I-linking, as the following

examples illustrate:

L-linking

verbs of applying
Anne wa pan ni bata o tsuke-ta.
(46) Anne TOP bread ALL butter ACC spread-PAST
Anne spread butter on toast.

verbs of scattering
Chichi wa michi ni gravel o mai-ta.
(47) father TOP road ALL water ACC sprinkle-PAST
Father sprinkled gravel on the road.

*I-linking

*Anne wa pan o bataa de tsuke-ta.

(48) Anne TOP bread ACC butter INST spread-PAST
?Anne spread the bread with butter.

180
 Chapter 5 Verbs of putting

*Chichi wa michi o jari de mai-ta.

(49) father TOP road ACC gravel INST sprinkle-PAST
*Father sprinkled the road with gravel.

Most of the verbs of scattering may omit the goal phrase and exhibit FA-linking. In

addition, these verbs can also take a source phrase and a path pattern.

FA-linking
Chichi wa jari o mai-ta.
(50) father TOP gravel ACC sprinkle-PAST
Father sprinkled gravel.

source phrase (path pattern)

Chichi wa mado kara {michi ni} jari o mai-ta.
(51) father TOP window ABL {road ALL} gravel ACC sprinkle-PAST
Father sprinkled gravel from the window {onto the road}.

5.4.6.4. Causal and aspectual patterns

The agent acts on the figure to move it to a surface of the ground. For verbs of

applying, the causation type can be extended; for example, one action of applying

butter on bread may be prolonged. For verbs of scattering, the action can only be

punctual; one action of scattering (ejecting) gravel on the road is punctual. Only by

repeating each action of ejecting the figure, will the event be extended. However, the

basic/unmarked interpretation with te-iru is activity in progress in both types.

Anne wa pan ni bataa o tsuke-te-iru.

(52) Anne TOP bread ALL butter ACC spread-TE-IRU
Anne is spreading butter on toast. (activity in progress)

Chichi wa michi ni jari o mai-te-iru.

(53) father TOP road ALL water ACC sprinkle-TE-IRU
Father is sprinkling gravel on the road. (activity in progress)

Because of the nature of the figure, we can apply it to the ground many times; being

an uncountable object, it allows an action to repeat itself. For verbs of applying like

spreading butter over bread, one action of applying it is also extended. Or, the act of

spreading can even be repeated an indefinite number of times.

181
 Chapter 5 Verbs of putting

5.4.6.5. Semantic representations

Two semantic representations are proposed for tsukeru and maku as follows: one is for

the undirected activity type and the other for the accomplishment type. The undirected

activity type is for an event of the figure being applied to a surface repeatedly (cf.

Dowty (1991) and Croft (2000)). Theoretically, we can keep on spreading butter on

bread many times, and also sprinkling gravel on the road. In this sense, they are each

treated as a undirected activity event of ejecting the figure to the ground. The activity

type representation is as follows:

Figure 12. undirected activity construal of L-linking of tsukeru ‘spread’ and maku
‘scatter’ in (46) and (47)

bread/road
exist
ni

butter/gravel move

tsukeru/maku
‘spread/scatter’

Anne/ apply/emit
father
Δ

This Figure is based on the figure of undirected activity construal of a directly affected

patient in Croft (2000: 69, Figure 18, §3.5.2). The aspectual contours of the agent and

the figure allow the possibility of applying the figure repeatedly without any particular

end. This is a case of what Croft (2000) describes as an emission verb plus locative

182
 Chapter 5 Verbs of putting

complement. When the goal phrase is omitted, the representation does not contain the

third (most upper) aspectual contour.

Figure 13. undirected activity construal of FA-linking of maku ‘scatter’ in (50)

gravel move

maku
‘scatter’

emit
father
Δ

An accomplishment construal for spreading butter on toast is possible. In the

normal situation, we stop the act of applying the figure (butter) to the surface of the

ground (bread) when the surface is covered with butter. The acceptability of the

following sentence provides evidence of this:

Anne wa pan ni bataa o tochuumade tsuke-ta.

(54) Anne TOP bread ALL butter ACC halfway spread-PAST
Anne spread butter on half of the bread.

The accomplishment reading is not possible with verbs of scattering. First of all, they

are more like emission verbs and secondly, the purpose of the events denoted by these

verbs is normally not to cover or fill the ground. Tochuumade ‘halfway’ sounds

ungrammatical with this type of verbs.

183
 Chapter 5 Verbs of putting

*Chichi wa douro ni jari o tochuumade mai-ta.

(55) father TOP road ALL gravel ACC halfway scatter-PAST
*Father scattered gravel on the road halfway.

Figure 14. accomplishment construal of L-linking of tsukeru ‘spread’ in (46)

bread exist
ni
be located

butter
tsukeru ‘spread’

Anne apply
Δ

This is the same type of representation as that for the locative alternation of spray

verbs in English. Now, we can generalise the semantic structures of the

accomplishment construal of L-linking. If we compare the accomplishment construals

of type 4 sosogu ‘pour into’ in Figure 8, of type 5 haru ‘fill’ in Figure 11, and Figure

14 above, they look similar except for the participants and labelling. What is

interesting is that in all these cases (example (36), (45), (54)) the ground acts as a

delimiter of the event. The event should finish when the cup is filled (36), when the

bathtub is filled (45), and the bread is covered (54). As a result, the amount of the

figure that is moved or applied is set. Actually, the tochuumade ‘halfway’ phrase,

which implies that the event in question is an accomplishment, modifies the state of

the ground; the cup is half filled, the bathtub is half filled, and the bread is half

covered. One might argue that the ground could be a verbal scale. However, I would

184
 Chapter 5 Verbs of putting

say that the figure is associated with the verbal scale as sukoshizutsu ‘little by little’

modifies the amount of the figure in the following sentences.

Haha wa chawan ni oyu o sukoshizutsu sosoi-da.

(56) mother TOP cup ALL hot.water ACC little.by.little pour-PAST
Mother poured hot water into the cup little by little.

Mary wa ohuro ni mizu o sukoshizutsu hat-ta.

(57) Mary TOP bathtub ALL water ACC little.by.little fill-PAST
Mary filled the bathtub with water little by little.

Anne wa pan ni bataa o sukoshizutsu tsuke-ta.

(58) Anne TOP bread ALL butter ACC little.by.little spread-PAST
Anne spread the butter on the toast little by little.

Sukoshizutsu ‘little by little’ is an indicator of directed activity (cf. §3). The fact

that the amount of the figure entailed is little rather than the amount of the ground that

is filled or covered in the above examples suggests that it is the figure that undergoes

directed activity. Thus, the tochuumade ‘halfway’ test should be treated as an

indicator of an accomplishment construal of the event in question as a whole, and then,

the sukoshizutsu ‘little by little’ test should be applied to see which argument is the

holistic theme.

5.4.6.6. Digression on pair compounds

Before proceeding to the next type, I shalll show the representation of pair

compounds. Chapter 4 indicates that V1 and V2 components in pair compounds have

identical argument patterns as well as synonymous meanings. When they are

combined, the resulting argument pattern for the compound verb is identical to those

of each of the component verbs. We have seen that some pair compounds have

appeared in the subclasses discussed so far, though not many. A few examples of pair

compounds are found verbs of scattering; they include maki-chirasu (sprinkle-scatter)

and hane-kakeru (splash-pour/sprinkle).

185
 Chapter 5 Verbs of putting

V1:
John wa douro ni jari o mai-ta.
(59a) John TOP road ALL gravel ACC scatter-PAST
John scattered gravel on the road.

V2:
John wa douro ni jari o chirashi-ta.
(59b) John TOP road ALL gravel ACC scatter-PAST
John scattered gravel on the road.

compound:
John wa douro ni jari o maki-chirashi-ta.
(59c) John TOP road ALL gravel ACC scatter-scatter-PAST
John scattered gravel on the road.

The linking patterns of maku (V1), and chirasu (V2) are L-linking, and the resulting

compound preserves this linking pattern. The semantic structures of these three verbs

are illustrated below. Here I use the simplest aspectual contour of achievement for the

compounds. However, they can be construed as undirected activities.

186
 Chapter 5 Verbs of putting

Figure 15. semantic structure for the L-linking of maki-chirasu ‘scatter-scatter’ in (59)

V1 V2

road road
exist exist
ni ni

gravel be located gravel be located

maku chirasu
‘scatter’ ‘scatter’

John emit John emit

Δ Δ

t t

Compound

road exist
ni
be located

gravel
maki-chirasu maku chirasu
‘scatter-scatter’

actor
emit
Δ

187
 Chapter 5 Verbs of putting

Langacker (1987: 277-327) discusses the grammatical valence relations which

hold between symbolic structures that are combined to form a more specified

expression. The composite structure of the elaborated expression is yielded by

integrating the component structures. When there are shared elements/substructures

between the two component structures (correspondences), the specifications of these

elements/substructures are merged by superimposing the shared

elements/substructures of one component structure onto the other. Since the two

component verbs in the above example are accepted as synonymous, with their

arguments being referentially identical (even though they are not perfectly

synonymous by reference to Langacker’s idea that different forms have different or

even subtly different meanings) and have the same causal-aspectual representations

concerning argument linking, they share exactly the same structures. Also the

resultant composite structure is the same as the component ones. That is, one

component structure as a whole is simply ‘superimposed’ on the other to form the

integrated structure of the compound expression. However, we do not have proper

grounds for determining which one is superimposed on which; one component

structure is not dependent on the other and both structures have equal status in

contributing to the composite structure.

Matsumoto (1996: 202) points out that the argument structure of pair compounds

is potentially ambiguous concerning headedness as the argument structures of the

component verbs are identical. He takes the position that they are likely to be right-

headed because of a lack of negative evidence and the fact that most compound verbs

are right-headed in Japanese. This is in contrast to Langacker (1987: 288), who takes

the position that none of the component structures determine the semantic structure

when there is no asymmetry between them. I shall come back to this issue later.

188
 Chapter 5 Verbs of putting

Even though the semantic structures of the two components are wholly

overlapping, this does not mean either of them is ‘meaningless’ (Langacker 1987:

297). At least, for pair compounds, this is true. According to Matsumoto (1996: 198),

pair compounds indicate ‘the repetitiveness or intensity of the described process’. The

compound verb, maki-chirasu, above has a meaning of scattering extensively all over

the surface or everywhere, and sometimes has a negative connotation that the

spreading causes a nuisance to others (JSD p. 86 and GJD p. 2045), which the two

component verbs lack. This is in accordance with Langacker’s view that an integrated

concept of the composite structure is experientially distinct from the recognition of

individual components plus instructions for their integration, and it may involve

entities and specifications beyond those provided by the components’ (Langacker

1987: 281). Even maku and chirasu are slightly different; the former indicates

‘sprinkle’ as in pouring water and the latter means to make something disperse and

scatter (GJD p. 2046 and p.1414).

5.4.7. Type 7 Tsurusu ‘hang’ verbs: verbs of putting in a spatial configuration

5.4.7.1. Members

The following are members of this verb type:

subtype 1:
tsurusu ‘hang’, tarasu ‘suspend, hang’, burasageru ‘suspend’, sageru ‘hang’,
tsuri-sageru ‘hang-hang’

subtype 2:
tateru ‘stand’, yokotaeru ‘lay’

5.4.7.2. Semantics

These verbs refer to events of putting an entity (the figure) into location (the ground)

like type 1 verbs of putting. A unique characteristic of verbs in this category is that

189
 Chapter 5 Verbs of putting

they specify the spatial configuration that the figure has relative to the ground.

Subclass 1 denotes attaching the figure in a high position such that the lower part is

free without touching the ground or letting the figure which is attached at the top fall

as a result of gravity. Thus, the figure maintains the vertical position in respect to the

ground. It may have a parallel position if the ground is also vertical to the earth as in

the case of a wall. In subclass 2, the figure keeps an upright position (‘stand’) or a

lying position (‘lay’).

5.4.7.3. Syntactic patterns

Subclass 1 takes L-linking. The allative ni can be replaced by the ablative kara when

the figure literally maintains the vertical position in respect to the ground, which is

high enough to keep the figure from touching the ground. In other cases, replacement

of the allative by the ablative form is prohibited. Subclass 2 takes only L-linking and

neither of the subclasses takes I-linking.

L-linking
(subclass 1)
Otouto wa mobiiru o tenjou ni/kara tsurushi-ta.
(60) brother TOP mobile ACC ceiling ALL/ABL hang-PAST
My brother hung/suspended a mobile from the ceiling.
(The ablative form is also used instead of the allative referring to the same location.)

Haha wa kaaten o mado ni/*kara tsurushi-ta.

(61) mother TOP curtain ACC window ALL hang-PAST
My mother hung the curtain over the window.

(subclass 2)
Gakuseitachi wa koutei ni hata o tate-ta.
(62) students TOP campus ALL flag ACC stand-PAST
Students stood the flag in/on the campus.

*I-linking

*Otouto wa mobiiru de tenjou o tsurushi-ta.

(63) brother TOP mobile INST ceiling ACC hang-PAST
*My brother hung the ceiling with a mobile.

190
 Chapter 5 Verbs of putting

*Gakuseitachi wa koutei o hata de tate-ta.

(64) students TOP campus ACC flag INST stand-PAST
*Students stood the campus with the flag.

Subclass 2 takes FA-linking as FNI, indicating only the figure’s spatial configuration

without the reference point. On the other hand, subclass 1 omits the ground in a

limited context, where we can imagine what sort of ground is used (example (65)).

This is because in order to stand something we do not necessarily need something

special to stand it on; we can stand an entity on the ground. However, to hang or

suspend an entity, we need a certain tool or condition. That is, the figure and ground

are more semantically bound in subclass 1 than in sublcass 2.

FA-linking
(subclass 1)
Haha wa sentakumono o tsurushi-ta.
(65) mother TOP washing ACC hang-PAST
My mother hung out the washing.

(subclass 2)
Otouto wa tamago o tate-ta.
(66) brother TOP egg ACC stand-PAST
My brother stood an egg on end.

5.4.7.4. Causal and aspectual patterns

The agent acts on the figure and locates it at the ground with a certain spatial

configuration. Without this unique characteristic, this type of verb is similar to type 1

verbs, causally and aspectually. They both take the retrospective reading with te-iru.

They may both get the activity reading in a runup achievement construal with te-iru

referring to the agent’s efforts to attach or locate entities to something in a certain

configuration or in an accomplishment construal with a derived verbal scale reading.

That is, in both types the verbs conventionally specify the final state of the figure

rather than its process of changing, which gives an achievement construal.

191
 Chapter 5 Verbs of putting

te-iru:
(subclass 1)
Otouto wa mobiiru o tenjou ni/kara tsurushi-te-iru.
(67) brother TOP mobile ACC ceiling ALL/ABL hang-TE-IRU
My brother has hung/suspended a mobile from the ceiling.
(My brother has attached a mobile to the ceiling.)
My brother is trying to hang a mobile from the ceiling. (for L-linking only).

(subclass 2)
Gakuseitachi wa koutei ni hata o tate-te-iru.
(68) students TOP campus ALL flag ACC stand-TE-IRU
Students has stood the flag in/on campus.
Students are trying to stand the flag in the campus.

What is interesting is that the ablative (kara) version of subclass 1 does not get the

activity in progress reading even in the runup achievement construal. It suggests that

the location is construed as the source from which the figure’s state of maintaining the

vertical position starts rather than the place where the efforts of attaching the figure to

the ground are taking place. On the other hand, L-linking merely refers to an event of

putting (attaching) an entity into a location with the extra meaning of the spatial

configuration. That is, the ground can be construed as the place the figure has a spatial

contact with as well as the place where it is attached and thus can be construed as the

place where the agent is acting on it (as type 1 verbs) 11.

5.4.7.5. Semantic representations

The semantic structure of subclass 1 (L-linking) is represented in the following Figure

16, in which ni and kara are represented as interchangeable. It should be noted that I

describe the spatial position in the labelling of Figure 16 on the right hand side to

11
Subclass 2 cannot have the ablative linking. Though situations denoted by subclass 1 and subclass 2
involve an inactive force at a certain point on the ground going against the natural force of gravity to
maintain the configuration, only subclass 1 allows the ablative linking as well as the more conventional
allative linking (L-linking) for verbs of putting. We may argue that the place of contact is different. For
subclass 1, it is the topmost part of the figure that has contact with the ground. On the other hand, it is
the lowest part of the figure for subclass 2. I shall leave the reason why subclass 1 allows the ablative
linking for future study.

192
 Chapter 5 Verbs of putting

differentiate it from type 1 ‘put’ oku verbs.

Figure 16. achievement construal of L-linking of tsurusu ‘hang’ in (60)

ceiling
exist
ni/kara
be located
with vertical (‘hung’) position

mobile tsurusu ‘hang’

brother cause become

Δ located

5.4.8. Type 8 Tsumu ‘load’ verbs

5.4.8.1. Members

The following are the three members of this verb type:

tsumu ‘load’, noseru ‘load’, tousai-suru ‘mount’

5.4.8.2. Semantics

The first two verbs above refer to an event of loading an entity onto a means of

transportation such as a truck or a ship. Tousai-suru ‘mount’ is used for an extended

sense of loading such as loading/installing a certain performance onto/into the ground

(e.g. loading software into a computer, installing the hi-fi system into the car.)

193
 Chapter 5 Verbs of putting

5.4.8.3. Syntactic patterns

Type 8 verbs take L-linking but not I-linking. They do not allow a source phrase

instead of a goal phrase. A path pattern is only very marginally acceptable.

L-linking
Watashitachi wa John no kuruma ni kagu o tsun-da.
(69) we TOP John GEN car ALL furniture ACC load-PAST
We loaded the furniture onto John’s car.

*I-linking
*Watashitachi wa John no kuruma o kagu de tsun-da.
(70) we TOP John GEN car ACC furniture INST load-PAST
We loaded John’s car with furniture.

*source-phrase
*Watashitachi wa John no ie kara kagu o tsun-da.
(71)
we TOP John GEN house ABL furniture ACC load-PAST
We loaded furniture from John’s house.

??path-pattern
??Watashitachi wa ie kara kuruma ni kagu o tsun-da.
(72) we TOP house ABL car ALL furniture ACC load-PAST
We loaded furniture from the house to the car.

What is peculiar with this type of verb is that the ground (the transportation means)

can appear as the subject in the Te-iru construction, referring to a present situation

where the ground has the figure inside it. In this case te-iru has the resultative reading.

John no kuruma wa kagu o tsun-de-iru.

(73) John GEN car TOP furniture ACC load-TE-IRU
John’s car is loaded with furniture.

I call the above construction as Ground-subject linking (GS-linking) construction

with te-iru.

5.4.8.4. Causal and aspectual patterns

With this type of verb, the agent moves the figure and puts it onto the ground.

Aspectually, the verbs have the activity in progress reading with te-iru, and that

activity can be an accomplishment as the following sentences show.

194
 Chapter 5 Verbs of putting

Karera wa John no kuruma ni kagu o tsun-de-iru.

(74) they TOP John GEN car ALL furniture ACC load-TE-IRU
They are loading funiture onto John’s car. (activity in progress)

Karera wa John no kuruma ni kagu o tochuumade tsun-da.

(75) they TOP John GEN car ALL furniture ACC halfway load-PAST
They half loaded the funiture into John’s car.

Here, furniture can be interpreted either as singular or plural. In both cases, the

activity in progress reading and the adverbial phrase are possible.

However, tou.sai-suru ‘mount’ conventionally favours the achievement reading, as

the event denoted by the verb is normally more ‘abstract’ loading. It is at best

interpreted as runup achievement with the adverbial support of ima ‘now’:

Ani wa ima shinsha ni haifai o tou.sai-shi-te-iru.

(76) brother TOP now new.car ALL Hi-Fi ACC mount-TE-IRU
My brother is now trying to install the Hi-Fi into his new car.

5.4.8.5. Semantic representations

The following are the semantic structures for tsumu with L-linking and with the GS-

pattern with te-iru.

Figure 17. accomplishment construal of L-linking of tsumu ‘load’ in (69) and (74)

John’s car exist

ni
travel

furniture
tsumu ‘load’

we cause move
Δ

195
 Chapter 5 Verbs of putting

Figure 18. semantic structure for GS-linking of tsumu ‘load’ in (73)

furniture

exist

hold as a result of
John’s being loaded
car
Δ

A holding event implies unprototypical force dynamic relations in that the force

being applied does not cause the change of motion or state though it is definitely

applied on the direct object by the subject (‘inactive action’). Being non-prototypical,

there is a cross-linguistic variation concerning the encoding of the event (Croft 2000,

§3.5.1). In Japanese, the holding-held relation is encoded as a result state of a

particular action (in this case ‘loading’); te-iru refers to the resultative state of the

car’s holding furniture inside after it has been loaded with furniture. In the

representation, however, the force dynamic relations between the two participants is

still illustrated by the two solid arrows which represent the extended causal link of the

inactive force.

196
 Chapter 5 Verbs of putting

5.4.9. Type 9 Oou 'cover' verbs

5.4.9.1. Members

Members of this class are exemplified as follows:

verbs of covering:
oou 'cover'
verbs of surrounding:
kakomu 'surround', kakou 'enclose', tori-maku 'surround'
verb of blocking:
husagu 'block'

5.4.9.2. Semantics

The verbs of this category relate to covering a surface of an entity with something else

(verbs of covering), to enclosing something within something else (verbs of

surrounding), and to filling a place with something (verb of blocking).

5.4.9.3. Syntactic patterns

These verbs take I-linking and not L-linking. Some verbs may allow the omission of

the instrumental phrase in certain contexts, in which case they show GA-linking.

I-linking
oou 'cover'
John wa aisha o shiito de oot-ta.
(77) John TOP own.car ACC sheet INST cover-PAST
John covered his own car with a sheet.

kakou 'surround'
Chichi wa ikegaki de ie o kakot-ta.
(78) father TOP hedge INST house ACC surround/enclose-PAST
Father surrounded the house with a hedge.

*L-linking
oou 'cover'
*John wa aisha ni shiito o oot-ta.
(79) John TOP own.car ALL sheet ACC cover-PAST
*John covered a sheet on his own car.

197
 Chapter 5 Verbs of putting

kakou 'surround'
*Chichi wa ikegaki o ie ni kakot-ta.
(80) father TOP hedge ACC house ALL surround/enclose-PAST
*Father enclosed the hedge to a house.

GA-linking (some)
John wa mimi o husai-da.
(81) John TOP ear ACC block-PAST
John covered his ears.

Verbs of this type also appear with Instrumental-Subject linking (IS-linking),

which means that the covering object, surrounding object or blocking object can

appear as the subject, yielding transitive two-argument linking with subject and direct

object. However, Japanese stylistically disfavours using inanimate subjects such as a

covering or surrounding object. Among inanimate entities, a natural phenomenon is

most likely to be in the subject position. The Te-iru form (84) sounds slightly better as

it can refer to a resultative state; because the inanimate subject is not construed as

causing change, it sounds more natural. Stylistically, the passive form as in (85) is

slightly more natural than (82):

IS-linking:
oou 'cover'
Hukai kiri ga machi o oot-ta.
(82) deep fog NOM city ACC cover-PAST
Deep fog covered the city.
(Fog is a covering entity.)

kakou 'surround'
Ikegaki ga sono ie o kakon-de-iru.
(83) hedge NOM the house ACC enclose-TE-IRU
The hedge surrounds the house.
(Hedge is a surrounding entity.)

Te-iru plus IS-linking:

Hukai kiri ga machi o oot-te-iru.
(84) deep fog NOM city ACC cover-TE-IRU
The city is covered with deep fog.

Passive form:
Machi wa hukai kiri ni oow-are-ta.
(85)
city TOP deep fog PA cover-PASS-PAST
The city got covered with deep fog.

198
 Chapter 5 Verbs of putting

5.4.9.4. Causal and aspectual patterns

The agent acts on the figure (covering object, surrounding object, filling object) to

cover, surround, and fill another entity, which is the ground. The aspectual pattern is

either punctual or extended. For example, the verb oou 'cover' has, at least, two

patterns.

John wa ryoute de kao o oot-te-iru.

(86) John TOP both.hands INST face ACC cover-TE-IRU
John covered his face with his hands. (resultative)

John wa aisha o shiito de oot-te-iru.

(87) John TOP own.car ACC sheet INST cover-TE-IRU
John is covering his own car with a sheet. (activity in progress)

Example (86) is punctual because covering one's face with one’s hands does not

take much time in a normal situation; it cannot be construed as extended over time.

Moreover, John is perceptible as the agent who covers his face (we see John's

covering his face and, of course, know that he is the one in charge of the event). This

favours the resultative reading with te-iru. On the other hand, in (87), it is possible to

construe the event as of the accomplishment type; the surface of the car is a verbal

scale that measures out the whole event. The following examples embody the

tochuumade ‘halfway’ and sukoshizutsu ‘little by little’ test:

John wa aisha o shiito de tochuu-made oot-ta.

(88) John TOP own.car ACC sheet INST halfway cover-PAST
John covered his own car with a sheet halfway.

John wa aisha o shiito de sukoshizutsu oot-ta.

(89) John TOP own.car ACC sheet INST little.by.little cover-PAST
John covered his own car with a sheet gradually.

‘Little by little’ modifies a part of the car which is being covered.

199
 Chapter 5 Verbs of putting

5.4.9.5. Semantic representations

I would say that the aspectual pattern for the verbs of this category can be either

punctual or extended. The situations they denote can be construed as accomplishments

with the adverbs 'little by little' and ‘halfway’. The following Figures show

representations of the verb oou to cover with its various types of linking. Figure 19

shows a representation of the achievement version with I-linking, Figure 20 that of the

accomplishment version with I-linking, Figure 21 that of the achievement version with

GA-linking and Figure 22 that of the verb with IS-lilnking.

Figure 19: achievement construal of I-linking of oou 'cover' as in (86) (without te-iru)

covered

face

move

both hands
oou ‘cover’

cover
John
Δ

200
 Chapter 5 Verbs of putting

Figure 20. accomplishment construal of I-linking of oou ‘cover’ in (77) and (87)

be covered

car

sheet
move

John
cover
Δ

Figure 21. achievement construal of GA-linking of oou ‘cover’ in (81)

covered

ears oou ‘cover’

John apply (something) to cover

Δ

201
 Chapter 5 Verbs of putting

Figure 22. achievement construal of IS-linking of oou ‘cover’ in (82)

covered

city oou ‘cover’

fog cover by changing

location
Δ

5.4.10. Type 10 Yogosu ‘dirty’ verbs

5.4.10.1. Members

The following are some members of the verb type:

verbs of contaminating; yogosu ‘dirty’, kegasu ‘dirty, soil’

verbs of dyeing; someru ‘dye’
verbs of covering; ho.sou-suru ‘pave surface’, to.sou-suru ‘paint, coat with paint’
verbs of decorating; sou.shoku-suru ‘decorate’

5.4.10.2. Semantics

These verbs denote an event where the ground can be interpreted to be affected by

having the figure positioned on it by the agent. The ground changes from not being

dirty to dirty in yogosu ‘dirty’, it changes color in someru ‘dye’, or changes from not

being to being covered/decorated. In that way, these can be thought of as change of

state verbs. They are also used in the metaphorical sense of causing a change of state

by disgracing somebody (yogosu ‘dirty’) or being involved with crime (someru ‘dye’).

202
 Chapter 5 Verbs of putting

5.4.10.3. Syntactic patterns

Verbs of this type do not take L-linking but do take I-linking.

*L-linking
*Patrick wa atarashii zubon ni wain o yogoshi-ta.
(90) Patrick TOP new trousers ALL wine ACC stain-PAST
*Patrick stained wine to his new trousers.

I-linking
Patrick wa atarashii zubon o wain de yogoshi-ta.
(91) Patrick TOP new trousers ACC wine INST stain-PAST
Patrick stained his new trousers with wine.

Chichi wa kuruma o penki de to.sou-shi-ta.

(92) father TOP car ACC paint INST covering-do-PAST
My father covered his car with the paint.

Often, the figure (what is put on) is omitted in FNI, in which case they manifest GA-

linking. The following examples illustrate:

GA-linking
Patrick wa atarashii zubon o yogoshi-ta.
(93) Patrick TOP new trousers ACC stain-PAST
Patrick stained his new trousers.

Chichi wa kuruma o to.sou-shi-ta.

(94) father TOP car ACC covering-do-PAST
My father covered his car (with the paint).

Jacqui wa kami o some-ta.

(95) Jacqui TOP hair ACC dye-PAST
Jacqui dyed her hair.

However, unlike type 9 oou ‘cover’ verbs, they do not appear in an IS-linking pattern,

which differentiates the two types even though semantically they refer to similar

events.

5.4.10.4. Causal and aspectual patterns

The agent acts on the figure, which as a result is to be located on the ground.

Aspectually, there are two types. Yogosu ‘dirty’ (verbs of contaminating) favours the

achievement reading and someru ‘dye’ or other verbs of decorating and covering,

203
 Chapter 5 Verbs of putting

favour the accomplishment reading in the most conventional scenario they encode.

People do not make something dirty on purpose, so the action of the verb is seldom

interpreted as an activity or accomplishment, though, of course, we can think of a

situation where somebody is deliberately making something like a wall dirty by

smearing mud on it (in which case, it is possible to have the activity in progress

reading of iterated achievements). The verb, interestingly, does not necessarily entail

that the ground is fully covered with something dirty. For example, only one stain is

enough to ‘yogosu’ (‘dirty’) the ground. It would be better to say that the verb denotes

the change of state from cleanliess to non-cleanliness. Yogosu ‘dirty’ gets the

retrospective reading with te-iru, as the following example shows:

Patrick wa atarashii zubon o wain de yogoshi-te-iru.

(96) Patrick TOP new trousers ACC wine INST stain-TE-IRU
Patrick has stained his new trousers with wine.

Situations denoted by someru ‘dye’ and other ‘painting to cover’ verbs are enacted

by a volitional agent in a non-marked situation, so giving the activity reading12.

Because of the nature of event the verbs encodes, they are conventionally construed as

accomplishment. Dyeing hair or covering a car with paint has a natural endpoint,

which suggests delimitedness.

Jacqui wa kami o akaku some-te-iru.

(97) Jacqui TOP hair ACC red dye-TE-IRU
Jacqui is dying her hair. (activity in progress reading)
Jacqui has dyed her hair. (Her hair is red now). (resultative reading).

Jacqui wa kami o sukoshizutsu some-ta.

(98) Jacqui TOP hair ACC little.by.little dye-PAST
Jacqui dyed her hair little by little.

12
In the example of dyeing one’s hair, it can also have the resultative reading with te-iru as the agent is
the possessor of hair, so, the subject manifests the perceptible result of dyeing. This has two senses, that
of changing color (focusing on the change to the final state) and that of applying colour (focusing on
the activity of changing colour). The former favours the resultative reading and the latter favours the
activity in progress reading with te-iru.

204
 Chapter 5 Verbs of putting

5.4.10.5. Semantic representations

The achievement reading of the I-linking of yogosu ‘dirty’ and the GA-linking of

someru ‘dye’ (accomplishment) are represented as follows:

Figure 23. achievement construal of I-linking of yogosu ‘dirty’ in (91)

not clean

new trousers

be located

wine
yogosu ‘dirty’

cause become
Patrick located
Δ

t
Figure 24. accomplishment construal of GA-linking of someru ‘dye’ in (95)

dyed
(colour changed)

hair
someru ‘dye’

Jacqui apply colour

205
 Chapter 5 Verbs of putting

5.4.11. Type 11 Kazaru ‘decorate’ verbs: verbs of locative alternation

5.4.11.1. Members

The following are the members of this verb type:

kazaru 'decorate', nuru 'smear', tsumeru 'pack', mitasu 'fill'

5.4.11.2. Semantics

Matsumoto (1997: 172) classifies the verbs of this category as those which

incorporate/entail the change of state caused by the event of caused-motion; as a result

of the motion of the figure, its attachment to the goal has resulted. The semantic

properties of the four verbs vary. They relate to decorating, covering surfaces, putting

an entity somewhere and filling a container. However, they share the important

syntactic behaviour of allowing the locative alternation.

5.4.11.3. Syntactic patterns

Verbs of this category exhibit L-linking, I-linking, and mostly GA-linking. They do

not allow source phrases nor a path pattern. The following examples illustrate:

L-linking
John wa kabe ni penki o nut-ta.
(99) John TOP wall ALL paint ACC smear-PAST
John smeared paint on the wall.

I-linking
John wa kabe o penki de nut-ta.
(100) John TOP wall ACC paint INST smear-PAST
John smeared the wall with paint.

GA-linking
John wa kabe o nut-ta.
(101) John TOP wall ACC smear
John covered the wall (by painting).

206
 Chapter 5 Verbs of putting

*source phrase
*John wa kan kara penki o nut-ta.
(102) John TOP tin ABL paint ACC smear-PAST
*John smeared paint from the tin.

This type of verbs is similar to the English spray/load verbs in that they allow the

locative alternation. There are only four simple verbs (as far as I can find) having the

two linking types. The number of verbs that have the locative alternation is much

smaller than the number in English (cf. Levin’s 9.7 class).

5.4.11.4. Causal and aspectual patterns

The agent acts on the figure to move it to the ground, which can be either a location, a

surface, or a container. Causal/aspectual patterns vary according to the verbs and

situations. As the semantics of this category vary and also the locative alternation is

one of the central issues in this study, I shall look at the four verbs in terms of

aspectual pattern and semantic representations in turn, arguing that they bear different

senses according to the different linking types.

5.4.11.4.1. Kazaru 'decorate'

The following examples illustrate the verb with L-linking and its te-iru version, and

that with I-linking and its te-iru version.

L-linking:
Jacqui wa heya ni hana o kazat-ta.
(103) Jacqui TOP room ALL flower ACC decorate-PAST
Jacqui decorated flower(s) in her room. (=Jacqui put flowers in her room.)

Jacqui wa heya ni ip pon no hana o kazat-ta.

(104) Jacqui TOP room ALL one CLSS GEN flower ACC decorate-PAST
Jacqui decorated a flower in her room.

207
 Chapter 5 Verbs of putting

(te-iru)
Jacqui wa heya ni hana o kazat-te-iru.
(105) Jacqui TOP room ALL flower ACC decorate-TE-IRU
Jacqui is displyaing flowers in her room. (activity in progress)
Jacqui has decorated a flower in her room. (retrospective or resultative)
(=Jacqui has put a flower in her room.)

Jacqui wa heya ni ip pon no hana o kazat-te-iru.

(106) Jacqui TOP room ALL one CLSS GEN flower ACC decorate-TE-IRU
Jacqui has decorated a flower in her room. (retrospective or resultative)

I-linking:
Jacqui wa heya o hana de kazat-ta.
(107) Jacqui TOP room ACC flower ISNT decorate-PAST
Jacqui decorated her room with flowers.

(te-iru)
Jacqui wa heya o hana de kazat-te-iru.
(108)
Jacqui TOP room ACC flower INST decorate-TE-IRU
Jacqui is decorating her room with flowers. (activity in progress)

The L-linking of kazaru ‘decorate’ has the same effect as with oku ‘put’ verbs, giving

the retrospective or resultative reading with te-iru13. The event of example (103) is

ambiguous in terms of aspect as the direct object ‘flower’ is quite indeterminate

concerning number in Japanese and does not always distinguish singular and plural as

in the English language. Therefore, its te-iru version (105) can be interpreted as

having an activity in progress reading with a multiple achievements event of putting

flowers one by one bunch at a each time or the retrospective/resultative reading.

However, if the number is disambiguated, specifying ‘one flower’ as in (104), then its

te-iru form (106) only has the retrospective/resultative reading. This is a characteristic

of achievement verbs. Kazaru ‘decorate’ in this sense manifests the ‘display’ sense of

the type 1 verb class. The semantic representation is presented in Figure 25.

The I-linking of the verb also naturally favours the activity in progress reading

with te-iru when the direct object is not clear about the number. A Japanese speaker

would interpret hana 'flower(s)' as plural. What is different between L-linking and I-

208
 Chapter 5 Verbs of putting

linking of kazaru ‘decorate’ is the holistic interpretation of the ground in I-linking as

in the English with variant. With I-linking, the ground is construed as 'wholly'

affected; that is, it undergoes a change of state from not being beautified to being

beautified. In other words, kazaru 'decorate' in the I-linking sense is to make the

ground more beautiful by putting/displaying the figure there (the decorate sense).

Therefore, at least, we intuitively interpret that there are more flowers (even leaving

no space) in the room in the I-linking pattern with example (107) than in the L-linking

pattern of (103). The event of example (108) is interpreted as the activity of Jacqui's

engagement in the activity of applying decoration in the accomplishment type event of

making the room beautiful14.

As evidence that indicates that the ground undergoes a change of state with I-

linking, I present the following examples.

Jacqui wa heya ni hana o kazat-ta ga mada kazari-tari-nai.

(109) Jacqui TOP room ALL flower ACC decorate-PAST but yet decorate-be.sufficient-not.
Jacqui displayed flowers in the room, but not yet sufficiently.

??Jacqui wa heya o hana de kazat-ta ga mada kazari-tari-nai.

(110) Jacqui TOP room ACC flower INST decorate-PAST but yet decorate-be.sufficient-not.
Jacqui decorated the room with flowers, but not yet sufficiently.

Example (110) sounds odd compared to (109) as it implies the room became

decorated after the event denoted by the first phrase.

The delimitability of the event denoted by I-linking is motivated by the

tochuumade ‘halfway’ test and the incrementality of the room is verified by the

sukoshizutsu ‘little by little’ test.

13
For the distinction between retrospective and resultative, see the discussion of oku 'put' verbs.
14
Ippon no hana 'one flower' can be an instrumental argument with I-linking as well. It gets the
punctual construal as L-linking does. However, in this case, a speaker exaggerates by saying that the
flower is so gorgeous that it can effectively make the room look nice. In this punctual construal, the
direct object is trivial incremental theme.

209
 Chapter 5 Verbs of putting

Jacqui wa heya o hana de tochuumade kazat-ta.

(111) Jacqui TOP room ACC flower INST halfway decorate-PAST
Jacqui decorated helf the room with flowers.

Jacqui wa heya o hana de sukoshizutsu kazat-ta.

(112) Jacqui TOP room ACC flower INST little.by.little decorate-PAST
Jacqui decorated the room with flowers little by little.
(‘Little by little’ modifies the state of the room.)

The semantic representations of the accomplishment reading of I-linking and its finer-

grained version are represented in Figure 26 and Figure 27.

Kazaru ‘decorate’ is the only Japanese verb which manifests the locative

alternation but whose English equivalent does not.

5.4.11.4.2. Nuru ‘smear/paint’

Nuru is to apply and put liquid (or cream) on the surface of another entity. Examples

(99) and (100) show the past tense use of the verb.

L-linking (te-iru):
John wa kabe ni penki o nut-te-iru.
(113) John TOP wall ALL paint ACC smear-TE-IRU
John is smearing paint on the wall. (activity in progress)

I-linking (te-iru):
John wa kabe o penki de nut-te-iru.
(114) John TOP wall ACC paint INST smear-TE-IRU
John is smearing the wall with paint. (activity in progress)

John wa kabe o penki de sukoshizutsu nut-ta.

(115) John TOP wall ACC paint INST little.by.little smear-PAST
John smeared the wall little by little with paint.
John painted the wall little by little.

L-linking with te-iru form gives an activity in progress reading, which encodes

multiple applications of paint or one long application of paint on the wall. The

semantic representation is similar to that of verbs of applying (tsukeru) (activity

without any upper bound) (cf. Figure 12), which does not specify any concrete

endpoint of the activity, or that of the accomplishment reading (cf. Figure 14) when

210
 Chapter 5 Verbs of putting

the wall acts as the delimiter15. The L-linking of nuru can be said to have the apply

sense. The undirected activity construal of L-linking is represented in Figure 28.

On the other hand, I-linking of the verb is interpreted as an accomplishment event

type since te-iru gives the activity in progress reading and the adverbial sukoshizutsu

‘little by little’ refers to the status of the wall and not the paint (John might have used

a huge amount of paint for one application to cover a small area of the wall)16. In this

case, the purpose of performing the act is to affect the wall and not only to apply the

paint. The wall is represented as undergoing the transition (change) in Figure 29. I-

linking has the cover (with paint) sense.

Interestingly, I-linking is only applied when an event refers to covering surfaces to

change the state of the ground. It cannot be used to refer to applying cream to the face

or spreading butter on toast, for example, because these events do not change the

nature of the face or the toast. These situations only allow the L-linking of nuru

‘smear’.

*I-linking
*Jacqui wa kao o kuriimu de nut-ta.
(116) Jacqui TOP face ACC cream INST smear-PAST
(= Jacqui applied face with cream.)

15
The wall acts as a delimiter as a normal situation of smearing the paint on the wall is to give it a
coating or colour and not to consume paint. When the wall is covered, the event will be finished.
16
Sukoshizutsu ‘little by little’ can modify the L-linking version of nuru ‘smear’.
John wa kabe ni penki o sukoshizutsu nut-ta.
(i)
John TOP wall ALL paint ACC little.by.little smear-PAST
John smeared the paint little by little on the wall.

Unlike in example (115), it is paint that is consumed little by little, and this results in the figure being a
verbal scale in L-linking as in sosogu ‘pour’, haru ‘fill’, tsukeru ‘spread’ from the previous examples.

211
 Chapter 5 Verbs of putting

5.4.11.4.3. Tsumeru ‘pack/cram’

The following are examples of tsumeru ‘pack’:

L-linking:
Julie wa nimotsu o suutsukeesu ni tsume-ta.
(117) Julie TOP stuff ACC suitcase ALL pack-PAST
Julie packed her stuff into the suitcase.

Kare wa kabe no ana ni shinbunshi o tsume-ta.

(118) he TOP wall GEN hole ALL newspaper ACC cram-PAST
He crammed newspaper into a hole in the wall.

I-linking:
Kare wa kabe no ana o shinbunshi de tsume-ta.
(119) he TOP wall GEN hole ACC newspaper INST cram-PAST
He filled a hole in the wall with newspaper.

The events denoted by L-linking of the verb entail changes of the location of the

figure. It can give an accomplishment construal. Tsumeru entails that entities are

moved to a container in a way that there is no unnecessary space between them17 or

that they are simply put into the container nicely/properly. Unlike with nuru ‘smear’,

the main purpose of packing the figure into the ground in example (117) is to carry the

stuff around and not to fill the suitcase, and in that way, normally the agent already

knows how much stuff he/she is going to put in, so, there is telicity in the event.

Alternatively, at least, there is a natural boundary for the container. The container only

has a limited space and the end of the space marks the end of the action. Unlike in the

case of applying paint on the wall, the action of packing stuff into the suitcase cannot

be repeated an infinite number of times. The te-iru version and the sukoshizutsu ‘little

by little’ test reveal that the figure (stuff) is a verbal scale in the accomplishment

reading. In example (121), ‘little by little’ modifies the amount of stuff.

17
L-linking of the verb does not strictly denote that the container should be filled though it is an
implication. An example to support this is that it is possible to utter a Japanese sentence equivalent to ‘I
packed my stuff into the suitcase, but there is still some space left’ in English.

212
 Chapter 5 Verbs of putting

Julie wa nimotsu o suutsukeesu ni tsume-te-iru.

(120) Julie TOP stuff ACC suitcase ALL pack-TE-IRU
Julie is packing her stuff into the suitcase. (activity in progress)

Julie wa nimotsu o suutsukeesu ni sukoshizutsu tsume-ta.

(121) Julie TOP stuff ACC suitcase ALL little.by.little pack-PAST
Julie packed her stuff into the suitcase little by little.

The previous example (118) is also modified with sukoshizutsu ‘little by little’ as

the following example illustrates:

Kare wa kabe no ana ni shinbunshi o sukoshizutsu tsume-ta.

(122) he TOP wall GEN hole ALL newspaper ACC little.by.little cram-PAST
He crammed newspaper little by little to a hole on the wall.

Again, the figure is still a verbal scale since it is the newspaper which is thought to be

put little by little in (122). As in the above examples (120) and (121), the space of the

ground is limited, which naturally delimits the figure. The representation of the L-

linking of tsumeru ‘pack’ is shown in Figure 30.

Compared to L-linking, the number of contexts where I-linking is allowed is

limited. It is used only when the purpose is to fill something up (which is normally

unfavorable) like a space which is not supposed to be there (example (119) above).

Another example that also allows both linkings is filling a tyre with air. As the

favorable/normal state of the tyre is to be filled with air, the situation is realised by I-

linking with te-iru as well as by L-linking18. Self-evidently, the I-linking in these

examples means that the ground is construed as a holistic theme; its space is being

filled. Sukoshizutsu ‘little by little’ modifies the ground, not the figure.

Kare wa kabe no ana o shinbunshi de sukoshizutsu tsume-ta.

(123) he TOP wall GEN hole ACC newspaper INST little.by.little cram-PAST
He filled a hole on the wall with newspaper little by little.

18
Packing the stuff into the suitcase can be rephrased into I-linking (though not naturally). However,
acceptability is lower in this case and we need a special context where the purpose of the situation is to
fill the suitcase.

213
 Chapter 5 Verbs of putting

The representation of I-linking is shown in Figure 31. We could say that L-linking of

the verb has the putting-into sense and I-linking has the filling sense.

5.4.11.4.4. Mitasu ‘fill’

The following are examples of mitasu ‘fill’ with L-linking and I-linking:

L-linking:
John wa koppu ni mizu o mitashi-ta.
(124) John TOP cup ALL water ACC fill-PAST
John poured water into the cup to the full.

I-linking:
John wa koppu o mizu de mitashi-ta.
(125) John TOP cup ACC water INST fill-PAST
John filled the cup with water.

Semantically, this verb is similar to tsumeru ‘pack’ above in that the ground which is

used with the verb sets the natural boundary for the figure. A subtle difference is that

mitasu ‘fill’ requires the ground to be filled in both linkings. I would still analyse the

verb in the same way as tsumeru, however; the figure is holistic theme in L-linking

and the ground is a holistic theme in I-linking. In both casese, sukoshizutsu ‘little by

little’ modifies the direct object as illustrated below:

John wa koppu ni mizu o sukoshizutsu mitashi-ta.

(126) John TOP cup ALL water ACC little.by.little fill-PAST
John poured water into the cup to the full little by little.
(Water was poured gradually.)

John wa koppu o mizu de sukoshizutsu mitashi-ta.

(127) John TOP cup ACC water INST little.by.little fill-PAST
John filled the cup with water little by little.
(The cup was filled gradually.)

The difference between these two examples is subtle because the two acts of pouring

water and filling the cup are so close; it is difficult to imagine a context where pouring

a large amount of water will gradually fill a cup. Thus, this is a significant example to

show that a difference in linking reflects a different construal of the same event. I

214
 Chapter 5 Verbs of putting

would say that L-linking of the verb gives the pouring-to-the-full sense and I-linking

gives the filling sense. The representations are similar to those for tsumeru ‘pack’.

5.4.11.5. Semantic representations

The following are the semantic representations of kazaru ‘decorate’, nuru ‘smear’ and

tsumeru ‘pack’.

Figure 25. achievement construal of L-linking of kazaru ‘decorate’ in (103)

room

exist
ni
displayed

one flower kazaru ‘decorate’

Jacqui display

215
 Chapter 5 Verbs of putting

Figure 26. accomplishment construal of I-linking of kazaru ‘decorate’ in (107)

be decorated

room

kazaru ‘decorate’

flower
change location

Jacqui
apply
decoration
Δ

Figure 27. fine-grained19 version of figure 26

decorated

room

kazaru ‘decorate’

flower
change location

Jacqui
apply decoration
Δ

19
In this figure, the activity contours of Jacqui and the flowers are disintegrated into multiple
achievement events.

216
 Chapter 5 Verbs of putting

Figure 28. undirected activity construal of L-linking of nuru ‘smear’ in (99)

wall
exist
ni

paint move

nuru ‘smear’

John apply painting

Figure 29. accomplishment construal of I-linking of nuru ‘smear’ in (100)

be covered

wall

nuru ‘smear’

paint
cover

John
apply
Δ

217
 Chapter 5 Verbs of putting

Figure 30. accomplishment construal of L-linking of tsumeru ‘pack’ in (117) and

(121)

suitcase exist
ni
be located

stuff
tsumeru ‘pack’

Julie cause become located

Δ

Figure 31. accomplishment construal of I-linking of tsumeru ‘pack’ in (119) and (122)

be filled

hole

tsumeru ‘pack’

newspaper
move

kare
cause move
Δ

Japanese has many fewer verbs that allow the two linkings. The four verbs above

allow the locative alternation, and I have mentioned that they have differences in

meanings according to the two linkings. Here is a summary of them:

218
 Chapter 5 Verbs of putting

Table 1. Summary of different senses between L-linking and I-linking

L-linking I-linking
(basic aspectual pattern) (basic aspectual pattern)
kazaru ‘decorate’ display sense decorate sense
(achievement) (accomplishment)
nuru ‘smear’ apply sense cover sense
(activity) (accomplishment)
tsumeru ‘pack’ put-into sense fill sense
(accomplishment) (accomplishment)
mitasu ‘fill’ pour-to-the-full sense fill sense
(accomplishment) (accomplishment)

The verbs denote different events. However, what is common is that with I-linking the

ground undergoes the change of state, and the more conventional construal in these

circumstances is accomplishment. Also the ground serves as a verbal scale. The

change of state of the ground naturally sets the telicity either in a concrete way or in

an abstract way. By ‘concrete’, I refer to the last three verbs where the telicity can be

easily seen in the ground (we can objectively see that the wall is painted, the container

is filled). By ‘abstract’ I refer to the case of kazaru ‘decorate’, where the completion

of decoration can be subjective depending on people’s point of view. The aspectual

construal of L-linking varies. The reason why tsumeru ‘pack’ and mitasu ‘fill’ have an

accomplishment reading with L-linking is that the space of the ground (the container

or the location) can set the endpoint to the amount of the figure. Both the figure and

the ground can be associated with a verbal scale and the subtle difference between the

two linkings is ascribed to the fact of which argument holistic theme is assigned to.

These two verbs are the ones which directly follow the semantic representation of the

spray/load verbs in English.

Moreover, we have seen that there are fewer contexts where I-linking is allowed

than were L-linking is allowed20. That is, L-linking is used more neutrally and is a

20
This is consistent with the general fact that Japanese has many more verbs which allow L-linking

219
 Chapter 5 Verbs of putting

default case, while I-linking is rather a special case. This is explained by saying that I-

linking requires not only that the figure is moved to the ground but also that the

ground undergoes some change of state, which is an additional meaning. This

additional meaning makes I-linking more special and limits the contexts where it

occurs21.

5.4.12. Type 12 Tsutsumu ‘wrap’ verbs

5.4.12.1. Members

The following are the four members of this verb type:

verbs of wrapping:
tsutsumu ‘wrap’, kurumu ‘wrap, tuck’ , maku ‘roll, wrap’
verb of hiding:
kakusu ‘hide, cover’

5.4.12.2. Semantics

These verbs typically refer to putting an entity (clothes, cloths, or paper) over another

to cover it. Some verbs (such as tsutsumu ‘wrap’, kakusu ‘hide’) entail that an object

is entirely covered up, while others (such as kurumu ‘wrap, tuck’) do not necessarily

entail complete coverage.

5.4.12.3. Syntactic patterns

The verbs manifest L-linking as example (128) shows. A peculiar characteristic of the

verbs of this type is that they also take ‘reversed’ instrumental linking as in example

(129):

rather than I-linking.

21
Fukui et al. (1985) are partly right in saying that verbs that allow the locative alternation have a
component of “changing ground”. However, to be more specific,we should say that this component
only appears with I-linking.

220
 Chapter 5 Verbs of putting

L-linking:
Jacqui wa okashi o shinbunshi ni tsutsun-da.
(128) Jacqui TOP sweets ACC newspaper ALL wrap-PAST
Jacqui wrapped the sweets in newspaper.

I-linking (Reversed instrumental linking)22:

Jacqui wa okashi o shinbunshi de tsutsun-da.
(129) Jacqui TOP sweets ACC newspaper INST wrap-PAST
Jacqui wrapped the sweets with newspaper.

The covered object (‘sweets’) is marked as accusative in both patterns and the

covering object (‘newspaper’) has two syntactic realisations, ni in L-linking and de in

I-linking. As the English translations and Levin’s description of (v) of 9.8 Fill verbs

indicates that English also has two patterns, in and with, to mark the covering object.

I call this alternation the FG construal alternation, for the reasons discussed below.

This alternation is different from that with the previous verb type that shows the

Japanese locative alternation, which alternates between L-linking and I-linking

without changing figure and ground construal.

Let us now consider the use of kurumu ‘wrap’ in examples (130) and (131) as well

as that of tsutsumu ‘wrap’ above.

Kurumu examples:

Watashi wa akanbo o mouhu ni kurun-da.

(130) I TOP baby ACC blanket ALL wrap-PAST
I tucked a baby in a blanket.

Watashi wa akanbo o mouhu de kurun-da.

(131) I TOP baby ACC blanket INST wrap-PAST
I tucked a baby with a blanket.

Examples (128) and (130) manifest conventional L-linking with the covered object

as the figure and the covering object as the ground. That is, the covering object is

22
Note that this linking is the same as normal I-linking in that the figure is marked as the antecedent
oblique and the ground is marked as the direct object. I have used the term “reversed” for the sake of
the convenience in that the figure and ground construal is different from L-linking of the verbs of the
category.

221
 Chapter 5 Verbs of putting

construed as a place where the covered object is located. Examples (129) and (131)

manifest I-linking, but this time the covered object is the ground and the covering

object is the figure. That is, the covering object is construed as the instrument to

‘cover’ the covered object. The verbs of this type show the two linking patterns with

the construal of the ground and the figure in one pattern reversed in the other.

We could attribute the possibility of the two construals above to the fact that the

semantic properties of both participants are ambiguous in determining the figure and

the ground. Talmy (1978: 627) defines the figure as a moving or conceptually

movable point and the ground as a reference point which has a stationary setting. The

covering object is larger than the covered object, which means that the former is more

likely to be construed as a reference point (the ground). On the other hand, the

covering object is moving while the covered object is stationary, which favours the

latter being the ground. (When we wrap something, we move a covering object and

not a covered object.) Each has the potential to be construed as either ground or figure,

thus making the figure/ground (FG) construal alternation possible. This is a feature of

this verb type as we cannot always change a figure construal for a ground construal.

For example, nuru ‘smear’ in penki o kabe ni nuru (‘paint ACC wall ALL smear’)

cannot have the alternative construal because wall is not mobile and is bigger in a

normal conceptualisation. .

One verb of hiding, kakusu, has L-linking and I-linking, in a slightly different way

from verbs of wrapping, which can refer to the same situation of wrapping with each

the two different linking patterns. In kakusu ‘hide’, two different linkings (i.e. two

different figure/ground construals) are in complementary distribution. The verb

originally meant to hide or conceal something such that it is not visible or perceivable.

One way to do this is to put that entity into a container or put it somewhere where

222
 Chapter 5 Verbs of putting

nobody sees it and the other way is to cover the entity with something. L-linking is

used to entail the former way and I-linking is used to entail the latter way as the

following examples illustrate:

Jacqui wa atarashii kutsu o hikidashi ni kakushi-ta.

(132a) Jacqui TOP new shoes ACC drawer ALL hide-PAST
Jacqui hid the new pair of shoes in the drawer.

Jacqui wa atarashii kutsu o mouhu de kakushi-ta.

(132b) Jacqui TOP new shoes ACC blanket INST hide-PAST
Jacqui hid the new pair of shoes by covering them with the blanket.

Thus, the hidden entity (the pair of shoes) is the direct object in the two linkings; it is

construed as the figure with L-linking and as the ground with I-linking. I call the

former use of kakusu the ‘hiding by putting-in(to)’ sense and the latter use the ‘hiding

by covering’ sense.

5.4.12.4. Causal and aspectual patterns

The causation pattern reflects the ambiguity of the figure and ground construals. The

agent acts on the figure to wrap the ground or to be wrapped by the ground. This is

extended causation as it takes a certain amount of time to wrap something and the

agent needs to exert force during that time. With te-iru, both patterns have an activity

in progress meaning, indicating that the agent is engaged in an activity of wrapping to

cover the entity.

Jacqui wa okashi o shinbunshi ni tsutsun-de-iru.

(133) Jacqui TOP sweets ACC newspaper ALL wrap-TE-IRU
Jacqui is wrapping the sweets in newspaper.

Jacqui wa okashi o shinbunshi de tsutsun-de-iru.

(134) Jacqui TOP sweets ACC newspaper INST wrap-TE-IRU
Jacqui is wrapping the sweets with newspaper.

To be more precise, the event is an accomplishment with the covered object as the

verbal scale. The sukoshizutsu ‘little by little’ test in examples (135) and (136) shows

223
 Chapter 5 Verbs of putting

that the sweets (the covered object) are the holistic theme in both L-linking and I-

linking.

L-linking
Jacqui wa okashi o shinbunshi ni sukoshizutsu tsutsun-da.
(135) Jacqui TOP sweets ACC newspaper ALL little.by.little wrap-PAST
Jacqui wrapped the sweets in newspaper little by little.

I-linking
Jacqui wa okashi o shinbunshi de sukoshizutsu tsutsun-da.
(136) Jacqui TOP sweets ACC newspaper INST little.by.little wrap-PAST
Jacqui wrapped the sweets with newspaper little by little.

This is not surprising because the event of wrapping an object is only measured by

how much the object is covered and not by how much paper or cloth is used.

5.4.12.5. Semantic representations

Semantic representations for L-linking and I-linking are proposed in the following

Figures. Both of them capture the fact that the covered object is associated with a

verbal scale. The order of causal chain is opposite in the two diagrams reflecting the

different figure and ground construals.

224
 Chapter 5 Verbs of putting

Figure 32. accomplishment construal of L-linking of tsutsumu ‘wrap’ in (128)

newspaper
exist
ni

be located in

sweets tsutsumu ‘wrap’

Jacqui cause become located

Δ

Figure 33. accomplishment construal of I-linking of tsutsumu ‘wrap’ in (129)

be wrapped
sweets

tsutsumu ‘wrap’

newspaper wrap

Jacqui
apply
Δ

225
 Chapter 5 Verbs of putting

5.4.13. Type 13 Hou.boku-suru ‘graze’ verbs: ground incorporation

5.4.13.1. Members

The following are some members of this verb type:

tou.goku-suru (throw’into.prison-do = ‘imprisonment-do’ = prison),

hou.boku-suru (release.pasture-do = ‘graze’)

5.4.13.2. Semantics

The number of verbs in this category is small compared to the ground incorporation

verbs of English (9.10 pocket verbs in Levin’s classification). They are VN-suru

verbs of Chinese origin. The first part refers to the VN of causing motion (verb) and

the second part to the ground (noun), where the figure is caused to move to.

5.4.13.3. Syntactic patterns

They take the figure as a direct object (complete incorporation) and manifest FA-

linking.

FA-linking
Hokkaido de-wa natsu ni ushi o hou.boku-suru.
(137) Hokkaido LOC-TOP summer LOC cattle ACC graze
In Hokkaido, they put the cattle out to pasture in summer.

They can also take the ground in the oblique position. In this case the verbs manifest

L-linking and this specifies additional meanings of the ground (incomplete

incorporation). Also the verbs do not take a source phrase.

L-linking
Hokkaido de-wa natsu ni ushi o sanpuku ni hou.boku-suru.
(138) Hokkaido LOC-TOP summer LOC cattle ACC hillside ALL graze
In Hokkaido, they put the cattle out to pasture on the hillside in summer.

*Source phrase
*Hokkaido de-wa natsu ni ushi o koya kara hou.boku-suru.
(139) Hokkaido LOC-TOP summer LOC cattle ACC pen ABL graze
*In Hokkaido, they put the cattle out to pasture from the pen in summer.

226
 Chapter 5 Verbs of putting

5.4.13.4. Causal and aspectual patterns

The agent acts on the figure and causes it to change its location. The aspect construal

of these verbs is basically punctual as te-iru does not give the activity in progress

sense but the retrospective sense or the habitual reading.

Hokkaido de-wa natsu ni ushi o hou.boku-shi-te-iru.

(140) Hokkaido LOC-TOP summer LOC cattle ACC graze-TE-IRU
In Hokkaido, they put the cattle out to pasture in summer.

5.4.13.5. Semantic representations

The two representations are proposed as follows: Figure 34 for complete ground

incorporation and Figure 35 for incomplete ground incorporation.

Figure 34. achievement construal of FA-linking of hou.boku-suru ‘graze’ in (137)

be located in pasture

cattle

hou.boku-suru ‘graze’

they release
Δ

227
 Chapter 5 Verbs of putting

Figure 35. achievement construal of L-linking of hou.boku-suru ‘graze’ in (138)

hillside
exist
ni
be located in pasture

cattle hou.boku-suru ‘graze’

they

Δ release

5.4.14. Type 14 San.sui-suru ‘sprinkle.water-do’ verbs: figure incorporation 1

5.4.14.1. Members23

The following are some member of this verb type:

pattern 1
hou.sui-suru (release/give’off.water-fo = spray water)
san.sui-suru (scatter.water-do = spray/sprinkle water)
pattern 2
shup.pin-suru (take’out.goods-do = exhibit)
nyuu.ka-suru (put’in.load-do = receive a supply of ~ )
pattern 3
shoku.ju-suru (plant.tree-do = plant a tree)

23
There are three patterns according to the type of verbs used as the first element in the VN verbs. The
first pattern is verbs of the scattering type, the second pattern is type 2 (caused motion action with
specified direction) and the third pattern is other type of verbs of putting. However, I do not
differentiate them into different ‘subtypes’ but only as having different ‘patterns’ which do not have
very outstanding differences in argument linking.

228
 Chapter 5 Verbs of putting

5.4.14.2. Semantics

The verbs are figure incorporation verbs (similar to butter verbs in Levin’s

classification). The first part of VN refers to the caused motion action and the second

part to the figure, which is caused to move or change its location.

5.4.14.3. Syntactic patterns

Not unlike in the ground incorporation pattern, there are two variations: complete and

incomplete incorporation. An argument which gives an additional meaning to the type

of figure will appear syntactically as the direct object in incomplete incorporation. A

significant difference from 9.9 butter verbs (figure incorporation verbs) in English is

that in Japanese the ground is realised as oblique (GO-linking) and only the figure can

be realised as the direct object (L-linking) while the ground is realised as the direct

object in English (cf. Matsumoto (1997: 176-177)), such as He watered the garden.

GO-linking (complete incorporation)

Kare wa niwa ni san.sui-shi-ta.
(141) he TOP garden ALL scatter.water-do-PAST
He watered the garden. (Jap. He watered to the garden.)

*GA-linking
*Kare wa niwa o san.sui-shi-ta.
(142) he TOP garden ACC scatter.water-do-PAST
Jap. *He watered the garden.

L-linking (incomplete incorporation)

Kare wa niwa ni idomizu o san.sui-shi-ta.
(143) he TOP garden ALL well.water ACC scatter.water-do-PAST
He watered the garden with water from the well.
(Jap. He watered water from the well to the garden.)
(This example is from Matsumoto 1997: 176)

229
 Chapter 5 Verbs of putting

source-phrase (some) 24
?Kare wa oke kara idomizu o san.sui-shi-ta.
(144) he TOP bucket ABL well.water ACC scatter.water-do-PAST
?He scattered well water from the bucket.

The examples show that the ground appears as oblique in both complete and

incomplete figure incorporation patterns in Japanese. The Japanese translation

equivalent of He watered the garden (with water from the well) is ungrammatical.

Some of the verbs in this category may appear with FA-linking as well.

Kare wa idomizu o san.sui-shi-ta.

(145) he TOP well.water ACC scatter.water-do-PAST
He scattered water from the well.

5.4.14.4. Causal and aspectual patterns

The causal pattern is that the agent acts on the figure and causes it to move or change

its location. The aspectual properties vary according to individual verbs25. Pattern 1

bears the activity in progress sense with te-iru indicating that somebody is in the

middle of spraying/scattering water. Pattern 2 cannot have the activity in progress

sense (except with a special context like runup achievement), so the event is punctual.

Pattern 3 can bear the activity in progress sense with te-iru indicating a runup

achievement event (when the figure is singular) or repeated events of achievement

(when the figure is plural).

24
With figure incorporation verbs having ‘put into’ and ‘take out’ verbs (pattern 2) or verbs of
scattering (pattern 1) as the first element, it is also possible to specify the source with Ablative case.
This conforms to the characteristics of these verbs already discussed.
25
The aspectual patterns of verbs of this category basically comply with that of the first character of
VN. For example, the first verbal part is equivalent to ‘put.in’ or ‘take.out’ (type 2) verbs, and therefore
characteristics of aspectual patterns of these verbs apply to them. If the first part is a verb of scattering,
then the VN follows the aspectual pattern of verbs of scattering.

230
 Chapter 5 Verbs of putting

5.4.14.5. Semantic representations

The complete/incomplete incorporation versions of san.sui-suru ‘scattering water’ are

presented in the Figures below.

Figure 36. undirected activity construal of GO-linking of san.sui-suru ‘scatter water’

in (141)

garden exist
ni

san.sui-suru
‘scatter water’

he emit water
Δ

Figure 37. undirected activity construal of L-linking of san.sui-suru ‘scatter water’ in

(143)

garden
exist
ni

well water move

san.sui-suru
‘scatter water’

he emit
Δ

231
 Chapter 5 Verbs of putting

5.4.15. Type 15 Shoku.rin-suru ‘afforest’ verbs: figure incorporation 2

5.4.15.1. Members

The following are the three members of this verb type:

shoku.rin-suru (plant.woods-do= afforest/plant land with trees)

juu.den-suru (fill.electricity-do = charge)
chaku.shoku-suru (dye.colour-do = dye/put colour)

5.4.15.2. Semantics

As with the previous type, the first part of the VN is the verb of putting and the second

part specifies the figure. The figure can be fully incorporated, in which case it does

not appear syntactically, or it can be partially incorporated, in which case, the figure

argument that adds specific meaning to the type of figure appears in the syntax.

5.4.15.3. Syntactic patterns

This category of figure incorporation assigns the direct object position to both figure

and ground. When the figure is completely incorporated, the ground is marked either

with Accusative (GA-linking) or with oblique case (GO-linking) as follows:

GA-linking
Hageyama o shoku.rin-suru.
(146) bald.mountain ACC afforest
They afforest the bald mountain.

GO-linking
Hageyama ni shoku.rin-suru.
(147) bald.mountain ALL afforest
They afforest the bald mountain.

Then, where the figure is partially incorporated, this type of verb has two linking

patterns (L-linking and I-linking) as follows:

L-linking
Hageyama ni sugi no ki o shoku.rin-suru.
(148) bald.mountain ALL Japanese.cedar GEN tree ACC afforest
They plant Japanese cedars in the bald mountain.

232
 Chapter 5 Verbs of putting

I-linking
Hageyama o sugi no ki de shoku.rin-suru.
(149) bald.mountain ACC Japanese.cedar GEN tree INST afforest
They afforest the bald mountain with Japanese cedars.

A source phrase is not so acceptable, and the path pattern is marginally grammatical

as following examples illustrate:

*source-phrase
*Mt.Fuji kara sugi no ki o shoku.rin-suru.
(150) Mt.Fuji ABL Japanese.cedars GEN tree ACC afforest
They plant Japanese cedars from Mt. Fuji.

?Path pattern
?Mt.Fuji kara hageyama ni sugi no ki o shoku.rin-suru.
(151) Mt.Fuji ABL bald.mountain ALL Jap.cedar GEN tree ACC afforest
They plant Japanese cedars from Mt. Fuji to the bald mountain.

5.4.15.4. Causal and aspectual patterns

The causal pattern for this type of verbs follows that of the previous types (type 14).

The aspectual pattern is accomplishment; events of charging a battery or afforesting

land extends over the t-scale. These verbs are quite similar to type 11 decorate verbs;

shoku.rin-suru ‘afforesting’ is making a certain place a forest by planting a lot of

trees26 or filling a place with trees and juu.den-suru ‘charging a battery’ is filling a

battery with electricity. The only difference is that the figure is incorporated in this

category. This type is especially close to tsumeru ‘pack’ and mitasu ‘fill’ of the

decorate type verbs; The ground naturally delimits the events and also sets the natural

limit on how much of the figure will be used. Therefore, either the ground or the

figure can be associated with a verbal scale. As we can expect, the difference between

L-linking and I-linking is very subtle. The interpretation of the verbal scale correlates

26
It would be interesting to contrast shoku.ju-suru (plant tree) of type 14 and shoku.rin-suru (plant
forest= afforest). Shoku.ju-suru can be used for planting only one tree. On the other hand, shoku.rin-
suru entails planting a certain number of trees as we cannot make a forest with one tree.

233
 Chapter 5 Verbs of putting

with the two syntactic patterns which reflect the different construals. Again, the

sukoshizutsu ‘little by little’ test with the above examples reveals that it is the direct

object which is the verbal scale. Example (152) and (153) illustrate:

L-linking
Hageyama ni sugi no ki o sukoshizutsu shoku.rin-shi-ta.
(152) bald.mountain ALL Jap.cedar GEN tree ACC little.by.little afforest-PAST
They planted Japanese cedars in the bald mountain little by little.
(Trees are planted little by little.)

I-linking
Hageyama o sugi no ki de sukoshizutsu shoku.rin-shi-ta.
(153) bald.mountain ACC Jap.cedar GEN tree INST little.by.little afforest-PAST
They afforest the bald mountain with Japanese cedars.
(The mountain is covered little by little.)

The subtle difference can be found in GA-linking and GO-linking as well as the

following examples illustrate:

GA-linking
Karera wa hageyama o sukoshizutsu shoku.rin-shi-ta
(154) they TOP bald.mountain ACC little.by.little afforest-PAST
They afforested the bald mountain little by little.

GO-linking
Karera wa hageyama ni sukoshizutsu shoku.rin-shi-ta
(155) they TOP bald.mountain ALL little.by.little afforest-PAST
They planted (trees) on the bald mountain little by little.

In GA-linking, the ground is associated with a verbal scale as ‘little by little’ modifies

the state of the mountain being gradually covered with trees in example (154). On the

other hand, the adverb indicates that the trees are planted little by little (155). The

semantic representations in Figures 38 and 39 capture the difference.

5.4.15.5. Semantic representations

The semantic representation for incomplete incorporation version is similar to those of

tsumeru ‘pack’. I therefore present only the figures for complete incorporation version.

234
 Chapter 5 Verbs of putting

Figure 38. accomplishment construal of GA-linking of shoku.rin-suru ‘afforest’ in

(154)

be covered
with trees/afforested

mountain
shoku.rin-suru
‘afforest’
cover by
they planting trees

Figure 39. accomplishment construal of GO-linking of shoku.rin-suru ‘afforest’ in

(155)

mountain
exist
ni

shoku.rin-suru
‘afforest’

plant trees
they
Δ

One may intuitively feel that it is strange to associate a verbal scale with the agent in

Figure 39. However, it is the only way to illustrate the above reading. As its subevent

denotes planting “trees”, it is still allowed to have a directed activity construal. The

235
 Chapter 5 Verbs of putting

contour of the agent in Figure 39 is a course-grained representation of the iterated

cycle of achievement of planting one tree.

5.4.16. Type 16 Hame-komu ‘fit.into-put.into’ verbs: means compound 1 with

specified direction

5.4.16.1. Members

The following are some of the members of this verb type27:

pattern 1:
hame-komu ‘put -put into’, shimai-komu ‘put.away-put.into’, tsume-komu ‘pack-
put.into’, sosogi-ireru ‘pour.into-put.into’

pattern 2:
hiki-komu ‘pull-put.into’, oshi-komu ‘push-put.into’, mochi-ageru ‘take-raise’, nui-
komu ‘sew-put.into’

VN-suru: chuu.nyuu-suru (‘pour.put.into-do’ = inject)

5.4.16.2. Semantics

Verbs of this type are either compounds or VN-suru verbs. As the title of 5.4.16

indicates, more than one element is added to the simple meaning of caused-motion. In

compounds, the V1 components indicate a sort of means by which the caused-motion

is carried out. The V2 components are verbs that are subcategorised as type 2, and

which specify the direction or location of caused-motion, such as ireru ‘put into’ and

ageru ‘raise’. Komu28 ‘put into’ is the most productive V2 component of verbs of

27
The compounds are classified into two patterns according to the way their component structures are
integrated. This will be discussed later. Their syntactic behaviours are not drastically different, so I do
not differentiate them as different subtypes.
28
The verb komu does not appear in type 2 verbs as it does not mean caused-motion when used as a full
verb. It means that a specific place is full of something as follows:
Densha ga komu.
(i)
train NOM be.full.of.people
The train is full of people.

236
 Chapter 5 Verbs of putting

putting in the data. Compound verbs of this type are very extensive and productive.

Various kinds of verb classes can appear as a V1-means component: verbs of putting,

verbs of contact by impact, verbs of throwing, verbs of sending and carrying, verbs of

exerting force, verbs of holding, and so on29. For Chinese-origin VN verbs, both of the

two elements of VN are verbs; the first is equivalent to V1 (specifying means) and

second to V2 (specifying caused motion and direction/location).

5.4.16.3. Syntactic patterns

Th linking patterns follow those of type 2 verbs. The verbs can take a path function

which specifies a spatial goal (marked with ni) and a spatial source (marked with

kara), though one of them is often omitted. Therefore, compounds with this type as

V2 may appear as verbs of removing as well. As has been discussed in relation to type

2 verbs, dasu ‘take out’ tends to omit the goal more than the others and therefore, is

more likely to appear as a verb of removing. Komu ‘put into’ predominantly takes the

goal rather than the source argument.

path pattern
Wayne wa omoi iwa o jimen kara zujou ni mochi-age-ta.
(156) Wayne TOP heavy rock ACC ground ABL overhead ALL take-raise-PAST
Wayne lifted a heavy rock from the ground to above his head.

L-linking:
Jacqui wa takusan no youhuku o kuroozetto ni oshi-kon-da.
(157) Jacqui TOP many GEN dresses ACC closet ALL push-put.into-PAST
Jacqui pushed many dresses into the closet.

When it is used as V2 in a compound, it bears a different function which refers to direction, that is, the
transfer of something to inside a certain entity. As V2, the verb is translated as ‘go in/into’ (intransitive)
or ‘put in/into’ (transitive).
29
There is a constraint that V1 verbs should be transitive verbs as V2 verbs here are transitive. The
constraint follows the principle of transitivity harmony referred to in (Kageyama 1993: 117), who
observes that the combination of compounds is possible only with transitive plus transitive, unergative
plus unergative, transitive plus unergative, and unaccusative plus unaccusative.

237
 Chapter 5 Verbs of putting

They do not take I-linking.

*I-linking
*Jacqui wa takusan no youhuku de kuroozetto o oshi-kon-da.
(158) Jacqui TOP many GEN dress INST closet ACC push-put.into-PAST
*Jacqui pushed the closet with many dresses.

5.4.16.4. Causal and aspectual patterns

The causal pattern and aspectual pattern follow those of type 2 verbs; the agent acts on

the figure and causes it to move to the ground. Ireru ‘put into’, dasu ‘put out’, or komu

‘put into’ conventionally take an achievement construal but they can also take an

accomplishment construal if the figure can be construed as a holistic theme associated

with a verbal scale such that it is crossing the boundary line, or it is plural or refers to

a certain amount of an uncountable entity. Ageru ‘raise’ and sageru ‘lower’ basically

have a path function and so can have an accomplishment construal but are often

construed as achievements (trivial verbal scale).

5.4.16.5. Semantic representations

For means compounds discussed in this chapter, the formation of compound verbs is

basically right-headed. Right-headed means that the argument structure of the V2 (left

component) is inherited by the compound verbs. That is the reason why the causal

pattern and syntactic pattern of type 2 verbs determine those of the compounds of this

category. I shall examine how V1 and V2 are compounded in semantic representation.

Here I use the simple aspectual construal of achievement unless otherwise specified.

There are several patterns in terms of semantic structure and argument structure of V1

components, thus the way V1 and V2 are integrated varies. I classify them into the

two main patterns; one pattern is that V1 has the same argument structure with V2 (I

call this a ‘middle’ compound), and the second pattern is that V1 has a different

238
 Chapter 5 Verbs of putting

argument linking structure from that of V2 (I call this a ‘pure means’ compound). Let

us observe some examples mainly involving komu ‘put into’. The semantic

representation of komu as V2 is supposed to be like ireru 'put in/into' which denotes

an action on the part of the agent that causes a change of location of the figure to the

ground.

5.4.16.5.1. Middle compounds: hame-komu ‘fit.into-put.into’

Let us observe the first pattern, where a V1 has the same argument structure as V2 and

also inherently has a meaning of change of location.

Haha wa waku ni paneru o hame-ta.

(159a) mother TOP frame ALL panel ACC fit.into-PAST
Mother fit the panel into the frame. (V1—L-linking)

Haha wa waku ni paneru o hame-kon-da.

(159b) mother TOP frame ALL panel ACC put.into-put.in-PAST
Mother put the panel into the frame. (Compound—L-linking)

Hameru 'fit, put into' itself is a verb of putting which takes L-linking. One might

think that hameru and ~komu are almost synonymous, and therefore that the verb in

example (159b) is a pair compound. However, hameru ‘put,fit into’ implies an

additional meaning to just ‘putting into’. It generally refers to the action of putting an

entity into/onto another entity so that the former entity fits into the latter. (cf. JSD:

1047) such as putting a watch onto one's wrist (Eng. wear a watch on the wrist)),

putting hands into gloves, or putting an entity into the frame etc. That is, we can

conclude that hameru has the putting-into sense and the putting-onto sense. I also

classified this verb as type 5 in the previous discussion. V1 has a certain additional

meaning as well as a basic cause-motion meaning, so it can be the means which

modifies how/in what way the caused motion events denoted by V2 have happened. In

the following representation, I add that means meaning to the labelling of the contour.

239
 Chapter 5 Verbs of putting

In the following semantic structure of the compound, the thick black lines indicate the

verb profile of the compound verb, and the arrows on the right hand side of the figure

with the component verbs in the middle indicate the frame/background of the structure

that is entailed by each component verb. The frame includes the unprofiled parts as

well as the profiled ones. I shall continue this convention for the semantic

representation of the other compounds I discuss in the thesis.

240
 Chapter 5 Verbs of putting

Figure 40. Semantic structure for L-linking of hame-komu ‘fit into’ in (159)

V1 V2

frame frame
exist exist
ni ni
be located
panel be set (properly) panel be located in
hameru komu
‘fit,put into’ ‘put into’

cause located cause located

Mother Mother
Δ Δ

t t

Compound

frame
exist
ni
be fitted into

panel hameru komu

hame-komu
‘fit-put.into’

Mother
cause located
Δ

241
 Chapter 5 Verbs of putting

The semantic representations of V1 and the compound are identical to that of oku

'put' verbs. Middle compound verbs of this type have the same problem as pair

compounds regarding how to decide the argument structure of the component that is

inherited into the compound; headedness is ambiguous because V1 and V2

supposedly have the same linking patterns. Just as in the case of pair compounds (cf.

Matsumoto 1996), we could regard them as right-headed as long as there is no

counterevidence. However, there is a partial supporting evidence that the argument

structure of V2 determines that of V1. Tsumeru ‘pack, fill’ is a verb of the locative

alternation (cf. § 5.4.11.) When it is compounded with komu as in tsume-komu ‘pack-

put into’, only L-linking is allowed and not I-linking. Thus, at least, we can conclude

that the V2 component has more effect in determining the argument structure of the

compound.

V1 components of this type are themselves already caused-motion three-argument

verbs and semantically include the caused-motion sense (but without a specified

direction). So, what is the point of V1 being compounded with V2? I would say that

komu is used to emphasise the meaning of locational transfer, especially in a specified

direction. That is, the putting-into sense is emphasised. JSD (p. 1047-8) explains the

subtle semantic difference between hameru ‘put, fit in’ and hame-komu ‘fit into’ by

saying that, in the latter, the meaning of putting something inside a certain entity is

stressed. We can also expect that contexts where hame-komu is used are more limited

because of this additional stress on the original meaning of hameru. An instance of

hameru in such a context is given by the following interesting example from JSD (p.

1047-8) of putting a ring on one's finger.

Haha wa yubi ni yubiwa o hame-ta.

(160) mother TOP finger ALL ring ACC put-PAST
Mother fit the ring on her finger.

242
 Chapter 5 Verbs of putting

Hame-komu, which emphasises putting an entity inside something, cannot be used in

the above sense, which is the putting-onto (cover) sense. Thus, example (161) is

ungrammatical as the ring does not go inside the finger30.

*Haha wa yubi ni yubiwa o hame-kon-da.

(161) mother TOP finger ALL ring ACC put-put.in-PAST
Mother fit the ring into her finger.

The example of tsume-komu ‘pack-put.into’ also shows that V2 emphasises the

‘putting into’ sense. According to the JSD, the compound means ‘pack forcibly’, a

meaning which V1 lacks.

In summary, in middle compounds, the argument structures of V1 and V2 are

identical and also participants are referentially identical. Therefore, it is difficult to see

which argument structure is inherited by compounds. However, at least, there is one

piece of evidence that V2 is a determinant factor. Though V1 has two linking patterns,

either one or the other is suppressed when it is combined with V2 (ex. tsume-komu).

Middle compounds are different from pair compounds in that V1 describes means so

it has some additional ‘subtle’ meaning or restriction on how the action of caused

motion has happened and V2 emphasises the meaning of V1 by further specifying the

directional sense it has.

30
Interestingly, the JSD points out that the following is possible.
Haha wa yubiwa ni yubi o {muriyari} hame-kon-da.
(i)
mother TOP ring ALL finger ACC {by.force} put-put.in-PAST
Mother put her finger into the ring forcibly.

Elements construed as the ground and the figure in example (161) are reversed. It is natural to say (i)
since physically the finger is inside the ring. The default context of wearing a ring is taking up a ring
and cause it to move to fit onto a finger, thus example (160) is grammatical. With this example of an
unconventional scenario, the speaker construes the sentence as if the finger is moving inside the ring
rather than the ring moving onto the finger.

243
 Chapter 5 Verbs of putting

5.4.16.5.2. Pure means compounds

Now we examine the other type of compounds where V1 has different argument

structure from V2.

5.4.16.5.2.1. Oshi-komu ‘push-put.into’

This compound, which is a means compound with extended causation, is easily

confused with manner compounds. Means compounds with extended causation have

an ambiguous status concerning the ni-yotte/~nagara ‘by~/while~’ tests (cf. § 4);.they

pass both of the tests. As the agent continues to apply force to the figure and causes it

to move, the means compound can be interpreted as indicating the ‘manner’. A whole

event is accompanied by a certain action of the agent. However, I define the ‘manner’

of caused-motion compound as that of the moving entity (the figure) not that of the

agent and I would treat caused-motion compounds with extended causation as means

compounds. This position is not contrary to Matsumoto (1996), who treats verbs in

this category as means compounds.

In the case of the second pattern, V1 verbs do not have the caused-motion sense as

the following examples indicate:

*Ane wa takusan no doresu o oshi-ta.

(162a) sister TOP many GEN dress ACC push-PAST
*My sister pushed many dresses. (V1)

Ane wa kabe o oshi-ta.

(162b) sister TOP wall ACC push-PAST
My sister pushed the wall. (V1)

Ane wa takusan no doresu o kuroozetto ni oshi-kon-da.

(162c) sister TOP many GEN dress ACC closet ALL push-put.into-PAST
My sister pushed many dresses into the closet. (Compound—L-linking)

*Ane wa kabe o dokoka ni oshi-kon-da.

(162d) sister TOP wall ACC somewhere ALL push-put.into-PAST
*My sister pushed the wall into somewhere. (Compound—L-linking)

244
 Chapter 5 Verbs of putting

Verbs of exerting force (hiku ‘pull’ and osu ‘push’) are analysed as having the

semantic representation of Figure 41 of V1 since pushing and pulling do not indicate

that the patient has to undergo changes (cf. Croft 2000, §3.5.2). When they are

compounded with komu, however, the meaning of change (of location) of the figure is

added to the original part of the representation of the V1 verb as well as the caused-

action part being maintained and emphasised by the V2. The following figure

represents komu and the compound in an accomplishment construal to make clear that

the causation of the agent continues from the beginning to the end until the dresses

reach inside the closet. Like other type 2 verbs, it can also be construed as an

achievement alternatively.

245
 Chapter 5 Verbs of putting

Figure 41. Semantic structure for L-linking of oshi-komu ‘push-put.into’ in (162)

V1 V2

wall GRD
*dresses (closet) exist
ni
exist be located

osu ‘push’ FIG komu

apply pushing ‘put into’
sister force cause become
located
Δ Δ AGT

t t

Compound

closet
exist
ni
be located by
pushing komu
dress
oshi-komu
‘push-put.into’
osu
sister
Δ push

246
 Chapter 5 Verbs of putting

V1 and V2 have different argument structures, but the resulting argument structure

of the compound is identical to that of V2, being right-headed. Thinking of the

integration of compounds in the cognitive and constructional approach, the semantic

structure of V2 is thought to be the construction which is instantiated by the semantic

structure of V1. The V2 construction of Figure 41 is superimposed onto that of V1 in

order to be instantiated. That is, the V2 construction can be an elaboration-site (e-site)

in Langacker’s terms (1987: 304-306), where the structure of V1 elaborates the shared

substructure or entities of V2 to specify its rather schematic representation with

specifying how the caused-motion happens. As the result of the superimposition, the

composite structure is given. I shall come back to this in §5.5.

Interestingly, the semantic constraints of the direct object are different between the

V1 and the compound ((162)). Self-evidently, the direct object should be a movable

entity in the compound because it indicates the causing of motion. In the case of osu

‘push’, an entity that cannot resist a pushing force is not preferred as its direct object,

thus it is strange to say “push the dresses”.

5.4.16.5.2.2. Tataki-komu ‘hit-put.into’

This is another examples of pure means compound. This verb designates punctual

causation.

Robert wa booru o tatai-ta.

(163a) Robert TOP ball ACC hit-PAST
Robert hit the ball. (V1)

Robert wa booru o uyoku ni tataki-kon-da.

(163b) Robert TOP ball ACC right.stand ALL hit-put.into-PAST
Robert hit the ball into the right stand. (Compound—L-linking)

The following is the proposed representation of the compound:

247
 Chapter 5 Verbs of putting

Figure 42. Semantic structure for L-linking of tataki-komu ‘hit-put.into’ in (163)

V1 V2

GRD
exist
ni
ball hit be located

FIG
tataku ‘hit’ komu ‘put into’

Robert AGT cause become

impact located
Δ Δ

t t

Compound

right stand
exist
ni
be located by
ball being hit
tataki-komu
‘hit into’ tataku komu

Robert impact

248
 Chapter 5 Verbs of putting

The process of combining the two verbs is similar to that in the previous example

(push plus put into case) in terms of argument linking. Tataku ‘hit’ was originally a

verb of contact by impact and does not take the third argument. The goal phrase is

introduced to the compound one by V2 komu ‘put into’, though it is not profiled by

the verb. The subevent the ball undergoes is also aspectually different. With tataku

‘hit’, the ball’s change of state is not entailed by the verb; in the representation above,

it is affected only for the moment of the contact (construed as a cyclic achievement).

The change of state, therefore, is also introduced by the V2. The semantic structure of

the V2 is superimposed onto that of the V1, which, in turn, instantiates V2.

The direct object of verbs of contact by impact can be either a mobile entity (such

as the ball in the above example) or a immobile one (such as a wall), as illustrated in

the following example.

Robert wa kabe o tatai-ta.

(164) Robert TOP wall ACC hit-PAST
Robert hit the wall. (V1)

As in the case of oshi-komu (push-put.into), only the mobile entity can be combined

with verbs of caused-motion with specified direction as V2. Verbs of contact by

impact with the immobile entity may be represented differently. Since an immobile

entity is less affected by the act of the agent, it can be represented in “less-affected”

way as follows (cf. Croft 2000 Figure 15b, §4.3):

249
 Chapter 5 Verbs of putting

Figure 43. Less affected direct object of tataku ‘hit’ in (164)

wall
exist

tataku’hit’

Robert impact
Δ

In this reprsentation, the wall is represented merely as a place of contact.

In summary, in pure means compounds, the change of location part is newly

introduced by V2, which has a different argument structure from V1. The semantic

restrictions on arguments (the figure) of V1 may be altered so as to be integrated into

compounds. A part of argument linking structure of V1 is preserved by virtue of the

subject and the direct object being assigned in the same way in compounds; the agent

appears as the subject and the entity acted on appears as the direct object.

5.4.17. Type 17 Nuri-tsukeru ‘smear-apply’ verbs: means compound 2

5.4.17.1. Members

The following are some members of this verb type:

(middle compound):
nuri-tsukeru (‘smear-apply’),
hari-tsukeru (‘stick-attach’),
ue-tsukeru (‘plant-attach’)

(pure means compound)

nui-tsukeru (‘sew-attach’),
huri-kakeru (‘shake-sprinkle’),
huri-maku (‘shake-scatter’)

250
 Chapter 5 Verbs of putting

5.4.17.2 Semantics

Most of the verbs of this class are compound verbs whose V1 component indicates

means and whose V2 component is a verb that entails an event of caused-motion

without specifying direction, unlike type 2 verbs. Rather, according to Matsumoto

(1997: 162-165), the V2 verbs of this category entail/incorporate a certain change of

state incidental to the caused-motion. (These are attaching/applying tsukeru verb, or

the kazaru ‘decorate’ type). I shall now examine some examples of means compounds

and suggest semantic representations for them.

5.4.17.3. Syntactic patterns

This type of verbs takes L-linking but not I-linking, as the following examples

illustrate:

L-linking
Mary wa kabe ni penki o nuri-tsuke-ta.
(165) Mary TOP wall ALL paint ACC smear-apply-PAST
Mary smeared paint on the wall.

Haha wa bajji o yunifoomu ni nui-tsuke-ta.

(166) mother TOP badge ACC uniform ALL sew-attach-PAST
Mother sewed the badge onto the uniform.

*I-linking
*Mary wa kabe o penki de nuri-tsuke-ta.
(167) Mary TOP wall ACC paint INST smear-apply-PAST
Mary smeared the wall with paint.

This type takes neither the source phrase nor the path pattern.

*source phrase (*path pattern)

*Mary wa baketsu kara {kabe ni} penki o nuri-tsuke-ta.
(168) Mary TOP bucket ABL {wall ALL} paint ACC smear-apply-PAST
Mary smeared paint from the bucket {onto the wall}.

Only a few of them (depending on V2) may allow FA-linking.

Jacqui wa kousui o huri-mai-ta.

(169)
Jacqui TOP perfume ACC shake-scatter-PAST
Jacqui sprinkled perfume.

251
 Chapter 5 Verbs of putting

5.4.17.4. Causal and aspectual patterns

The whole meaning of compounds entails that the agent acts on the figure to change

the location to the ground. The aspectual pattern may vary. An event can be construed

as an achievement or accomplishment.

5.4.17.5. Semantic representations

To see how the argument structures of V1 and V2 are integrated into that of

compound, I use examples of ~tsukeru ‘apply/attach’ compounds, which occur

frequently in the data. As a full verb and as a V2 verb, it has several senses such as

applying, sticking, putting on, and attaching. It may also appear in other verb classes

such as verbs of attaching. I discuss the verb as a verb of putting here as the various

senses can be reduced to one single meaning of making two entities have a contact (a

prolonged one not an instanteous contact) physically or more abstractly (in an

extended sense) by moving one entity to the other.

Again, in terms of argument structure, compounds with ~tsukeru ‘apply/attach’ as

V2 can be subdivided into two types; one is that which results from the combination

with a V1 whose argument structure is the same as that of tsukeru (middle comounds),

and the other is that which results from the combination with a V1 whose argument

structure is different (pure means compounds).

252
 Chapter 5 Verbs of putting

5.5.17.5.1. Middle compounds: Nuri-tsukeru ‘smear-apply’

Tsukeru is combined with nuru ‘smear’, which is an alternating verb (type 11). The

following examples illustrate.

Mary wa kabe ni penki o nut-ta.

(170a) Mary TOP wall ALL paint ACC smear-PAST
Mary smeared paint on the wall. (V1—L-linking)

Mary wa kabe ni penki o tsuke-ta.

(170b) Mary TOP wall ALL paint ACC apply/attach-PAST
Mary put paint on the wall31. (V2—L-linking)

Mary wa kabe ni penki o nuri-tsuke-ta.

(170c) Mary TOP wall ALL paint ACC smear-apply-PAST
Mary smeared paint on the wall. (Compound—L-linking)

Mary wa kabe o penki de nut-ta.

(170d) Mary TOP wall ACC paint INST smear-PAST
Mary smeared the wall with paint. (V1—I-linking)

When tsukeru ‘apply/attach’ and nuru ‘smear’ are compounded, the resulting

compound only has L-linking even though nuru has the two linking patterns. This is

an example of the right-headedness of compounds. Both tsukeru and nuru have an

identical aspectual contour as that represented in Figure 44. (It is represented as the

achievement construal.)

31
This sounds as if Mary dirties the wall with paint without having the purpose of covering it.

253
 Chapter 5 Verbs of putting

Figure 44. Semantic structure for L-linking of nuri-tsukeru ‘smear-apply’ in (170)

V1 V2

wall wall
exist exist
ni ni
be located be located

paint nuru paint tsukeru

‘smear’ ‘apply,attach’

cause become
Mary apply Mary located
Δ Δ

t t

Compound

wall exist
ni
be located
by being applied
paint
nuri-tsukeru nuru tsukeru
‘smear-apply’

Mary cause become

located by applying
Δ

254
 Chapter 5 Verbs of putting

The two verbs profile the same part of the semantic representation. V1 simply

specifies how the event denoted by V2 is carried out since the semantic structure of

V1 instantiates that of V2.

5.4.17.5.2. Pure means compounds: Nui-tsukeru ‘sew-attach’

Let us see the case of nui-tsukeru ‘sew-attach’, for which some examples follow:

L-linking
Haha wa yunifoomu o nut-ta.
(171a) mother TOP uniform ACC sew-PAST
Mother sewed the/a uniform. (V1)

Haha wa bajji o yunifoomu ni tsuke-ta.

(171b)
mother TOP badge ACC uniform ALL attach-PAST
Mother attached the badge to the uniform. (V2—L-linking)

Haha wa bajji o yunifoomu ni nui-tsuke-ta.

(171c) mother TOP badge ACC uniform ALL sew-attach-PAST
Mother sewed the badge onto the uniform. (Compound—L-linking)

*Haha wa bajji o nut-ta.

(171d) mother TOP badge ACC sew-PAST
*Mother sewed the badge. (V1)

Nuu ‘sew’ does not denote any caused-motion event in itself. It does not allow L-

linking.

*Haha wa bajji o yunifoomu ni nut-ta.

(171e) mother TOP badge ACC uniform ALL sew-PAST
Mother sewed the badge onto the uniform.

It has two senses; the creation of clothing by sewing and mending by sewing32. Both

senses conventionally construe an event as an accomplishment and entail the activity

construal on the agent.

32
Croft points out (pers.comm.) that there are three kinds of transitive meaning of sewing in English as
well as the intransitive meaning. As the intransitive, ‘sewing’ refers to a general activity of sewing and
as the transitive, it has the meanings of (A) creation of clothing by sewing, (B) mending the defect by
sewing, and (C) attachment by sewing. The following are examples:
intransitive (general activity of sewing):
(i) What is she doing?
255
 Chapter 5 Verbs of putting

(172a) creation sense:

Haha wa atarashii yunifoomu o nut-te-iru.
mother TOP new uniform ACC sew-TE-IRU
Mother is sewing a new uniform.

(172b) mending sense:

Haha/isha wa hokorobi/kizuguchi o nut-te-iru.
mother/doctor TOP open.seam/cut ACC sew-TE-IRU
Mother is mending an open seam by sewing./Doctor is sewing up the cut.

I propose a semantic representation of nui-tsukeru ‘sew-attach’ as in Figure 45.

—I think she is sewing.

transitive (creation by sewing):

(ii) She is sewing a new dress for the Christmas party.

transitive (mending by sewing):

(iii) She is sewing her torn skirt.

transitive (attachment by sewing):

(iv) She is sewing the badge on the uniform.

On the other hand, Japanese sew has only two uses; creation and mending, which are equivalent to (ii)
and (iii), respectively. Japanese uses a different phrase and construction for the intransitive use of
sewing; nui-mono o suru (sewing-stuff/thing ACC do). Moreover, nuu ‘sew’ cannot be a three-
argument verb as in English (iv). We need it to be compounded with tsukeru ‘attach’ to refer to
attachment by sewing (see example (171c)). Nui-tsukeru cannot be replaced by nuu (see example
(171e)).

256
 Chapter 5 Verbs of putting

Figure 45. Semantic structure of L-linking of nui-tsukeru ‘sew-attach’ in (171)

exist
uniform ni
created be located

uniform nuu ‘sew’ badge tsukeru

‘attach’

sewing mother cause

mother activity become located
Δ Δ

t t

uniform exist
ni
be located
by being sewn
badge nui-tsukeru
‘sew-attach’ tsukeru
nuu
cause become
located by sewing
Mother activity

257
 Chapter 5 Verbs of putting

The semantic representations of tsukeru (V2) and nui-tsukeru (compound) are

accomplishment version of the caused-motion event. Nuu (V1) is represented as

having accomplishment construal of the creation sense. Actually, it does not matter for

the current purpose whether one uses either the representation of the creation sense or

the mending sense as these senses are not directly reflected in the compound version.

What is important is that the verb is a two-argument verb and its subject has the

activity aspectual contour33. Actually, even what nuu ‘sew’ selects as direct object is

semantically different from what the compound selects as direct object (of (171a) and

(171d)). (This happens to other compounds as well cf. oshi-komu ‘push-put.into’). In

this way, the agent’s sewing activity counts in the representation of the compound,

where sewing specifies the way of attachment. Nuu ‘sew’ is a two-argument verb and

the act of sewing is carried upon the badge in the compound, so I represent the verb

profile of nuu ‘sew’ as from the agent to the direct object in the model of the

compound.

5.4.18. Type 18 Ooi-kakusu ‘cover-hide’ verbs: means compounds 3

5.4.18.1. Members

The following are the members of this verb type:

ooi-kakusu (cover-hide)
ooi-tsutsumu (cover-wrap)

5.4.18.2. Semantics

These verbs are compounds whose first part specifies the means by which the action

of change of location denoted by the second part occurrs as in other means compounds.

V2 components are verbs of putting that take I-linking.

33
It may be more appropriate to represent nuu as a simple intransitive activity though it does have this
use as a full verb.

258
 Chapter 5 Verbs of putting

5.4.18.3. Syntactic patterns

This type takes I-linking and not L-linking. The compound in the following example

is more likely to be used to cover non-spatial things such as the truth, or bribery

scandals, in which case, GA-linking is used. (The hidden entity is realised as

accusative).

I-linking
John wa hon no hyoushi o shinbunshi de ooi-kakushi-ta.
(173) John TOP book GEN cover ACC newspaper INST cover-hide-PAST
John hid the cover of the book with newspaper.

*L-linking
*John wa hon no hyoushi ni shinbunshi o ooi-kakushi-ta.
(174) John TOP book GEN cover ALL newspaper ACC cover-hide-PAST
John hid the cover of the book into newspaper.

GA-linking
Seihu wa jijitsu o ooi-kakushi-ta.
(175) government TOP truth ACC cover-hide-PAST
The government hid the truth.

5.4.18.4. Causal and aspectual patterns

The causal pattern is that the agent acts on the figure to change the location to the

ground. Not unlike with other covering verbs, the ground is construed to undergo

changes by the figure being located in it and also the conventional aspectual construal

of the verb is accomplishment. Concerning ooi-kakusu, oou ‘cover’ has the I-linking

and kakusu ‘hide’ itself has L-linking (‘putting in(to) sense’) and I-linking (‘cover

sense’) as discussed in type 12. Oou ‘cover’ as the means part specifies which sense of

kakusu ‘hide’ is compatible with it. In this case, of course, the cover sense is

approriate, so kakusu also has I-linking34.

34
Note that it is not that the V1 verb selects the argument linking of V2. It only specifies the means. As
we see, kakusu ‘hide’ has two meanings with two different construals according to the two linking
patterns. Oou ‘cover’ simply limit one of the meanings, which happens to have a certain linking pattern.
It is strange to say ‘putting an entity into something by covering it’, but it is not impossible to say
259
 Chapter 5 Verbs of putting

5.4.18.5. Semantic representations

The following are the example sentences with V1, V2, and the resultant compound

ooi-kakusu ‘cover-hide’.

John wa hon no hyoushi o shinbunshi de oot-ta.

(176a) John TOP book GEN cover ACC newspaper INST cover-PAST
John covered the cover of the book with a newspaper. (V1—I-linking)

John wa hon no hyoushi o shinbunshi de kakushi-ta.

(176b) John TOP book GEN cover ACC newspaper INST hide-PAST
John hid the cover of the book with a newspaper. (V2—I-linking)

John wa hon no hyoushi o shinbunshi de ooi-kakushi-ta.

(176c) John TOP book GEN cover ACC newspaper INST cover-hide-PAST
John hid the cover of the book with a newspaper. (Compound—I-linking)

The causal-aspectual analysis of ooi-kakusu ‘cover-hide’ is represented in Figure 46.

This is a middle compound, where V1 and V2 have the same argument structure. V1

specifies the way the action denoted by V2 is carried out, as we can see that the sense

that kakusu has is specified as the ‘cover’ sense by the V1. I represent the structure as

accomplishment, as example (176c) can get the directed activity reading with te-iru

though I do not give an example sentence here.

‘covering something by covering it’.

260
 Chapter 5 Verbs of putting

Figure 46. Semantic structure of I-linking of ooi-kakusu ‘cover-hide’ in (176)

V1 V2

covered hidden
cv. of cv. of
book book kakusu
oou ‘hide’
news- ‘cover’ news-
paper move paper move

John John
Δ apply Δ apply

t t

Compound

hidden-by-
covering
cv of
book ooi-kakusu
‘cover-hide’
oou kakusu
news- move
paper

John apply
Δ

261
 Chapter 5 Verbs of putting

5.5. Digression on compounds

We have seen the integration of the semantic structures of V1 and V2 into compounds

in the above analysis. I am now going to provide a tentative cognitive and

constructional way of analysing the integration of compounds. This is a detailed

version of the discussion entered into in §5.4.16.

Means compound verbs with type 2 verbs as V2 are equivalent to the English Path

construction (Croft 2000: 85, §4.2) or the caused-motion construction (Goldberg

1995). Actually, type 2 verbs and also the productive V2 verb, tsukeru ‘apply, attach’,

behave like English prepositions. The following are the proposed syntactic frames of

into in English and of ~komu ‘put into’ in Japanese.

(177) English: [SBJ Verb OBJ into OBL]

ex. John pushed/pressed many clothes into his bag.

Japanese: [SBJ wa/ga OBJ o AdpOBJ ni V1 ~komu]

John wa ooku no ihuku o kaban ni oshi-kon-da.

ex. John TOP many GEN clothes ACC bag ALL push-put.into-PAST
John pushed many clothes into his bag.

Other spatial prepositions such as out of or away from can replace into in English, and

other verbs of caused-motion can replace ~komu with a possible alternation of the ni

part with kara ‘from’ depending on the verbs in Japanese. The ‘Verb’ in English

corresponds to the V1 verb in Japanese and into corresponds to ~komu ‘put into’ with

its spatial adpositional particle ni. Various types of verbs can appear as the ‘Verb’ in

English and in ‘V1’ verbs in Japanese. In brief, English ‘into OBL’ corresponds to

‘OBL ni ~komu’ in Japanese. Since the English caused-motion verbs have been

discussed extensively, I limit myself to discussing Japanese compounds with ~komu

as an example.

262
 Chapter 5 Verbs of putting

Komu ‘put into’ as the V2 component takes various kinds of verbs as V1. The

examples we have seen are verbs of putting, verbs of exerting force, and verbs of

creation. Others are verbs of throwing, sending and carrying, verbs of contact by

impact, or activity verbs though I have not presented all of them here because of the

limitation of the space. This means that verbs with different argument structures or

different valency, or different causal structures can appear as V1. The argument

structure of means compounds is inherited from V2 no matter what argument structure

V1 gets35.

The right-headedness of the argument structure and the semantic fact that V1

specifies the way V2 is carried out in means compounds in the earlier discussion are

good reasons for regarding V2 as the construction where V1 is realised. We can call

‘~komu construction’, and also can use ‘the Path construction’ or ‘caused-motion

construction’ to cover other V2 verbs which entail caused-motion (such as ireru

‘put.into’ or ageru ‘raise/lift’). Also the construction can be instantiated by V1 by

being superimposed onto it. In §5.4.16., I mentioned that V2 is an e-site in

Langacker’s terminology, which the structure of V1 elaborates. I now present a more

abstract schematic representation36 of ~komu based on the convention in Croft (2000,

§3 and 4).

35
I do not mean that any kind of verb can be V1. There should be some syntactic or semantic restriction
as well. I leave the descriptions of constraints for further study. At least, there is a fundamental
constraint which follows the principle of transitivity harmony (Kageyama 1993: 116-126). That is, a V1
which is compatible with ~komu or other verbs that entail the caused-motion sense should be a
transitive verb.
36
It is schematic in that it is abstracted away from extent on both t and Δ (See Croft 2000, §3.4).

263
 Chapter 5 Verbs of putting

Figure 47. tentative analysis of semantic structure for V2 ~komu as construction

exist
S.Obl
ni
be located in /travel
by ‘V1’
Obj
V1-komu

Sbj cause move/located

by ‘V1’
Δ

I leave the representation as schematic as possible. The aspectual contour of the

subject is undirected and it goes back to the rest state after the event. The schematic

striped box accommodates process (both punctual and extended). This is the same for

the contour of the direct object. However, it undergoes the directed achievement after

the event. There is only one arrow pointing from the subject to object indicating the

causation. The two schematic striped boxes and one pointed arrow indicate that the

event can be both extended causation and punctual causation. (In the case of extended

causation, another arrow is needed between the right ends of the two striped boxes.)

The grey vertical line to the contour of oblique means that the direct object ends up in

a non-causal relation with the oblique. (In the case of extended causation, the non-

causal relation should also be extended and indicated by other grey lines.) As a whole,

it allows the achievement construal, the accomplishment construal, and the runup

achievement construal. The dotted thick box is an e-site for V1 verb profiles. That is,

the box is the minimum requirement for V1 indicating that V1 verbs should be

264
 Chapter 5 Verbs of putting

transitive (cf. the principle of transitivity harmony of compounds in Kageyama 1993:

116-126).

In the case of V1 verbs with a different argument structure from that of V2 as in

pure means compound types, their semantic structures can be altered by

superimposition including a semantic restriction or an aspectual contour on the direct

object. As Langacker (1987: 305) mentions, there is a construal operation imposed by

the semantic structure of V1 which elaborates the V2 construction whose substructure

it comes to share.

The schematic constructional approach of ~komu can be valid with other V2 verbs

of type 2 and at least, tsukeru. They are so productive and frequently used, that we

could regard them as the grammatical constructions. This approach, actually,

presupposes the right-headedness of compounds since the semantic structure

(including the causal structure that is related to argument linking) of V2 is taken as the

construction itself. This means that the argument structure of V2 is a determinant

factor in the argument linking of compounds. If I extend this idea and specify the

semantic structures of every V2 verb as constructions, it becomes more apparent that

there is not even a clear distinction between the pair compounds and the middle

compounds above. In middle compounds, V1 and V2 are just not semantically close

enough to form pair compounds though their argument linking structure is identical

and their arguments are referentially identical. Even the V1 of middle compounds

normally entail the caused-motion sense itself. Actually, I have found it difficult to

categorise some means compound verbs in my data because sometimes the distinction

depends on the intuition about whether V1 has additional, specified, or subtly different

meanings from V2. In that way, the definition of pair compounds of Matsumoto

(1996: 202) and Kageyama (1993: 99) are not well articulated; the former only states

265
 Chapter 5 Verbs of putting

that ‘pair compounds are composed of two component verbs that are identical in

argument structure’ and the latter says that ‘two elements whose contents are

equivalent are combined’. As I mentioned in relation to pair compounds of type 6,

there are always subtle semantic differences, including semantic restrictions on

arguments between the two different verbs (even though they are close enough to be

recognised as synonymous), from the descriptive usage of JSD or GJD. That is, there

is a gradient scale between pair compounds and what I call middle compounds in

terms of the semantic closeness of V1 to V2.

Finally, at the opposite end of the constructional approach to compounds, there are

frozen/idiomatic compounds, where the composite structure designates a totally

different entity from those of each component strucutre.

I only presented the possibility of the constructional analysis of Japanese

compound verbs here. Close examination of each V2 verb will be left for the future

study.

5.6. Summary

In this chapter, I have classified verbs of putting in Japanese into eighteen categories,

in terms of syntactic behaviour, mainly by argument linking patterns. I present a brief

table describing each category in Appendix C. The table shows (1) the class number

(CLS), and (2) the subclass (Scls) number where there is one, (3) the name of the

types, (4) the number of members, (5) syntactic patterns, (6) the acceptability of

source phrase, (7) the other syntactic patterns they manifest, (8) aspectual patterns,

and (9) other prominent characterstics. For (5) syntactic patterns, there is an FA-type

column (L-linking, FA-linking), a GA-type column (I-linking, GA-linking), a GO-

linking; ‘FA-type’ indicates whether the verbs in question take FA-type linking or not,

266
 Chapter 5 Verbs of putting

‘GA-type’ indicates if they take GA-type linking or not. ‘Source’ indicates if verbs a

source phrase instead of a goal phrase, and ‘Other syntactic patterns’ specifies other

special syntactic patterns they take. In ‘aspect’, I only specify if verbs favour a

‘punctual’ or ‘extended’ construal conventionally, as situations and contexts allow a

variety of aspectual patterns. When they get the activity in progress reading with te-iru

(except as the runup achievement reading), I specify them as ‘extended’ and if not,

then I specify them as ‘punctual’. With some types of verbs which aspectual pattern is

more conventional is very ambiguous; verbs of the same category can have different

patterns or the same verb shows an ambiguity. In that case, I put both ‘punctual’ and

‘extended’.

Appendix D indicates what verbs belong to what class.

5.6.1. Locative alternation

Japanese has a limited number of verbs of putting which manifest the locative

alternation. These include four simple verbs of type 11 decorate verbs (four simple

verbs plus two compound verbs that are related to simple verbs), type 12 wrap verbs

(four verbs), and type 15 afforest verbs (three verbs). Type 12 verbs manifest two

linking patterns because they have two different figure and ground construal. In type

11 and type 15, figure and ground alignment is identical in L-linking and I-linking.

The subtle semantic differences between the two syntactic patterns discussed in the

relevant sections are captured in the differences of semantic structures including

verbal profile. Most of type 11 and type 15 verbs relate to events of filling and

covering and have an accomplishment construal. A verbal scale is associated with the

figure in L-linking and with the ground in I-linking in accomplishment construals.

Also, a ‘trivial verbal scale’ in the achievement reading is associated with the figure in

267
 Chapter 5 Verbs of putting

L-linking and with the ground in I-linking. In both cases, the ground is construed as a

mere reference point where the figure is located or moves to in L-linking and as a

participant undergoing a change of state in I-linking. Thus, the holistic interpretation

of the ground in the I-linking in Japanese (with-variant in English) is well illustrated in

the semantic representation of verbs.

5.6.2. Semantic structures of each argument linking construction

In the discussion of alternating verbs, I have shown that these verbs get different

senses in the two linking patterns. That is, specific construction contribute different

construals of the events denoted by particular verbs, though if a certain verb can occur

in a construction is a verb-specific conventional property. This conforms to the basic

idea of cognitive linguistics that the constructions also have semantic import.

Therefore, it is necessary to give a semantic representation for each linking

construction, after having seen a lot of example Figures. The following are schematic

representations of each construction.

268
 Chapter 5 Verbs of putting

5.6.2.1. L-linking construction

The semantic structure of L-linking is shown as follows:

Figure 48. Semantic structure of L-linking

S.Obl exist

ni particle

Obj be with
Verb (by moving)

Sbj cause
Δ

This is the abstract representation minimising the information to accommodate

various occurences in actual examples. The striped boxes for Obj and Sbj

accommodate process (activity and achievement). The single arrow from Sbj to Obj

indicates that the causal relation holds between the two; the subject applies some force

on the direct object. The grey line indicates that there is a non-causal relation between

the direct object and the adpositional oblique, at least, which is true at a point in time.

The adpositional object is treated as a reference point. This abstract representation

allows the achievement construal (as in Figure 2), the accomplishment construal (as in

Figure 3), the undirected activity construal (as in Figure 28), and also allows the

possibility of the runup achievement construal though I have not illustrated one of

these in the Figures so far. The grey part is profiled by the subsequent oblique marker

ni and the thick dotted box is profiled by the verb which elaborates the construction.

269
 Chapter 5 Verbs of putting

The labelling for each contour is also very abstract. The subject only causes the

direct object to be in spatial contact with the ground. The ways of causing can be

various. The direct object ends up by being together with the subsequent oblique,

which in turn serves as a mere reference point. The L-linking is predominant in verbs

of putting in Japanese and is even used for verbs of filling.

5.6.2.2. I-linking construction

The following is the semantic structure of I-linking:

Figure 49. Semantic structure of I-linking

be in a certain state
Obj by being located with A.Obl

de particle
A.Obl move/be located

Verb

Sbj cause
Δ

This is again an abstract representation. The striped boxes for each contour allow

different aspectual construal for verbs with this construction. The thick arrows

indicate that the subject acts on the adpositional object, which, in turn, acts on the

direct object. What is definite with this construction is that the direct object undergoes

the change of state at the end of the event, which is indicated by the broken line on the

right top part of the aspectual contour of the direct object. The thick dotted box is an

270
 Chapter 5 Verbs of putting

e-site for the verbal profile. The adpositional profile (indicated by the grey dotted box)

overlaps with that of the verbal profile. It is represented by the adpositional object

contour plus the arrow from it to the direct object. (cf. English with of Croft 2000: 65,

Figure 12, §3.4). It accommodates the achievement construal as in Figure 19 and the

accomplishment construal as in Figure 20.

The labelling for the subevent of the direct object shows that it undergoes a certain

‘unspecific’ change. For the antecedent oblique, there are two choices: move or be

located. The labelling of the subject remains abstract as “cause”.

Verbs which appear with I-linking are verbs of covering, blocking, and filling, and

“change-of-state” verbs.

5.6.2.3. FA-linking construction

As the main focus in the thesis is three-argument linking, I have not said a lot about

the two-argument linking. The following, however, is a tentative illustration of the

semantic structure for FA-linking.

Figure 50. Semantic structure of FA-linking with verbs of putting

move or
be located
Obj

Verb

Sbj cause
Δ

271
 Chapter 5 Verbs of putting

The striped boxes indicate process and the single arrow indicates the force-dynamic

relation between the subject and direct object. After the event, the subject goes back to

the rest state, while it is not clear whether the direct object undergoes the directed

process or goes back to the rest state. The schematic representation accommodates the

achievement construal (as in Figure 6) and the undirected activity construal (as in

Figure 13).

5.6.2.4. GA-linking construction

Schematically, GA-linking is similar to FA-linking except for some minor points.

Figure 51. Semantic structure of GA-linking with verbs of putting

be in a certain state

Obj

Verb

Sbj cause
Δ

The only schematic difference from FA-linking is that the direct object undergoes a

change of state in its subevent. Correspondingly, the labelling for its subevent is also

different. The schematic representation accommodates the achievement construal as in

Figure 21 and the accomplishment construal as in Figure 24.

272
 Chapter 5 Verbs of putting

5.6.2.5. GO-linking construction

GO-linking appears only in figure incorporation types (type 14 and type 15). The

following is the semantic structure of GO-linking:

Figure 52. Semantic structure of GO-linking

exist

S.Obl ni particle

Verb

Sbj cause move entity(something)

Δ to

The subject’s act is a process of causing a certain entity to move. Whether it

undergoes the directed transition after the process is not known. The entity assigned to

the subsequent oblique is a mere location where the entity is moving to and located

with. It accommodates the undirected activity construal as in Figure 36 and the

accomplishment construal as in Figure 39.

5.6.3. Comparison with English

Unlike English, Japanese does not have a lot of verbs of putting with GA-type

linking. This also leads to the drastically small number of locative alternation verbs.

The three simple verbs of type 11 (except kazaru ‘decorate’) and the three figure

incorporation verbs which manifest alternation (type 15) refer to events of covering

and filling whose conventional aspectual pattern is accomplishment. Only kazaru is

different and the English equivalent to this verb ‘decorate’ does not alternate. All

273
 Chapter 5 Verbs of putting

these verbs with alternation are explained to have different senses in the two (L and I)

linking constructions which are caused by two construals. They differ in relation to

which of the participants the verbal scale is associated with. When it is assigned to the

figure, L-linking is realised. When it is assigned to the ground contour, it takes I-

linking.

The tendency of Japanese toward L-linking is shown in figure incorporation type 1

(type 14). Even in the complete incorporation pattern, verbs of this type assign oblique

case to a sole argument (ground) other than the agent. On the other hand, English

verbs of figure incorporation type take GA-linking equivalent (the ground is realised

as Accusative).

In Japanese, there are also verbs of filling, covering, and surrounding that takes L-

linking only; haru ‘fill (pour to the full)’, tataeru ‘fill’, kabuseru ‘cover (put over)’,

and megurasu ‘surround (put around)’. All these classes of verbs have the with-variant

or the locative alternation (with the locative variant) in English. That is, even verbs

which semantically refer to the event of causing a change of state on the part of the

ground can be realised in L-linking in Japanese.

Another interesting comparison is that Japanese is rich in compounds and makes a

lot of caused-motion compounds. As I discussed above, the Japanese caused-motion

verbs with specified direction (type 16) are productive and roughly correspond to

English locative prepositions in their use. Therefore, I categorises Japanese caused-

motion verbs according to the various kinds of V2 verb used in compounds. English,

on the other hand, categorises them differently according to main verbs which

designate means (cf. Levin (1993)). I have also proposed the way to analyse the

structure of Japanese compound verbs using Langacker’s valence relation.

274
 Chapter 5 Verbs of putting

Finally, in collecting Japanese equivalents of English verbs of putting, verbs of

other classes were found. There are some recurring classes; not suprisingly they are

three-argument verbs37. This suggests that even boundaries between classes of three-

argument verbs may vary between English and Japanese. This requires me to look into

many more classes of verbs, but I shall leave this for future study.

37
These verbs are represented in the appendix of verbs of putting as Type 20-1 to Type 20-6,
respectively in Appendix D.

275
6 VERBS OF REMOVING

6.1. Introduction

This chapter discusses verbs of removing, which are semantically the opposite of

verbs of putting; they relate to removing an entity from a location. However, they do

not always show opposite syntactic behaviour because of the asymmetry in the world

knowledge of putting and removing, which will be discussed later. Verbs of removing

in English are classified in chapter of 10 in Levin’s (1993) book. For her detailed

classification, see Appendix E.

6.2. Basic linking patterns

The three arguments of verbs of removing are the agent (A), the figure (F), and the

ground (G) as in the case of verbs of putting; in verbs of removing, the figure

basically refers to an entity, a person, a possession, or an inalienable part of something

which is removed and the ground refers to a location, a place or a possessor from

which the figure is removed. There are two main argument linking types in Japanese

verbs of removing. These are the figure-accusative (FA) type and the ground-

accusative (GA) type depending on which of the arguments appears as the direct

object.

FA-type constructions (3 linking patterns):

A-linking construction (Ablative-linking):

figure—o (ACC), ground—kara (OBL)

Oliver wa zubon kara yogore o otoshi-ta

(1) Oliver TOP trousers ABL stain ACC remove-PAST
Oliver removed the stain from his trousers.

This is similar to the L-linking of verbs of putting in that the figure is the direct object

and the ground is realised as the oblique. Ablative kara is used instead of Allative ni. I

276
 Chapter 6 Verbs of removing

call this A(blative)-linking1.

Another type of FA type linking is G(enitive)-linking.

G-linking construction:
ground—no (GEN), figure—o (ACC)

Graham wa ringo no kawa o mui-ta

(2) Graham TOP apple GEN skin ACC peel-PAST
Graham pared an apple.

The figure is marked as Accusative and the ground is marked with Genitive, which

modifies the figure2.

Finally, FA-linking marks the figure as accusative as the only syntactic argument

other than the subject.

FA-linking construction:
figure—o (ACC)

John wa otsuri o gomakashi-ta

(3) John TOP change ACC pocket-PAST
John pocketed the change.

GA-type construction (ground-accusative):

Japanese is said to lack an equivalent construction to the English of-variant of the

locative alternation as in Doug cleared the table of dishes3. Croft (2000: 85, 4.2)

calls this the antecedent oblique Of construction. When the ground is the direct object

in verbs of removing, it is the only argument other than the subject.

1
In spatial removal or motions, kara can be replaced by yori ‘from’.
Jacqui wa tori no su yori tamago o tot-ta.
(i)
Jacqui TOP bird GEN nest ABL egg ACC steal-PAST
Jacqui stole eggs from the birds’ nest.
2
Nihongo Goi Taikei [Japanese Vocabulary Compendium] lists this linking as an independent
construction.
3
I shall argue later that there is Reversed A-linking in Japanese and this will be equivalent to this
English construction.

277
 Chapter 6 Verbs of removing

GA-linking construction
ground—o (ACC), figure—omitted

Taro wa teeburu o katazuke-ta

(4) Taro TOP table ACC clear-PAST
Taro cleared the table.

Taroo wa zubon o arat-ta.

(5) Taroo TOP trousers ACC wash-PAST
Taroo washed the trousers.

In the above examples, both the trousers and the table are the ‘ground’ in that they are

the location from which something is removed. This ‘something’ is the unexplicit

‘figure’ which is implied in the verbs’ lexical meanings.

6.3. Japanese Genitive no and the semantic relation between figure and ground

The compatibility of G-linking with verbs is related to the semantic relation between

the figure and the ground since the Genitive marker is the one which links two NPs.

However, this does not mean that the verbs’ meaning is not related to the linking at all.

It is ‘indirectly’ related in that a certain event denoted by a verb is likely to set

constraints on the semantic properties of its arguments. For example, the situation

denoted by muku ‘peel’ entails removing the ‘covering entity’ in a whole-part relation

such as the skin of an apple. The relation between skin (the figure and a part of the

ground) and apple (the ground) fits in to the core use of Genitive no in Japanese: ‘A

modifies B and indicates a specific member(s) of B among all the members of B’

(Makino 1986: 314) in A no B. The skin of an apple is a ‘member' of an apple.

Therefore, we can predict that muku ‘peel’ will get the Genitive linking.

The combination of a noun plus Genitive can modify nouns quite extensively.

When we have ‘A (noun) no (GEN) B (noun),’ the relations that can hold between the

278
 Chapter 6 Verbs of removing

two nouns, A and B, can be as follows:

(6) some common uses of Genitive no (A no B):

(A) A is the possessor of B.
ex. watashi no hon (I GEN book=my book)
(B) A is the location where B exists.
ex. tana no hon (shelf GEN book= a book on the shelf)
(C) B is about/on A.
ex. rekishi no hon (history GEN book= a book on history)
(D) A is a specific kind of B.
ex. haadokabaa no hon (hardcover GEN book = a hardcover book)
(E) A is the object and B is the subject.
ex. nihongo no sensei (Japanese GEN teacher = teacher of Japanese)
(F) A created B.
ex. Picasso no e (Picasso GEN painting = painting of Picasso)
(G) A is an attribute of B.
ex. shachou no Tanaka san (president GEN Mr. Tanaka)
(H) B is made of /from A.
ex. Hitsuji no miruku (sheep GEN milk = sheep’s milk)

(A Dictionary of Basic Japanese Grammar 1986: 314)

Unlike verbs of putting where the figure and ground are entities of separate

existence, those in verbs of removing show various ways in which the two entities are

‘attached’ before one is taken away. As some verbs lexically specify the semantic

relation of the figure and the ground and this is also related to the use of G-linking, I

shall propose some semantic types that the figure and ground hold in verbs of

removing. The relations between the figure and the ground may be locational, a

whole-part relation, possessional or identificational. The conditions applying to these

are outlined in the following:

(7) The relations between the figure and the ground

1. Locational
Figure and ground are separate entities. The ground is a location where the figure
exists. In this relation, we can say ‘figure wa ground ni aru/iru’ (figure TOP
ground LOC exist). The following are some subtypes of locational relations.

1.1 Pure locational

As the term shows, the relation is purely locational. Examples of this are ‘weed
and lawn’, ‘water and bucket’ or ‘John and the house’.

279
 Chapter 6 Verbs of removing

1.2 Impurities locational

The figure is an unfavorable or unnecessary thing that should be removed. It
does not normally have an ‘independent’ status of existence normally. ‘Dust on
the table’ or a ‘stain on the shirt’ will diminish after being removed.

1.3 Attachment
This is a marginal case. The degree of closeness between the figure and the
ground is between a ‘pure location’ and ‘a whole-part relation’, which will be
discussed next. ‘The wallpaper on the wall’ can be thought to be attached to the
wall forever unless a certain external force is applied though the two were
originally independent entities. It is awkward to say that ‘the wallpaper exists
on the wall’ in Japanese.

1.4 Extended locational

The notion of location can have the extended sense of referring to more
abstract relations such as ‘John in Lloyds Bank’ or ‘Steve in the dance society’.
The places do not necessarily indicate that the person is in these places, rather
they refer to organisations or groups that people belong to.

2 Whole-part relation
This relation is exemplified by ‘skin and apple’ or ‘juice and lemon’. The figure is
a part of the ground and the two are fully integrated as one entity. (Naturally, there
are not apples without skin nor lemons without juice).

3 Possessional
This refers to the relation of possessor and possessum as in ‘John and his car’ or
‘Jacqui and her money’. This relation can be extended to a more abstract
‘possessional’ notion such as ‘John and his job’ or ‘Jacqui and her duty’ (extended
possessional).

4 Identificational
This refers to the relation between a person and his/her identification such as ‘Dr.
Tanaka as a professor’ and ‘Mr. Woodall as the president of the company’.

I shall use these terms in discussing verbs of removing.

Thus, the grammaticality of G-linking with a certain verb of removing should be

measured by the compatibility between the semantic structure of the G-linking

construction and that of the verb and how the semantic relation between figure and

ground which are required by the verb fits into the descriptive use of the Genitive in

the examples in (6).

280
 Chapter 6 Verbs of removing

6.4. Classification

As in the previous chapter, verbs of removing in Japanese are classified and discussed

in terms of members, semantics, syntactic patterns, causal and aspectual patterns, and

semantic representation. Concerning syntactic patterns, the main concern is whether

verbs take FA-type linking or GA-type linking or both. Some verbs of a same class

may behave slightly differently regarding G-linking or FA-linking. In this case, I may

subcategorise them as different subtypes of the same class. A table of the

classifications and a list of verbs of removing are attached in Appendices F and G.

6.4.1. Type 1 Tori-nozoku ‘remove’ verbs

6.4.1.1. Members

The following are some members of this verb type:

subtype 1: tori-nozoku ‘remove’ type

tori-nozoku ‘remove’, nozoku 'remove', toru ‘remove, take’, nuku ‘remove’,
harau ‘remove’, hazusu ‘detach’, tep.pai-suru ‘abolish’,
boku.metsu-suru ‘exterminate’, habuku ‘omit’,
kesu ‘delete’, mas.shou-suru ‘delete, cross out’

subtype 2: hiku ‘subtract’ type

hiku ‘subtract’, sashi-hiku ‘deduct’, kou.jo-suru ‘subtract’

6.4.1.2. Semantics

Verbs of this type are the general ‘removing’ type. They indicate removing an entity

away from another, taking an entity away from another entity, a location, or a person,

or eliminating, deleting, subtracting, omitting an entity from something. Some verbs

(such as nozoku ‘remove’ or toru ‘remove’) denote ‘abstract’ removing situations like

removing feelings of worry/pain from somebody’s heart. The figure is normally an

unnecessary or unfavorable thing. Some of the simple verbs of this category are

extensively used as V2 components of compound verbs of removing (cf. 6.4.18).

281
 Chapter 6 Verbs of removing

6.4.1.3. Syntactic patterns

Type 1 verbs take A-linking ((8)-(10)) but not GA-linking (17). Depending on

contexts (specifically, on the semantic relations between the figure and the ground),

G-linking is either acceptable or less acceptable. ‘Locational’ relation, ‘whole-part’

relation, ‘attachment’, ‘possessional’ relations are more acceptable in G-linking. Most

verbs of this category may appear in FA-linking especially when the figure is of the

impurities locational type (even in an extended abstract sense such as ‘remove

violence from the town’) and can disappear after being removed. Only subtracting

verbs such as hiku ‘subtract’ strictly prohibit FA-linking (15) and G-linking (13),

which are the reasons these are categorised as another subtype. Examples of type 1

verbs with various linking patterns follow:

A-linking
Jacqui wa shibahu kara zassou o tori-nozoi-ta.
(8) Jacqui NOM lawn ABL weeds ACC remove-remove-PAST
Jacqui weeded the lawn.

Karera wa kaiin meibo kara John no namae o keshi-ta.

(9) they TOP membership list ABL John GEN name ACC delete-PAST
They eliminated John’s name from the membership list.

Kodomo wa hyaku kara juugo o hii-ta.

(10) child TOP one.hundred ABL fifteen ACC substract-PAST
The child subtracted fifteen from one hundred.

G-linking (depends)
Jacqui wa shibahu no zassou o tori-nozoi-ta.
(11) Jacqui TOP lawn GEN weeds ACC remove-remove-PAST
Jacqui weeded the lawn.

?Karera wa kaiin meibo no John no namae o keshi-ta.

(12) they TOP membership list GEN John GEN name ACC delete-PAST
They eliminated John’s name in the membership list.

*Kodomo wa hyaku no juugo o hii-ta.

(13) child TOP one.hundred GEN fifteen ACC subtract-PAST
The child substracted fifteen of one hundred.

282
 Chapter 6 Verbs of removing

FA-linking (depends)

Jacqui wa zassou o tori-nozoi-ta.

(14) Jacqui TOP weeds ACC remove-remove-PAST
Jacqui removed weeds.

*Kodomo wa juugo o hii-ta.

(15) child TOP fifteen ACC subtract-PAST
*The child substracted fifteen.

*goal phrase instead of source phrase

*Karera wa dokoka ni John no namae o keshi-ta.
(16) they TOP somewhere ALL John GEN name ACC delete-PAST
*They eliminated John’s name to somewhere.

*GA-linking
*Jacqui wa {zassou de} shibahu o tori-nozoi-ta.
(17) Jacqui TOP {weeds INST} lawn ACC remove-remove-PAST
*Jacqui removed the lawn of weeds.

For situations of physically/spatially removing something away, the adverbial phrase

kirei ni ‘clean RESULT,’ whose meaning of leaving no marks or dirt, may be added

and changes the situation from a directed activity to an accomplishment.

Jacqui wa shibahu kara kirei ni zassou o tori-nozoi-ta.

(18) Jacqui TOP lawn ABL clean RST weeds ACC remove-remove-PAST
Jacqui weeded the lawn completely.

6.4.1.4. Causal and aspectual patterns

Verbs of this type refer to an event of the agent’s acting on the figure and taking it

away from the ground. They vary aspectually, being either punctual (achievement) or

extended (directed activity, runup achievement or accomplishment). The removing of

weeds as in (8) is interpreted as directed activity or accomplishment as the

sukoshizutsu ‘little by little’ test (test for directed process) and the tochuumade

‘halfway’ test (test for accomplishment reading of event) in the following examples

show.

283
 Chapter 6 Verbs of removing

sukoshizutsu ‘little by little’ test

Jacqui wa shibahu kara zassou o sukoshizutsu tori-nozoi-te-iru.
(19) Jacqui TOP lawn ABL weeds ACC little.by.little remove-remove-TE-IRU
Jacqui was weeding the lawn little by little.

tochuumade ‘halfway’ test

Jacqui wa shibahu kara zassou o tochuumade kirei ni tori-nozoi-ta.
(20) Jacqui TOP lawn ABL weeds ACC halfway clean RST remove-remove-PAST
Jacqui weeded half the lawn.

Sentence (19) does not necessarily entail the event as an accomplishment even though

in the real context, we cannot keep on removing weeds from the lawn. The quantity of

weeds which exists in the lawn is already determined though it is not described in a

linguistic expression. However, the agent can still engage in the activity of moving

weeds away without causing an apparent change of state of the ground (the lawn). The

following sentence is better analysed as having a directed activity reading.

Jacqui wa shibahu kara zassou o tori-nozoi-ta ga mada sukoshi nokot-te-iru.

(21)Jacqui TOP lawn ABL weeds ACC remove but still some remain-TE-IRU
Jacqui removed weeds from the lawn, but there is still some remaining.

Other more ‘abstract’ events of removing/eliminating like eliminating a name from

a list are better construed as punctual as it is difficult to see intermediate stages of the

events.

6.4.1.5. Semantic representations

The semantic structures for the directed activity construal, the accomplishment

construal and the achievement construal of type 1 verbs are represented in Figures 1-4.

The spatial source (a location from which something is moved) is expressed in verbs

of removing instead of the goal as in verbs of putting. The agent applies force on the

figure, which is indicated by force-dynamic solid arrows, and the relations between

the figure and the ground are represented by the non-force-dynamic grey lines. In G-

linking, the number of participants is reduced to two as is the number of contours.

284
 Chapter 6 Verbs of removing

Figure 1. directed activity construal of A-linking of tori-nozoku ‘remove’ in (8)

lawn exist
kara
move away

weeds
tori-nozoku
‘remove’

Jacqui cause move away

Δ

Figure 2. accomplishment construal of A-linking of tori-nozoku ‘remove’ in (8) and

(20)

lawn exist
kara
eliminated

weeds
tori-nozoku
‘remove’

Jacqui cause move away

Δ

285
 Chapter 6 Verbs of removing

Figure 3. accomplishment construal of G-linking of tori-nozoku ‘remove’ in (11)

eliminated and separated

(move away)
weeds in
the lawn
tori-nozoku
‘remove’

Jacqui cause move away

Figure 4. achievement construal of A-linking of kesu ‘delete’ in (9)

membership
list exist
kara
eliminated

John’s name kesu ‘delete’

they eliminate
Δ

286
 Chapter 6 Verbs of removing

6.4.2. Type 2 Dasu ‘take out’ verbs: spatial caused-transfer verbs with a specified
direction

6.4.2.1. Members

The following are members of this verb type:

subtype 2-1(spatial direction):

dasu ‘take out’, modosu ‘return’, kaesu ‘return’
subtype 2-2 (removing of spatial transfer):
nokeru ‘put aside/away’, dokeru ‘put aside/away’

6.4.2.2. Semantics

Verbs of this type refer to a spatial transfer of an entity from one location to another.

The first subclass of the category is equivalent to type 2 verbs of putting which entail

the direction of caused-motion. Verbs can have an overt expression of goal as well as

source (kara). That is, they have L-linking (goal phrase only) or a path pattern (goal

phrase and source phrase). The second subclass refers to taking away/removing

something unfavorable or improper from the ground. These verbs do not specify

direction but are still better categorised as indicating spatial transfer because they can

take a goal phrase (with/without a source phrase) specifying where the figure ended

up after being taken away/removed from the original source. The relation between the

figure and the ground is locational.

6.4.2.3. Syntactic patterns

Type 2 verbs allow A-linking, and also a path pattern (with a goal phrase), but do not

take GA-linking. The following examples illustrate:

A-linking
John wa kobune o kishi kara dashi-ta.
(22) John TOP boat ACC shore ABL take.out-PAST
John took a boat off the shore.

287
 Chapter 6 Verbs of removing

Patrick wa heya kara isu o noke-ta.

(23) Patrick TOP room ABL chair ACC put.aside-PAST
Patrick put the chair away from the room.

with goal phrase (L-linking or Path pattern)

John wa kobune o {kishi kara} oki ni dashi-ta.
(24) John TOP boat ACC {shore ABL} offing ALL take.out-PAST
John took a boat {off the shore} to the offing.

Patrick wa {heya kara} ima ni isu o noke-ta.

(25) Patrick TOP {room ABL} living.room ALL chair ACC put.aside-PAST
Patrick put the chair away (from the room) to the living room.

*GA-linking
*John wa kishi o {kobune de} dashi-ta.
(26) John TOP shore ACC {boat INST} take.out-PAST
*John took out the shore with a boat.

*Patrick wa heya o {isu de} noke-ta.

(27) Patrick TOP room ACC {chair INST} put.aside-PAST
*Patrick put the room away with the chair.

?G-linking
?John wa kishi no kobune o dashi-ta.
(28) John TOP shore GEN boat ACC take.out-PAST
John took a boat on the shore out.

?Patrick wa heya no isu o noke-ta.

(29) Patrick TOP room GEN chair ACC put.aside-PAST
Patrick put the chair away in the room.

I have placed a question mark against the above sentences of G-linking as they are

“not necessarily” equivalent to what the A-linking of the verbs entails. Also they do

not sound perfectly felicitous. This is related to the use of the Genitive and also the

semantics of the verbs in this category. In the Genitive construction ‘A no B’, A can

modify B as the location of B exists but it can also modify B other than as a location.

For example, A can be the place where B originally exists or is supposed to exist as

one of the properties of B. Thus, it is possible to say the following, with A specifying

a characteristic of B and being used in the spatial transfer:

288
 Chapter 6 Verbs of removing

John wa heya no isu o ima kara niwa ni noke-ta.

(30) John TOP room GEN chair ACC living.room ABL garden ALL put.aside-PAST
John put the chair (in/for the room) away from the living room into the garden.

That is, the chair is characterised by the genitive phrase as one for the room (inside)

and not for outside, and so it is possible to have a source phrase that specifies where

the chair exists (temporarily). The first noun of the genitive phrase does not always

have to be the ground. This is why A-linking and G-linking do not necessarily mean

‘identical’ things in the example sentences. G-linking may entail what A-linking

means but not necessarily.

However, (28) and (29) are most likely to be interpreted as having the genitive

modifier as the ground, as these verbs very rarely accept FA-linking (of the FNI).

??/*FA-linking

*John wa kobune o dashi-ta.

(31) John TOP boat ACC take-out
*John take out the boat.

??Patrick wa isu o noke-ta.

(32) Patrick TOP chair ACC put.aside-PAST
??Patrick put aside the chair.

Since it requires either a goal phrase or a source phrase, when it is realised in G-

linking (without any other oblique phrases), the modifying noun of the Genitive

construction is plausibly interpreted as the ground. The complexity here is a good

indicator that the Genitive is required by the verb and the argument structure

construction.

6.4.2.4. Causal and aspectual patterns

The agent acts on the figure and causes it to move from the original location, which is

construed as the ground. The aspectual pattern for subtype 2-1 varies (see the

289
 Chapter 6 Verbs of removing

discussion of type 2 verbs of putting). Situations can be construed as achievements or

accomplishments (with the verbal scale that indicates the trajectory of motion of/on

the figure or with an iterated/multiple events reading). For example, John’s act of

taking the boat off the shore can have the retrospective reading or the activity in

progress reading with te-iru. In the latter case, it is construed as an accomplishment

because of the compatibility of tochuumade ‘halfway’. Sukohiszutsu ‘little by little’

modifies the motion of the boat whose imaginary trajectory line is on the surface

crossing the water from on the shore to off the shore. That is, the figure is associated

with a verbal scale.

John wa kobune o kishi kara dashi-te-iru.

(33) John TOP boat ACC shore ABL take.out-TE-IRU
John has taken a boat off the shore. (retrospective)
John is taking a boat off the shore. (activity in progress)

John wa kobune o tochuumade kishi kara dashi-ta.

(34) John TOP boat ACC halfway shore ABL take.out-PAST
John took a boat halfway off the shore.

John wa kobune o kishi kara sukoshizutsu dashi-te-iru.

(35) John TOP boat ACC shore ABL little.by.little take.out-TE-IRU
John is taking a boat off the shore little by little.

Subtype 2-2 verbs do not get the activity in progress reading unless a situation is

construed as repeated events of putting aside. That is, they represent achievements.

Patrick wa heya kara isu o noke-te-iru.

(36) Patrick TOP room ABL chair ACC put.aside-TE-IRU
Patrick is putting chairs away from the room.
(activity in progress as iterated activity)

6.4.2.5. Semantic representations

The achievement construal of nokeru ‘put aside’ and the accomplishment construal of

dasu ‘take out’ are as follows.

290
 Chapter 6 Verbs of removing

Figure 5. achievement construal of A-linking of nokeru ‘put aside’ in (23)

room
exist
kara
dislocated

chair nokeru ‘put aside’

Patrick cause move away

∆

Figure 6. accomplishment construal of A-linking of dasu ‘take out’ in (22) and (35)

shore exist
kara
move away

boat dasu ‘take out’

John cause move

Δ away

291
 Chapter 6 Verbs of removing

6.4.3. Type 3 Katazukeru ‘clear’ verb

6.4.3.1. Member

The following verb is the only member of this verb type:

katazukeru ‘clear’

6.4.3.2. Semantics

I have only found one member here. It refers to an activity of emptying/clearing a

place by moving an entity to another place, which is normally thought to be proper for

the entity (the cupboard for dishes, for example). The relation between the figure and

the ground is locational.

6.4.3.3. Syntactic patterns

This category appears with both FA-type and GA-type linking. This is similar to the

locative alternation of English verbs of removing. It is also compatible with the phrase

kirei ni ‘clean RST’ indicating that the ground is completely cleared.

A-linking
Joe wa teeburu kara sara o katazuke-ta.
(37) Joe TOP table ABL dish ACC clear-PAST
Joe cleared the dishes from the table.

G-linking
Joe wa teeburu no sara o katazuke-ta.
(38) Joe TOP table GEN dish ACC clear-PAST
Joe cleared the dishes on the table.

FA-linking
Joe wa sara o katazuke-ta.
(39) Joe TOP dish ACC clear-PAST
Joe cleared the dishes away.

GA-linking
Joe wa teeburu o katazuke-ta.
(40) Joe TOP table ACC clear-PAST
Joe cleared the table.

292
 Chapter 6 Verbs of removing

Kirei ni (clean RST):

Joe wa teeburu o kirei ni katazuke-ta.
(41) Joe TOP table ACC clean RST clear-PAST
Joe cleared the table completely (without leaving anything).

6.4.3.4. Causal and aspectual patterns

The agent acts on the figure, removes it from the ground, and makes it clean or

ensures nothing is left there. The figure is omitted in an FNI-way and the event can be

construed as the agent’s acting directly on the ground. Te-iru has the activity in

progress sense with the verb. The holistic theme is the figure with A-linking, as

sukoshizutsu ‘little by little’ modifies the dishes, indicating the removing of the dishes

little by little while the ground has GA-linking indicating that the surface of the table

is cleared little by little, changing to being cleared of everything.

Joe wa teeburu kara sara o sukoshizutsu katazuke-te-iru.

(42) Joe TOP table ABL dish ACC little.by.little clear-TE-IRU
Joe is clearing the dishes from the table little by little.

Joe wa teeburu o sukoshizutsu katazuke-te-iru.

(43) Joe TOP table ACC little.by.little clear-TE-IRU
Joe is clearing the table little by little.

In FA-type linkings, the event can have the directed activity reading without

specifying the definite change of state as the following example shows.

Joe wa sara o teeburu kara katazuke-ta ga mada ichi mai nokot-te-iru.

(44) Joe TOP dish ACC table ABL clear-PAST but still one CLSS remain-TE-IRU
Joe cleared the dishes from the table, but there is one (dish) left there.

On the other hand, GA-type linking does not allow this, and thus ensures the

change of state reading on the ground in GA-linking, as the unacceptable of example

(45) shows:

*Joe wa teeburu o katazuke-ta ga mada sara ga nokot-te-iru.

(45) Joe TOP table ACC clear-PAST but still dish NOM remian-TE-IRU
*Joe cleared the table, but there are still some dishes left there.

293
 Chapter 6 Verbs of removing

That is, FA-type linkings entail the figure being dislocated while GA-linking entails

the ground becoming cleared.

6.4.3.5. Semantic representations

The following are the representations of A-linking and GA-linking of the verb.

Holistic theme is assigned differently between the two representations.

Figure 7. accomplishment construal of A-linking of katazukeru ‘clear’ in (37)

table exist
kara
dislocated

katazukeru
dishes ‘clear’

Joe cause dislocated

Δ

294
 Chapter 6 Verbs of removing

Figure 8. accomplishment construal of GA-linking of katazukeru ‘clear’ in (40)

be clean/cleared

table
katazukeru
‘clear’

Joe make clean by removing

6.4.4. Type 4 Haku ‘sweep’ verbs

6.4.4.1. Members

The following are some members of this verb type:

haku ‘sweep’, huku ‘wipe’, nuguu ‘wipe’

6.4.4.2. Semantics

Type 4 verbs ‘typically’ refer to cleaning a surface of an entity by removing dust or

rubbish from the surface. The surface is construed as the ground and the unnecessary

entities (normally dust, stain etc.) as the figure. The relation between figure and

ground is impurities locational. These verbs are also similar to manner verbs in

Levin’s classification in that they entail a certain manner in which the figure (dust or

rubbish) is removed. The entailment of manner is what distinguishes verbs of this type

from type 3 katazukeru ‘clear’ verb.

295
 Chapter 6 Verbs of removing

6.4.4.3. Syntactic patterns

Like type 3 katazukeru ‘clear’ verb, type 4 verbs may have both FA and GA-linking.

However, A-linking is not so acceptable. The following examples illustrate:

??A-linking
??Joel wa yuka kara gomi o hai-ta.
(46) Joel TOP floor ABL dust ACC sweep-PAST
Joel swept the dust off the floor.

G-linking
Joel wa yuka no gomi o hai-ta.
(47) Joel TOP floor GEN dust ACC sweep-PAST
Joel swept the dust on the floor.

FA-linking
Joel wa gomi o hai-ta.
(48)
Joel TOP dust ACC sweep-PAST
Joel swept the dust away.

GA-linking
Joel wa yuka o hai-ta.
(49) Joel TOP sweep ACC floor-PAST
Joel swept the floor.

The kirei ni phrase is compatible with verbs of this category.

Kirei ni (clean RST):

Joel wa yuka o kirei ni hai-ta.
(50) Joel TOP floor ACC clean RST sweep-PAST
Joel swept the floor clean.

Another property of the verbs is that they are often used as a V1 in compounds with

type 1 or type 2-1 verbs as V2. (cf. 6.4.17 -18).

6.4.4.4. Causal and aspectual patterns

Basically, these are activitiy verbs of making the ground clean by removing impurities.

They denote situations which are construed either as activity or accomplishment. They

get the activity in progress reading with te-iru, but do not always entail that the ground

becomes clean (see example (52)).

296
 Chapter 6 Verbs of removing

GA-linking
Joel wa yuka o hai-te-iru.
(51) Joel TOP floor ACC sweep-TE-IRU
Joel is sweeping the floor. (activity in progress)

Joel wa yuka o hai-ta ga zenzen kirei ni nara-nakat-ta.

(52a) Joel TOP floor ACC sweep-PAST but not.at.all clean RST become-not-PAST
Joel swept the floor, but it did not become clean at all.

Joel wa yuka o hai-ta ga mada gomi ga nokot-te-iru.

(52b) Joel TOP floor ACC sweep-PAST but still dust NOM remain-TE-IRU
Joel swept the floor, but still there is some dust remaining.

Joel wa yuka o sukoshizutsu hai-ta.

(53) Joel TOP floor ACC little.by.little sweep-PAST
Joel swept the floor little by little.
(the surface of the floor is verbal scale.)

G-linking
Joel wa yuka no gomi o hai-te-iru.
(54) Joel TOP floor GEN dust ACC sweep-TE-IRU
Joel is sweeping the dust on the floor. (activity in progress)

Joel wa yuka no gomi o sukoshizutsu hai-ta.

(55) Joel TOP floor GEN dust ACC little.by.little sweep-PAST
Joel swept the dust on the floor little by little.

The possibility of negation as in (52) suggests that the act of sweeping can be an

activity of repeated contact with the surface without causing any change; one can even

sweep or wipe something that is already clean. However, as the incrementality in (53)

suggests, it also has the directed activity reading of making the ground clean by

removing something from its surface.

In example (54) (G-linking), dust is an uncountable entity and the event is

conventionally construed as one that comprises repeated events of sweeping dust and

therefore activity. Concerning the delimitedness, even though the sentence does not

syntactically specify how much dust was on the floor, there is a natural endpoint, that

is, if every grain of dust is gone, there is nothing more to sweep off. Therefore, the

event is accomplishment and the verbal scale is dust which is moving away gradually

297
 Chapter 6 Verbs of removing

till there is nothing left. This reading is enforced by the kirei ni ‘clean RST’ example.

Joel wa yuka no gomi o kirei ni hai-ta.

(56) Joel TOP floor GEN dust ACC clean RST sweep-PAST
Joel swept the dust on the floor completely.

6.4.4.5. Semantic representations

There follow semantic structure representations for the undirected activity construal

and the accomplishment construal with GA-linking, and the accomplishment construal

with G-linking.

Figure 9. undirected activity construal of GA-linking of haku ‘sweep’ in (49) and

(52a)

floor exist

haku ‘sweep’

Joel sweeping
activity
Δ
t

298
 Chapter 6 Verbs of removing

Figure 10. accomplishment construal of GA-linking of haku ‘sweep’ in (49)

cleaned

floor
haku ‘sweep’

clean with
Joel sweeping activity

Figure 11. accomplishment construal of G-linking of haku ‘sweep’ in (47) and (56)

move away

dust on the
floor
haku ‘sweep’

Joel cause move by

sweeping activity

299
 Chapter 6 Verbs of removing

6.4.5. Type 5 Muku ‘peel’ verbs

6.4.5.1. Members

The following are some members of this verb type:

muku ‘peel’, hagu ‘tear, skin, bark’, soru ‘shave’

6.4.5.2. Semantics

Verbs of this type refer to a separation of two entities, one of which is a part of the

other (a whole-part relation). Or, at least, two entities are in close contact as in one

entity stuck to another with glue (attachment). What is removed is normally thought to

be an unfavorable, or, at least, a useless part of the whole4. Unlike type 1 verbs and

not unlike type 4 verbs, they entail a manner or means how the act of separation

happens.

6.4.5.3. Syntactic patterns

These verbs appear in both FA and GA linking types. A-linking may be only

marginally acceptable, and in most cases, it sounds more natural to use G-linking

rather than A-linking. FA-linking is basically not so acceptable.

??A-linking
??Haha wa jagaimo kara kawa o mui-ta.
(57) mother TOP potato ABL skin ACC peel-PAST
Mother peeled the skin from a potato.

4
I include verbs of digging horu and sai.kutsu-suru in this category (see appendix G). They are not
prototypical members of this category in that they normally do denote the relation of location. Also,
they are totally OK with A-linking. Semantically, they can be categorised as in the type 6 (I would say
that they are inbetween of the two categories.) I classify them here under this category, however,
because they share an important syntactic behaviour which is the grammaticality with FA and GA type
constructions. Moreover, semantically, digging is an action which forced upon the surface that is
construed as the ground. The main difference from the type 6 is that the verbs of digging are totally
grammatical with GA-linking.

300
 Chapter 6 Verbs of removing

G-linking
Haha wa jagaimo no kawa o mui-ta.
(58) mother TOP potato GEN skin ACC peel-PAST
Mother peeled a potato.

*FA-linking (some OK)

*Haha wa kawa o mui-ta.
(59) mother TOP skin ACC peel-PAST
??Mother peeled the skin.

Chichi wa hige o sot-ta.

(60) father TOP beard/mustach ACC shave-PAST
Father shaved.

The reason why (60) is acceptable is that a beard is a part of the subject, which can

also be the ground where the figure is located. Shave puts more constraints on the

direct object which is semantically more specified than ‘skin’. The beard is a person’s

beard. On the other hand, the skin in (59) can be any skin (of a potato, apple, chicken,

etc.) Therefore, (59) is not acceptable without a proper context.

Most of these verbs allow GA-linking.

GA-linking
Haha wa jagaimo o mui-ta.
(61) mother TOP potato ACC peel-PAST
Mother peeled a potato.

These verbs are also compatible with kirei ni ‘clean RST,’ a phrase indicating that

no marks or no remains are left on the ground and they are often also used as the V1

in compounds with type 1 verbs.

Haha wa jagaimo no kawa o kirei ni mui-ta.

(62a) mother TOP potato GEN skin ACC clean RST peel-PAST
Mother peeled the skin of a potato completely.

Haha wa jagaimo o kirei ni mui-ta.

(62b) mother TOP potato ACC clean RST peel-PAST
Mother peeled a potato completely.

301
 Chapter 6 Verbs of removing

6.3.5.4. Causal and aspectual patterns

Causally, these verbs denote that the agent takes or removes the figure away from the

ground. Aspectually, they bear the activity in progress reading with te-iru. They refer

to the processual activity of separating a part from a whole.

Haha wa jagaimo no kawa o mui-te-iru.

(63a) mother TOP potato GEN skin ACC peel-TE-IRU
Mother is peeling the skin of the potato.

Haha wa jagaimo o mui-te-iru.

(63b) mother TOP potato ACC peel-TE-IRU
Mother is peeling the potato.

In the events denoted by these verbs, either the figure, as in (63a), or the ground, as in

(63b), serves as a verbal scale of directed activity. Conventionally the verbs have the

directed activity construal and the accomplishment construal. As the following

example shows, the sentences can be modified by tochuumade ‘halfway’ or

sukoshizutsu ‘little by little’:

Haha wa jagaimo no kawa o tochuumade/sukoshizutsu mui-ta.

(64a) mother TOP potato GEN skin ACC halfway/little.by.little peel-PAST
Mother peeled a potato halfway/little by little.

Haha wa jagaimo o tochuumade/sukoshizutsu mui-ta.

(64b) mother TOP potato ACC halway/little.by.little peel-PAST
Mother peeled a potato halfway/little by little.

Because the amount of skin that covers a potato is already fixed, the event has the

natural end point in both G-linking and GA-linking. Sukoshizutsu ‘little by little’

modifies the skin in G-linking and the potato in GA-linking. In the former, if the skin

is all gone, then the act of peeling a skin from a potato is finished, and in the latter, if

the potato becomes bare, then the act of peeling the potato is done. Thus, the aspectual

status of figure or ground reflects that of the whole event. The accomplishment

reading is reinforced with the use of kirei ni ‘clean RST’ as in (62).

302
 Chapter 6 Verbs of removing

The directed activity reading is made clear in the sentences, as examples (65) show

that the accomplishment reading can be prohibited.

Kodomo wa jagaimo no kawa o mui-ta ga

(65a) child TOP potato GEN skin ACC peel-PAST but

mada sukoshi kawa ga nokot-te-iru.

still some skin NOM remain-TE-IRU

The child peeled the skin of a potato, but there is still some skin remaining.

Kodomo wa jagaimo o mui-ta ga

(65b) child TOP potato ACC peel-PAST but

mada sukoshi kawa ga nokot-te-iru.

still some skin NOM remain-TE-IRU

The child peeled a potato, but there is still some skin remaining.

However, the negation (as in the previous type) is not allowed, as illustrated by the

following:

*Kodomo wa jagaimo no kawa o mui-ta ga zenzen muke-nakat-ta.

(66a) child TOP potato GEN skin ACC peel-PAST but not.at.all peel-not-PAST
The child peeled the skin of the potato, but could not peel at all.

*Kodomo wa jagaimo o mui-ta ga zenzen muke-nakat-ta.

(66b) child TOP potato ACC peel-PAST but not.at.all peel-not-PAST
The child peeled the potato, but could not peel at all.

That is, the undirected activity reading is not conventionally allowed.

6.4.5.5. Semantic representations

The accomplishment construals for the verb muku ‘peel’ for G-linking and GA-linking

are represented as follows. The two representations reflect the differences of assigning

the verbal scale to different participants.

303
 Chapter 6 Verbs of removing

Figure 12. accomplishment construal of G-linking of muku ‘peel’ in (58) and (62a)

move away and

be separated
potato’s
skin
muku ‘peel’

Taro cause move by

peeling activity

Figure 13. accomplishment construal of GA-linking of muku ‘peel’ in (61) and (62b)

be bare

potato

muku ‘peel’

Taro make bare by

peeling activity

304
 Chapter 6 Verbs of removing

6.4.6. Type 6 Hagasu ‘peel’ verbs

6.4.6.1. Members

The following are some members of this verb type:

hagasu ‘peel’, nuku ‘pull out’

6.4.6.2. Semantics

Verbs of this type refer to the physical separation of two independent entities. The

relation between the figure and the ground is attachment. Verbs of this type can

accommodate a lot of examples like type 1. However, unlike type 1, they entail a

certain manner/means of removal.

6.4.6.3. Syntactic patterns

These verbs manifest A-linking, G-linking, and FA-linking but not GA-linking as the

following examples illustrate:

A-linking
Mary wa kabe kara hurui kabegami o hagashi-ta.
(67) Mary TOP wall ABL old wallpaper ACC peel-PAST
Mary peeled the old wallpaper off the wall.

G-linking
Mary wa kabe no hurui kabegami o hagashi-ta.
(68) Mary TOP wall GEN old wallpaper ACC peel-PAST
Mary peeled the old wallpaper on the wall off.

FA-linking
Mary wa kabegami o hagashi-ta.
(69) Mary TOP wall ACC peel-PAST
Mary peeled the old wallpaper

*GA-linking
*Mary wa kabe o hagashi-ta.
(70) Mary TOP wall ACC peel-PAST
*Mary peeled the wall.

They are compatible with the phrase kirei ni.

305
 Chapter 6 Verbs of removing

kirei ni
Mary wa kabe kara hurui kabegami o kirei ni hagashi-ta.
(71) Mary TOP wall ABL old wallpaper ACC clean RST peel-PAST
Mary peeled the old wallpaper off the wall completely .

Also they are often compounded as V1 with type 1 or type 2-1 verbs (§ 6.4.17-18).

6.4.6.4. Causal and aspectual patterns

With these verbs the agent acts on the figure and takes/removes it away from the

ground. Aspectually, these verbs conventionally get the activity in progress sense with

te-iru, which motivates the processual analysis of events denoted by them.

Mary wa kabe kara hurui kabegami o hagashi-te-iru.

(72a)
Mary TOP wall ABL old wallpaper ACC peel-TE-IRU
Mary is peeling the old wallpaper off the wall.

Mary wa kabe no huri kabegami o hagashi-te-iru.

(72b) Mary TOP wall GEN old wallpaper ACC peel-TE-IRU
Mary is peeling the old wallpaper on the wall off.

The figure (old wallpaper) in (72) is associated with a verbal scale because its

aspectual status reflects that of the whole event. Sukoshizutsu ‘little by little’ modifies

the status of wallpaper, which is taken away little by little with both linkings, though I

only show the example of A-linking below.

Mary wa kabe kara hurui kabegami o sukoshizutsu hagashi-ta.

(73) Mary TOP wall ABL old wallpaper ACC little.by.little peel-PAST
Mary peeled the old wallpaper off the wall little by little.

However, in different contexts, a runup achievement construal is also possible, for

example where Mary was trying to peel off the poster stuck on the wall with glue, it

did not come off for a while and suddenly the poster came off instantaneously. This is

a runup achievement. The achievement construal is also possible in such a case as

peeling a stamp off an envelope. The event is conventionally construed as an

achievement though an accomplishment reading is possible.

306
 Chapter 6 Verbs of removing

6.4.6.5. Semantic representations

Figure 14 and 15 are the proposed aspectual representations for (67) and (68). Figure

16 is one for a runup achievement.

Figure 14. accomplishment construal of A-linking of hagasu ‘peel’ in (67)

wall exist
kara
move away

wallpaper
hagasu ‘peel’

Mary cause move

Δ by peeling activity

Figure 15. accomplishment construal of G-linking of hagasu ‘peel’ in (68)

move away and be separated

wallpaper
on wall
hagasu ‘peel’

Mary cause move

by peeling activity

307
 Chapter 6 Verbs of removing

Figure 16. runup achievement construal of A-linking of hagasu ‘peel’ in (67)

wall
exist
kara
dislocated

hagasu ‘peel’

wallpaper

Mary
dislocate
Δ by peeling activity

6.4.7. Type 7 Arau ‘wash’ verbs

6.4.7.1. Members

The following are some members of this verb type:

subtype 7-1 arau ‘wash’, sou.ji-suru ‘clean’

subtype 7-2 kiyomeru ‘purify, cleanse’, jou.ka-suru ‘purify’

6.4.7.2. Semantics

Subtype 7-1 verbs denote an activity of cleaning something (the ground) by removing

impurities. Subtype 7-2 verbs denote an event of making the ground clean by

removing impurities. The relation between the figure and the ground is the impurities

locational type though the figure is not syntactically realised.

308
 Chapter 6 Verbs of removing

6.4.7.3. Syntactic patterns

These verbs appear with GA-linking. They do not manifest any FA-type linkings such

as A-linking or G-linking. That is, the figure is not realised (FNI) though there is a

consensus about the kind of things removed: dust or a stain, something that makes an

entity dirty. These properties of the figure are lexically entailed in the verbs, which

makes only GA-linking possible. As the event focuses on what has happened to the

ground, these verbs are similar to the change of state verbs.

GA-linking:
Patrick wa zubon o arat-ta.
(74) Patrick TOP trouser ACC wash-PAST
Patrick washed the trousers.

Joel wa heya o sou.ji-shi-ta.

(75) Joel TOP room ACC clean-PAST
Joel cleaned the room.

Sono onna wa mi o kiyome-ta.

(76) the woman TOP body ACC cleanse-PAST
The woman cleansed her body.

*A-linking:
*Patrick wa zubon kara yogore o arat-ta.
(77) Patrick TOP trousers ABL stain ACC wash-PAST
Patrick washed the stain from the trousers.

*G-linking:
*Patrick wa zubon no yogore o arat-ta.
(78) Patrick TOP trousers GEN stain ACC wash-PAST
Patrick washed the stain on the trousers.

Subtype 7-1 verbs are also compatible with the phrase kirei ni ‘clean RST’, which

indicates that the ground becomes completely clean as in example (79). However, in

example (80) with subtype 7-2, this sounds less acceptable as the verbs themselves

entail that the ground becomes clean completely.

Patrick wa zubon o kirei ni arat-ta.

(79) Patrick TOP trouser ACC clean RST wash-PAST
Patrick washed the trousers clean.

309
 Chapter 6 Verbs of removing

??Sono onna wa mi o kirei ni kiyome-ta.

(80) the woman TOP body ACC clean RST cleanse-PAST
The woman made her body clean.

6.4.7.4. Causal and aspectual patterns

With these verbs the agent acts on an entity, which is supposed to be ground with the

intention of causing changes by removing something attached to it. Type 7-1 verbs

have the activity in progress reading with te-iru, as indicated below:

Patrick wa zubon o arat-te-iru.

(81) Patrick TOP trouser ACC wash-TE-IRU
Patrick is washing the trousers.

Joel wa heya o sou.ji-shi-te-iru.

(82) Joel TOP room ACC clean-TE-IRU
Joel is cleaning the room.

They do not necessarily entail that the ground becomes clean, as the following

sentences show.

Patrick wa zubon o arat-ta ga zenzen kirei ni nar-anakat-ta.

(83) Patrick TOP trousers ACC wash-PAST but not.at.all clean RST become-not-PAST
Patrick washed the trousers, but they did not become clean at all.

Joel wa heya o sou.ji-shi-ta ga zenzen kirei ni nar-anakat-ta.

(84) Joel TOP room ACC clean-PAST but not.at.all clean RST become-not-PAST
Joel cleaned the room, but it did not become clean at all.

This suggests that, not unlike type 4 verbs, they can have an undirected activity

construal without referring to the change of state of the ground. In this construal, the

agent acts on the ground, which does not undergo any directed changes (see Figure 17

below).

The other subclass of verbs of this type, such as kiyomeru ‘purify’, behaves

differently in terms of delimitedness. With te-iru, it favours the

retrospective/resultative reading referring to the clean state of the ground or at most,

310
 Chapter 6 Verbs of removing

the activity in progress reading in the runup achievement. It also does not allow

negation.

Sono onna wa mi o kiyome-te-iru.

(85) the woman TOP body ACC cleanse-TE-IRU
The woman has cleansed her body.
The woman is cleansing her body. (the runup achievement)

*Sono onna wa mi o kiyome-ta ga kirei ni nar-anakat-ta.

(86) the woman TOP body ACC cleanse-PAST but clean RST become-not-PAST
The woman cleansed her body, but it did not become clean.

This behaviour suggests that kiyomeru (the subtype 7-2) is an achievement verb which

profiles the final transition of the ground from not being clean to being clean. We can

say this is the corresponding verb to yogosu ‘dirty’ in verbs of putting.

The incrementality of the event varies according to verbs and situations. Arau

‘wash’ is rarely acceptable with sukoshizutsu ‘little by little’ unless the direct object is

plural and construed as a derived verbal scale. Kiyomeru ‘purify’ is the same.

??Patrick wa sukoshizutsu zubon o arat-ta.

(87) Patrick TOP trouser little.by.little ACC wash-PAST
Patrick washed the trousers little by little.
(??when the trousers is interpreted as one pair)

On the other hand, sou.ji-suru ‘clean’ and jou.ka-suru ‘purify’ can be construed as

directed activities or accomplishment, as the following examples show:

Joel wa heya o sukoshizutsu sou.ji-shi-ta.

(88) Joel TOP room ACC little.by.little clean-PAST
Joel cleaned the room little by little.

Karera wa osui o sukoshizutsu jou.ka-shi-ta.

(89) they TOP dirty.water ACC little.by.little purify-PAST
They purified dirty water little by little.

It is possible to clean the room on part at a time, as we can imagine from the way we

hoover a room for example. We can do it gradually and achieve more space in the

room to be cleaned. If we hoover half the room and it becomes clean, we can see the

cleaning event of the room as half done as the cleanliness of half the room is

311
 Chapter 6 Verbs of removing

perceptible. Thus, the verb has the accomplishment construal with the ground as its

verbal scale. On the other hand, the act of washing is not measured incrementally

because of the way we wash things. We do not wash half of the shirt at one time and

wash the rest at another time. Also, we see the cleanliness of the shirt at the end of

washing as a whole. It is not possible in our conventionalised situation that half the

shirt becomes clean in the middle of washing and the rest of it becomes clean at the

end. Therefore, arau ‘wash’ in this example gets the runup achievement interpretation.

In the case of jou.ka-suru ‘purify’, it is possible to purify the ground little by little

when the ground is uncountable entity like water.

6.4.7.5. Semantic representations

Below are the Figures for the undirected activity construal, the accomplishment

construal, the runup achievement construal, the achievement construal, and the

directed activity construal of this category:

312
 Chapter 6 Verbs of removing

Figure 17. undirected activity construal of GA-linking of arau ‘wash’ in (74)5 and (83)

trousers exist

arau ‘wash’

washing
Patrick activity
Δ
t

Figure 18. accomplishment construal of GA-linking of sou.ji-suru ‘clean’ in (75)

be clean

room
sou.ji-suru ‘clean’

Joel cleaning activity

5
Unlike in Figure 9 of type 4, the aspectual contour of the ground is represented as undirected activities.
This is to show the difference in affectedness of the ground; the floor remains exactly the same even
though it is swept, but the trousers have soap applied to them, are put into water, squeezed etc in the
process of washing. That is, they can be interpreted to undergo certain undirected activities.

313
 Chapter 6 Verbs of removing

Figure 19. runup achievement construal of GA-linking of arau ‘wash’ in (74) and (81)

clean

trousers arau ‘wash’

Patrick washing activity

Figure 20. achievement construal of GA-linking of kiyomeru ‘purify’ in (76)

be clean

body
kiyomeru ‘purify’

woman clean

314
 Chapter 6 Verbs of removing

Figure 21. directed activity construal of GA-linking of jou.ka-suru ‘purify’ in (89)

be clean

dirty water
jou.ka-suru ‘purify’

they cleaning activity

Δ

6.4.8. Type 8 Nusumu ‘steal’ verbs6

6.4.8.1. Members

The following are some members of this verb type:

subtype 8-1
ubau ‘steal’, nusumu ‘steal’, tou.you-suru ‘plagiarize’
subtype 8-2
kai.huku-suru ‘regain’

6.4.8.2. Semantics

Verbs of this type express events in which an actor deprives people or organisations

(ground) of some entities (figure). The relation between the figure and the ground is

possessional.

6
Possessional deprivation verbs are treated as ‘spatial transfer’ and not captured as referring to
possessional transfer relations (cf. Levin (1993), Croft (2000: 104, 109, §5.1)). In English, these verbs
have the same alternation as verbs of removing (the Path construction and the antecedent of
construction).

315
 Chapter 6 Verbs of removing

6.4.8.3. Syntactic patterns

These verbs manifest A-linking, G-linking and FA-linking.

A-linking
Dorobou ga John kara kane o nusun-da.
(90) thief NOM John ABL money ACC steal-PAST
The thief stole money from John.

G-linking
Dorobou ga John no kane o nusun-da.
(91) thief NOM John GEN money ACC steal-PAST
The thief stole John’s money.

FA-linking
Dorobou ga kane o nusun-da.
(92) thief NOM money ACC steal-PAST
The thief stole the money.

G-linking and FA-lilnking are grammatical with verbs of this category. However, G-

linking does not necessarily denote the same event as A-linking. A-linking favours a

locational (even extended way) reading, and G-linking a possessional reading between

figure and ground with verbs of stealing. In (90), the ground is a person, so it is

natural to interpret it as a possessor and (90) and (91) are equivalent. In another case,

where money is stolen from a bank, A-linking favours interpreting the bank as the

‘location’ and G-linking still favours the possessional reading. The location and the

possessor can be different entities. The bank in A-linking may merely specify the

place where the money is taken from but the money’s possessor can be somebody else

like the bank’s customer. Thus, the following is possible.

Dorobou wa ginkou kara John no kane o nusun-da.

(93) thief TOP bank ABL John GEN money ACC steal-PAST
The thief stole John’s money from the bank.

These verbs do not manifest GA linking as indicated by the inacceptability of the

following example.

316
 Chapter 6 Verbs of removing

*GA-linking
*Dorobou ga John o nusun-da.
(94) thief NOM John ACC steal-PAST
*The thief stole John.

Another subclass of this category, verbs of regaining (dakkan-suru ‘recapture’ or

dak.kai-suru ‘recapture’, kaishuu-suru ‘recover, withdraw’), which refer to events of

retrieving, regaining, or returning, are not compatible with G-linking at all. These

verbs mean that some property of someone or something which is kept in a certain

location are regained by a ‘claimed’ original or proper possessor. In other words, what

is taken away is returned to the original place. The speaker, when using these verbs

conceptualises the figure as not being a proper possession or proper possessum of the

ground. This is why G-linking is not allowed with these verbs. The following

examples illustrate:

A-linking
Wareware wa tsuini tekigun kara shima o dak.kan-shi-ta.
(95) we TOP finally enemy ABL island ACC recapture-do-PAST
We finally retook the island from the enemy.

*G-linking
*Wareware wa tsuini tekigun no shima o dak.kan-shi-ta.
(96) we TOP finally enemy GEN island ACC recapture-do-PAST
*We finally retook the island of the enemy.

6.4.8.4. Causal and aspectual patterns

Causally, these verbs refer to an event where the agent takes the figure away from the

ground. They get the retrospective reading with te-iru, which suggests that they denote

punctual causation.

John wa yuujin kara kane o nusun-de-iru.

(97) John TOP friend ABL money ACC steal-TE-IRU
John has stolen money from his friends (in the past).

These verbs may have the activity in progress reading with te-iru in runup

317
 Chapter 6 Verbs of removing

achievement construals though the contexts are limited. The following examples are

possible if somebody is witnessing the robber stealing a wallet from John.

Dorobou ga John kara saihu o nusun-de-iru!

(98a) thief NOM John ABL wallet ACC steal-TE-IRU
The thief is stealing the wallet from John!

Dorobou ga John no saihu o nusun-de-iru!

(98b) thief NOM John GEN wallet ACC steal-TE-IRU
The robber is stealing John’s wallet!

There is not an intermediate stage of stealing a wallet. One cannot steal the wallet

halfway or gradually. Also the act of stealing is finished at the last moment the robber

successfully gets the wallet. In the above sentences, the witness is watching the

robber’s attempt to steal and not the completion; if it has been completed, the speaker

would use the past tense. Therefore, the activity in progress sense here is better

understood as processual activity in a runup achievement.

6.4.8.5. Semantic representations

Here are the proposed semantic structure for the achievement reading of A-linking, G-

linking, and the runup achievement construal of G-linking.

318
 Chapter 6 Verbs of removing

Figure 22. achievement construal of A-linking of nusumu ‘steal’ in (90)

John
exist
kara
enter physical possession

nusumu
money ‘steal’

thief gain physical possession

The representation above is slightly different from those of pure spatial removal. As

Croft (2000: 109) argues, the thief’s obtaining money clearly affects the ground

(victim) in that it causes the loss of the possession. Therefore, the relation between

money and John above is represented by the causal arrow (based on Croft 2000: 104,

Figure 2, § 5.1). However, the interpretation of possessional deprivation as spatial

transfer is still maintained by assigning the subevent of “exist” to the victim.

319
 Chapter 6 Verbs of removing

Figure 23. achievement construal of G-linking of nusumu ‘steal’ in (91)

enter physical possession

and separate away
John’s
money
nusumu ‘steal’

thief gain physical possession

Δ

Figure 24. runup achievement construal of G-linking of nusumu ‘steal’ in (91)

enter physical possession

John’s
money
nusumu ‘steal’

thief gain physical possession

320
 Chapter 6 Verbs of removing

6.4.9. Type 9 Tsui.hou-suru ‘expel’ verbs

6.4.9.1. Members

The following are some members of this verb type:

verbs of expelling
tsu.ihou-suru ‘expel, banish’, oi-dasu ‘chase-put out = drive out’
verbs of release
hanasu ‘set free, release’

6.4.9.2. Semantics

Verbs of this type refer to an event of removing people or animal away from a

location by driving them away or by setting them free. The relation between the figure

and the ground is locational. The two subtypes differ in terms of how a person or

animal is ‘removed’ and the willingness of the figure to be removed. However, their

main syntactic behaviour is identical. Verbs of expelling include many compound

verbs whose V1 is ou ‘chase’. However, the V2 varies and they are frozen expressions

where the meaning of the V2 and even that of the V1 do not contribute to the meaning

of the compound. So, I categorise them as idiomatic compounds here rather than

means compounds.

6.4.9.3. Syntactic patterns

These verbs manifest A-linking and G-linking. However, as discussed, G-linking does

not always indicate the same situation as A-linking because of the diverse use of

Genitive linking. The following examples illustrate the use of A-linking and G-linking.

A-linking
Karera wa ou o huransu kara tsui.hou-shi-ta.
(99) they TOP king ACC France ABL expel-do-PAST
They expelled the king from France.

321
 Chapter 6 Verbs of removing

Jacqui wa kago kara tori o hanashi-ta.

(100) Jacqui TOP cage ABL bird ACC release-PAST
Jacqui released the bird from the cage.

G-linking
Karera wa huransu no ou o tsui.hou-shi-ta.
(101) they TOP France GEN king ACC expel-do-PAST
They expelled the king of France.

Jacqui wa kago no tori o hanashi-ta.

(102) Jacqui TOP cage GEN bird ACC release-PAST
Jacqui released the bird in the cage.

In example (101), ‘France’ can be interpreted favourably as the nationality of the

king rather than the place where he temporarily stays. In that case, the construal given

by G-linking is not quite equivalent to that given by A-linking. It is possible to use

another phrase indicating a place from which the French king is expelled, as in the

following:

Karera wa sono kuni kara huransu no ou o tsui.hou-shi-ta.

(103) they TOP that country ABL France GEN king ACC expel-do-PAST
They expelled the king of France from that country.

I would say that the difference in acceptability of G-linking here actually has

something to do with the relations between the figure and the ground, rather than with

verbs themselves. As we can think of more than one relation between ‘France’ and

‘king’ (nationality of the king as well as location of the king), the G-linking becomes

ambiguous for the above examples.

Verbs of this type also allow FA-linking.

FA-linking

Karera wa ou o tsui.hou-shi-ta.
(104) they TOP king ACC expel-do-PAST
They expelled the king.

Jacqui wa tori o hanashi-ta.

(105) Jacqui TOP bird ACC release-PAST
Jacqui released the bird.

322
 Chapter 6 Verbs of removing

Instead of the source phrase, the goal phrase can appear, indicating where a

person/animal ends up after being driven away or released from the original place.

The following examples illustrate:

path pattern (Goal phrase)

Karera wa ou o {huransu kara} osutoria ni tsui.hou-shi-ta.
(106) they TOP king ACC {France ABL} Austria ALL expel-do-PAST
They expelled the king {from France} to Austria.

Jacqui wa {kago kara} oozora ni tori o hanashi-ta.

(107) Jacqui TOP cage ABL sky ALL bird ACC release-PAST
Jacqui released the bird to the air {from the cage}.

This type of verbs does not, however, allow GA-linking.

*GA-linking
*Karera wa huransu o tsui.hou-shi-ta.
(108) they TOP France ACC expel-do-PAST
*They expelled France.

*Jacqui wa kago o hanashi-ta.

(109) Jacqui TOP cage ACC release-PAST
*Jacqui released the cage.

6.4.9.4. Causal and aspectual patterns

With these verbs, the agent acts on the figure (person or animals) and makes it leave

the ground (location). They have basically an achievement construal as they

conventionally get the retrospective reading with te-iru, and there are no intermediate

stages of performing the action, unless the direct object is construed as a derived

verbal scale.

Karera wa ou o huransu kara tsui.hou-shi-te-iru.

(110) they TOP king ACC France ABL expel-do-TE-IRU
They have expelled the king from France. (existential use)

*Karera wa ou o huransu kara sukoshizutsu tsui.hou-shi-ta.

(111) they TOP king ACC France ABL little.by.little expel-do-PAST
*They expelled the king from France little by little.

323
 Chapter 6 Verbs of removing

Jacqui wa kago kara tori o hanashi-te-iru.

(112) Jacqui TOP cage ABL bird ACC release-TE-IRU
Jacqui has released the bird from the cage. (retrospective)

6.4.9.5. Semantic representations

The semantic representation for the A-linking of type 9 verbs is shown in the

following Figure.

Figure 25. achievement construal of A-linking of tsui.hou-suru ‘expel’ in (99)

France
exist
kara
expelled

king tsui.hou-suru ‘expel’

they expel
Δ

6.4.10. Type 10 Ha.mon-suru ‘excommunicate’ verbs

6.4.10.1. Members

The following are some members of this verb type:

verbs of abstract “expel”

ha.mon-suru ‘excommunicate’, jo.mei-suru ‘oust, expel’7
shirizokeru ‘remove, oust’

7
Jo.mei-suru was originally a figure-incorporation type as jo is ‘remove’ and mei is ‘name’; removing a
name. However, it is now used as a frozen expression meaning ‘oust’ or ‘remove’.

324
 Chapter 6 Verbs of removing

6.4.10.2. Semantics

Type 10 verbs refer to an event of expelling someone from a social organisation, a

group, a sect, a position in a company or society. The relation between the figure and

the ground is generally that of the extended locational type but can be identificational

for a small number of verbs (e.g. shirizokeru ‘oust’). The ground is an organisation or

group for the former type of relation while it is the name of a position for the latter.

6.4.10.3. Syntactic patterns

These verbs manifest A-linking. G-linking, as with the previous types, can be

ambiguous and marginal. These verbs are similar to the previous type having an

extended reading of the spatial removal of somebody from an abstract location. The

following examples illustrate.

A-linking

Karera wa Steve o Katorikku kyoukai kara ha.mon-shi-ta.

(113) they TOP Steve ACC Catholic church ABL excommunicate-do-PAST
They excommunicated Steve from the Catholic Church.

Kaisha wa shitenchou kara John o shirizoke-ta.

(114) company TOP branch.manager ABL John ACC oust-PAST
?The company ousted John from branch manager.

??G-linking
??Karera wa Katorikku kyoukai no Steve o ha.mon-shi-ta.
(115) they TOP Catholic church GEN Steve ACC excommunicate-do-PAST
They excommunicated Steve of the Catholic Church.

??Kaisha wa shitenchou no John o shirizoke-ta.

(116) company TOP branch.manager GEN John ACC oust-PAST
?The company ousted John of the branch manager.

However, a difference is that these verbs do no occur with a goal phrase, as the

following examples show.

325
 Chapter 6 Verbs of removing

*Goal-phrase
*Karera wa Steve o Purotesutanto kyoukai ni ha.mon-shi-ta.
(117) they TOP Steve ACC Protestant church ALL excommunicate-do-PAST
They excommunicated Steve to the Catholic Church.

*Kaisha wa hirashain ni John o shirizoke-ta.

(118) company TOP employee ALL John ACC oust-PAST
*The company ousted John to an employee.

FA-linking and GA-linking are also not possible with this type of verbs.

*FA-linking
*Karera wa Steve o ha.mon-shi-ta.
(119) they TOP Steve ACC excommunicate-do-PAST
They excommunicated Steve.

*Kaisha wa John o shirizoke-ta.

(120) company TOP John ACC oust-PAST
The company ousted John.

*GA-linking
*Karera wa Katorikku kyoukai o ha.mon-shi-ta.
(121) they TOP Catholic church ACC excommunicate-do-PAST
*They excommunicated the Catholic Church.

*Kaisha wa shitenchou o shirizoke-ta.

(122) company TOP branch.manager ACC oust-PAST
*The company ousted the branch manager.

6.4.10.4. Causal and aspectual patterns

Verbs of this type bear the retrospective meaning in the te-iru form, which suggests

that events denoted by them are punctual. The folllowing examples illustrate:

Karera wa Steve o Katorikku kyoukai kara ha.mon-shi-te-iru.

(123) they TOP Steve ACC Catholic church ABL excommunicate-do-TE-IRU
They have excommunicated Steve from the Catholic Church. (existential)

Kaisha wa shitenchou kara John o shirizoke-te-iru.

(124) company TOP branch.manager ABL John ACC oust-TE-IRU
The company has ousted John from being branch manager.

326
 Chapter 6 Verbs of removing

6.4.10.5. Semantic representations

The semantic structure for the achievement reading of the verb is as follows.

Figure 26. achievement construal of A-linking of ha.mon-suru ‘excommunicate’ in

(113)

Catholic
church exist
kara
excommunicated

Steve ha.mon-suru ‘excommunicate’

they
Δ excommunicate

6.4.11. Type 11 Kai.hou-suru ‘liberate’ verbs

6.4.11.1. Members

kai.hou-suru ‘release, liberate’8, hou.men-suru ‘ release, acquit’

6.4.11.2. Semantics

Type 11 verbs refer to the removal of an abstract notion from a person. This includes

an event of setting somebody free from an abstract notion of suppression,

responsibility, burden, or sin etc. The relation between the figure and the ground is

very abstract.

8
The verb is also used as ‘spatial’ release as type 9-2 as well.

327
 Chapter 6 Verbs of removing

6.4.11.3. Syntactic patterns

These verbs take A-linking.

A-linking
Seihu wa shuukyouteki yokuatsu kara hitobito o kai.hou-shi-ta.
(125) government TOP religious oppression ABL people ACC liberate-do-PAST
The government liberated people from the religious oppression.

However, neither do they take G-linking, FA-linking, GA-linking, nor a goal phrase.

*G-linking
*Seihu wa shuukyouteki yokuatsu no hitobito o kai.hou-shi-ta.
(126) government TOP religious oppression GEN people ACC liberate-do-PAST
The government liberated people of/in the religious oppression.

*FA-linking9
Seihu wa hitobito o kai.hou-shi-ta.
(127) government TOP people ACC liberate-do-PAST
The government liberated people.

*GA-linking
*Seihu wa shuukyouteki yokuatsu o kai.hou-shi-ta.
(128) government TOP religious oppression ACC liberate-do-PAST
*The government liberated the religious oppression.

*goal phrase
*Seihu wa jiyuu ni hitobito o kai.hou-shi-ta.
(129) government TOP freedom ALL people ACC liberate-do-PAST
The government liberated people to freedom.

Genitive linking is not grammatical in this category as the relation between the figure

and the ground is very abstract and the genitive marker is not used to cover the

abstract notion that refers to something that is mentally preoccupied or suppressed by

an abstract thing.

6.4.11.4. Causal and aspectual patterns

As one of the entities which are denoted by the figure and the ground becomes more

abstract, and so do the events denoted by verbs of this category, we need to pay

9
The sentence is acceptable if it refers to spatial release of people such as from prison.

328
 Chapter 6 Verbs of removing

attention to the figure and the ground construals. So far, the verbs I have discussed

denote pure spatial relations or quasi-spatial relations and it is self-evident which

entity can be construed as the figure and which as the ground from their size and

mobility. Being in an abstract relation, however, each of the two entities in this

category can be construed either as the figure or the ground.

The difference in FG-construal is what differentiates verbs of this type and those

of the next type. Semantically, they are similar; they have an agent, a freed entity (in

most cases human), and an abstract notion which is to be removed. For this type, I

would argue that ‘the abstract notion’ is the ground and ‘the person’ is the figure. That

is, the event denoted by the verb is merely thought of as a metaphorical extension of

spatial removal, and the semantic relation between the two bears ‘extended locational’,

with the abstract notion being construed as the location where the person is confined.

Motivation for this comes from the passive form. Comparing the passive form of

(130) and those of purely (quasi-)spatial events of removal, they all have a [Figure

TOP Ground ABL Verb-PASS] frame.

Passive (Type 11)

Hitobito wa shuukyouteki yokuatsu kara kai.hou-s-are-ta.
(130) people TOP religious oppression ABL liberate-do-PASS-PAST
People were liberated from the religious oppression.

Passive (Type 1)
Zassou wa shibahu kara tori-nozok-are-ta.
(131) weeds TOP lawn ABL remove-remove-PASS-PAST
Weeds were removed from the lawn.

Passive (Type 2)
Kobune wa kishi kara das-are-ta.
(132) boat TOP shore ABL take.out-PASS-PAST
The boat was taken out off the shore.

329
 Chapter 6 Verbs of removing

Passive (Type 8)

Taikin ga ginkou kara nusum-are-ta.

(133) huge.sum.of.money NOM bank ABL steal-PASS-PAST
A huge sum of money was stolen from the bank.

Passive (Type 9)

Ou wa huransu kara tsui.hou-s-are-ta.

(134) king TOP France ABL expel-do-PASS-PAST
The king was expelled from France.

In all the examples from other types of verbs, the figure which is the direct object in

the active sentences10 becomes the subject in the passive, and the ground is marked as

oblique in both types of sentences. The ground cannot be marked as the direct object

in the passive: *[Figure TOP Ground ACC Verb-PASS]. Here is one example,

*Zassou wa shibahu o tori-nozok-are-ta.

(135) weeds TOP lawn ACC remove-remove-PASS-PAST
Weeds were removed from the lawn.

The same is true with type 11 verbs.

*Hitobito wa shuukyouteki yokuatsu o kai.hou-s-are-ta.

(136) people TOP religious oppression ACC liberate-do-PASS-PAST
People were liberated from the religious oppression.

Thus, syntactically, there is good reason to think that the abstract notion is the ground

and the person is the figure in verbs of this category. The abstract notion to be

removed is construed as a metaphorical location where people’s minds are confined.

Moreover, in this analysis, the ungrammaticality with G-linking can be explained

more clearly. It is not because the relation is abstract but because it is an extended

(abstract) locational relation. As the figure and the ground are in the extended

locational relation, G-linking is not so acceptable (cf. see type 10).

Aspectually, the achievement construal is most conventional, as the verbs get the

10
For active sentences, refer to the examples in the discussion of each verb type.

330
 Chapter 6 Verbs of removing

retrospective reading with te-iru as a default reading. That is, the event is an

achievement.

Seihu wa shuukyouteki yokuatsu kara hitobito o kai.hou-shi-te-iru.

(137) government TOP religious oppression ABL people ACC liberate-do-TE-IRU
The government has liberated people from the religious oppression.

6.4.11.5. Semantic representations

The achievement construal of type 11 verbs is represented as follows:

Figure 27. achievement construal of A-linking of kai.hou-suru ‘release’ in (125)

religious
oppression exist
kara
be free

people kai.hou-suru ‘liberate’

government
Δ set free

6.4.12. Type 12 Toku ‘relieve’ verbs

6.4.12.1. Members

The following are some members of this verb type:

toku ‘relieve’, men.jo-suru ‘exempt’, menzuru ‘exempt’

331
 Chapter 6 Verbs of removing

6.4.12.2. Semantics

Like the previous type, verbs of this category refer to removing an abstract notion

such as a duty, responsibility, guilt, or oppression from a person. The relation between

the figure and the ground is abstract.

6.4.12.3. Syntactic patterns

As with wrap type of verbs of putting, the syntactic configuration becomes

complicated as this is related to the FG-construal itself. They manifest linking which

is similar to A-linking, with a freed person encoded as direct object and with an

abstract entity as oblique. I shall explain this linking in the following sub-section.

Reversed A-linking ((similar to) A-linking) [person = ACC, abstract notion = OBL]
Kaisha wa yatto juu.seki kara kare o toi-ta.
(138) company TOP finally heavey.responsibility SRC he ACC relieve-PAST
The company finally relieved him of heavy responsibility.

G-linking
*[abstract notion GEN person]
*Kaisha wa yatto juuseki no kare o toi-ta.
(139a) company TOP finally heavey.responsibility GEN he ACC relieve-PAST

The company finally relieved him of heavy responsibility.

[person GEN abstract notion]

Kaisha wa yatto kare no juuseki o toi-ta.
(139b) company TOP finally he GEN heavy.responsibility ACC relieve-PAST
The company finally relieved his heavy responsibility.

*FA-linking
*Kaisha wa yatto juuseki o toi-ta.
(140) company TOP finally heavy.responsibility ACC relieve-PAST
*The company finally relieved the heavy responsibility.

*GA-linking
*Kaisha wa yatto kare o toi-ta.
(141) company TOP finally he ACC relieve-PAST
*The company finally relieved him.

The complexity resides in the naming of the term. These syntactic linkings are named

332
 Chapter 6 Verbs of removing

according to how the figure and the ground are linked to the syntactic expression, but

they do not refer to which entity is construed as the figure or the ground. In most cases,

the FG-construal is self-evident and does not cause problems. However, in these

abstract events of removal, we need to take an FG-construal into account.

6.4.12.4. Causal and aspectual patterns

It is clear that the agent comes first in the causal representation. However, the causal

order of the person and the abstract notion to be removed is not so straightforward.

Either the person comes first or the abstract notion comes first.

Let us think of the first case. From the compatibility with ‘A-linking’ and with

analogy to other spatial removal types that allow A-linking, we can assume that toku

‘relieve’ has the semantic structure like that of the previous type (Figure 27), with the

freed person as the figure and the abstract notion as the ground. The whole event is

seen as a metaphorical extension of spatial removal. The representation of this version

is as in Figure 28a in the next sub-section.

However, there are two things which cannot be explained by this first stipulation

of the causal order. First of all, toku ‘relieve’ does not behave like the previous type or

other types of spatial removal in the passive form.

*Kare wa juuseki kara tok-are-ta.

(142) he TOP heavy.responsibility ABL relieve-PASS-PAST
He was relieved from heavy responsibility.

Kare wa juuseki o tok-are-ta.

(143) he TOP heavy.responsibility ACC relieve-PASS-PAST
He was relieved heavy responsibility.

The person who is temporarily supposed to be the figure is realised as the subject. The

abstract notion which is supposed to be the ground does not appear in the ablative

333
 Chapter 6 Verbs of removing

phrase, but in the accusative phrase in the passive form, which is different from the

other verb types.

Secondly, the behaviour concerning G-linking is also different between this type

and other types of spatial removal. In the previous type, the “ground GEN figure”

phrase is not allowed and I explained that it is because of the abstract/extended

location between the figure and the ground. The ungrammaticality of (139a) can be

explained in the same way. However, the grammaticality of (139b) cannot be

explained. The equivalent expression with the previous type of verb is not acceptable.

*Seihu wa hitobito no shuukyouteki yokuatsu o kai.hou-shi-ta.

(144) government TOP people GEN religious oppression ACC release-do-PAST
The government released people’s religious oppression.

In the above, ‘figure GEN ground’ is not grammatical. In (139b), if we assume the

abstract notion as the ground and the freed person as the figure, ‘figure GEN ground’

pattern should be somehow acceptable here.

The above two syntactic differences between this type and the other (quasi-)spatial

removal types suggest that the semantic structure of this type should not be proposed

in the same way as the previous one, i.e. ‘extended locational reading’. The second

possibility is that the abstract notion to be removed comes first in the causal structure.

That is, causally, the agent acts on the abstract notion and removes it from the person;

with the abstract notion (heavy responsibilities) as the figure and the freed person (he)

as the ground.

This alternative causal order is similar to that with nusumu ‘steal’ verbs, which

manifest possessional relation between the possessor (ground) and the possessum (the

figure). Let us compare these two types. Actually, ‘steal’ type verbs and this type

behave similarly in the passive form.

334
 Chapter 6 Verbs of removing

nusumu type:
A-linking [ground-ABL figure-ACC]
Dorobou wa roizu ginkou kara okane o nusun-da.
(145) thief TOP Lloyds bank ABL money ACC steal-PAST
The thief stole money from Lloyds Bank.

Passive
*[Ground TOP Figure ABL Verb-PASS]
*Roizu ginkou wa okane kara nusum-are-ta.
(146) Lloyds bank TOP money ABL steal-PASS-PAST
*Lloyds Bank was stolen from money.

Passive
[Ground TOP Figure ACC Verb-PASS]
Roize ginkou wa okane o nusum-are-ta.
(147) Lloyds bank TOP money ACC steal-PASS-PAST
Lloyds Bank had its money stolen.

Comparing (146) and (147) with (142) and (143), we can see similarities. We can

conclude that Lloyds bank in (146) and (147) and he in (142) and (143) are both the

ground and the possessor. When they are realised as the subject in the passive

sentence, the figure is marked as accusative and not oblique 11 . This alternative

construal explains the first difference between the behaviour of type 12 verbs and

other spatial removal type verbs in the passive form.

Moreover, the second difference between the behaviour of these two groups of

verbs (type 12 and other spatial removal types) is explained by the alternative causal

order. By putting the abstract notion first in the causal order, the event is construed as

the freed person being the possessor of that abstract notion, which will be taken away

by the agent. The genitive case can indicate the possessor-possessum relation in ‘A

GEN B’, so the genitive phrase ‘his heavy responsibilities’ in (139b) is validated. That

is, the genitive linking is compatible with an extended possessional relation (but not

11
Note that in nusumu ‘steal’ type, which is treated as an extended meaning of spatial removal, when
figure is the subject of the passive sentence, ground is marked as oblique as in cases of the other spatial
types and type 11 (see the examples of passives in the discussion of type 11).

335
 Chapter 6 Verbs of removing

an extended locational relation).

Thus, concerning the order of participants, there is good motivation for treating the

causal order of type 12 like the poseessional relation between the person and the

abstract notion, through comparison with the nusumu ‘steal’ type of verbs, as they

behave syntactically in a similar way in relation to the passive. One notable difference

between the two types, however, is the realisation of A-linking; in the nusumu ‘steal’

type, the possessor is realised as oblique and the possessum as accusative, while in

toku ‘relieve’, the possessor (the freed person) is realised as accusative and the

possessum (the abstract notion) is realised as oblique. That is, the A-linking in (138) is

actually Reversed A-linking since the figure and the ground are realised syntactically

in the opposite way to how they would be with normal A-linking.

The semantic representation of type 12 should be altered accordingly. As the

ground is realised as the direct object and also because of Figure-first coercion (cf.

Croft 1991: 198-206), the verb should profile all of the three participants. The ground,

which is marked as the direct object, is no more a spatial reference point, but is a

participant that undergoes changes. The figure is located before the direct object,

which means the postposition kara is not a subsequent oblique but an antecedent

oblique. In 4.4.1.2.4, I list the use of kara. It can be an antecedent oblique marker as

source. It may be plausible to say that kara in reversed A-linking is the source so that

the figure here (heavy responsibilities) can also be thought to be an abstract source

from which the freed person is freed as well as the possessum. The reversed A-linking

corresponds to the English antecedent oblique of construction. In Japanese, the

distinction between normal A-linking and the reversed one is not immediately

apparent as they use the same marker kara, unlike English, which uses the different

prepositions from and of.

336
 Chapter 6 Verbs of removing

The alternative and more explanatory semantic structure of the type is presented as

Figure 28b in the next subsection. It represents the achievement construal. The verb

has the retrospective reading with te-iru.

Kaisha wa juuseki kara kare o toi-te-iru.

(148) company TOP heavy.responsibility ABL he ACC relieve-TE-IRU
The company has relieved him of heavy responsibility. (existential)

337
 Chapter 6 Verbs of removing

6.4.12.5. Semantic representations

Figure 28a*. achievement construal of A-linking of toku ‘relieve’ in (138): spatial

transfer reading

heavy
responsibility exist
kara
be free

he toku ‘relieve’

company release
Δ

Figure 28b. achievement construal of Reversed A-linking of toku ‘relieve’ in (138):

possessional transfer reading

not in possession/
without responsibility

he

taken away

heavy
responsibility toku ‘relieve’

company take away

Δ

338
 Chapter 6 Verbs of removing

6.4.13. Type 13 Kai.nin-suru ‘dismiss’ verbs

6.4.13.1. Members

The following are some members of this verb type.

kai.nin-suru ‘relieving.duty-do = dismiss’,

kai.shoku-suru ‘relieving.job-do = dismiss’,
men.shoku-suru ‘exempting.job-do = dismiss’,
hi.men-suru ‘stop.relieve = dismiss, remove’,
kai.ko-suru ‘relieving.employment-do = dismiss, discharge’

6.4.13.2. Semantics

The verbs of this category all refer to events of removing somebody from his/her

position in an organisation like a company or a ministry. With the exception of

hi.men-suru, they are all figure incorporation verbs; the first part corresponds to verbs

of type 12 (kai and toku (relieve), men and menjiru (exempt)), which has reversed A-

linking, and the second part refers to the figure, which is to be removed (abstract

notion). In this case, it is the position of a person. The relation between a person and a

position is abstract. However, in this connection, previous discussion of verbs of type

12 should be recalled, where I discuss that an abstract notion to be removed is

construed as the figure and a person from whom it is taken away is construed as the

ground. As the first part of the VN verbs is a type 12 verb, I also assign the function

the ground to the person who is dismissed and the figure to the position from which

one is removed. The relation between the figure and the ground is identificational (a

filler to its role as Steve = branch manager).

6.4.13.3. Syntactic patterns

These verbs do not permit A-linking. They also sound unnatural with reversed A-

linking. The fact that the figure and the ground are in an identificational relation may

339
 Chapter 6 Verbs of removing

make the reversed A-linking difficult. As they refer to the same person by using the

two nouns (the names of a person and his position), the linking which makes the two

asymmetrical by assigning one accusative and the other oblique is not felicitous.

A-linking
*Kaisha wa shitenchou o Steve kara kai.nin-shi-ta.
(149) company TOP branch.manager ACC Steve ABL relieve.duty-do-PAST
*The company dismissed branch manager from Steve.

??Reversed A-linking
(150)
??Kaisha wa Steve o shitenchou kara kai.nin-shi-ta.
company TOP Steve ACC branch.manager SRC relieve.duty-do-PAST
The company dismissed Steve from a branch manager.

On the other hand, G-linking is acceptable. The use of genitive case fits into the (G)

use of genitive in example (6) in §6.3; A is an attribute of B in ‘A no B’.

G-linking
Kaisha wa shitenchou no Steve o kai.nin-shi-ta.
(151) company TOP branch.manager GEN Steve ACC relieve.duty-do-PAST
The company dismissed Steve, whose is a branch manager.

The verbs of this category is grammatical with FA-linking but sounds unnatural with

GA-linking, as the following examples illustrate:

FA-linking

Kaisha wa shitenchou o kai.nin-shi-ta.

(152) company TOP branch.manager ACC relieve.duty-do-PAST
The company dismissed a branch manager.

??GA-linking

??Kaisha wa Steve o kai.nin-shi-ta.

(153) company TOP Steve ACC relieve.duty-do-PAST
The company dismissed Steve.

As we can see from the above examples, they sound better when the name of the

position appears as the direct object. Only kai.ko-suru ‘relieving.employment-do’ is

neutral; it can appear perfectly well with GA-linking. This suggests that most of the

340
 Chapter 6 Verbs of removing

verbs in this category require the figure to be specified more in the linking patterns

above than in the case of an incorporated figure such as a duty or a job. That is, the

verbs have only incomplete figure incorporation.

The behaviour of these verbs in relation to the passive forms as in (154) and (155)

confirms that the construal of the above with the role as the figure and the person as

the ground is correct12.

Passive:
Shitenchou wa kai.nin-s-are-ta.
(154) branch.manager TOP relieve.duty-do-PASS-PAST
The branch manager was dismissed.

Steve wa shitenchou o/*kara kai.nin-s-are-ta.

(155)
Steve TOP branch.manager ACC/*ABL relieve.duty-do-PASS-PAST
Steve was dimissed as the branch manager.

6.4.13.4. Causal and aspectual patterns

With these verbs, the agent removes the figure (the position) from the ground (person).

Conventionally, they have an achievement construal as they cannot have the activity

in progress reading with te-iru but the retrospective reading.

Kaisha wa shitenchou o kai.nin-shi-te-iru.

(156) company TOP branch.manager ACC relieve.duty-do-TE-IRU
The company has dismissed a branch manager.
(existential)

12
Compare (155) with the passive of type 10 verbs as follows.
John wa shitenchou kara/*o shirizoke-rare-ta.
(i)
John TOP branch.manager ABL/*ACC oust-PASS-PAST
John was ousted from the position of branch manager.

Both types refer to an event of dismissal. However, FG-construal is different as a person and a job is
identificational, therefore, abstract. In Type 10, the situation is construed as the extended location of a
person being expelled from the position, while it is construed as extended possession in type 13, where
the position is taken away from the person.

341
 Chapter 6 Verbs of removing

6.4.13.5. Semantic representations

Here are the semantic structures for FA-linking and G-linking with this type of verb.

In the following structure, the branch manager in FA-linking can be replaced by Steve,

who is the branch manager (figure GEN ground) in G-linking. It should also be noted

that the labelling of the two representations is slightly different. In Figure 29, the

event is construed as dismissing from a social role and in Figure 30, it is construed as

Steve’s losing identification.

Figure 29. achievement construal of FA-linking of kai.nin-suru ‘dismiss’ in (152)

dimissed

branch manager

kai.nin-suru ‘dismiss’

company dismiss
Δ

342
 Chapter 6 Verbs of removing

Figure 30. achievement construal of G-linking of kai.nin-suru ‘dismiss’ in (151)

lose identification

Steve of
branch manager
kai.nin-suru ‘dismiss’

company cause to lose identification

Δ

6.4.14. Type 14 Sen.patsu-suru ‘wash hair’ verbs: ground incorporation

6.4.14.1. Members

The following are the two members of this verb type:

sen.gan-suru ‘wash.face-do = face.washing-do’

sen.patsu-suru ‘wash.hair-do = hair.washing-do’

6.4.14.2. Semantics

The number of verbs in this category is less than that of ground incorporation verbs of

putting in English and Japanese. They are VN-suru verbs of Chinese origin. The first

part refers to a verb of removing and the second part to the ground (noun), from which

the figure is caused to move. Actually the two examples I have found consist of arau

‘clean’ verbs as the first part and the figure, which is thought to be dust or some other

impurity, syntactically is not realised.

343
 Chapter 6 Verbs of removing

6.4.14.3. Syntactic patterns

They appear as in intransitives without any direct object.

intransitive:
Kirstie wa maiasa sen.patsu-suru.
(157) Kirstie TOP every.morning washing.hair-do
Kirstie washes her hair every morning.

Washing (one’s) hair and washing (one’s) face are reflexive activities, and there is no

syntactic need for the ground to appear. However, when the activity is performed on

somebody else as in (158), the ground is marked as accusative.

GA-linking
Kirstie wa Jean no kami o sen.patsu-shi-ta.
(158) Kirstie TOP Jean GEN hair ACC wash.hair-do-PAST
Kirstie washed Jean’s hair.

Kirei ni the ‘clean RST’ phrase is compatible with making the event delimited.

Kirstie wa Jean no kami o kirei ni sen.patsu-shi-ta.

(159) Kirstie TOP Jean GEN hair ACC clean RST wash.hair-do-PAST
Kirstie washed Jean’s hair clean.

6.4.14.4. Causal and aspectual patterns

With these verbs the agent acts on the ground and makes it clean by removing

impurities. They get an activity in progress reading with te-iru.

Kirstie wa sen.patsu-shi-te-iru.
(160) Kirstie TOP wash.hair-do-TE-IRU
Kirstie is washing her hair.

Kirstie wa Jean no kami o sen.patsu-shi-te-iru.

(161) Kirstie TOP Jean GEN hair ACC wash.hair-do-TE-IRU
Kirstie is washing Jean’s hair.

The undirected activity construal is possible as the event of washing hair does not

necessarilly entail that the hair becomes clean.

Kirstie wa sen.patsu-shi-ta ga kirei ni nar-anakat-ta.

(162) Kirstie TOP wash.hair-do-PAST but clean RST become-not-PAST
Kirstie washed her hair, but it did not become clean at all.

344
 Chapter 6 Verbs of removing

The accomplishment construal is also possible with these verbs. We can wash our hair

or our face gradually, making it clean bit by bit, even though this is not so common.

Kirstie wa Jean no kami o sukoshizutsu sen.patsu-shi-ta.

(163) Kirstie TOP Jean GEN hair ACC little.by.little wash.hair-do-PAST
Kirstie washed Jean’s hair little by little.

6.4.14.5. Semantic representations

The activity construal of the intransitive version and the accomplishment construal of

GA-linking are presented as below.

Figure 31. undirected activity construal of intransitive linking of sen.patsu-suru ‘wash

hair’ in (157) (cf. Croft 2000: 59 Figure 5 § 3.4)

sen.patsu-suru
‘wash hair’
washing hair
Kirstie activity

345
 Chapter 6 Verbs of removing

Figure 32. accomplishment construal of GA-linking of sen.patsu-suru ‘wash hair’ in

(158) (cf. Croft 2000: 73, Figure 25 §3.5.3 similar to transitive dressing)

be clean

Jean’s hair
sen.patsu-suru
‘wash hair’

Kirstie washing activity

6.4.15. Type 15 Hai.sui-suru ‘drain water’ verbs: figure incorporation 1

6.4.15.1. Members

The following are the two members of this verb type:

hai.sui-suru ‘put away/remove-water-do = drain’

hai.ki-suru ‘put away/remove-air-do = exhaust gas’

6.4.15.2. Semantics

This category corresponds to type 14 figure-incorporation type 1 in verbs of putting13.

They are figure incorporation verbs (similar to pit verbs in Levin’s classification). The

first part of VN refers to the removing action and the second part to the figure, which

is caused to move or change its location. The verbs denote spatial removal.

13
There are three subpatterns for figure incorporation type 1 of verbs of putting (see footnote 23 of the
5). For verbs of removing, there is only one pattern which is similar to pattern 1 of figure
incorporation type of verbs of putting, which allow the path pattern.

346
 Chapter 6 Verbs of removing

6.4.15.3. Syntactic patterns

These verbs allow A-linking and also allow G-linking, where genitive case indicates

the locational relation between the figure and the ground, and also the goal phrase,

which indicates where the figure moves to.

A-linking
Haha wa senmenjo kara yogoreta mizu o hai.sui-shi-ta.
(164) mother TOP wash.basin ABL dirty water ACC remove.water-do-PAST
Mother drained the dirty water from the washbasin.

G-linking
Haha wa senmenjo no yogoreta mizu o hai.sui-shi-ta.
(165) mother TOP wash.basin GEN dirty water ACC remove.water-do-PAST
Mother drained the dirty water in the washbasin.

Goal-phrase
Haha wa mizo ni yogoreta mizu o hai.sui-shi-ta.
(166) mother TOP ditch ALL dirty water ACC remove.water-do-PAST
Mother drained the dirty water to the ditch.

path pattern
Haha wa senmenjo kara mizo ni yogoreta mizu o hai.sui-shi-ta.
(167) mother TOP wash.basin ABL ditch ALL dirty water ACC remove.water-do-PAST
Mother drained the dirty water from the wash basin to the ditch.

What is interesting is that even though they are figure incorporation verbs

compositionally, they do not take the ground only (without the figure) as the only

argument other than the agent. They do not take GA-linking. They do not allow even

GO-linking. That is, the verbs only partially incorporate the figure (incomplete

incorporation). This is very different from the figure incorporation verbs 1 of verbs of

putting, which allow GO-linking. One explanation is that the verbs of this type are

used almost as frozen expressions. They may be treated as type 2 or type 9 verbs of

removing.

*GA-linking
*Haha wa senmenjo o hai.sui-shi-ta.
(168) mother TOP wash.basin ACC remove.water-do-PAST
Mother drained the washbasin.

347
 Chapter 6 Verbs of removing

*GO-linking
*Haha wa senmenjo kara hai.sui-shi-ta.
(169) mother TOP wash.basin ABL remove.water-do-PAST
Mother drained from the washbasin.

They take FA-linking.

FA-linking
Haha wa yogoreta mizu o hai.sui-shi-ta.
(170) mother TOP dirty water ACC remove.water-do-PAST
Mother drained the dirty water.

6.4.15.4. Causal and aspectual patterns

With these verbs the agent removes the figure from the ground. Aspectually, they

conventionally select the accomplishment reading (or the directed activity reading

without any end point). They get the activity in progress reading with te-iru and are

also compatible with the adverbs, tochuumade ‘halfway’ and sukoshizutsu ‘little by

little’, as shown by the following example:

Haha wa senmenjo kara yogoreta mizu o hai.sui-shi-te-iru.

(171) mother TOP wash.basin ABL dirty water ACC remove.water-do-TE-IRU
Mother is draining dirty water from the washbasin. (activity in progress)

Haha wa senmenjo kara yogoreta mizu o tochuumade hai.sui-shi-ta.

(172) mother TOP wash.basin ABL dirty water ACC halfway remove.water-do-PAST
Mother drained dirty water from the washbasin halfway.

Haha wa senmenjo kara yogoreta mizu o sukoshizutsu hai.sui-shi-ta.

(173) mother TOP wash.basin ABL dirty water ACC little.by.little remove.water-do-PAST
Mother drained dirty water little by little from the washbasin.

348
 Chapter 6 Verbs of removing

6.4.15.5. Semantic representations

The proposed semantic structure follows:

Figure 33. accomplishment construal of G-linking of hai.sui-suru ‘drain water’ in

(165)

move away and

dirty water be separated
in washbasin

hai.sui-suru
‘drain water’

Mother cause move away

Δ
t

6.4.16. Type 16 Jo.setsu-suru ‘clear of snow’ verbs: figure incorporation 2

6.4.16.1. Members

The following are some members of this verb type:

jo.setsu-suru ‘removing.snow-do = clear of snow’,

jo.shitsu-suru ‘removing.moist-do = dehumidify’

6.4.16.2. Semantics

The first element of VN is a verb of removing and the second indicates the figure,

what is removed. Unlike in the previous type, the figure can be fully incorporated, in

which case it does not appear syntactically. It can also be partially incorporated, in

which case, the figure argument that adds specific meaning to the type of figure

appears in the syntax. From the data I have collected, the first element of these verbs

349
 Chapter 6 Verbs of removing

are all simple verbs of type 1 such as nozoku, nuku ‘remove’.

6.4.16.3. Syntactic patterns

Unlike the previous type, these verbs allow GA-type linking as well as FA-type

linking (such as A-linking). The following examples illustrate:

A-linking (incomplete incorporation)

Josetsuki wa hodou kara shinsetsu o jo.setsu-shi-ta.
(174) snowplough TOP pavement ABL fresh.snow ACC remove.snow-do-PAST
The snowplough removed the fresh snow from the pavement.

G-linking (locational)
Josetsuki wa hodou no shinsetsu o jo.setsu-shi-ta.
(175) snowplough TOP pavement GEN fresh.snow ACC remove.snow-do-PAST
The snowplough removed the fresh snow on the pavement.

FA-linking
Josetsuki wa shinsetsu o jo.setsu-shi-ta.
(176) snowplough TOP fresh.snow ACC remove.snow-do-PAST
The snowplough removed the fresh snow.

GA-linking (complete incorporation)

Josetsuki wa hodou o jo.setsu-shi-ta.
(177) snowplough TOP pavement ACC remove.snow-do-PAST
The snowplough cleared the pavement.

*GO-linking
*Josetsuki wa hodou kara jo.setsu-shi-ta.
(178) snowplough TOP pavement ABL remove.snow-do-PAST
The snowplough cleared of the pavement.

*Goal phrase
(179a)
*Josetsuki wa hodou ni shinsetsu o jo.setsu-shi-ta.
snowplough TOP pavement ALL fresh.snow ACC remove.snow-do-PAST
*The snowplough removed the fresh snow to the pavement.

*Path pattern

*Josetsuki wa hodou kara torakku ni shinsetsu o jo.setsu-shi-ta.

(179b) snowplough TOP pavement ABL truck ALL fresh.snow ACC remove.snow-do-PAST
*The snowplough removed the fresh snow from the pavement to the truck.

The unacceptability of GO-linking is an asymmetrical characteristic when we compare

350
 Chapter 6 Verbs of removing

the verbs of this category with the corresponding type in verbs of putting (type 15). I

shall return to this later.

6.4.16.4. Causal and aspectual patterns

The agent removes/takes away the figure from the ground. Therefore, the causal order

is the agent, the figure and the ground. However, as we can see, in complete figure

incorporation, the ground can be construed as being directly acted on by the agent

through the removal of the figure, which is located in the ground. Aspectually, the

verbs have a directed activity reading or an accomplishment reading since the figure

which is incorporated into these verbs refers to an uncountable entity which it is

possible to remove little by little and also the ground serves as a natural end point for

the event; when the ground becomes clear, then the event ends. Kirei ni ‘clean RST’

enforces the accomplishment reading.

Josetsuki wa hodou kara shinsetsu o jo.setsu-shi-te-iru.

(180) snowplough TOP pavement ABL fresh.snow ACC remove.snow-do-TE-IRU
The snowplough is removing the fresh snow from the pavement.

Josetsuki wa hodou kara shinsetsu o tochuumade jo.setsu-shi-ta.

(181) snowplough TOP pavement ABL fresh.snow ACC halfway remove.snow-do-PAST
The snowplough removed the fresh snow from half the pavement.

Josetsuki wa hodou kara shinsetsu o sukoshizutsu jo.setsu-shi-ta.

(182) snowplough TOP pavement ABL fresh.snow ACC little.by.little remove.snow-do-PAST
The snowplough removed the fresh snow little by little from the pavement.

351
 Chapter 6 Verbs of removing

6.4.16.5. Semantic representations

The following are the accomplishment construal of A-linking and GA-linking of

jo.setsu-suru ‘clear snow’.

Figure 34. accomplishment construal of A-linking of jo.setsu-suru ‘clear snow’ in

(174) and (182)

pavement exist
kara
move away

fresh snow
jo.setsu-suru
‘clear snow’

snowplough removed
Δ

t
Figure 35. accomplishment construal of GA-linking of jo.setsu-suru ‘clear snow’ in
(177)

cleared (of snow)

pavement

jo.setsu-suru
‘clear snow’

snowplough make clean

by removing snow
Δ

352
 Chapter 6 Verbs of removing

As we have seen, there is no encoding for GO-linking of the verb. Thus, the following

representation is prohibited.

Figure 36. accomplishment construal of *GO-linking (ungrammatical) of jo.setsu-

suru ‘clear snow’ in (178)

pavement exist

kara

jo.setsu-suru
‘remove snow’

remove snow
snowplough
Δ

353
 Chapter 6 Verbs of removing

6.4.17. Type 17 Nuki-dasu ‘pull.out-take.out’ verbs: means compound 1 with

specified direction

6.4.17.1. Members

The two categories below are classified according to the way component verbs are

integrated into compounds (cf. §5.4.16-17).

(middle compounds)
tori-dasu ‘take-take out = take out’,
nuki-dasu ‘pull out-take out = pull out’,
tori-modosu ‘take-return = take back’,
shibori-dasu ‘squeeze-take out = squeeze out’

(pure means compounds)

pattern 1:
mochi-dasu ‘take-take out = take out’,
tsumami-dasu ‘pinch/pick-take out = pick out’,
kiki-dasu ‘listen/hear-take out = find out’,
sasoi-dasu ‘invite/lure-take out’,
haki-dasu ‘sweep-take out’

pattern 2
arai-dasu ‘dig up one’s past’,
shime-dasu ‘lock out’

VN-suru
hou.shutsu-suru ‘release.take out-do’

6.4.17.2. Semantics

Verbs of this category include compounds or VN-suru verbs. For compounds, V1

specifies how the spatial (extended spatial) transfer of an entity is carried out. V2

indicates the caused-motion with specified direction. In verbs of removing, dasu ‘take

out’ (type 2-1 verbs of removing) predominantly appears as V2. This type corresponds

to the 5.4.16 verb of putting means compound 1 pattern. Compound verbs of this

type (especially with dasu) appear very extensively and are productive. Various kinds

of verb classes can appear as the V1-means component as we can see in the

354
 Chapter 6 Verbs of removing

corresponding types of compound verbs of putting. For Chinese-origin VN verbs, both

of the two parts of the VN are verbs; the first is equivalent to V1 (specifying means)

and the second to V2 (specifying caused-motion and direction/location). Theoretically,

there can be Japanese VN verbs, but I have been totally unable to find data in the so

far and further efforts will be left to future study.

6.4.17.3. Syntactic patterns

The syntactic patterns this category manifests correspond to those of type 2 verbs. The

means compounds I am treating here are all right-headed; the argument structure of

V2 verbs is carried onto that of compound verbs. Therefore, they take A-linking,

allow the goal phrase, but prohibit GA-linking. G-linking depends on the relations

between the figure and the ground and varies according to the verb. The following

examples illustrate:

A-linking
Jimuin wa tana kara shorui o nuki-dashi-ta.
(183) clerk TOP shelf ABL document ACC pull.out-take.out-PAST
The clerk took the document off the shelf by pulling it.

?G-linking (varies)
Jimuin wa tana no shorui o nuki-dashi-ta.
(184) clerk TOP shelf GEN document ACC pull.out-take.out-PAST
The clerk took the document on the shelf off of it.

??Chichi wa heya no neko o shime-dashi-ta.

(185) father TOP room GEN cat ACC shut-take.out-PAST
Father shut the cat out of the room.

*FA-linking
*Jimuin wa shorui o nuki-dashi-ta.
(186) clerk TOP document ACC pull.out-take.out-PAST (*as FNI)
The clerk took the document out.

??Chichi wa neko o shime-dashi-ta.

(187) father TOP cat ACC shut-take.out-PAST
Father shut the cat out.

355
 Chapter 6 Verbs of removing

Goal phrase
Jimuin wa tsukue ni shorui o nuki-dashi-ta.
(188) clerk TOP desk ALL document ACC pull.out-take.out-PAST
The clerk pulled and took the document out on the desk.

*GA-linking
*Jimuin wa tana o nuki-dashi-ta.
(189) clerk TOP shelf ACC pull.out-take.out-PAST
*The clerk took the shelf by pulling.

6.4.17.4. Causal and aspectual patterns

The compounds as a whole causally refer to an event of the agent moving the figure

away from the ground specifying a means. This is mostly identical to type 2 verbs as

the argument linking reflects the causal structure. However, the causal pattern is not

so easily defined in terms of the semantic structure of V1 and the way V1 and V2 are

combined. When V1 elaborates the schematic structure of V2, there may be a minor or

major adjustment in the semantic structure of V1. We shall see some examples in the

next sub-section. Aspect also varies according to the nature of the participants and

also according to the means events which are specified by V1. It should also be noted

that even V2 verbs themselves may have the achievement or accomplishment

construals.

356
 Chapter 6 Verbs of removing

6.4.17.5 Semantic representations

Concerning the integration of the semantic structures of component verbs, there are

mainly two types, the latter of which is further divided into two. One is what I call

‘middle’ compounds (see the previous chapter for discussion) and the other is pure

means compounds. In verbs of removing, I subclassify the latter further into two

patterns.

6.4.17.5.1. Middle compounds: Nuki-dasu ‘pull.out-take.out’

The following is an example of nuki-dasu with A-linking.

Jimuin wa tana kara shorui o nui-ta.

(190a) clerk TOP shelf ABL document ACC pull.out-PAST
The clerk pulled the document from the shelf. (V1—A-linking)

Jimuin wa tana kara shorui o dashi-ta.

(190b) clerk TOP shelf ABL document ACC take.out-PAST
The clerk took the document from the shelf. (V2—A-linking)

Jimuin wa tana kara shorui o nuki-dashi-ta.

(190c) clerk TOP shelf ABL document ACC pull.out-take.out-PAST
The clerk took the document off the shelf by pulling it out.
(compound—A-linking)

Figure 37 shows the semantic representation for the V1, the V2, and the compound.

357
 Chapter 6 Verbs of removing

Figure 37. semantic structure for achievement construal of A-linking of nuki-dasu

‘pull.out-take.out’ in (190)

V1 V2

shelf shelf
exist exist
kara kara
change change location
document location document

nuku ‘pull out’ dasu ‘take out’

grasp and cause change
Sue pull Sue location

Δ Δ

t t

Compound

shelf
exist
kara
change location
document
nuki-dasu nuku dasu
‘pull.out-take.out’

take out
Sue by pulling
Δ

358
 Chapter 6 Verbs of removing

In these middle compounds, V1 and V2 have identical argument structures with an

identical unprofiled part and they have identical referents as their participants. Both of

them have the caused-motion sense. However, V1 is more specific about how the

caused-motion is carried out. The more schematic structure of V2 is superimposed on

V1 and as a result, the argument structure of the compound is identical to both that of

of the V1 and that of the V2. The V1, V2 and the compound all profile the same part

of the event and also have the same unprofiled background. This first subtype

(‘middle compound’) is similar to pair compounds in terms of the integration of

semantic structures of composite verbs, in that it is not easy to define which argument

structure is inherited into the compound.

6.3.17.5.2. Pure means compounds

6.3.17.5.2.1. Mochi-dasu ‘take -take.out’

This is a compound that denotes extended causation and is the type which can be

confused with manner compounds. As in verbs of putting, I treat this means

compound as having extended causation.

Steve wa kane o mot-ta.

(191a) Steve TOP money ACC take-PAST
Steve took money. (V1)

Steve wa ie kara kane o dashi-ta.

(191b) Steve TOP house ABL money ACC take.out-PAST
Steve took money out of the house. (V2—A-linking)

Steve wa ie kara kane o mochi-dashi-ta.

(191c) Steve TOP house ABL money ACC take-take.out-PAST
Steve took money out of the house. (Compound—A-linking)

Figure 38 is the representation for the V1, the V2, and the compound verb.

359
 Chapter 6 Verbs of removing

Figure 38. semantic structure for A-linking of mochi-dasu ‘take-take.out’ in (191)

V1 V2

house
exist
be taken kara
money be unlocated

motsu money dasu ‘take out’

‘take’

John take (hold of) John cause

unlocated
Δ Δ

t t

Compound

house exist
kara
be unlocated

bag mochi-dasu
‘take-take.out’
motsu dasu

John cause unlocated

by holding
Δ

360
 Chapter 6 Verbs of removing

I represent the achievement reading of mochi-dasu here because it is difficult to think

of a situation where it gets the activity or accomplishment construal even with te-iru

unless the whole event is construed as a coarse-grained representation of

multiple/repeated events of taking money. It favours the retrospective reading with te-

iru and the event is achievement. The reason I call this extended causation is that the

agent has to hold the figure to take it out from the ground. Holding something

represents extended and “inactive” causation. The V1 verb, motsu, refers to the

achievement event of somebody’s taking hold of something. Therefore, it profiles the

inceptive phase with an unprofiled result process. Since taking something is not the

end of the event, it has the implication that the agent keeps hold of the entity by

applying “invisible” force to it. The Japanese equivalent of have in I have money is

entailed by motsu ‘take’ plus te-iru. Also, this te-iru is interpreted as the resultative

state of the agent’s taking hold of the money. The representation of V1 above refers to

this reading of motsu and only profiles the inceptive phase and the implication of the

agent’s continuous effort is represented as the direct change of state in the agent’s

contour.

In pure means compounds, V1 does not have the identical argument structure to

V2. In the example above, V1 is a two-argument verb that only specifies the agent and

what is acted on. The representation of V2 is superimposed on that of V1 by

overlapping the correponding shared entities (agent and figure). Also the composite

structure inherits and preserves the argument structure of V2. In the above example,

money is represented as what is acted on by the agent of V1 and V2; there is no

alternation in the causal order in V1.

361
 Chapter 6 Verbs of removing

6.4.17.5.2.2. Shime-dasu ‘shut-take.out’

The second pattern of pure means compounds is rarely found in compound verbs of

putting. The integration of component structures involves a major adjustment, that is,

in the causal order on V1.

Chichi wa *neko/??heya/doa o shime-ta.

(192a) father TOP *cat/??room/door ACC shut-PAST
My father shut *the cat/??room/door. (V1)

Chichi wa neko o heya kara dashi-ta.

(192b) father TOP cat ACC room ABL take.out-PAST
My father put the cat out of the room. (V2—A-linking)

Chichi wa neko o heya kara shime-dashi-ta.

(192c)
father TOP cat ACC room ABL shut-put.out-PAST
My father shut the cat out of the room. (Compound—A-linking)

Shimeru ‘shut’ is a two-argument verb and takes an exit or a kind of open place as

direct object. Apparently, its semantic restriction on the direct object does not match

with that of V2 or of the compound. An entrance of the entity which is construed as

the ground in the compound is a referential entity to the direct object of V1. That is, a

part of the ground of V2 or of the compound is construed as what is acted on directly

by the agent in V1. Thus, the semantic structure of V1 is altered in terms of the causal

order in that what comes after the agent in the causal order in V1 is forced to be the

last participant in the causal order of the compound. Figure 39 is the semantic

representation.

362
 Chapter 6 Verbs of removing

Figure 39. achievement construal of A-linking of shime-dasu ’shut-take.out’ in (192)

V1 V2

room exist
kara
shut
be unlocated
door (of cat
the room)
shimeru dasu ‘take out’
‘shut’ cause unlocated
father shut father
Δ Δ

t t

Compound

room
exist
kara
kept away from
cat

shime-dasu shimeru
dasu
‘shut-take.out’

keep out
father by shutting

363
 Chapter 6 Verbs of removing

Actually, even the compound itself refers to the situation where the agent shuts or

locks the entrance and does not let an animate thing in. That is, the agent can perform

the action without actively applying force to the figure. What the compound denotes is

also different from the V2 in terms of the force-dynamic relations, though the

compound inherits the argument structure of the V2. The V2 denotes the active

participation of the agent applying force on the figure. I differentiate the labelling of

each participant in the semantic structure of the compound. The force which is applied

to the figure is the act of keeping it away from a certain location. This force-dynamic

relation is what Talmy (1988: 66) categorises as the case where the Agonist tends to

have more force than the Antagonist (to come in, in this case). The force is applied to

keep an entity out no matter whether it wants to come in or not. This exemplifies the

argument of Langacker (1987: 281) that the composite structure may ‘involve entities

and specification beyond those provided by the components’. That is, compounds may

place totally different semantic restrictions on arguments, to those placed on them by

their component verbs.

In short, in this type of pure means compounds, there is an adjustment in the causal

order in the semantic structue of V1 when it is superimposed by V2 and it elaborates

an e-site of V2. Namely, the participant which is directly acted on by V1 verbs is

altered so as to become the last participant in the compounds.

364
 Chapter 6 Verbs of removing

6.4.18. Type 18 Arai-otosu ‘wash-remove’ verbs: means compound 2

6.4.18.1. Members

The following are some members of this verb type:

(middle compound)
nusumi-toru ‘steal-take’
huki-toru ‘wipe-remove’ = remove by wiping
(pure means compound)
subpattern 1
kiri-toru ‘cut-take/remove’ = take/remove by cutting,
sori-otosu ‘shave-remove’ = remove by shaving

subpattern 2
damashi-toru ‘deceive-take’ = take by deceiving
arai-otosu ‘wash-remove’ = wash by removing
toki-hanatsu ‘relieve-release’ = release

VN-suru
setsu.jo-suru ‘cut.remove-do’

6.4.18.2. Semantics

Most of the verbs in this class are compound verbs. V1 components indicate means

and V2 components are verbs of removing which take A-linking. Unlike the previous

type, they do not specify the direction of the move. Rather, they entail a certain

change of state incidental to the caused-motion. This category corresponds to type 17

of verbs of putting. There is one Chinese VN verb; the first element represents means

and the second represents removal. Again, there are two main groups of compounds

according to the way V1 and V2 are integrated. The second group can be further sub-

divided into two patterns. I show one example for each from compounds whose V2

elements are very productive; toru (take, remove) and otosu (remove) in the section of

semantic representations (§6.4.18.5).

365
 Chapter 6 Verbs of removing

6.4.18.3. Syntactic patterns

They take FA-type linking such as A-linking or G-linking but not GA-linking.

A-linking
Ani wa kao kara hige o sori-otoshi-ta.
(193) brother TOP face ABL beard ACC shave-remove-PAST
My brother shaved the beard from his face.

G-linking
Otoko wa rojin no kane o nusumi-tot-ta.
(194) man TOP old.man GEN money ACC steal-take-PAST
The man stole the money of the old man.

FA-linking
Ani wa hige o sori-otoshi-ta.
(195) brother TOP beard ACC shave-remove-PAST
My brother shaved off his beard.

*GA-linking
*Otoko wa rojin o nusumi-tot-ta.
(196) man TOP old.man ACC steal-take-PAST
*The man stole the old man.

6.4.18.4. Causal and aspectual patterns

These comounds refer to events where the agent takes/removes the figure away from

the ground. The aspectual pattern varies between achievement and accomplishment

depending on the semantic nature of the figure and the ground and also on what kind

of V1 is included in the compound. V2 verbs of removing (most of which are type 1)

are more general in meaning than the V1, having less semantic restrictions on the

kinds of arguments they take. The V2 is more schematic than the V1.

6.4.18.5. Semantic representations

I shall now show the semantic representation for three types: the middle compound,

and the pure means compound with and without the adjustment on the causal order.

366
 Chapter 6 Verbs of removing

6.4.18.5.1. Middle compounds: Nusumi-toru ‘steal-take’

This middle compound takes A-linking, so do the V1 and V2.

All A-linking
Otoko ga roujin kara kane o nusun-da.
(197a) man NOM old.man ABL money ACC steal-PAST
The man stole money from the old man. (V1)

Otoko ga roujin kara kane o tot-ta.

(197b) man NOM old.man ABL money ACC take-PAST
The man stole money from the old man. (V2)

Otoko ga roujin kara kane o nusumi-tot-ta.

(197c) man NOM old.man ABL money ACC steal-take-PAST
The man stole money from the old man. (Compound)

The semantic representation of nusumi-toru is represented in Figure 40.

367
 Chapter 6 Verbs of removing

Figure 40. semantic structure for A-linking of nusumi-toru ‘steal-take’ in (197)

V1 V2

old man old man

exist exist
kara kara
enter possession taken away

money nusumu money toru

‘steal’ ‘take’

get
John possession John take

Δ Δ

t t

Compound

old man exist

kara
taken away
and enter possession
money nusumi-toru nusumu toru
‘steal-take’

John take and

get possession

368
 Chapter 6 Verbs of removing

The identical causal structures of the V1 and the V2 are integrated by superimposing

the structure of the V2 onto that of the V1. As a result, the semantic structure of the

compound is identical to those of both that of the V1 and V2.

6.4.18.5.2. Pure means compounds

6.4.18.5.2.1. Sori-otosu ‘shave-remove’

Here is an example of a pure means compound without the causal order alignment.

Ani wa {kao no} hige o sot-ta.

(198a) brother TOP {face GEN} beard ACC shave-PAST
My brother shaved his beard {on the face}. (V1—FA-linking)

Ani wa kao kara hige o otoshi-ta.

(198b)
brother TOP face ABL beard ACC remove-PAST
My brother removed his beard from his face. (V2—A-linking)

Ani wa kao kara hige o sori-otoshi-ta.

(198c) brother TOP face ABL beard ACC shave-remove-PAST
My brother removed his beard from his face by shaving.
(Compound—A-linking)

The semantic representation for this compound is proposed in Figure 41. Soru ‘shave’

takes FA-linking and G-linking but sounds strange (though not totally ungrammatical)

in A-linking. It does not favour the argument linking of three participants. However, it

is combined with otosu ‘remove’, and the resulting composite structure takes A-

linking perfectly felicitously. The structure of V1 elaborates a more schematic V2

structure, so the V1 can be either a two or three-argument verb.

369
 Chapter 6 Verbs of removing

Figure 41. semantic structure for accomplishment construal of A-linking of sori-otosu

‘shave-remove’ in (198)

V1 V2

face exist
kara

beard move away move away

(disappear)
soru ’shave’ beard otosu
‘remove’
father shaving father
activity remove
Δ Δ

t t

face exist
kara

move away
sori-otosu
beard ‘shave-remove’ soru otosu

remove by
shaving
father

370
 Chapter 6 Verbs of removing

6.4.18.5.2.2. Arai-otosu ‘wash-remove’ and damashi-toru ‘deceive-take’

In the following two examples, there is a more basic adjustment of the semantic

structure of V1; the arguments that appear as the direct object get oblique status in the

compounds.

Arai-otosu: (wash-remove)

GA-linking
Wayne wa huku o arat-ta.
(199a) Wayne TOP clothes ACC wash-PAST
Wayne washed clothes. (V1—GA-linking)

A-linking
Wayne wa huku kara shimi o otoshi-ta.
(199b) Wayne TOP clothes GEN stain ACC remove-PAST
Wayne removed the stain from the clothes. (V2—A-linking)

A-linking
Wayne wa huku kara shimi o arai-otoshi-ta.
(199c) Wayne TOP clothes GEN stain ACC wash-remove-PAST
Wayne washed the stain off the clothes. (Compound—A-linking)

GA-linking
*Wayne wa huku o arai-otoshi-ta.
(199d) Wayne TOP clothes ACC wash-remove-PAST
Wayne washed the clothes away. (Compound--*GA-linking)

The semantic representation for arai-otosu ‘wash-remove’ is shown in Figure 42.

371
 Chapter 6 Verbs of removing

Figure 42. semantic structure for acheivement construal of A-linking of arai-otosu

‘wash-remove’ in (199)
V1 V2

clothes
exist
kara

clothes move away

exist (disappear)
stain
arau’wash’ otosu ‘remove’
washing
mother activity mother remove
Δ Δ

t t

clothes exist
move away

arai-otosu
stain ‘wash-remove’
arau otosu

mother remove by
washing
Δ

372
 Chapter 6 Verbs of removing

The semantic structure for arau ‘wash’ represents it as a processual activity

without any specified change on the direct object since it is not necessary for it to

become clean after washing even though becoming clean is a normally expected result.

The event denoted by otosu ‘remove’ is represented as an achievement for

convenience’ sake, as it is difficult to imagine the accomplishment construal from the

examples above, though they may get the accomplishment construal with different

kinds of figure and ground. The semantic structure of the V1, which on its own takes

the ‘ground’ (location) as the direct object, undergoes major changes in the causal

order and is realised as an oblique in the composite structure. It should also be noted

that the original linking of the V1 is prohibited in compound (see (199d)).

Here is another example where the acted on participant of the V1 is put at the end

of the causal order in the compound.

Damashi-toru (deceive-take):

Oliver wa yuujin o damashi-ta.

(200a) Oliver TOP friend ACC deceive-PAST
Oliver deceived his friend. (V1—GA-linking)

Oliver wa yuujin kara kane o tot-ta.

(200b)Oliver TOP friend ABL money ACC take-PAST
Oliver took money from his friend. (V2—A-linking)

Oliver wa yuujin kara kane o damashi-tot-ta.

(200c) Oliver TOP friend ABL money ACC deceive-take-PAST
Oliver took money from his friend by deceiving him/her.
(Compound—A-linking)

*Oliver wa yuujin o damashi-tot-ta.

(200d) Oliver TOP friend ACC deceive-take-PAST
Oliver deceived and took his friend. (Compound—*GA-linking)

Damasu ‘deceive’ is a two-argument verb to which a source phrase is not added. It

takes as the direct object a person who is capable of the mental activity of being

deceived. The representation for the compound is shown in Figure 43.

373
 Chapter 6 Verbs of removing

Figure 43. semantic structure for the achievement construal of A-linking of damashi-
toru ‘deceive-take’ in (200)

V1 V2

friend
exist
be deceived kara
friend taken away

money
damasu toru ‘take’
‘deceive’
Oliver deceive Oliver
take away
Δ Δ

t t

friend
exist
kara
taken away

money
damashi-toru damasu toru
‘deceive-take’
take away
Sue by deceiving
Δ

374
 Chapter 6 Verbs of removing

The V1 refers to the event of the agent’s making a friend believe what he says by

mentally acting on him. The V2 is a general ‘taking away’ action. As a result, the

compound itself refers to the event which is causally complicated as it does not mean

the physical taking away by force, unlike nusumi-toru ‘steal-take’, which does not

entail the possessor’s willingness. The agent acts on a person (the figure) mentally,

that person decides to part with the possession (the ground) and acts on that

possession to change the possessor; it goes to the agent and he takes it. However, in

the syntax, the chain of these subevents is represented as the simple argument

structure of the compound. It is still construed as the extended sense of spatial

removal rather than the event of possessional transfer. The argument linking does not

encode the victim’s role as the deceived person giving money to the deceiver.

Interestingly, even though in the real situation, the agent acts on the victim (the

ground) first, the latter (the ground) is represented as the last participant in the causal

order of the compound, from which the figure is taken away. Again, the example

shows that the V2 is the determinant factor in terms of argument linking and the

causal order. When the V1 elaborates the V2, the first acted upon participant of the V1

is placed at the end of the causal order of the compound. What is realised as the direct

object of the V1 is marked as oblique in the compound, which prohibits the original

linking of the V1 (see the example (200d) above). This “hybrid” type (mixture of

verbs of FA-type linking and GA-type linking respectively) of means compounds is

rarely seen in verbs of putting14.

14
One might say nui-tsukeru ‘sew-attach’ is similar to the type (Figure 45, §5.4.17.5.2). However, nuu
‘sew’ has different senses as an V1 and in a compound, so we cannot say it is the alternate causal order
case. (In verbs of removing, the ground of V1 and V2 are the same).

375
 Chapter 6 Verbs of removing

6.4.19. Type 19 Haki-kiyomeru ‘sweep-cleanse’ verbs: means compound 3

6.4.19.1. Members

The following are the two members of this verb type:

arai-kiyomeru ‘wash-cleanse’ = make clean by washing

haki-kiyomeru ‘sweep-cleanse’ = make clean by sweeping

6.4.19.2. Semantics

Like other means compounds, the V2 elements refer to events of removing (in a

general way) and the V1 elements specify the means of the action. In this subclass, the

V2s are verbs that take GA-linking.

6.4.19.3. Syntactic patterns

Inheriting the argument linking pattern from their V2, the compounds take GA-linking,

but not FA-linking type (such as A-linking or G-linking). The figure is interpreted

implicitly as an entity that is removed away, such as dust.

GA-linking
Soujihu wa butai o haki-kiyome-ta.
(201) cleaner TOP stage ACC sweep-cleanse-PAST
The cleaner made the stage clean by sweeping.

*A-linking
*Soujihu wa butai kara hokori o haki-kiyome-ta.
(202) cleaner TOP stage ABL dust ACC sweep-cleanse-PAST
*The cleaner made the dust clean from stage by sweeping.

*G-linking
*Soujihu wa butai no hokori o haki-kiyome-ta.
(203) cleaner TOP stage GEN dust ACC sweep-cleanse-PAST
*The cleaner made the dust from stage clean by sweeping.

376
 Chapter 6 Verbs of removing

6.4.19.4. Causal and aspectual patterns

With these compounds, the agent acts on the surface (the ground) to clear it of

impurities. The impurities do not appear syntactically. The aspectual pattern follows

that of kiyomeru ‘cleanse’. This belongs to a subclass of type 7, and it does not have

the activity in progress reading with te-iru unless in the runup achievement reading. It

favours the retrospective reading, and also does not take the adverbial sukoshizutsu

‘little by little’ or tochuumade ‘in halfway’15. It is an achievement verb specifying the

final change of state of an entity. Compounds with this verb have similar

characteristics. The following examples illustrates:

??Soujihu wa butai o haki-kiyome-te-iru.

(204) cleaner TOP stage ACC sweep-cleanse-TE-IRU
??The cleaner is making the stage clean by sweeping.
The cleaner has cleaned the stage by sweeping. (run-up achievement)

??Soujihu wa butai o tochuumade haki-kiyome-ta.

(205) cleaner TOP stage ACC halfway sweep-cleanse-PAST
The cleaner cleaned the stage by sweeping in halfway.

6.4.19.5. Semantic representations

The following examples are the two linking patterns (GA and G-linkings) of the V1,

GA-linking of the V2 and GA-linking of the compound.

Soujihu wa butai o hai-ta.

(206a) cleaner TOP stage ACC sweep-PAST
The cleaner swept the stage. (V1—GA-linking)

Soujihu wa butai no gomi o hai-ta.

(206b) cleaner TOP stage GEN dust ACC sweep-PAST
The cleaner swept dust on the stage. (V1—G-linking)

Soujihu wa butai o kiyome-ta.

(206c) cleaner TOP stage ACC cleanse-PAST
The cleaner cleansed the stage. (V2—GA-linking)

15
I have marked the example sentences with double question marks. As the V1 has an activity construal,
the compound sounds slightly better in the sentences than kiyomeru ‘purify’ does on its own.

377
 Chapter 6 Verbs of removing

Soujihu wa butai o haki-kiyome-ta.

(206d) cleaner TOP stage ACC sweep-cleanse-PAST
The cleaner made the stage clean by sweeping. (Compound—GA-linking)

The proposed representation for haki-kiyomeru ‘sweep-purify’ is Figure 44.

378
 Chapter 6 Verbs of removing

Figure 44. semantic structure for achievement construal of GA-linking of haki-

kiyomeru ‘sweep-purify’ in (206)

V1 V2

stage be clean
exist stage
haku
‘sweep’ kiyomeru ‘purify’
sweeping
cleaner activity cleaner clean
Δ Δ

t t

clean
stage
haki-kiyomeru
‘sweep-clean’
haku kiyomeru
clean by
cleaner sweeping activity
Δ

379
 Chapter 6 Verbs of removing

Haku2 ‘sweep’ can have both GA-type linking and FA-type linking (G-linking). I take

the view that in this case, the semantic structure of the GA-linking of the V1 is

superimposed by that of V2 with concomitant change in aspectual pattern from

undirected activity to achievement.

6.4.20. Other verb type

As in the case of verbs of putting, there are quite a few verbs in the data from other

classes such as transfer of possession. Among them, there is one interesting verb class

which is treated as verbs of removing in English, but should be categorised under

another class in Japanese. I would call this ‘mental request’. There are five verbs in

this class.

nedaru (A beg F from G, A beg G of/for F)

sebiru (A pester G for F)
takaru (A bum F off G)
kyou.you-suru (A exact F from G)
you.kyuu-suru (A deman F from/off G)

In Japanese, the above five verbs require the figure (an entity that is asked for) to be

marked as o (Accusative) and the ground (a person who is asked) to be marked as ni.

Otouto wa ryoushin ni atarashii jitensha o nedat-ta.

(207) brother TOP parents ALL new bicycle ACC beg-PAST
My brother begged a new bicycle from my parents.

English construes these situations as the act of an agent’s getting or trying to get the

figure from the ground (provider). Japanese construes them as the acts of the agent’s

making a request to the ground. I am not sure about the proper causal structure for this

at this moment. However, this clearly shows that similar situations are denoted by

different classes of verbs in English and Japanese.

380
 Chapter 6 Verbs of removing

6.5. Summary

Verbs of removing are classified into 19 categories according to syntactic behaviours

and semantics. The summary of the classification is in Appendix F. Like the table of

verbs of putting (Appendix C), it shows (1) the class number, and (2) subclass number,

(3) the name of the types, (4) the number of the members, (5) syntactic patterns, (6)

acceptability of the goal phrase, (7) other syntactic patterns manifested, (8) aspectual

patterns, and (9) other prominent characteristics. In syntactic patterns of verbs of

removing, there are A-linking, G-linking, and FA-linking as FA-types, and RA-

linking and GA-linking as GA-types. Besides the above features, I also present (10)

the semantic relations between the figure and the ground.

Japanese does not have the locative alternation as in some English verbs of

removing which alternate between the Path construction and the antecedent Oblique

Of construction (cf. Croft 2000, §4.2). That is, no verbs of removing in Japanese

alternate between the A-linking construction and the RA-linking construction. Instead,

some verbs alternate between FA-type linking and GA-linking. A typical example is

type 3 the katazukeru ‘clear’ verb. As in the discussion in verbs of putting, I analyse

how different linkings reflect different construals. When an event in question is an

accomplishment or directed activity, the verbal scale is associated with different

participants; the figure with FA type linkings and with the ground with GA type

linking. When there is an achievement construal, the figure is construed as the one that

undergoes changes with FA-type linkings and the ground so construed with GA-type

linkings. With the undirected activity construal, the participant that is acted on in FA-

type linkings is the figure while that acted on in GA-type linkings is the ground.

381
 Chapter 6 Verbs of removing

6.5.1. Semantic structures of each argument linking construction

In summary, I shall show a schematic semantic structure for each construction.

6.5.1.1. A-linking construction

A-linking corresponds to L-linking in verbs of putting with the same ‘Path

construction’. It is the predominantly used construction for verbs of removing in

Japanese. However, it sounds rather awkward to use A-linking for the whole-part

relation or the impurities locational relation between figure and ground (see type 4 and

5) as the construction basically denotes ‘spatial’ removal. The whole-part relation is

much closer than the locational relation. It is intuitively less acceptable to think that

the skin of an apple is located on the apple or the dust on the floor.

The schematic representation is similar to that of L-linking except for the

description of the subevent of the direct object.

Figure 45. Semantic structure for A-linking

S.Obl exist

kara particle

Obj be dislocated with

Verb (be away with)

Sbj cause
Δ

382
 Chapter 6 Verbs of removing

The striped boxes allow both punctual and extended causation. The direct object either

moves away or become dislocated from the ground, which serves as a reference point.

The relationship can be causal as in verbs of possessional deprivation (Figure 22,

§6.4.8.5 type 8). Therefore, the possibility of the presence of an arrow (the causal

relation) is illustrated with the dotted grey arrow. The representation above

accommodates the directed activity construal as in Figure 1, the accomplishment

construal as in Figure 2, the achievement construal as in Figure 4, or the runup

achievement construal as in Figure 16.

6.5.2.2. G-linking construction

This construction is also frequently used. The compatibility with this construction of a

certain verb depends on the relation between the figure and the ground. G-linking is

used for locational relation, possessional relation, the whole-part relation, the

(abstract) possessional, identificational but not for the extended locational relation.

We have also seen that there is a complexity in G-linking since verbs’ lexical

meaning may also interfere with the acceptability of this construction. First of all, it

does not always designate the same relation between the figure and the ground with

A-linking. As A-linking favours the spatial removal interpretation and G-linking

denotes other relations as well, with verbs that allow FA-linking, such as type 8-1 and

type 9 verbs, it is possible for G-linking to be further embedded in A-linking

(examples (93) and (103)). Actually, G-linking is a special type of FA-linking

construction. The difference is that the figure is modified by the ground in the

Genitive case. Therefore, it is not surprising that G-linking, which is a subtype of FA-

linking, is more specific and adds more information in the A-linking construction.

Secondly, G-linking is not completely acceptable with some verbs, even though the

383
 Chapter 6 Verbs of removing

semantic relation between the figure and the ground conforms to one of the uses of the

Genitive case presented in (6). For example, type 10 ‘excommunicate’ verbs do not

allow G-linking despite the fact that the figure and the ground are in an

identificational and locational relation which no can modify (See (115) and (116)).

This means that whether verbs are acceptable with G-linking also depends on the type

of verb.

The following is the proposed analysis of the G-linking construction.

Figure 46. Semantic structure for G-linking

Gen.Obj separate away

no
Obj

Verb

Sbj cause separation

Δ

The striped boxes allow punctual and extended causation. The subject goes back to

the rest state in its subevent contour while the direct object with a Genitive marker

undergoes the change of state. I describe its subevent as “separate away” indicating

that two entities which are close together in a spatial or abstract way, or in a whole-

part relationship, are disconnected at the end of the event.

The schematic representation accommodates the accomplishment construal as in

Figure 3, the achievement construal as in Figure 23, and the runup achievement

384
 Chapter 6 Verbs of removing

construal as in Figure 24.

6.5.1.3. RA-linking construction

Reversed-A-linking corresponds to I-linking in verbs of putting. However, this

construction only occurs with type 12 Verbs. These verbs refer to very abstract

notions of removing, so, the events are punctual. The verbs of type 12, when they

occur with this construction, refer to the event of forgiving or pardoning somebody by

taking away something painful.

Figure 47. Semantic structure for Reversed-A-linking

be without A.Obl

Obj
kara particle

be taken away from

A.Obl
Verb

Sbj cause
Δ

The representation is very similar to that of I-linking, except for the aspectual

punctuality and the description of the subevents of the direct object and antecedent

oblique. The black dotted box is the profile of the verb and the grey dotted box is that

of the antecedent oblique marker.

385
 Chapter 6 Verbs of removing

6.5.1.4. FA-linking construction

Again, there are not so many examples of FA-linking constructions. The following is

the proposed semantic structure for FA-linking in verbs of removing.

Figure 48. Semantic structure for FA-linking in verbs of removing

move away or
eliminated
Obj

Verb

Sbj cause
Δ

The representation above accommodates the achievement construal as in Figure 29.

In the main text, I have not often illustrated FA-linking. Readers are asked to imagine

that the semantic structure of FA-linking is similar to that of A-linking without the

ground contour. That is, FA-linking can be realised in the accomplishment linking

(Figure 2 without the ground contour) or the directed activity reading (Figure 1

without the ground contour).

The representation above is slightly different from that of FA-linking in verbs of

putting in that the direct object contour ends up with a new state in Figure 48 above.

With verbs of putting, when the figure is material such as paint, it is possible to have

an undirected activity construal as we can add that material to somewhere over and

over again without limit. However, in verbs of removing, the amount we can take or

move away is already fixed even though it is not specifically realised. Removing

386
 Chapter 6 Verbs of removing

weeds from the garden has the natural end point; we cannot removing weeds

indefinitely unless the garden is interpreted as a derived verbal scale or unless new

weeds grow. That is, the removing/taking away of something is always directed away

from the rest state.

6.5.1.5. GA-linking construction

Unlike with FA-linking, there are a lot of examples of GA-linking with verbs of

removing. Here is the proposed semantic structure for GA-linking.

Figure 49. Semantic structure for GA-linking of verbs of removing

acted on

Obj

Verb

Sbj act on
Δ

The white box in the direct object contour indicates the profile of the event of

process and state. This is to allow the undirected activity construal as in Figure 9,

where the ground does not undergo any changes and the event is construed as mere

existence. We cannot specify whether the direct object goes back to the original state

or changes into a new state. The labelling for the subevent of the two participants also

needs to be vague reflecting the various occurrences in different construals of GA-

387
 Chapter 6 Verbs of removing

linking.

The representation accommodates an undirected activity construal as in Figures 9

and 17, the accomplishment construal as in Figure 10, the achievement construal as in

Figure 20, and the directed activity construal as in Figure 21.

6.5.2. Differences between English and Japanese

The fact that Japanese favours the Path construction is made further apparent by

reference to verbs of removing. The equivalent to the English antecedent oblique of

construction is not frequent (there are only five verbs which I found with it). The

difference in number is quite striking. English, according to Levin (1993), has fifty-

two members that have this construction. (Even the existence of the construction in

Japanese has not been discussed (Fukui et al. (1985))). Instead, most of the time, the

figure is omitted so that the events are realised with GA-linking. In these cases, it is

clear what is removed and the situation is construed as an uncovering event (as in

peeling a potato in type 5), or as a clearing event (as in clearing the table in type 3), as

an activity whose purpose is to cleanse the ground (as in washing in type 7) or as an

activity of contact to clear the ground (as in sweeping in type 4).

As in the case of verbs of putting, we have found that Japanese uses compound

verbs where the V2 is equivalent to English prepositions.

6.5.3. Asymmetries between verbs of putting and removing found in Japanese

Finally, I discuss semantic asymmetries between “putting” event and “removing”

event that are reflected in differences in syntactic behaviour between verbs of putting

and removing in Japanese. Dowty (1991: 592) points out the asymmetry of the physics

388
 Chapter 6 Verbs of removing

of space in discussing verbs of putting and removing; in an event of removing, the

relation between the figure and the ground is one of correlation. Since the figure

occupies a certain place (the ground) and the ground is occupied with the figure,

removing the figure from the ground entails emptying the ground and emptying the

ground entails removing the figure from it. On the other hand, in an event of putting,

we can put the total amount of the figure in/on the ground without filling/covering it

or apply the figure to the ground indefinitely after the ground is covered (such as paint

on the wall). This difference is based on knowledge about the real world. Differences

of this kind are further illuminated by the analysis I have undertaken in Chapter 5 and

Chapter 6. Appendix H compares the two classes of verbs.

There are some correspondences between verbs of putting and verbs of removing

such as in the case of figure or ground incorporation verbs, and compound verbs.

Caused-spatial transfer with specified direction verbs of type 2 verbs or means

compound 1 types can be cross-listed. In the table, corresponding categories from two

verb classes are grouped together by thick black lines. P indicates verbs of putting and

R indicates verbs of removing. P2, for example, refers to type 2 verbs of putting. P2

and R2 are corresponding verb classes and therefore grouped together. I also include

the number of the members of each type though the list is not exhaustive.

P-verbs have L-linking and R-verbs have A-linking, but these two linking

constructions correspond in having the figure in the accusative and the ground as

oblique. G-linking is peculiar to R-verbs, and FA-linking is shared by the two classes.

Likewise, I-linking and RA-linking correspond to one another in that the ground is

accusative and the figure is oblique in both cases. The columns for GA-linking and

GO-linking are shared. The property of P-verbs occurring with a source phrase and

that of R-verbs occurring with a goal phrase are represented side by side in Appendix

389
 Chapter 6 Verbs of removing

H.

On the first page of Appendix H, P2 and R2 types are grouped together. Basically,

verbs of these classes can be cross-listed since they can take goal phrases only or

source phrases only and can appear with a path pattern as well. That is, verbs of the

P2-2 type can appear as verbs of removing and verbs of the R2-2 as verbs of putting.

However, as I discussed in the main text, some of them such as dasu ‘take out’ which

tend to appear with the source phrase, are categorised as R2-1 type. R2-2 type verbs

tend to occur with source phrases, so they are also categorised as verbs of removing.

What is peculiar is that G-linking may only be acceptable (though not perfectly

grammatical) between the ‘source’ ground and the figure and not between the ‘goal’

ground and the figure. It may be possible to say (6.28) John took a boat in the shore

out (I put one question mark), but it is definitely ungrammatical to link the ‘goal’

ground and the figure, as the ungrammaticality of the following sentence shows.

*John wa kishi no kobune o ire-ta.

(208) John TOP shore GEN boat ACC put.into
*John put a boat in the shore in.

(208) is ungrammatical in the sense that kishi ‘shore’ cannot be interpreted as the

place the boat is put into but its original place.

P13 and R14 correspond in that they are both ground incorporation verbs.

However, there is a big contrast in their syntactic behaviour. P13 selects FA-type

linking and R14 selects GA-type linking. This is because the first element of P13

verbs is a verb that takes L-linking (such as ‘put into’ and ‘throw’) while that of R14

is a verb that takes GA-linking (‘wash’). Interestingly, I have not so far found ground

incorporated verbs of putting whose first element is a GA-linking type verb (such as

huku.sha-suru meaning ‘cover.car-do’ (cover car)) or ground incorporated verbs of

390
 Chapter 6 Verbs of removing

removing whose first element is an FA-linking type verb (such as shuk.koku-suru 16

meaning ‘take out.country-do’ (remove a person from a country)).

P14 and R15 correspond in that they are figure incorporation verbs which allow

only FA type linkings, and P15 and R16 correspond as figure incorporated verbs that

allow both FA-type and GA-type linkings. The difference is that the P-class takes GO-

linking while the R-class does not. P15 and R16 also differ in their acceptance of FA-

linking.

P16 and R17 are means compounds with specified direction. Basically, some verbs

in these categories can be cross-listed. Some of them such as ~komu ‘put into’ or

~dasu ‘take out’, however, predominantly belong to either of the categories. Types

P17 and R18 are both means compounds that take FA-type linking, and P18 and R19

are both means compounds that take GA-type linking. R18 allows FA-linking

constructions while P17 does not. In the means compounds section (type 17 and 18

verbs of removing), we found some ‘hybrid’ compounds whose component verbs have

opposed linking patterns such as with V1 taking GA-linking type linking and V2

taking FA-type linking. In verbs of putting, V1s mostly combine with V2s when the

two have the same argument linking type. V1s may have different argument linking

types from V2s in compound verbs of putting. Even in this case, however, V1s are

limited to verbs which denote contact in some way (such as type 20-4 or type 20-6 in

verbs of putting). It is rare to see means compounds of verbs of putting whose V1 is

an I-linking type of verbs of putting and whose V2 is an L-linking type of verb (the

only example in the data is ooi-kabuseru ‘cover-put on’). On the other hand, verbs of

removing include more of this hybrid type of means compound (arai-otosu ‘wash-

remove’ and damashi-toru ‘deceive-take’).

16
Japanese has this verb but it is only used in an intransitive verb meaning ‘go out of the country’.

391
 Chapter 6 Verbs of removing

The P-class and R-class verbs that take FA-type linking are P1, P3, P4, P5, P6, P7, P8,

R1, R6, R8, R9, R10, R11, and R13. Comparing these types, there are more types and

numbers of verbs that allow FA-linking in verbs of removing. Verbs of putting have

only a limited number of verbs of scattering (such as maku3 ‘scatter’) and verbs with a

specified configuration that take FA-linking. FA-linking in the latter is possible

because verbs can focus on the configurational state of the figure only, without

specifying where it is.

The verbs that take GA-type linking only are P9, P10, R7. In terms of syntactic

linking patterns, they are almost equivalent (cf. Appendix H). The change of state

verbs of yogosu ‘dirty’ and kiyomeru ‘cleanse’ correspond semantically; one is the

event of making an entity not clean and the other is the event of making it clean. The

two verbs are punctual (achievement). Taking the two opposite events of non-

cleaning/cleaning in terms of ‘putting dirt’ and ‘removing dirt’, there is an interesting

aspectual constrast. As yogosu ‘dirty’ refers to causing something to change from a

clean to a non-clean state, one act of putting dirt is enough to make it dirty. On the

other hand, kiyomeru ‘cleanse’, which refers to causing something to change from a

non-clean to a clean state, requires all dirt to be removed from that entity. This

difference is exemplified in the interpretation with te-iru. As achievement verbs, they

need construals of events that get the activity reading with te-iru. Yogosu ‘dirty’ only

allows an iterated achievement reading, since the first act of putting dirt completes an

event. Kiyomeru ‘cleanse’ only allows the runup achievement reading, since the event

is not completed till the last dirt on something is removed.

392
 Chapter 6 Verbs of removing

Finally, the following are the verbs that allow both FA and GA-type linking: P11,

P12, R3, R4, R5, and R12. There are more types and numbers of verbs that allow both

these linking types in verbs of removing than in verbs of putting.

Differences observed between verbs of putting and verbs of removing are

summarised as follows:

(i) the presence of G-linking in removing

(ii) more acceptance of FA-linking in removing
(iii) more verbs that have alternative linkings in removing
(iv) more hybrid compounds in removing (GA type + FA type)
(v) much less RA-linking compared to I-linking
(vi) the presence of GO-linking in putting
(vii) difference in ground incorporation verbs

First of all, the togetherness of the figure and the ground, both in purely spatial

terms and more abstract terms, in verbs of removing reflects most of the differences

above. Self-evidently, the starting rest state of an event of removing and putting is

different. For an event of removing, the figure and the ground are supposed to be

together; they might be in a whole-part relation, a locational relation, or an

identificational relation etc. On the other hand, in an event of putting, the figure and

the ground are supposed to be separate. Because of the togetherness, G-linking is

possible with verbs of removing. The higher acceptability of FA-linking in verbs of

removing is also explained by the togetherness. Because the figure is supposed to be

in spatial or abstract contact with the ground, it can be thought merely to exist in the

world (togetherness with the world). That is, an act of removing without specifying a

goal phrase can be interpreted as just the elimination or non-existence of the figure in

this world. Actually, verbs of removing include verbs of elimination (type R1-1).

However, verbs of putting do not include any ‘creation’ verbs as the figure’s existence

is already presupposed in verbs of putting. On the other hand, the putting class has the

393
 Chapter 6 Verbs of removing

type 7 verbs of specified configuration which focus on how the figure comes to relate

spatially to the ground. Verbs of removing do not have this counterpart since the kind

of spatial relation the figure is in relative to the ground before being removed is

already presupposed/included in the rest state which verbs of removing do not profile.

If one wants to express it, one has to use a periphrastic expression that refers to the

configurational position of the figure in relation to the ground.

The semantic notion of togetherness also explains why there are more verbs that

can occur with either FA-type linkings or GA-type linkings and more hybrid

compound types in verbs of removing. As the figure and the ground are spatially or

sometimes abstractly very close (they even comprise a same entity), an event of

removing is in reality acting on the figure and the ground at the same time and

therefore it is construed as either acting on the figure or on the ground. For example,

in peeling the skin from a potato, one is acting on the skin and the potato which has

the skin as its part and in sweeping dust on the floor, one is acting on the dust and the

floor simultaneously. If one wants to take something (the figure) away from a person,

one can act on that person (like deceiving) first, to get the figure. On the other hand, in

situations denoted by verbs of putting, one has to act on the figure first. In the real

world, it is impossible for an agent to act on the ground first to let the figure be

located or moved to the ground. One cannot act on the wall, for example, to make

paint be smeared there. That is, in verbs of removing, which participant the agent acts

on first is very ambiguous because of the togetherness of the figure and the ground;

the causal relation can be very blurred.

There is a further interesting contrast between putting and removing concerning

togetherness. It is relatively easier to separate two things which are together rather

than to put separate things together. For example, we can remove a skin from a potato

394
 Chapter 6 Verbs of removing

but we cannot put them back together again. Actually, there are more types of verbs of

putting with reversible events than there are verbs of removing with them. We can see

by comparing the Figures representing semantic structures. It is a lot more likely that

we can separate again what we artificially or causally locate together than that we can

put together again what we separate. We can put the vase in the living room and can

take it away again, but we cannot remove weeds from the lawn and put them back

again. The difficulty of “putting” things together is further illustrated in a more

abstract relation of the figure and the ground. Verbs of removing include

“possessional deprivation verbs” where the agent forcibly takes the possessum away

from the possessor. Verbs of putting do not have these counterparts which refer to an

event where the agent forcibly locates the figure onto the ground (‘intended

possessor’). We do not need any causal contribution from the “possessor” to deprive it

of something, although, we need cooperation on his part if we want the ‘intended’

possessor (the ground) to have the ‘intended’ possessum (the figure). In this case, an

event cannot be categorised as a putting event but a transfer of possession event

because of the causal contribution of the ground.

The fact that the RA-linking construction is much rarer than I-linking is specific to

Japanese, as the antecedent oblique of construction (equivalent to RA-linking) often

appears in English. I would argue that this is because it is intuitively more difficult to

construe a removing event as affecting the ground by using the figure than to construe

a putting event as such based on the assumption that the figure is a kind of instrument

or means to cause changes in the ground in both the RA-linking construction and the

I-linking construction. It is easier to think of the figure as an instrument to be applied

to the ground in verbs of putting than in verbs of removing because of the

“togetherness” of the figure and the ground in the latter. For example, if we compare

395
 Chapter 6 Verbs of removing

filling a bathtub using water (by putting it in) and emptying a bathtub using water (by

removing it), the latter is intuitively more strange. Japanese conventionally has a

construction to express the former but not the latter.

Secondly, there is a different degree of focus on the goal and the source in spatial

transfer. Theoretically, it is possible for both phrases to appear with types P2 and R2

(verbs with specified direction), with types P14 and R15 (figure incorporation 1), or

with types P16 and R17 (means compound 1). As the number of these categories show,

there are more in verbs of putting (especially, compare P14, R15 and P16, R17).

Though they can be cross-listed, I have noticed that these verbs of the spatial transfer

with specified direction types are more likely to choose the goal phrase over the

source phrase (the exception is only dasu ‘put out, take out’). I would say that it is

because we naturally focus on the result state (the goal) as the result of transfer rather

than the rest state (the source), which specifies the original state before the spatial

transfer. This asymmetry may be related to the absence of ground incorporation verbs

whose ground indicates a referential place from which the figure moves or is removed.

Even the VN verbs whose first element is classified as a verb of removing such as hou

‘release’ in hou.boku-suru ‘release.pasture-do’ are linked with the ‘goal’ ground rather

than the ‘source’ ground. Verbs that can occur with a path pattern generally are more

likely to take the goal phrase than the source phrase.

Finally, the presence of GO-linking in verbs of putting may be explained by the

different degree of affectedness. The affectedness the ground gets with spatial transfer

verbs in being located with something like filling may be higher than when it is

cleared of something. Therefore, it is possible that only the place marked as oblique

can appear syntactically as the sole argument other than the subject in figure-

incorporation verbs of putting. Another explanation may be that it is something to do

396
 Chapter 6 Verbs of removing

with the use of ni. This can be used for the object which is acted on, not unlike in the

English conative construction as in ‘chew at the bone’. Japanese uses ni for a

participant to which English assigns accusative such as kare ni au (he OBL meet =

‘meet him’), where it indicates a participant that the subject action is directed to (cf. §

4.4.1.2.1). On the other hand, kara ‘from’ does not have this use. That may be a

reason GO-linking (with ni) is present in verbs of putting.

Thus, verbs of putting and removing, which can be thought of as semantic

counterparts, do not behave exactly in the opposite way syntactically. This is because

they reflect the way the world is, where an event of putting and removing in close

investigation are not symmetrical in terms of ‘togetherness’, ‘affectedness’ or the

tendency to focus on the ‘result’ goal.

397
 Chapter 7 Conclusion

7. CONCLUSION

In this thesis, verbal semantics have been represented in terms of the two important

dimensions, aspect and argument structure, in the causal-aspectual model (Croft 2000),

which has been applied to analyze verbs of putting and removing in Japanese.

The dimension of aspect is represented by Croft (2000) as a two-dimensional

model, which has the time scale on the abscissa and the qualitative scale on the

ordinate. The two dimensional model is motivated by the fact that a construal

operation of a speaker may require adjustments on both scales as in the case of

construals of events that are habitual.

Aspect is argued to have more types than the four presented by Vendler (1967).

States, activities, accomplishments, and achievements are further classified into

fourteen types according to possible aspectual contours which are represented in the

two-dimensional aspectual model. The fourteen types are built up from a set of basic

phase types (i-state, t-state, p-state, d-transition, r-transition, u-process, d-process).

These fourteen types are considered to be universal aspectual types. Aspectual types

are distinguished from aspectual classes. Aspectual classes are language-specific and

divided in terms of their distributional behaviour with TA-constructions which are

available in each language.

In Chapter 3, forty-eight situation types of Japanese predicates are classified into

forty aspectual classes according to their behaviour in relation to the Present, the

Te-iru, and the Past constructions. With the closer analysis of situation types of

predicates that occur in these TA constructions, sometimes in combination with

adverbials, the different senses of the three constructions are also revealed. In the

event, the Present construction was shown to have four senses, the Te-iru construction

398
 Chapter 7 Conclusion

to have eight senses, and the Past construction to have eight senses. A single unified

definition to cover the different senses of each construction has also been proposed

briefly in Chapter 3. The next step would be to represent a single meaning of each

construction as a schematic aspectual type in the two-dimensional representation. I

leave this for future study.

Argument structure, which is the other dimension of verbal semantic structure,

determines argument linking. It is argued in Croft (1990, 1991, 1993, 1994ab, 1995ab,

1998a, 1999a) to derive from the event structure, which is represented by the causal

structure of events. Chapter 2 presented his causal structure model. He claims that the

force-dynamic relationship between participants in events has a crucial role in

determining argument linking. Participants which are ranked in the causal order are

mapped into syntactic arguments via the linking rules. Croft also uses the notion of

verbal profile to explain subtle semantic differences between variants of alternations

(such as the locative variant and the with variant in the spray/load alternation). In the

case of the spray/load alternation, for example, it is argued that since the with variant

includes the ground in the verbal profile, the variant has the holistic interpretation.

In Chapter 4, I introduced the new verbal semantic representation of Croft (2000),

which integrates causal structure and aspectual structure (the causal-aspectual model).

In this new model, each participant in a situation has its own aspectual contour, and

the subevents are causally ordered. The causal-aspectual model can specify the

aspectual properties of each subevent a participant undergoes, as well as representing

the causal structure which shows the force-dynamic relationship between participants

that determines argument linking. Moreover, each subevent must be described

specifically. The causal-aspectual model can articulate and describe events that verbs

denote more clearly than the old model of the causal structure. The locative variant

399
 Chapter 7 Conclusion

and the with variant of spray, for example, are explained not only to have two

different verbal profiles as the old causal analysis did, but also to have different

subevents. Also, the verbal scale is associated with different participants in the two

variants. All these differences now capture the fact that spray has two senses

according to the different variants, an “emitting” sense in the locative variant and a

“covering” sense in the with variant.

In Chapter 5 and Chapter 6, verbs of putting and removing are analysed in terms of

the causal-aspectual model. Verbs are subcategorised into small classes according to

their semantics and the linking patterns they take. For verbs of putting, five argument

linking constructions are used: the L-linking, I-linking, FA-linking, GA-linking and

the GO-linking constructions. For verbs of removing, five argument linking

constructions are used, only two of which overlap with those for verbs of putting: the

A-linking, G-linking, RA-linking, FA-linking and the GA-linking constructions. The

semantic structures of verbs occurring with each construction have been represented

in the causal-aspectual model. A single verb that occurs in different constructions is

argued to have different semantic structures in the causal-aspectual model, reflecting

different construals of the situation by speakers. After examining various occurrences

of verbs with constructions, I proposed a semantic structure in the causal-aspectual

model for each construction. These semantic structures for the constructions are

schematic and abstract reflecting the nature of the constructional meanings, which are

not as concrete and not as rich in encyclopaedic definitions as those of verbs.

Through analysing verbs of putting and removing in Japanese, some systematic

differences between English and Japanese were found. Japanese predominantly

favours the locative variant (the L-linking and A-linking constructions), which marks

the figure as the direct object and the ground as the subsequent path oblique. This fact

400
 Chapter 7 Conclusion

is reflected by the number of Japanese verbs that take the with-variant and the

antecedent of variant (I-linking and RA-linking in Japanese). The number is very

small compared to that in English. Moreover, there are many fewer verbs that take the

locative alternation in Japanese than there are in English. Even in some figure

incorporation verbs, Japanese marks its single argument other than the subject, the

ground, as oblique, which further illuminates the characteristic of the language

favouring the locative type construction. Another notable difference is that Japanese

frequently uses compound verbs for the caused-motion situations that English encodes

as a verb plus a prepositional phrase. A path expression that is represented by an

English preposition is introduced by a V2 component which denotes the

caused-spatial transfer situation in Japanese. Various kinds of verbs can appear as V1

in Japanese just as English main verbs vary in the caused-motion construction. How

V1 and V2 are combined into a compound is also shown in terms of the

causal-aspectual model.

Finally, syntactic asymmetries between verbs of putting and removing such as the

presence/absence of a certain construction have been discussed in terms of the

differences between the semantic natures of events that the two verb classes denote.

The thesis presents and attempts to apply the causal-aspectual model (Croft 2000)

to the Japanese language. However, this would be a monumental task; the dissertation

has a much narrower scope. It has limited itself to examining in some detail the three

TA constructions (the Present, the Te-iru, and the Past constructions), forty-eight

situation types of predicates, the two classes of verbs (verbs of putting and removing)

and the related linking constructions such as L-linking, A-linking, I-linking and

RA-linking constructions. Further investigation is therefore appropriate in the

following areas: (i) defining meanings of other TA constructions such as the container

401
 Chapter 7 Conclusion

adverbial, the durative adverbial, and the Te-ita (Te-iru plus Past) constructions in

terms of aspectual representation, (ii) investigating more situation types of predicates

in terms of their aspectual behaviour in relation to various TA constructions, and (iii)

examining more verb classes with three arguments such as verbs of possessional

transfer, verbs of combining and attaching, verbs of separating, verbs of contact by

impact, verbs of sending and carrying, and touch/poke verbs in terms of what kind of

linking constructions they take. Such investigation will be left to future research.

402
REFERENCES

Anderson, Stephen R. 1971. On the role of deep structure in semantic interpretation.

Foundations of Language 7. 387-396.

Croft, William. 1990. Possible verbs and event structure. Meanings and prototypes:
Studies on linguistic categorization, ed. S. L. Tsohatzidis, 48-73. London:
Routledge.

Croft, William. 1991. Syntactic categories and grammatical relations: The cognitive
organization of information. Chicago: University of Chicago Press.

Croft, William. 1993. Case marking and the semantics of mental verbs. Semantics and
the lexicon, ed. James Pustejovsky, 55-72. Dordrecht: Kluwer Academic.

Croft, William. 1994a. The semantics of subjecthood. Subjecthood and subjectivity,

ed. Maria Yaguello. Paris: Ophrys.

Croft, William. 1994b. Voice: beyond control and affectedness. Voice: form and
function, ed. Paul Hopper and Barbara Fox, 89-117. Amsterdam: John Benjamins.

Croft, William. 1995a. Argument linking and event structure. Paper presented to the
Department of Linguistics, University of Pavia, Pavia, Italy.

Croft, William. 1995b. Construal of Events and Argument Linking. Paper presented at
the Functional Approaches to Grammar Conference, Albuquerque NM, USA.

Croft, William. 1998a. Event structure in argument linking. The projection of

arguments: lexical and compositional factors, ed. Miriam Butt and Wilhelm
Geuder, 21-63. Stanford: Center for the Study of Language and Information.

Croft, William. 1998b. Linguistic evidence and mental representations. Cognitive

Linguistics 9. 151-73.

Croft, William. 1999a. Event structure and the grammar of verbs. Paper presented in
the Workshop on Language and Thought, Department of Psychology, University
of Chicago, Chicago USA.

Croft, William. 1999b. Lexical rules vs constructions: a false dichotomy. To appear

for a festschrift in 1999.

Croft, William. 2000. Verbs: aspect and argument structure. Manuscript, July 19,
2000.

Croft, William, Chiaki Taoka and Esther J. Wood. 2000. Argument linking and the
commercial transaction frame in English, Russian and Japanese. To appear in
Language Sciences.

DDWin CD-ROM Dictionary version 1.56. 1996. Version 1.56. Tokyo: NIFTY-Serve.
403
Dowty, David. 1979. Word meaning and Montague grammar. Dordrecht: Reidel.

Dowty, David. 1991. Thematic proto-roles and argument selection. Language 67. 547-
619.

Fauconnier, Gilles. 1985. Mental spaces: aspects of meaning construction in natural

language. Cambridge: the MIT Press.

Fillmore, Charles J. 1968. The case for case. Universals in linguistic theory, eds.
E.Bach and R.T Harms. 1-88. New York: Holt, Rinehart and Winston.

Fillmore, Charles J. 1970. The grammar of hitting and breaking. Readings in English
transformational grammar, eds. R. Jacobs and P. Rosenbaum. Waltham, MA:
Ginn.

Fillmore, Charles J. 1977. The case for case reopened. Syntax and semantics 8:
grammatical relations. 59-81. New York: Academic Press.

Fillmore, Charles J. 1976. Frame semantics and the nature of language. Origins and
evolutions of language and speech, eds. S. Harnad, H. Steklis, and J. Lancaster,
20-32. New York: NY Academy of Sciences.

Fillmore, Charles J. and Paul Kay. 1993. Construction Grammar Coursebook,

Chapter 1 through11 (Reading Materials for Ling. X20). University of California,
Berkeley.

Fukui, N., S. Miyagawa and C. Tenny. 1985. Verb classes in English and Japanese: a
case study in the interaction of syntax, morphology and semantics. Lexicon Project
Working Papers 3, Cambridge, MA: Center for Cognitive Science, MIT.

Fujii, Tadashi. 1966. “Doosi + te-iru” no imi [The meaning of “verb + te-iru”].
Reprinted in H. Kindaichi (ed.), Nihongo doosi no asupekuto [Aspect in Japanese
verbs] (1976), 97-116. Tokyo: Mugi Shoboo.

Genius English-Japanese Dictionary CD-ROM version 1.11. 1997. Tokyo: Taishukan.

Goldberg, Adele E. 1995. Constructions: A construction grammar approach to

argument structure. Chicago: University of Chicago Press.

Goldberg, Adele E. 1996. Jackendoff and construction-based grammar. Cognitive

Linguistics 7. 3-19.

Gropen, Jess, Steven Pinker, Michelle Hollander, and Richard Goldberg. 1991.
Affectedness and direct objects: the role of lexical semantics in the acquisition of
verb argument structure. Cognition 41. 153-195.

Hay, Jennifer, Christopher Kennedy and Beth Levin. 1999. Scaler structure underlies
telicity in “degree achievements”. Proceedings of SALT 9, 127-44.

404
Hopper, Paul J., and Sandra A. Thompson. 1980. Transitivity in grammar and
discourse. Language 56. 251-299.

Ikegami, Yoshihiko. 1981. Suru to naru no gengogaku. [Linguistics of “do” and

“become”]. Tokyo: Taishuukan.

Ikegami, Yoshihiko. 1985. ‘Activity’—‘Accomplishment’—‘Achievement’—A

language that can’t say ‘I burned it but it did not burn’ and one that can.
Linguistics and Philosophy: essays in honor of Rulon S. Wells, ed. A. Makkai and
A. K. Melby, 265-304. Amsterdam: John Benjamins.

Iwaoka, Toyoko and Kiwami Okamoto. 1993. Japanese for Foreigners: Innovative
Workbooks in Japanese 3—Verbs. Tokyo: Aratake Shuppan.

Jackendoff, Ray. 1983. Semantics and Cognition. Cambridge, MA: MIT Press.

Jackendoff, Ray. 1990. Semantic structures. Cambridge, Mass.: MIT Press.

Jackendoff, Ray. 1996a. Conceptual semantics and cognitive linguistics. Cognitive

Linguistics 7. 93-129.

Jackendoff, Ray. 1996b. The proper treatment of measuring out, telicity, and perhaps
even quantification in English. Natural Language and Linguistic Theory 14.
305-354.

Jackendoff, Ray. 1997. Twistin’ the night away. Language 73. 534-59.

Jacobsen, Wesley M. 1989. Tadousei to purototaipu ron [Transitivity and prototype].

Nihongogaku no Shintenkai [The new development of the studies of the Japanese
langauge], eds. Akira Kuno and Masayoshi Shibatani. Tokyo: Kurosio Publishers.

Jacobsen, Wesley M. 1992. The transitive structure of events in Japanese. Tokyo:

Kurosio Publishers.

Kageyama, Taro. 1980. The role of thematic relations in the spray paint hypallage.
Papers in Japanese Linguistics 7: 35-64.

Kageyama, Taro. 1993. Bunpo to Go Keisei. [Grammar and Word Formation.] Tokyo:
Hitsuji Shoboo.

Kageyama, Taro. 1996. Doushi imiron. [Semantics of verbs.] Tokyo: Kurosio

Publishers.

Kawakami, Seisaku. 1996. Ninchigengogaku no kiso. [An Introduction to Cognitive

Linguistics.] Tokyo: Kenkyusha.

Kageyama, Taro and Yoko Yumoto. 1997. Go Keisei to Gainen Kouzou. [Word
Formation and Conceptual structure]. Tokyo: Kenkyusha.

405
Kenkyusha’s New Collegiate Japanese-English Dictionary. 1983. 3rd ed. Tokyo:
Kenkyusha.

Kenkyusha’s New Japanese-English Dictionary. 1974. 4th ed. Tokyo: Kenkyusha.

Kindaichi, Haruhiko. 1950. Kokugo dooshi no ichibunrui. [A classification of

Japanese verbs.] Reprinted in H. Kindaichi (ed.), Nihongo doosi no asupekuto
[Aspect in Japanese verbs] (1976). 5-26. Tokyo: Mugi Shoboo.

Kindaichi, Haruhiko. 1955. Nihongo doosi no tensu to asupekuto. [Tense and aspect
in Japanese verbs]. Reprinted in H. Kindaichi (ed.), Nihongo doosi no asupekuto
[Aspect in Japanese verbs] (1976). 27-62. Tokyo: Mugi Shoboo.

Kitagawa, Chisato, Osamu Kamada and Iguchi Atsuo. 1988. Japanese for foreigners:
Innovative Workbooks in Japanese 7—Particles. Tokyo: Aratake Shuppan.

Kouji-en. [Japanese extensive dictionary]. 1991. 4th ed. Tokyo: Iwanami Shoten.

Kudou, Mayumi. 1995. Aspekuto/tensu taikei to tekusuto. [Aspect/Tense system and

text.] Tokyo: Hitsuji Shoboo.

Kuno, Susumu. 1973. The structure of the Japanese language. Cambridge, Mass.:
MIT press.

Lakoff, George. 1987. Women, fire, and dangerous things: what categories reveal
about the mind. Chicago: the University of Chicago Press.

Langacker, Ronald W. 1976. Semantic representation and the linguistic relativity

hypothesis. Foundations of Language 14(3). 307-357.

Langacker, Ronald W. 1987. Foundations of cognitive grammar, vol I: Theoretical

prerequisites. Stanford: Stanford University Press.

Langacker, Ronald W. 1991a. Concept, image, and symbol: The cognitive basis of
grammar. Berlin and New York: Mouton de Gruyter.

Langacker, Ronald W. 1991b. Foundations of cognitive grammar, vol II: Descriptive

application. Stanford: Stanford University Press.

Levin, Beth. 1993. English verb classes and alternation: a preliminary investigation.
Chicago: University of Chicago Press.

Levin, Beth and Malka Rappaport Hovav. 1991. Wiping the slate clean: a lexical
semantic exploration. Cognition 41. 123-151.

Levin, Beth and Malka Rappaport Hovav. 1995. Unaccusativity: at the syntax-lexical
semantics interface. Cambridge, Massachusetts: The MIT Press.

406
Levin, Beth and Malka Rappaport Hovav. To appear. From lexical semantics to
argument realization. Handbook of morphosyntax and argument structure, ed.
Hagit Barer. Dordrecht: Kluwer Academic Publishers.

Machida, Ken. 1989. Nihongo no jisei to asupekuto. [Tense and aspect in Japanese.]
Tokyo: Aruku.

Makino, Seiichi and Michio Tsutsui. 1986. A dictionary of basic Japanese grammar.
Tokyo: The Japan Times.

Makino, Seiichi and Michio Tsutsui. 1995. A dictionary of intermediate Japanese

grammar. Tokyo: The Japan Times.

Masuoka, Takashi and Yukinori Takubo. 1992. Kiso Nihongo Bunpou. [Basic
Japanese Grammar.] Tokyo: Kurosio Publishers.

Matsumoto, Yo. 1996. Complex predicates in Japanese: A syntactic and semantic

study of the notion ‘word’. California and Tokyo: CSLI Publications and Kurosio
Publishers.

Matsumoto, Yo. 1998. Semantic change in the grammaticalization of verbs into

postpositions in Japanese. Studies in Japanese grammaticalization: cognitive and
discourse perspectives, ed. Toshio Ohori. Tokyo: Kurosio Publishers.

McClain, Yuko Matsuoka. 1981. Handbook of Modern Japanese Grammar. Tokyo:

Hakuseido Press.

Morikawa, Hiromi. 1997. Acquisition of case marking and argument structures in

Japanese. Tokyo: Kurosio Publishers.

Nihongo Goi Taikei vol.5 Koubun Taikei. [Japanese vocabularly compendium vol. 5
system of syntactic construction.] 1997. Tokyo: Iwanami Shoten.

Nihongo Kihon Doushi Youhou Jiten. [Dictionary for Usage of Japanese Basic Verbs.]
1989. Tokyo: Taishukan Shoten.

Nihongo Sakabiki Jiten. [Reverse Dictionary of Japanese.] 1990. Tokyo: Taishukan

Shoten.

Niimi, Kazuaki, Youichi Yamaura and Tokuko Utsuno. 1987. Japanese for
foreigners:
Innovative Workbooks in Japanese 4—compound verbs. Tokyo: Aratake Shuppan.

Nitta, Yoshio. 1997. Nihongo bunpo kenkyuu josetsu. [Introduction to study of

Japanese grammar.] Tokyo: Kurosio Publishers.

Palmer, Robert F. 1987. The English verb. 2nd ed. London: Longman.

Palmer, Robert F. 1994. Grammatical roles and relations. Cambridge: Cambridge

University Press.

407
Progressive Japanese-English Medium Dictionary CD-ROM version 1.11. 1993.
Tokyo: Shogakukan

Rappaport Hovav, Malka and Beth Levin. 1988. What to do with theta-roles. Syntax
and semantics 21: thematic relations, ed. Wendy Wilkins, 7-37. NewYork:
Academic Press.

Rappaport Hovav, Malka, Beth Levin, and Mary Laughren. 1993. Levels of lexical
representation. Semantics and the lexicon, ed. J. Pustejovsky, 37-54. Netherlands:
Kluwer Academic Publishers.

Rappaport Hovav, Malka and Beth Levin. 1998. Building verb meanings. The
projection of arguments: lexical and compositional factors, ed. Miriam Butt and
Wilhelm Geuder, 97-134. Stanford: Center for the Study of Language and
Information.

Rappaport Hovav, Malka and Beth Levin. 1999. Two types of compositionally
derived events. Manuscript, 1999.

Sadakane, Kumi and Masatoshi Koizumi. 1995. On the nature of the “dative” particle
ni in Japanese. Linguistics 33. 5-33.

Sanseido’s Daily Concise English-Japanese Japanese-English Dictionary. 1990. 4th

ed. Tokyo: Sanseido.

Shirota, Shun. 1993. Bunpo kaku to hukushi kaku. [syntactic case and adjunctive
case], Nihongo no kaku o megutte [On Japanese case], ed. Yoshio Nitta. Tokyo:
Kurosio Publishers.

Shogakukan Progressive English-Japanese Dictionary. 1987. 2nd ed. Tokyo:

Shogakukan.

Sugamoto, Nobuko. 1982. Transitivity and Objecthood in Japanese. Syntax and

Semantics, vol. 15: Studies in Transitivity, ed. Paul J. Hopper and Sandra A.
Thompson. New York: Academic Press.

Talmy, Leonard. 1976. Semantic causative types. The Grammar of Causative

Constructions. (Syntax and Semantics, Vol. 6.), ed. Masayoshi Shibatani, 43-116.
New York: Academic Press.

Talmy, Leonard. 1978. Figure and ground in complex sentences. Universals of Human
Language, vol. 4: syntax, ed. Joseph H. Greenberg, 625-649. Stanford: Stanford
University Press.

Talmy, Leonard. 1988. Force dynamics in language and cognition. Cognitive Science
12. 49-100.

Tanaka, Shigenori and Yo Matsumoto. 1997. Kuukan to idou no hyougen.

[Expressions of space and motion]. Tokyo: Kenkyusha.

408
Taoka, Chiaki. 1995. On mental verbs in Japanese. MA dissertation, Department of
Linguistics, University of Manchester. .

Taylor, John R. 1996. On running and jogging. Cognitive Linguistics 7. 21-34.

Tenny, Carol. 1992. The aspectual interface hypothesis. Lexical matters. eds. I. Sag
and A. Szabolcsi, 1-28. Stanford: CSLI Publications.

Teramura, Hideo. 1982. Nihongo no shintakusu to imi, vol. I. [Syntax and Semantics
in Japanese, vol. I.] Tokyo: Kurosio Publishers.

The Great Japanese Dictionary. 1995. 2nd ed. Tokyo: Kodansha.

Tsujimura, Natsuko. 1996. An introduction to Japanese linguistics. Massachusetts and

Oxford: Blackwell Publishers.

Tsukaikata no wakaru ruigo reikai jiten: A dictionary of synonyms in Japanese. 1994.

Tokyo: Shogakukan.

Uehara, Satoshi. 1998. Syntactic categories in Japanese: a cognitive and typological

introduction. Tokyo: Kurosio Publishers.

Vendler, Zeno. 1967. Linguistics in Philosophy. Ithaca: Cornell University Press.

Washio, Ryuuichi, and Ken-ichi Mihara. 1997. Boisu to asupekuto. [Voice and
aspect.] Tokyo: Kenkyusha.

Yoshikawa, Chizuko. 1995. Nichiei hikaku doushi no bunpo. [Japanese-English

comparative grammar of verbs.] Tokyo: Kurosio Publishers.

Yourei Kensaku Sohuto 2. [Reference soft for examples 2]. 1997-1998. Osaka:
Sharp Corporation.

409
APPENDIX A. Senses of the constructions and situation types

1. Situation types that take 1I

[1-1Natural Kinds]
Sono ishi wa daiamondo da.
(1I) that stone TOP diamond DA
That stone is a diamond.

[1-2 Ethnicity]
Kanojo wa igirisujin da.
(1I) she TOP English DA
She is English.

[1-3 Biological Kinds]

Are wa tokage da.
(1I) that TOP lizard DA
That is a lizard.

[1-4 Physical Properties]

Kare wa se ga takai.
(1I) he TOP height NOM tall
He is tall.

[2-1 Existence]
Chikyuu ni-wa mizu ga aru.
(1I) earth in-TOP water NOM exist
There is water on the earth.

[2-2 Necessity]
Sono shokubutsu ni-wa juubunna eiyou ga iru2.
(1I) the plant to-TOP enough nutrition NOM need/necessary
The plant needs a lot of nutrition.

[2-3 Dispositions]
Jack wa shinsetsu da.
(1I) Jack TOP kind DA
Jack is kind.

[4 Physical States]
Sono mizuumi no mizu wa tsumetai.
(1I) the lake GEN water TOP cold
The water in the lake is cold.

[6 Relation 1]
Gengogaku wa bunkengaku to-wa kotonaru.
(1I) Linguistics TOP philology with-TOP differ
Linguistics differs from philology.

410
[9 Perception]
Kono basho kara wa Huji-san ga mieru.
(1I) this place from TOP Mt.Fuji NOM be.visible
Mt.Fuji is visible from this place.

[11 Cognition 2]
Watashi wa kami o shinjiru.
(1I) I TOP God ACC believe
I believe in God.

[15 Attitudes]
Kare wa chottoshita koto de sugu hinekureru.
(1I) he TOP trivial thing with easily become.warped
He gets warped easily with trivial things.

[17 Change of Physical Properties]

Kono keito no uwagi wa kantanni chijimu.
(1I) this wool GEN jacket TOP easily shrink
This woolen jacket shrinks easily.

[18 Blooming]
Sono hana wa umibe ni saku.
(1I) the flower TOP seaside at bloom
The flower blooms at the seaside.

[20 Open/Close]
Sono mesu wa kantanni kanbu o hiraku.
(1I) the surgical.knife TOP easily diseased.part ACC open
The surgical knife opens diseased parts easily.

[21-1 Light Emission]

Sono dentou wa sukoshi no enerugi de hikaru.
(1I) the light TOP little GEN energy with shine
The light shines with little energy.

[21-2 Sound Emission]

Sono tori wa totemo utsukushiku naku.
(1I) the bird TOP very beautifully cry
The bird cries/sings very beautifully.

[24 Manner of Motion 1]

Taro wa totemo hayaku hashiru.
(1I) Taro TOP very fast run
Taro runs very fast.

[25 Manner of Motion 2]

Hanako wa totemo jouzuni oyogu.
(1I) Hanako TOP very well swim
Hanako swims very well.

411
[26 Motion/Movement]
Sono omocha wa denchi de ugoku.
(1I) the toy TOP battery with move
The toy moves with a battery.

[27 Performance]
Markus wa warutsu o jouzuni odoru.
(1I) Markus TOP waltz ACC well dance
Markus dances the waltz very well.

[28 Create Mark/Defect]

Uchi no neko wa tsume de nandemo hikkaku.
(1I) home GEN cat TOP claw with anything scratch
Our cat scratches anything with its claws.

[29 Consumption]
Watashi wa tamago o taberu.
(1I) I TOP egg ACC eat
I eat eggs. (I am not allergic.)

[31 Creation]
Haha wa jouzuni doresu o tsukuru.
(1I) mother TOP well dress ACC make
My mother makes dresses very well.

[32 Dismantle]
Sono burudoozaa wa tatemono o kantanni torikowasu.
(1I) the bulldozer TOP building ACC easily dismantle
The bulldozer dismantles building easily.

[33 Dyeing and Shaving]

Kanojo wa jouzuni kami o someru.
(1I) she TOP well hair ACC dye
She dyes her hair very well.

[34 Dressing]
Jacqui wa kimono o jouzuni kiru.
(1I) Jacqui TOP Japanese.dress ACC well dress
Jacqui puts on Japanese dress very well.

[36 Fracture/Breaking (intra)]

Kono mado wa chottoshita shougeki de kantanni wareru.
(1I) this window TOP small impact with easily break
This window breaks easily with a small impact.

[37 Attachment]
Sono teepu wa kantanni te ni kuttsuku.
(1I) the tape TOP easily hand to stick
The tape easily sticks to hands.

412
[38 Killing]
Sono misairu wa 100man-nin demo korosu.
(1I) that missile TOP 100.ten.thousand-CLSS even kill
The missile kills even one million people.

2. Situation types that take 1T

[2-1 Existence]
Sono hon ga tsukue no ue ni aru.
(1T) the book NOM desk GEN surface at be
The book is on the desk.

[2-2 Necessity]
Watashi ni-wa ima kane ga iru2
(1T) I to-TOP now money NOM need/necessary
I need money now.

[2-3 Dispositions]
Jack wa kyou wa shinsetsu da.
(1T) Jack TOP today TOP kind DA
Jack is kind today.

[3 Social Role]
John wa shachou da.
(1T) John TOP president DA
John is president.

[4 Physical States]
Sono mizuumi no mizu wa ima wa tsumetai.
(1T) the lake GEN water TOP now TOP cold
The water in the lake is cold now.

[5 Bodily States]
Kare wa byouki da.
(1T) he TOP ill DA.
He is ill.

[9 Perception]
Kyou wa mezurashiku Huji-san ga mieru.
(1T) today TOP unusually Mt.Fuji NOM be.visible
Mt.Fuji is visible today, which rarely happens.

[11 Cognition 2]
Watashi wa kyou dake kami o shinjiru.
(1T) I TOP today only God ACC believe
I believe in God only today.

413
3. Situation types that take 1H

[2-1 Existence]
John wa kanojo no iru tokoro ni-wa itsumo iru.
(1H) John TOP she GEN be place at-TOP always be
John is always there wherever she is.

[2-2 Necessity]
Kare ni-wa itsumo kane ga iru2
(1H) he to-TOP always money NOM need/necessary
He always needs money.

[2-3 Dispositions]
Jack wa tanomigoto ga aru toki dake shinsetsu da.
(1H) Jack TOP favor NOM be when only kind DA
Jack is kind only when he receives a favour.

[4 Physical States]
Sono mizuumi no mizu wa maitoshi nigatsu ni-wa tsumetai.
(1H) the lake GEN water TOP every.year February in-TOP cold
The water of the lake is cold every February.

[5 Bodily States]
Kare wa huyu ni-wa kimatte byouki da.
(1H) he TOP winter in-TOP always ill DA
He is always ill in winter.

[8 Posture 2]
Watashi wa densha no naka de yoku suwaru.
(1H) I TOP train GEN inside in often sit
I often take a seat in the train.

[9 Perception]
Hare no hi wa itsumo koko kara Huji-san ga mieru.
(1H) fine.weather GEN day TOP always here from Mt.Fuji NOM be.visible
Mt.Fuji is always visible here when it is fine.

[10 Cognition 1]
Taro wa itsumo hito no kokoro o yoku rikaisuru.
(1H) Taro TOP always person GEN heart ACC well understand
Taro always understands people’s feelings well.

[11 Cognition 2]
Watashi wa uranai o ii toki dake shinjiru.
(1H) I TOP fortune.telling ACC good when only believe
I believe fortune telling only when it is good.

414
[12 Emotional Activity]
Janet wa doubutsu o miru to yorokobu.
(1H) Janet TOP animal ACC see when rejoice
Janet gets happy when she sees animals.

[13 Sleeping]
Chichi wa maiban 11ji ni nemuru.
(1H) father TOP every.night 11o'clock at sleep
My father goes to bed at 11 o’clock every night.

[14 Perceptual Activity]

Chichi wa maiasa terebi no nyuusu o miru.
(1H) father TOP every.morning TV GEN news ACC watch
My father watches the news on TV every morning.

[15 Attitudes]
Kare wa shocchuu hinekureru.
(1H) he TOP always get.warped
He always gets warped.

[16 Change of Weight]

Watashi wa maitoshi natsu ni hutoru.
(1H) I TOP every.year summer in gain.weight
I gain weight every summer.

[17 Change of Physical Properties]

Kono keito no uwagi wa sentaku-suru tabi ni chijimu.
(1H) this wool GEN jacket TOP wash time at shrink
This woolen jacket shrinks every time I wash it.

[18 Blooming]
Maitoshi gogatsu ni niwa ni bara no hana ga saku.
(1H) every.year May in garden at rose GEN flower NOM bloom
Roses bloom in our garden every May.

[19 Directed Motion]

Sono inu wa dareka kuru to sugu koya ni hairu
(1H) the dog TOP somebody come when quickly kennel ALL go.into
The dog goes into its kennel quickly when a stranger comes.

[20 Open/Close]
Sono mise wa maiasa 8ji ni hiraku.
(1H) the shop TOP every.morning 8o'clock at open
The shop opens at 8 every morning.

[21-1 Light Emission]

Sono dentou wa maiban 6ji ni hikaru.
(1H) the light TOP every.night 6o'clock at shine
The light shines at 6 o’clock every night.

415
[21-2 Sound Emission]
Sono tori wa asa ni kimatte naku.
(1H) the bird TOP morning at always cry
The bird cries/sings every morning.

[22-1 Contact]
Jack wa yoku jibun no seito o tataku.
(1H) Jack TOP often own GEN student ACC hit
Jack often hits his students.

[22-2 Bodily Motion]

Mary wa wakareru toki itsumo John ni te o huru.
(1H) Mary TOP separate when always John to hand ACC wave
Mary always waves to John when she says goodbye to him.

[23 Touching]
Nina wa neko o miru to sugu te o hureru.
(1H) Nina TOP cat ACC see when quickly hand ACC touch
Nina touches a cat quickly with her hand whenever she sees one.

[24 Manner of Motion 1]

Taro wa maiasa kouen o hashiru.
(1H) Taro TOP every.morning park ACC run
Taro runs in the park every morning.

[25 Manner of Motion 2]

Hanako wa maiasa 2kilo o oyogu.
(1H) Hanako TOP every.morning 2kilometers ACC swim
Hanako swims 2 kilometres every morning.

[26 Motion/Movement]
Sono omocha wa shindou ga aru tabi ni ugoku.
(1H) the toy TOP shock NOM occur time at move
The toy moves every time there is a shock.

[27 Performance]
Markus wa maishuu kayoubi warutsu o odoru.
(1H) Markus TOP every.week Tuesday waltz ACC dance
Markus dances the waltz every Tuesday.

[28 Create Mark/Defect]

Watashi wa yoku juuyouna shorui o ayamatte yaburu.
(1H) I TOP often important document ACC mistaking tear.up
I often tear up important documents by mistake.

[29 Consumption]
Watashi wa maiasa gohan o taberu.
(1H) I TOP every.morning.rice ACC eat
I eat rice every morning.

416
[30 Covering]
Hazukashigariya no kanojo wa itsumo te de kao o oou.
(1H) shy.person GEN she TOP always hand with face ACC cover
She, as a shy person, always covers her face with her hands.

[31 Creation]
Haha wa ikkagestsu ni ic-chaku doresu o tsukuru.
(1H) mother TOP one.month at one-CLSS dress ACC make
My mother makes one dress per month.

[32 Dismantle]
Sono kodomo wa tsuki ni ichido wa omocha o kowasu.
(1H) the child TOP month in once TOP toy ACC destroy
The child destroys his toys once a month.

[33 Dyeing and Shaving]

Kanojo wa maitsuki kami o someru.
(1H) she TOP every.month hair ACC dye
She dyes her hair every month.

[34 Dressing]
Jacqui wa maitoshi oshougatsu ni kimono o kiru.
(1H) Jacqui TOP every.year new.year's.day on Jap.dress ACC wear
Jacqui wears Japanese dress every New Year’s Day.

[37 Attachment]
Sono teepu wa atatamaru to te ni kuttsuku.
(1H) the tape TOP get.warm when hand to stick
The tape sticks to hands when it is warm.

[38 Killing]
Sono kakkazan wa maitoshi ooku no hito o korosu.
(1H) that active.volcano TOP every.year many GEN people ACC kill
That active volcano kills many people every year.

[39 Winning]
Mike wa itsumo daijina shiai ni makeru.
(1H) Mike TOP always important game OJCT lose
Mike always loses important games.

417
4. Situation types that take 1*

[7-1 Relation 2]

[7-2 Posture 1]

[35 Dying]

[40 Discovery]

5. Situation types that take 2U

[12 Emotional Activity]

Haha wa watashi no seikou o yorokon-de-iru.
(2U) mother TOP I GEN success ACC rejoice-TE-IRU
My mother is happy with my success.

[13 Sleeping]
Chichi wa ima shizukani nemut-te-iru.
(2U) father TOP now quietly sleep-TE-IRU
My father is sleeping now.

[14 Perceptual Activity]

Chichi wa ima terebi no nyuusu o mi-te-iru.
(2U) father TOP now TV GEN news ACC watch-TE-IRU
My father is watching news on TV now.

[21-1 Light Emission]

Sono dentou wa chikachika hikat-te-iru.
(2U) the light TOP ONMP2 shine-TE-IRU
The light is flashing.

[21-2 Sound Emission]

Nezumi ga chuchu nai-te-iru.
(2U) mouse NOM ONMP2 cry-TE-IRU
The mouse is squeaking.

[22-1 Contact]
Jack ga mado o tatai-te-iru.
(2U) Jack NOM window ACC knock-TE-IRU
Jack is knocking at the window.

[22-2 Bodily Motion]

Mary wa John ni te o hut-te-iru.
(2U) Mary TOP John to hand ACC wave-TE-IRU
Mary is waving her hand to John.

418
[23 Touching]
Nina wa atarashii garasu ni petapeta te o hure-te-iru.
(2U) Nina TOP new glass to ONMP2 hand ACC touch-TE-IRU
Nina is repeatedly touching the new glass.

[24 Manner of Motion 1]

Taro wa ima hashit-te-iru.
(2U) Taro TOP now run-TE-IRU
Taro is running now.

[25 Manner of Motion 2]

Hanako wa ima oyoi-de-iru.
(2U) Hanako TOP now swim-TE-IRU
Hanako is swimming now.

[26 Motion/Movement]
Jishin de tsukue ga ugoi-te-iru!!
(2U) earthquake with desk NOM move-TE-IRU
The desk is moving because of the earthquake!!

[27 Performance]
Jacqui wa ima utat-te-iru.
(2U) Jacqui TOP now sing-TE-IRU
Jacqui is singing now.

[28 Create Mark/Defect]

Kanojo wa sono tegami o biribiri yabui-te-iru.
(2U) she TOP the letter ACC ONMP2 tear-TE-IRU
She is tearing up the letter. (iterated tearing)

6. Situation types that take 2D

[14 Perceptual Activity]

Watashi wa sono bideo o sukoshizutsu mi-te-iru.
(2D) I TOP the video ACC little.by.little watch-TE-IRU
I am watching the video little by little.

[15 Attitudes]
Kare wa masumasu hinekure-te-iru.
(2D) he TOP more.and.more get.warped-TE-IRU
He is getting more and more warped.

[16 Change of Weight]

Mary wa sukoshizutsu hutot-te-iru.
(2D) Mary TOP little.by.little get.fat-TE-IRU
Mary is getting fatter little by little.

419
[17 Change of Physical Properties]
Suupu ga sukoshizutsu atatamat-te-iru.
(2D) soup NOM little.by.little get.warm-TE-IRU
The soup is getting warm.

[20 Open/Close]
Sono tsubomi wa sukoshizutsu hirai-te-iru.
(2D) the bud TOP little.by.little open-TE-IRU
The bud is opening little by little.

[26 Motion/Movement]
Jishin de tsukue ga sukoshizutsu minami ni ugoi-te-iru.
(2D) earthquake with desk NOM little.by.little south to move-TE-IRU
The desk is moving to the south because of the earthquake.

[27 Performance]
Jacqui wa ima Corrs no shinkyoku o utat-te-iru.
(2D) Jacqui TOP now Corrs GEN new.song ACC sing-TE-IRU
Jacqui is singing Corrs’ new song now.

[28 Create Mark/Defect]

Kanojo wa sono tegami o yabui-te-iru.
(2D) she TOP the letter ACC tear-TE-IRU
She is tearing the letter. (One tearing taking time)

[29 Consumption]
Joel wa yuushoku no pasuta o tabe-te-iru.
(2D) Joel TOP dinner GEN pasta ACC eat-TE-IRU
Joel is eating the pasta for his dinner.

[30 Covering]
Chichi ga sukoshizutsu kuruma o shiito de oot-te-iru.
(2D) father NOM little.by.little car ACC sheet with cover-TE-IRU
My father is covering his car with sheet little by little.

[31 Creation]
Haha wa watashi no doresu o tsukut-te-iru.
(2D) mother TOP I GEN dress ACC make-TE-IRU
My mother is making a dress for me.

[32 Dismantle]
Sono burudoozaa wa kyuu kousha o torikowashi-te-iru.
(2D) the bulldozer TOP old schoolhouse ACC dismantle-TE-IRU
The bulldozer is dismantling the old schoolhouse.

[33 Dyeing and Shaving]

Kanojo wa {sukoshizutsu} kami o some-te-iru.
(2D) she TOP {little.by.little} hair ACC dye-TE-IRU
She is dying her hair little by little.

420
7. Situation types that take 2T

[8 Posture 2]
Steve wa asoko ni tat-te-iru.
(2T) Steve TOP there at stand-TE-IRU
Steve is standing over there.

[9 Perception]
Koko kara Huji-san ga mie-te-iru!!
(2T) here from Mt.Fuji NOM be.visible-TE-IRU
I can see Mt. Fuji from here!!

[11 Cognition 2]
Watashi wa kami o shinji-te-iru.
(2T) I TOP God ACC believe-TE-IRU
I believe in God.

[15 Attitudes]
Kanojo wa kyou wa mezurashiku sumashi-te-iru.
(2T) she TOP today TOP unusually put.on.airs-TE-IRU
She puts on airs today, which is rather unusual for her.

[16 Change of Weight]

Mary wa mezurashiku ima hutot-te-iru.
(2T) Mary TOP unusually now become.fat-TE-IRU
Mary is fat now, which is rare.

[17 Change of Physical Properties]

Suupu ga atatamat-te-iru.
(2T) soup NOM get.warm-TE-IRU
The soup is warm.

[18 Blooming]
Ookina bara no hana ga ichi-rin niwa ni sai-te-iru.
(2T) big rose GEN flower NOM one-CLSS garden in bloom-TE-IRU
The one big rose is in bloom in our garden.

[19 Directed Motion]

John wa jibun no koya ni hait-te-iru.
(2T) John TOP own GEN kennel ALL go.into-PAST
John is in his kennel (as a result of going there).

[20 Open/Close]
Kare wa ashi o hirai-te-iru.
(2T) he TOP leg ACC open-TE-IRU
He left his legs open.

421
[23 Touching]
Ashi ga puuru no soko ni hure-te-iru.
(2T) foot NOM swimming.pool GEN bottom at touch-TE-IRU
My feet touch the bottom of the swimming pool.

[26 Motion/Movement]
Kinou no jishin de tsukue ga sukoshi ugoi-te-iru.
(2T) yesterday GEN earthquake with desk NOM a.little move-TE-IRU
The desk has moved a little because of yesterday’s earthquake.

[30 Covering]
Kiri ga machi o sukkari oot-te-iru.
(2T) fog NOM city ACC completely cover-TE-IRU
The fog completely covers the city.

[33 Dyeing and Shaving]

Kanojo wa kami o {kiiroku} some-te-iru.
(2T) she TOP hair ACC {yellow} dye-TE-IRU
She has dyed her hair yellow. (Now she has the yellow hair.)

[34 Dressing]
Jacqui wa akai doresu o ki-te-iru.
(2T) Jacqui TOP red dress ACC wear-TE-IRU
Jacqui is wearing a red dress.

[37 Attachment]
Doro ga zubon ni tsui-te-iru.
(2T) dirt NOM trousers to stick-TE-IRU
The dirt is sticking to the trousers.

8. Situation types that take 2I

[6 Relation 1]
Gengogaku wa bunkengaku to-wa kotonat-te-iru.
(2I) Linguistics TOP philology with-TOP differ-TE-IRU
Linguistics differs from philology.

[7-1 Relation 2]
Kare no ronbun wa kono ronbun yori sugure-te-iru.
(2I) he GEN article TOP this article than excel-TE-IRU
His dissertation is superior to this dissertation.

[7-2 Posture 1]
Huji-san ga me no mae ni sobie-te-iru.
(2I) Mt.Fuji NOM eye GEN front at tower-TE-IRU
Mt. Fuji towers high in front of us.

422
[10 Cognition 1]
Taro wa gengogaku o yoku rikaishi-te-iru.
(2I) Taro TOP Linguistics ACC well understand-TE-IRU
Taro understands Linguistic very well.

[15 Attitudes]
Kare wa tottemo hinekure-te-iru.
(2I) he TOP very become.warped-TE-IRU
He has such a warped disposition.

[16 Change of Weight]

Mary wa hutot-te-iru.
(2I) Mary TOP become.fat-TE-IRU
Mary is fat.

[24 Manner of Motion 1]

Kousokudouro ga machi no chuushin o hashi-te-iru.
(2I) highway NOM city GEN center ACC run-TE-IRU
The highway runs through the centre of the city.

[30 Covering]
Midori no kigi ga sono kuni o oot-te-iru.
(2I) green GEN trees NOM the country ACC cover-TE-IRU
Green trees cover the country.

[35 Dying]
Neko ga michibata de shin-de-iru.
(2I) cat NOM roadside on die-TE-IRU
The cat is dead on the roadside.

[36 Fracture/Breaking (intra)]

Heya no mado ga ware-te-iru!
(2I) room GEN window NOM break-TE-IRU
The window in the room is broken!

9. Situation types that take 2R

[34 Dressing]
Jacqui wa ima tonari no heya de kimono o ki-te-iru.
(2R) Jacqui TOP now next GEN room at Jap.dress ACC put.on-TE-IRU
Jacqui is putting on her Japanese dress in the next room.

[38 Killing]
Haha wa kanshasai ni shichimencho o kososhi-te-iru.
(2R) mother TOP thanksgiving.day for turkey ACC kill-TE-IRU
Mother is killing a turkey for Thanksgiving Day.

423
[39 Winning]
Nihon chiimu wa kankoku chiimu ni genzainotokoro kat-te-iru.
(2R) Japan team TOP Korea team against at.this.moment win-TE-IRU
The Japanese team is leading the Korean team at this moment.

10. Situation types that take 2H

[11 Cognition 2]
Watashi wa uranai o ii toki dake shinji-te-iru.
(2H) I TOP fortune.telling ACC good when only believe
I believe fortune telling only when it is good.

[13 Sleeping]
Chichi wa maiban 11ji ni nemut-te-iru.
(2H) father TOP every.night 11o'clock at sleep-TE-IRU
My father goes to bed at 11 o’clock every night.

[14 Perceptual Activity]

Chichi wa maiasa terebi no nyuusu o mi-te-iru.
(2H) father TOP every.moring TV GEN news ACC watch-TE-IRU
My father watches news on TV every morning.

[15 Attitudes]
Kanojo wa itsumo John no mae de sumashi-te-iru.
(2H) she TOP always John GEN front in put.on.airs-TE-IRU
She always puts on airs in front of John.

[19 Directed Motion]

Uchi no inu wa maitsuki huro ni hait-te-iru
(2H) home GEN dog TOP every.month bath ALL go.into-TE-IRU
Our dog gets into the bath every month.

[20 Open/Close]
Haha wa maiasa mise o 8ji ni hirai-te-iru.
(2H) mother TOP every.morning shop ACC 8o'clock at open-TE-IRU
My mother opens the shop at 8 o’clock every morning.

[21-1 Light Emission]

Sono dentou wa maiban chikachika hikat-te-iru.
(2H) the light TOP every.night ONMP2 shine-TE-IRU
The light flashes every night.

[21-2 Sound Emission]

?Sono tori wa maiasa nai-te-iru.
(2H) the bird TOP every.morning cry-TE-IRU
The bird cries/sings every morning.

424
[22-1 Contact]
Jack wa maiban chichi no kata o tatai-te-iru.
(2H) Jack TOP every.night father GEN shoulder ACC pat-TE-IRU
Jack pats his father on his shoulder every night.

[22-2 Bodily Motion]

Mary wa jugyouchuu itsumo John ni te o hut-te-iru.
(2H) Mary TOP in.the.class always John to hand ACC wave-TE-IRU
Mary always waves to John during the class.

[23 Touching]
Nina wa maiasa neko no ke ni te o hure-te-iru.
(2H) Nina TOP every.morning cat GEN fur to hand ACC touch-TE-IRU
Nina touches the cat’s fur with her hand every morning.

[24 Manner of Motion 1]

Taro wa maiasa kouen o hashit-te-iru.
(2H) Taro TOP every.morning park ACC run-TE-IRU
Taro runs in the park every morning.

[25 Manner of Motion 2]

Hanako wa maiasa 2kilo o oyoi-de-iru.
(2H) Hanako TOP every.morning 2kilometers ACC swim-TE-IRU
Hanako swims 2 kilometres every morning.

[27 Performance]
Markus wa maishuu kayoubi warutsu o odot-te-iru.
(2H) Markus TOP every.week Tuesday waltz ACC dance-TE-IRU
Markus dances the waltz every Tuesday.

[28 Create Mark/Defect]

Kaisha de-wa maishuu iranai shorui o yabui-te-iru.
(2H) company in-TOP every.week unnecessary document ACC tear-TE-IRU
We tear up unnecessary documents in the company every week.

[29 Consumption]
Watashi wa maiasa gohan o tabe-te-iru.
(2H) I TOP every.morning rice ACC eat-TE-IRU
I eat rice every morning.

[30 Covering]
Chichi wa maiban kuruma o shiito de oot-te-iru.
(2H) father TOP every.night car ACC sheet with cover-TE-IRU
My father covers his car with a sheet every night.

[31 Creation]
Haha wa ikkagetsu ni ic-chaku doresu o tsukut-te-iru.
(2H) mother TOP one.month at one-CLSS dress ACC make-TE-IRU
My mother makes one dress per month.

425
[32 Dismantle]
Sono kodomo wa maitsuki omocha o kowashi-te-iru.
(2H) the child TOP every.month toy ACC destroy-TE-IRU
The child destroys his toy every month.

[33 Dyeing and Shaving]

Kanojo wa maitsuki kami o some-te-iru.
(2H) she TOP every.month hair ACC dye-TE-IRU
She dyes her hair every month.

[34 Dressing]
Mary wa maiasa jibun de huku o ki-te-iru.
(2H) Mary TOP every.morning oneself with dress ACC put.on-TE-IRU
Mary’s child dresses herself every morning.

[39 Winning]
Nihon chiimu wa maitoshi kankoku chiimu ni kat-te-iru.
(2H) Japan team TOP every.year Korea team OJCT win-TE-IRU
The Japanese team wins against the Korean team every year.

11. Situation types that take 2F

[8 Posture 2]
Steve wa sono tokubetsu seki ni gokai suwat-te-iru.
(2F) Steve TOP the special seat at five.times sit-TE-IRU
Steve has taken that special seat five times.

[9 Perception]
Kono basho kara 1920nen ni ichido Huji-san ga mie-te-iru.
(2F) this place from 1920year in once Mt.Fuji NOM be.visible-TE-IRU
Mt. Fuji has been seen from here once, in 1920.

[10 Cognition 1]
?Taro wa kako ni ichido sono suuushiki o rikaishi-te-iru.
(2F) Taro TOP past in once the formula ACC understand-TE-IRU
Taro has understood the formula once in the past.

[11 Cognition 2]
?Chichi wa ichido dake isuramukyou o shinji-te-iru.
(2F) father TOP once only Muslim ACC believe-TE-IRU
My father has believed in the Muslim religion only once.

[12 Emotional Activity]

Chichi wa ichido dake watashi no seikou o yorokon-de-iru.
(2F) father TOP once only I GEN success ACC rejoice-TE-IRU
My father has rejoiced over my success only once (in the past).

426
[13 Sleeping]
Chichi wa kako ni ichido sabaku de nemut-te-iru.
(2F) father TOP past in once desert in sleep-TE-IRU
My father has slept in the desert once in the past.

[14 Perceptual Activity]

Chichi wa kako ni ichido yurei o mi-te-iru.
(2F) father TOP past in once ghost ACC see-TE-IRU
My father has seen a ghost once in the past.

[15 Attitudes]
?Kare wa kako ni ichido dake hinekure-te-iru.
(2F) he TOP past in once only get.warped-TE-IRU
He has got warped only once in the past.

[16 Change of Weight]

?Mary wa kako ni ichido hut-te-iru.
(2F) Mary TOP past in once get.fat-TE-IRU
Mary has been fat once in the past.

[17 Change of Physical Properties]

Sono uwagi wa izenni ichido chijin-de-iru.
(2F) that jacket TOP before once shrink-TE-IRU
That jacket has been shrunk once before.

[18 Blooming]
Niwa no bara no hana wa kyonen sai-te-iru.
(2F) garden GEN rose GEN flower TOP last.year bloom-TE-IRU
The rose in the garden bloomed last year.

[19 Directed Motion]

Yougisha wa 5ji ni sono mise ni hait-te-iru.
(2F) suspect TOP 5o'clock at that shop ALL go.into-TE-IRU
The suspect went into the shop at 5 o’clock.

[20 Open/Close]
Sono kaisha wa ichido Tokyo ni shiten o hirai-te-iru.
(2F) the company TOP once Tokyo in branch.office ACC open-TE-IRU
The company opened its branch in Tokyo once.

[21-1 Light Emission]

Sono dentou wa kako ni ichido dake chikat-to hikat-te-iru.
(2F) the light TOP past in once only ONMP1 shine-TE-IRU
The light has flashed once in the past.

[21-2 Sound Emission]

Sono tokushuna tori wa kako ni ichido dake nai-te-iru.
(2F) the special bird TOP past in once only cry-TE-IRU
The rare species of birds has sung only once in the past.

427
[22-1 Contact]
Jack wa kako ni kodomo o hidoku tatai-te-iru.
(2F) Jack TOP past in child ACC badly hit-TE-IRU
Jack has hit his child badly in the past.

[22-2 Bodily Motion]

Mary wa kako ni ichido dake John ni te o hut-te-iru.
(2F) Mary TOP past in once only John to hand ACC wave-TE-IRU
Mary has waved to John once in the past.

[23 Touching]
John wa kako ni ichido Mary no kami ni te o hure-te-iru.
(2F) John TOP past in once Mary GEN hair at hand ACC touch-TE-IRU
John has touched Mary’s hair once in the past.

[24 Manner of Motion 1]

Chichi wa ichido Honolulu marason de hashit-te-iru.
(2F) father TOP once Honolulu marathon in run-TE-IRU
Father has run once in the Honolulu marathon.

[25 Manner of Motion 2]

Haha wa ichido sekai taikai de oyoi-de-iru.
(2F) mother TOP once world competition in swim-TE-IRU
My mother has swum once in the world competition.

[26 Motion/Movement]
Sono kekkan sha wa ichido dake ikkagestsu mae ni ugoi-te-iru.
(2F) the defecitve car TOP once only one.month ago at move-TE-IRU
The defective car has moved only once, one month ago.

[27 Performance]
Haha wa 1960nen ni Blackpool de odot-te-iru.
(2F) mother TOP 1960year in Blackpool in dance-TE-IRU
My mother danced in Blackpool in 1960.

[28 Create Mark/Defect]

Kanojo wa ichido sono tegami o birit-to yabui-te-iru.
(2F) she TOP once the letter ACC ONMP1 tear-TE-IRU
She has torn the letter once.

[29 Consumption]
John wa kako ni ichido hebi o tabe-te-iru.
(2F) John TOP past in once snake ACC eat-TE-IRU
John ate a snake once in the past.

[30 Covering]
Hukai kiri ga kako ni ichido sono machi o oot-te-iru.
(2F) deep fog NOM past in once the city ACC cover-TE-IRU
The deep fog covered the city once in the past.

428
[31 Creation]
Haha wa kako ni ichido dake kimono o tsukut-te-iru.
(2F) mother TOP past in once only Japanese.dress ACC make-TE-IRU
My mother made a Japanese dress once in the past.

[32 Dismantle]
Sono daigaku wa kako ni honsha o torikowashi-te-iru.
(2F) the university TOP past in main.building ACC dismantle-TE-IRU
The university has dismantled its main building in the past.

[33 Dyeing and Shaving]

Kanojo wa kako ni ichido kami o some-te-iru.
(2F) she TOP past in once hair ACC dye-TE-IRU
She has dyed her hair once in the past.

[34 Dressing]
Mary wa kako ni ichido dake kimono o ki-te-iru.
(2F) Mary TOP past in once only Jap.dress ACC wear-TE-IRU
Mary has worn a Japanese dress once in the past.

[35 Dying]
Chichi wa 1990nen ni nakunat-te-iru.
(2F) father TOP 1990year in die-TE-IRU
My father has been dead since 1990.

[36 Fracture/Breaking (intra)]

Sono tsubo wa kyonen ichido ware-te-iru.
(2F) the vase TOP last.year once break-TE-IRU
The vase was broken once last year.

[37 Attachment]
Sono kimono ni-wa ichido gamu ga kuttsui-te-iru.
(2F) the Japanese.dress to-TOP once chewing.gum NOM stick-TE-IRU
Chewing gum has stuck to the Japanese dress once.

[38 Killing]
John wa kako ni ichido hito o koroshi-te-iru.
(2F) John TOP past in once man ACC kill-TE-IRU
John has killed a man once in his past.

[39 Winning]
Nihon chiimu wa kankoku chiimu ni 10nen mae ni kat-te-iru.
(2F) Japan team TOP Korea team against 10year ago at win-TE-IRU
The Japanese team beat the Korean team 10 years ago.

[40 Discovery]
Sono kenkyuu chiimu wa {kyonen} shinsei o hakkenshi-te-iru.
(2F) the research team TOP {last.year} new.star ACC discover-TE-IRU
The research team has discovered a new star (*last year).

429
12. Situation types that take 2*

[1-1Natural Kinds]

[1-2 Ethnicity]

[1-3 Biological Kinds]

[1-4 Physical Properties]

[2-1 Existence]

[2-2 Necessity]

[2-3 Dispositions]

[3 Social Role]

[4 Physical States]

[5 Bodily States]

13. Situation types that take 3P

[14 Perceptual Activity]

Watashi wa sono bideo o nijikan de mi-ta.
(3P) I TOP the video ACC two.hours in watch-PAST
I watched the video in two hours.

[16 Change of Weight]

Mary wa sankagestsu de {kirei ni} yase-ta.
(3P) Mary TOP three.months with {beautiful RST} become.thin-PAST
Mary became beautifully slender in three months.

[17 Change of Physical Properties]

Suupu wa gohun de atatamat-ta.
(3P) soup TOP five.minutes with get.warm-PAST
The soup got warm in five minutes.

[18 Blooming]
Niwa no bara no hana ga ichinichi de sai-ta.
(3P) garden GEN rose GEN flower NOM one.day in bloom-PAST
The rose in the garden bloomed in one day.

[19 Directed Motion]

Sono kodomo wa gohun de mizu ni hait-ta.
(3P) that child TOP five.minutes with water in go.into-PAST
The child went into the water in five minutes.
(The event happens gradually, a part of the child’s body enters the water little
by little.)

430
[20 Open/Close]
Sono tsubomi wa gojikan de hirai-ta.
(3P) the bud TOP five.hour in open-PAST
The bud opened in five hours.

[24 Manner of Motion 1]

Taro wa 10kiro o 50pun de hashit-ta.
(3P) Taro TOP 10kilometre ACC 50minutes in run-PAST
Taro ran 10 kilometres in 50 minutes.

[25 Manner of Motion 2]

Hanako wa 2kiro o 10pun de oyoi-da.
(3P) Hanako TOP 2kilometre ACC ten.minute in swim-PAST
Hanako swum 2 kilometres in ten minutes.

[27 Performance]
Jacqui wa Corrs no shinkhoku o gohun de utat-ta.
(3P)
Jacqui TOP Corrs GEN new.song ACC five.minute in sing-PAST
Jacqui sang Corrs’ new song in five minutes.

[29 Consumption]
Jack wa ookina sandoicchi o ippun de tabe-ta.
(3P) Jack TOP big sandwich ACC one.minute in eat-PAST
Jack ate the big sandwich in one minute.

[30 Covering]
Chichi wa juppun de kuruma o shiito de oot-ta.
(3P) father TOP ten.minute in car ACC sheet with cover-PAST
My father covered his car with a sheet in ten minutes.

[31 Creation]
Haha wa sono doresu o ikkagestsu de tsukut-ta.
(3P) mother TOP the dress ACC one.month in make-PAST
My mother made that dress in one month.

[32 Dismantle]
Sono burudoozaa wa ichinichi de kousha o torikowashi-ta.
(3P) the bulldozer TOP one.day in schoolhouse ACC dismantle-PAST
The bulldozer dismantled the old schoolhouse in one day.

[33 Dyeing and Shaving]

Kanojo wa ichijikan de kami o some-ta.
(3P) she TOP one.hour in hair ACC dye-PAST
She dyed her hair in one hour.

431
14. Situation types that take 3V

[8 Posture 2]
Steve wa totsuzen tat-ta.
(3V) Steve TOP suddenly stand-PAST
Steve stood up suddenly.

[9 Perception]
Huji-san ga totsuzen mie-ta.
(3V) Mt.Fuji NOM suddenly be.visible-PAST
Mt. Fuji suddenly became visible.

[10 Cognition 1]
Taro wa shunkanni sono suushiki o rikaishi-ta.
(3V) Taro TOP in.an.instant the formula ACC understand-PAST
Taro finally understood the formula in an instant.

[11 Cognition 2]
Watashi wa totsuzen [Hanako o kirei da to] omot-ta.
(3V) I TOP suddenly [Hanako is beautiful] think-PAST
I suddenly thought that Hanako was beautiful.

[12 Emotional Activity]

Haha wa watashi no seikou o totsuzen yorokon-da.
(3V) mother TOP I GEN success ACC suddenly rejoice-PAST
My mother suddenly became happy with my success.

[13 Sleeping]
Chichi wa saiminjutsu de shunkanni namut-ta.
(3V) father TOP hypnotism with in.an.instant sleep-PAST
My father fell asleep in an instant due to being hypnotised.

[14 Perceptual Activity]

Watashi wa sono shunkan yuurei o mi-ta.
(3V) I TOP the instant ghost ACC see-PAST
I saw a ghost in that instant.

[15 Attitudes]
Kanojo wa kyuuni sumashi-ta.
(3V) she TOP suddenly put.on.airs-PAST
She suddenly put on airs.

[17 Change of Physical Properties]

Sono kinzoku wa sono saishin gijutsu de shunkanni hie-ta.
(3V) that metal TOP that latest technology with in.an.instant cool-PAST
The metal cooled down in an instant with the latest technology.

432
[18 Blooming]
Niwa no bara no hana ga totsuzen sai-ta.
(3V) garden GEN rose GEN flower NOM suddenly bloom-PAST
The rose in the garden bloomed suddenly.

[19 Directed Motion]

John wa totsuzen koya ni hait-ta.
(3V) John TOP suddenly kennel ALL go.into-PAST
John suddenly went into his kennel.

[20 Open/Close]
Totsuzen mado ga hirai-ta.
(3V) suddenly window NOM open-PAST
Suddenly the window opened.

[21-1 Light Emission]

Sono dentou wa totsuzen chikachika hikat-ta.
(3V) the light TOP suddenly ONMP2 shine-PAST
The light flashed suddenly.

[21-2 Sound Emission]

Nezumi ga totsuzen chuchu nai-ta.
(3V) mouse NOM suddenly ONMP2 cry-PAST
The mouse suddenly squeaked.

[22-1 Contact]
Jack wa Mary no kata o totsuzen ponpon tatai-ta.
(3V) Jack TOP Mary GEN shoulder ACC suddenly ONMP2 pat-PAST
Jack patted Mary on the shoulder suddenly. (not semelfactive)

[22-2 Bodily Motion]

Mary wa John ni totsuzen te o hut-ta.
(3V) Mary TOP John to suddenly hand ACC wave-PAST
Mary suddenly waved to John.

[23 Touching]
John wa totsuzen Mary no kami ni te o hure-ta.
(3V) John TOP suddenly Mary GEN hair at hand ACC touch-PAST
John suddenly touched Mary’s hair with his hand.

[24 Manner of Motion 1]

?Taro wa totsuzen hashit-ta.
(3V) Taro TOP suddenly run-PAST
Taro suddenly ran. (Taro suddenly started to run.)

[25 Manner of Motion 2]

?Hanako wa totsuzen oyoi-da.
(3V) Hanako TOP suddenly swim-PAST
Hanako suddenly swam. (Hanako suddenly started to swim.)

433
[26 Motion/Movement]
Sono kuruma wa totsuzen ugoi-ta.
(3V) the car TOP suddenly move-PAST
The car suddenly moved.

[27 Performance]
?Jacqui wa totsuzen utat-ta.
(3V) Jacqui TOP suddenly sing-PAST
Jacqui suddenly sang. (Jacqui suddenly started to sing.)

[28 Create Mark/Defect]

Kanojo wa totsuzen tegami o yabui-ta.
(3V) she TOP suddenly letter ACC tear-PAST
She suddenly tore up the letter.

[29 Consumption]
?Jack wa shunkanni sandoicchi o hitokuchi de tabe-ta.
(3V) Jack TOP in.an.instant sandwich ACC one.gulp with eat-PAST
Jack ate the sandwich in one gulp in an instant.

[30 Covering]
Kiri ga machi o totsuzen oot-ta.
(3V) fog NOM city ACC suddenly cover-PAST
The fog suddenly covered the city.

[32 Dismantle]
Bakudan de totsuzen kyuu kousha o torikowashi-ta.
(3V) bomb with suddenly old schoolhouse ACC dismantle-PAST
They dismantled the old schoolhouse suddenly with the bomb.

[33 Dyeing and Shaving]

?Kanojo wa totsuzen kami o some-ta.
(3V) she TOP suddenly hair ACC dye-PAST
She suddenly dyed her hair. (She suddenly started to dye her hair.)

[34 Dressing]
Mary wa 5ji 5hun ni huku o ki-ta.
(3V) Mary TOP 5o'clock 5minutes at clothes ACC put.on-PAST
Mary dressed herself at 5:05.

[35 Dying]
Chichi wa totsuzen nakunat-ta.
(3V) father TOP suddenly die-PAST
My father died suddenly.

[36 Fracture/Breaking (intra)]

Sono tsubo wa totsuzen ware-ta.
(3V) the vase TOP suddenly break-PAST
The vase broke suddenly.

434
[37 Attachment]
Sono secchakuzai wa shunkanni kinzoku ni kuttsui-ta.
(3V) the adhesive.agent TOP in.an.instant metal to stick-PAST
The adhesive agent stuck to metal in an instant.

[38 Killing]
Chichi wa shunkanni shichimenchou o kososhi-ta.
(3V) father TOP in.an.instant turkey ACC kill-PAST
Father killed a turkey in an instant.

[39 Winning]
John wa 5ji 5hun ni kesshou.sen ni kat-ta.
(3V) John TOP 5o'clock 5minutes at final.game OJCT win-PAST
John won the final game at 5:05.

[40 Discovery]
Sono kenkyuu chiimu wa 5ji 5hun ni shinsei o hakkenshi-ta.
(3V) the research team TOP 5o'clock 5minutes at new.star ACC discover-PAST
The research team discovered a new star at 5: 05.

15. Situation types that take 3C

[21-1 Light Emission]

Sono dentou wa chikat-to ichido hikat-ta.
(3C) the light TOP ONMP1 once shine-PAST
The light flashed only once.

[21-2 Sound Emission]

Nezumi ga ichido chut-to nai-ta.
(3C) mouse NOM once ONMP1 cry-PAST
The mouse squeaked once.

[22-1 Contact]
Jack wa Mary no kata o totsuzen pon-to tatai-ta.
(3C) Jack TOP Mary GEN shoulder ACC suddenly ONMP1 pat-PAST
Jack suddenly patted Mary on the shoulder once.

[22-2 Bodily Motion]

Mary wa John ni ikkai te o hut-ta.
(3C) Mary TOP John to once hand ACC wave-PAST
Mary waved to John once.

[23 Touching]
John wa isshun Mary no kami ni te o hure-ta.
(3C) John TOP one.moment Mary GEN hair at hand ACC touch-PAST
John touched Mary’s hair with his hand just for a second.

435
[28 Create Mark/Defect]
Kanojo wa birit-to sono tegami o yabui-ta.
(3C) she TOP ONMP1 the letter ACC tear-PAST
She tore the letter.

16. Situation types that take 3R

[8 Posture 2]
Sono roujin wa gohun de tat-ta.
(3R) the old.man TOP five.minute in stand-PAST
The old man stood up in five minutes.

[9 Perception]
Huji-san ga juppun de mie-ta.
(3R) Mt.Fuji NOM ten.minute in be.visible-PAST
Mt. Fuji became visible in ten minutes (after ten minutes).

[10 Cognition 1]
Taro wa ikkagestsu de sono suushiki o rikaishi-ta.
(3R) Taro TOP one.month in the formula ACC understand-PAST
Taro understood the formula in one month (one month later).

[11 Cognition 2]
?Hanako wa kami o ikkagetsu de shinji-ta.
(3R) Hanako TOP God ACC one.month in believe-PAST
Hanako came to believe in God in one month (one month later).

[13 Sleeping]
Mary wa gohun de nemut-ta.
(3R) Mary TOP five.minute in sleep-PAST
Mary fell asleep in five minutes (five minutes later).

[14 Perceptual Activity]

?Watashi wa ippun de yuurei o mi-ta.
(3R) I TOP one.minute in ghost ACC see-PAST
I saw the ghost in one minute (one minute later).

[18 Blooming]
?Niwa no bara no hana ga ichinichi de sai-ta.
(3R) garden GEN rose GEN flower NOM one.day in bloom-PAST
The rose in the garden bloomed in one day (one day later).

[19 Directed Motion]

John wa ippun de koya ni hait-ta.
(3R) John TOP one.minute with kennel ALL go.into-PAST
John went into his kennel in one minute (one minute later).

436
[20 Open/Close]
?Sono doa wa ippun de hirai-ta.
(3R) the door TOP one.minute in open-PAST
The door opened in one minute (one minute later).

[21-1 Light Emission]

?Sono dentou wa ippun de hikat-ta.
(3R) the light TOP one.minute in shine-PAST
The light flashed in one minute (one minute later).

[21-2 Sound Emission]

?Sono tori wa ichijikan de nai-ta.
(3R) the bird TOP one.hour in cry-PAST
The bird sang in one hour (one hour later).

[22-1 Contact]
?Jack wa nihun de Brigitte no kata o tatai-ta.
(3R) Jack TOP two.minute in Brigitte GEN shoulder ACC pat-PAST
Jack patted Brigitte on the shoulder in two minutes. (It took two minutes.)

[22-2 Bodily Motion]

?Mary wa John ni ippun de te o hut-ta.
(3R) Mary TOP John to one.minute in hand ACC wave-PAST
Mary waved to John in one minute (after one minute).

[23 Touching]
?Nina wa juppun de buta ni hajimete hure-ta.
(3R) Nina TOP ten.minute in pig at for.the.first.time touch-PAST
Nina touched a pig in ten minutes for the first time (ten minutes later).

[24 Manner of Motion 1]

?Taro wa ippun de hashit-ta.
(3R) Taro TOP one.minute in run-PAST
Taro ran in one minute (one minute later).
(It took one minute before Taro started running.)

[25 Manner of Motion 2]

?Hanako wa ippun de oyoi-da.
(3R) Hanako TOP one.minute in swim-PAST
Hanako swam in one minute (one minute later).
(It took one minute before Hanako started swimming.)

[26 Motion/Movement]
Sono kuruma wa sanpun de ugoi-ta.
(3R) the car TOP three.minute in move-PAST
The car moved in three minutes (three minutes later).

437
[27 Performance]
?Jacqui wa gohun de utat-ta.
(3R) Jacqui TOP five.minutes in sing-PAST
Jacqui sang in five minutes (five minutes later).
(It took five minutes before she started to sing.)

[28 Create Mark/Defect]

?Kanojo wa gohun de sono tegami o yabui-ta.
(3R) she TOP five.minute in the letter ACC tear-PAST
She tore up the letter in five minutes (five minutes later).

[30 Covering]
?Juppun de kiri ga machi o oot-ta.
(3R) ten.minute in fog NOM city ACC cover-PAST
The fog covered the city in ten minutes (ten minutes later).

[32 Dismantle]
Sono kodomo wa juppun de sono omocha o kowashi-ta.
(3R)
the child TOP ten.minute in the toy ACC destroy-PAST
The child destroyed the toy in ten minutes (ten minutes later).

[34 Dressing]
Mary wa gohun de huku o ki-ta.
(3R) Mary TOP five.minute in clothes ACC put.on
Mary dressed herself in five minutes (five minutes later).

[35 Dying]
Chichi wa hatsubyou kara ichinen de nakunat-ta.
(3R) father TOP falling.ill from one.year in die-PAST
My father died in one year after he fell ill.

[36 Fracture/Breaking (intra)]

Sono tsubo wa gohun de ware-ta.
(3R) the vase TOP five.minutes in break-PAST
The vase broke in five minutes (five minutes later).

[37 Attachment]
Sono secchakuzai wa juppun de kinzoku ni kuttsui-ta.
(3R) the adhesive.agent TOP ten.minute in metal to stick-PAST
The adhesive agent stuck to metal in ten minutes (after ten minutes).

[38 Killing]
Chichi wa 10pun de shichimenchou o koroshi-ta.
(3R) father TOP 10minutes with turkey ACC kill-PAST
Father killed a turkey in ten minutes (ten minutes later).

[39 Winning]
John wa 10pun de kesshou.sen ni kat-ta.
(3R)
John TOP 10minutes with final.game OJCT win-PAST
John won the final game in ten minutes (ten minutes later).

438
[40 Discovery]
Sono kenkyuu chiimu wa ichinen de shinsei o hakkenshi-ta.
(3R) the research team TOP one.year in new.star ACC discover-PAST
The research team discovered a new star in one year (one year later).

17. Situation types that take 3U

[12 Emotional Activity]

Haha wa nagai aida watashi no seikou o yorokon-da.
(3U) mother TOP long period I GEN success ACC rejoice-PAST
My mother was happy with my success for a long time.

[13 Sleeping]
Watashi wa kinou rokujikan no aida nemut-ta.
(3U) I TOP yesterday six.hour GEN period sleep-PAST
I slept for six hours yesterday.

[14 Perceptual Activity]

Watashi wa kinou sanjikan no aida terebi o mi-ta.
(3U) I TOP yesterday three.hours GEN period TV ACC watch-PAST
I watched TV for three hours yesterday.

[18 Blooming]
?Niwa no bara no hana ga mikkakan no aida sai-ta.
(3U) garden GEN rose GEN flower NOM three.day GEN period bloom-PAST
?The rose in the garden was in bloom for three days.

[21-1 Light Emission]

Sono dentou wa ichijikan no aida hikat-ta.
(3U) the light TOP one.hour GEN period shine-PAST
The light flashed for one hour.

[21-2 Sound Emission]

Sono tori wa ichijikan no aida nai-ta.
(3U) the bird TOP one.hour GEN period cry-PAST
The bird sang for one hour.

[22-1 Contact]
Jack wa Brigitte no kata o ichijikan no aida tatai-ta.
(3U) Jack TOP Brigitte GEN shoulder ACC one.hour GEN period pat-PAST
Jack patted Brigitte on the shoulder for an hour.

[22-2 Bodily Motion]

Mary wa John ni gohunkan no aida te o hut-ta.
(3U) Mary TOP John to five.minutes GEN period hand ACC wave-PAST
Mary waved her hand to John for five minutes.

439
[23 Touching]
Nina wa neko ni te o ichijikan no aida hure-ta.
(3U) Nina TOP cat at hand ACC one.hour GEN period touch-PAST
Nina touched the cat’s fur with her hand for one hour.

[24 Manner of Motion 1]

Taro wa ichijikan no aida hashit-ta.
(3U) Taro TOP one.hour GEN period run-PAST
Taro ran for one hour.

[25 Manner of Motion 2]

Hanako wa ichijikan no aida oyoi-da.
(3U) Hanako TOP one.hour GEN period swim-PAST
Hanako ran for one hour.

[26 Motion/Movement]
Sono omocha wa shibaraku no aida ugoi-ta.
(3U) the toy TOP while GEN period move-PAST
The toy moved for a while.

[27 Performance]
Jacqui wa gohun no aida utat-ta.
(3U) Jacqui TOP five.minutes GEN period sing-PAST
Jacqui sang for five minutes. (without finishing a song)

[28 Create Mark/Defect]

Neko wa gohun no aida kabe o hikkai-ta.
(3U) cat TOP five.minute GEN period wall ACC scratch-PAST
The cat scratched the wall for five minutes.
(engaged in the activity for five minutes)

18. Situation types that take 3D

[14 Perceptual Activity]

Watashi wa sono bideo o ichijikan no aida mi-ta.
(3D) I TOP the video ACC one.hour GEN period watch-PAST
I watched the video for an hour.

[15 Attitudes]
Kare wa masumasu hinekure-ta.
(3D) he TOP more.and.more get.warped-PAST
He got more and more warped.

[16 Change of Weight]

Mary wa sukoshizutsu hutot-ta.
(3D) Mary TOP little.by.little become.fat-PAST
Mary became fat little by little.

440
[17 Change of Physical Properties]
Suupu wa sukoshizutsu atatamat-ta.
(3D) soup TOP little.by.little get.warm-PAST
The soup got warm little by little.

[20 Open/Close]
Sono doa wa sukoshizutsu hirai-ta.
(3D) the door TOP little.by.little open-PAST
The door opened little by little. (But we do not know if it opens fully)

[26 Motion/Movement]
Jishin de tsukue ga sukoshizutsu minami ni ugoi-ta.
(3D) earthquake INST desk NOM little.by.little south to move-PAST
The desk moved little by little towards the south because of the earthquake.

[27 Performance]
Jacqui wa gohun no aida sono shinkyoku o utat-ta.
(3D) Jacqui TOP five.minute GEN period the new.song ACC sing-PAST
Jacqui sang Corrs’ new song for five minutes.
(She did not necessarily finish it.)

[29 Consumption]
Joel wa gohunkan no aida yuushoku o tabe-ta.
(3D) Joel TOP five.minutes GEN period supper ACC eat-PAST
Joel ate his supper for five minutes.
(He did not necessarily finish his supper.)

[30 Covering]
Chichi wa kuruma o shiito de gohunkan no aida oot-ta.
(3D) father TOP car ACC sheet with five.minutes GEN period cover-PAST
My father covered his car with a sheet little by little for five minutes.

[31 Creation]
Haha wa nijikan no aida doresu o tsukut-ta.
(3D) mother TOP two.hour GEN period dress ACC make-PAST
My mother made (was making) the dress for two hours.
(engaged in the activity for two hours)

[32 Dismantle]
Sono kikai wa sanjikan no aida tatemono o torikowashi-ta.
(3D) the machine TOP three.hours GEN period building ACC dismantle-PAST
The machine dismantled (was dismantling) the building for three hours.
(does not have to be finished)

441
[33 Dyeing and Shaving]
Kanojo wa ichijikan no aida kami o some-te
(3D) she TOP one.hour GEN period hair ACC dye-and

sorekara shibaraku kyuukei-shi-ta.

then for.a.while take.rest-PAST

She dyed her hair for an hour, then took a rest.

(May not have finished dying her hair.)

19. Situation types that take 3T

[2-1 Existence]
Gohun mae wa soko ni hon ga at-ta.
(3T) five.minute ago TOP there at book NOM be-PAST
The book was there five minutes ago.

[2-2 Necessity] iru2 ‘need’

Kyonen kare ni-wa kane ga it2-ta.
(3T) last.year he to-TOP money NOM need/necessary-PAST
He needed money last year.

[2-3 Dispositions]
Jack wa sukoshi no aida shinsetsu dat-ta.
(3T) Jack TOP little GEN period kind DA-PAST
Jack was kind for a while.

[3 Social Role]
John wa gonenkan shachou dat-ta.
(3T) John TOP for.five.years president DA-PAST
John was president for five years.

[4 Physical States]
Sono mizuumi no mizu wa ikkagestsukan tsumetak-kat-ta.
(3T) the lake GEN water TOP for.a.month cold-inflection-PAST
The water of the lake was cold for one month.

[5 Bodily States]
Kare wa byouki dat-ta.
(3T) he TOP ill DA-PAST.
He was ill.

[9 Perception]
Kinou wa Huji-san ga gojikan no aida mie-ta.
(3T) yesterday TOP Mt.Fuji NOM five.hour GEN period be.visible-PAST
Mt. Fuji was visible for five hours yesterday.

442
20. Situation types that take 3*

[1-1Natural Kinds]
[1-2Ethnicity]
[1-3 Biological Kinds]
[1-4 Physical Properties]
[7-1 Relation 2]
[6 Relation 1]
[7-2 Posture 1]

443
APPENDIX B. Levin’s classification of verbs of putting
(Levin 1993: 111-122)

9.1 Put Verbs (install, place, put)

ex. I put the book on/under/near the table.

(i) they refer to putting an entity at some location

(ii) the location is expressed via a prepositional phrase headed by one of the
locative prepositions
(iii) Prepositional phrases cannot be headed by the goal to or the source
from
ex. *I put the book from Edna to Sally.

9.2 Verbs of Putting in a Spatial Configuration (hang, lean, stand)

ex. Cheryl stood the books on the shelf/next to the magazines.

(i) they refer to putting an entity at some location

(ii) they specify the particular spatial configuration that the placed entity
ends up in with respect to the location
(iii) they allow a variety of locative prepositional phrases, but not with
source or goal phrases
(iv) most of them have the causative alternation
ex. a. Cheryl stood the books on the table.
b. The books stood on the table.

9.3 Funnel Verbs (funnel, scrape, wipe)

ex. I funneled the mixture into the bottle.

?? I funneled the mixture in the bottle.
(i) they relate to putting an entity in some location in some manner
(ii) they usually involve putting entities in spatially confined locations
(iii) they prefer into rather than in and onto rather than on
(iv) they occur with a wide range of locative prepositions
(v) they do not occur with to but some of them occur with from
(vi) they are vague about the resulting spatial configuration of the entity
placed

9.4 Verbs of Putting with a Specified Direction (hoist, lift, lower)

ex. I lifted the books.

I lifted the book onto the table/out of the box.
*I lifted the book on the table.

(i) they relate to putting an entity somewhere, typically by moving it in a

specific direction (exerting a force against the action of gravity)
(ii) they show a preference for onto and into over on or in
(iii) they occur with a wide range of locative prepositions
(iv) some of them occur in source prepositions or with paths describing a
444
 trajectory of motion

9.5 Pour Verbs (dribble, pour, spew)

ex. Tamara poured water into the bowl/over the flowers.

Tamara poured water from/out of the pitcher.

(i) they relate to putting things—typically liquids—on surfaces or in

containers
(ii) they allow only the locative variant of the locative alternation
ex. a. Tamara poured water into the bowl.
b. *Tamara poured the bowl with water.
(iii) they allow from phrases
(iv) many of them participate in a causative alternation
ex. a. Tamara poured water onto the plants.
b. Water poured onto the plants.

9.6 Coil Verbs (coil, roll, twist)

ex. Cora coiled the rope around the post.

(i) they relate to putting something around something else.

(ii) they are most often found with the preposition around
(a limited range of preposition)
(iii) they can be used as intransitive verbs of manner of motion

9.7 Spray/Load Verbs (inject, load, spray)

ex. Jessica loaded boxes onto/into/under the wagon.

(i) they relate to covering surfaces and putting things into containers
(ii) they participate in the locative alternation
(holistic/partitive effect)
ex. a. Jessica loaded boxes on the wagon.
b. Jessica loaded the wagon with boxes.
(iii) verbs that take a liquid or a set of small particles as the typical direct
object in the locative variant are found in the conative alternation
ex. a. Jessica sprayed water at me.
b. *Jessica loaded boxes at the truck.

9.8 Fill Verbs (cover, decorate, staff)

ex. Leslie staffed the store with employees.

(i) they relate to covering surfaces and putting things into containers
(ii) they are only found in the with variant of the locative alternation
(iii) when the locatum is expressed as the subject, the sentence can be
understood as describing a state
(iv) they typically describe the resulting state of a location as a consequence
of putting something on it or in it

445
(v) verbs that involve covering with clothes or cloths, allow with to
alternate in
ex. Leigh swaddled the baby with/in blankets.

9.9 Butter Verbs (butter, plaster, sugar)

ex. Lora buttered the toast.

(i.e., Lora put butter on the toast.)

(i) they all have zero-related nominals(note 2); the related nouns refer to an
entity that is moved
(ii) the object of these verbs receives the “holistic” interpretation found in
the with variant of the locative alternation
ex. a. *Lora buttered unsalted butter on the toast.
b. Lora buttered the toast with unsalted butter.

9.10 Pocket Verbs (bottle, ground, pocket)

ex. Lydia pocketed the change.

(i.e., Lydia put the change in her pocket.)

(i) they all have zero-related nominals: the related nouns refer to a location
where things can be put

446
APPENDIX C. List of classes of verbs of putting in Japanese

447
CLS Scls Type Name No FA-type linkings GA-type linkings GO-linking Source Other s.pattern aspect prominent syntactic or semantic
L-linking FA-linking I-linking GA-linking characteristics
1 oku 'put' 15 OK * * * * * punctual (1) allow various locational phrases,
(2) putting an entity to a location

2 1 ireru 'put into' 5 OK * * * * OK path pattern punctual (1) are productive as V2 components,
(2) spatial transfer with a specified
direction
2 2 ageru 'raise' 3 OK OK * * * OK path pattern punctual/ (1) are productive as V2 components,
extended (2) spatial transfer with a specified
direction
3 hitasu 'soak' 18 OK * * * * * punctual/ (1) the locational phrase (inside) can
extended be used instead of ni. (2) putting
things in a confined place

4 sosogu 'pour into' 9 OK * * * * ?? (OK for path pattern punctual/ (1) putting or pouring an entity into a
some) extended container
5 kabuseru 'put on' 38 OK * * * * * punctual/ (1) similar to type 1 and type 3, but it
extended does not specify the figure is in/inside
the ground
6 1 tsukeru 'apply' 4 OK * * * * * extended (1) the figure is n uncountable entity,
(2) the ground is a surface
6 2 maku 'scatter' 14 OK OK * * * OK path (mostly) extended (1) the figure is uncountable entity,
(mostly) (mostly) (2) the ground is a surface
7 1 tsurusu 'hang' 6 OK OK * * * OK punctual (1) entail spatial configuration, (2) ni
can be replaced by kara specifying an
identical referrent (the ground)

7 2 tateru 'stand' 2 OK OK * * * * punctual (1) entail spatial configuration

448
CLS Scls Type Name No FA-type linkings GA-type linkings GO-linking Source Other s.pattern aspect prominent syntactic or semantic
L-linking FA-linking I-linking GA-linking characteristics
8 tsumu 'load' 3 OK * * * * * GS-linking punctual/ (1) denote event of loading, (2) take
with te-iru, extended GS-linking
??path
9 oou 'cover' 9* * OK OK (some) * * IS-linking punctual/ (1) take IS-linking
extended
10 yogosu 'dirty' 15 * * OK OK * * punctual/ (1) used as change of state verbs
extended
11 kazaru 'decorate' 6 OK * OK OK (some) * * extended (1) manifest the locative alternation
(punctual)
12 tsutsumu 'wrap' 4 OK * OK (RI) * * * extended (1) manifest the locative alternation
by FG construal alternation
13 hou.boku-suru 'graze' 6 OK OK * * * * punctual (1) ground incorporation

14 san.sui-suru 'scatter 12 OK OK (some) * * OK OK (some) path (some) punctual (1) figure incorporation
water' extended
15 shoku.rin-suru 3 OK * OK OK OK * path (some) extended (1) figure incorporation with the
'afforest' alternating linkings
16 hame-komu 'put in/on- 80 OK * * * * OK (some) path pattern punctual/ (1) means compound with specified
put.into' extended direction
17 nuri-tsukeru 'smear- 22 OK * * * * * punctual/ (1) means compound of FA-type
apply' extended linking
18 ooi-kakusu 'cover- 2* * OK OK * * extended (1) means compound of GA-type
hide' linking

449
APPENDIX D. List of verbs of putting in Japanese

The following are the major topics in the list.

(1) The major topics in the list1:

[CLS] Class of verbs (of putting)

[Scls] Subclass of verbs (of putting) (or type of events)
[Japanese] Japanese verbs
[Vtype] Type of Verb
[English] English translation equivalent
[Direct Translation] Direct and detailed translation
[English Levin] English Corresponding Verbs in Levin (1993)

[CLS] (and [Scls]): class of verbs

They correspond to the number of types (and subtypes) of events described and

discussed. If a certain verb of putting in Japanese is ‘1 put type,’ then the number the

verb is assigned will be ‘1.’ These numbers for CLS and Scls are the same as in the

summary of verbs of putting in the main text. I use Type 20-1 ~ Type 20-6 to refer to

other classes of three argument verbs which are related to verbs of putting.

Type 20-1—verbs of sending and carrying

Type 20-2—verbs of transfer of possession
Type 20-3—verbs of combining
Type 20-4—verbs of contact
Type 20-5—verbs of throwing
Type 20-6—verbs of removing

[Japanese]: Japanese verbs

I list the Japanese verbs. I do not distinguish homonymous verbs that have different

senses which can be categorised differently in my classification of verbs of putting or

polysemous verbs that have different but related senses.

1
Characters/words in the square brackets show an abbreviation or a title for each item in the list.

450
[Vtype]: Type of verb

Under this item, I clarify which type of verb (simple, compound, or VN-suru) a verb

in question belongs to. The following is the abbreviation for each verb type:

(2) Types of Verbs and abbreviation:

Simple verb—simple
Pair compound—pair
Means compound—means
Deverbalised V2 compound—dev2
Deverbalised V1 compound—dev1
Idiomatic (frozen) compound—frozen
Chinese VN suru—CHI
English VN suru—ENG

[English]: English translation equivalent

I have referred to the Japanese-English dictionaries listed in the methodology. I pick

up English verbs whose meaning is closest. However, I need to mention that these

translation equivalents are sometimes ‘rough’ or ‘coarse-grained’ in a strict sense.

What I intend to do is to investigate the semantics of verbs which are thought to be

relevant to argument linking and to propose a proper semantic representation. In this

aim, I look at each verb more closely than the dictionaries usually do. For example,

kazaru and its ‘rough’ English equivalent decorate do not exactly appear in the same

argument linking patterns.

[Direct Translations]: direct and detailed translation

This applies to non-simple verbs. A detailed translation is given with each of the

components of verbs translated into English. For example, each V1 and V2

component of compounds is translated.

451
[English Levin]: corresponding English verbs in Levin

This column lists corresponding English verbs in Levin’s classification. The numbers

in the parentheses correspond to the subsection where the verbs in question appear in

Levin’s subcategory. (‘1’ in the discussion of verbs of putting means that the verb is

classified as 9.1 Put verbs in Levin’s classification, for example.) In the case of verbs

which are found in A Dictionary of Synonyms in Japanese or other sources in the

process of ‘(4) looking for more members,’ (cf. § 4.3.4) I specify the source. I mark

with an asterisk verbs whose source is my knowledge.

452
CLS Scls Japanese V-type English Direct Translation English Levin
1 chin.retsu-suru CHI display, exhibit line.line-do bench(10)
1 hai.chi-suru CHI arrange, post line.put-do arrange(1), place(1), position(1), set(1)
1 hai.retsu-suru CHI arrange line.line-do arrange(1), place(1)
1 ichizukeru simple position set(1)
1 ko.tei-suru CHI fix, settle fix.settle-do mount(1), set(1)
1 noseru simple put put(1), set(1)
1 oku simple put, assign place(1), position(1), put(1), set(1), situate(1), lay(2), perch(2), rest(2), settle(in)(7),
intersperse(8), staff(8), bang(3), ground(10), saddle(9), stress(9), shelve(10)
1 sarasu simple expose pillory(10)
1 sou.chi-suru CHI mount, equip decorate.put-do mount(1)
1 sueru simple set place(1)
1 sue-tsukeru frozen install install.put-do install(1), mount(1), set(1)
1 tori-tsukeru dev1 install install(1), cap(9), trap(10)
1 tsukeru simple put dock(10)
1 tsukeru simple attach spread(7), string(7), fleck(8), splotch(8), trim(8), frame(8), trim(8), smudge(7),
soil(8), spot(8), stain(8), speckle(8), line(8), harness(9), rouge(9), buttonhole(9),
bait(9), feather(9), label(9), halter(9), ticket(9), frame(9), sulphur(9), string(10),
1 uchi-tateru frozen set hit-stand set(1)
2 1 dasu simple take out polish(9)
2 1 ireru simple put into dip(3), drop(4), pump(into)(7), funnel(3), pot(10), bottle(10), dock(10), poison(9),
garage(10), box(10), case(10), bag(10), house(10), kennel(10), jar(10), cage(10),
corral(10), hangar(10), drydock(10), jug(10), spice(9), water(9), jug(10), coop(10),
pen(10), leaven(9), pocket(10), sugar(9), warehouse(10), stable(10), load(7)
2 1 kaesu simple return Matsumoto (1997: 169)
2 1 modosu simple return, vomit spew(5)
2 1 otosu simple drop Matsumoto (1997: 169)
2 2 ageru simple raise lift(4), raise(4), hoist(4)
2 2 orosu simple lower, drop, unload lower(4), dump(3), drop(4), drip(5), land(10)
2 2 sageru simple lower JSD p. 1020

453
CLS Scls Japanese V-type English Direct Translation English Levin
3 cho.zou-suru CHI keep, store store.storehouse-do bin(10)
3 hitasu simple soak immerse(1), bathe(8), drench(8), saturate(8), soak(8), oil(9), dip(3)
3 ho.zon-suru CHI keep keep.keep-do bottle(10)
3 kakumau simple shelter shelter(10)
3 kan.kin-suru CHI confine control.forbid-do pen(10)
3 kou.chi-suru CHI detain capture.put-do jail(10)
3 osameru simple put away sheathe(10)
3 otoshi-ireru frozen entrap drop-put into entangle(8)
3 shizumeru simple sink immerse(1)
3 shuu.nou-suru CHI put away put in.keep-do house(10)
3 shuu.you-suru CHI accommodate, lock up put in.put into-do house(10)
3 sonaeru simple furnish, prepare stock(with)(7)
3 takuwaeru simple store, reserve cellar(10)
3 toji-komeru frozen confine shut-put into pen(10), coop(10), cloister(10)
3 tomeru simple give lodge lodge(10)
3 tsukeru simple soak immerse(1), soak(8), chrome(9)
3 umeru simple bury, fill stop up(8)
3 uzumeru simple bury, fill fill(8)
4 chuu.sha-suru CHI inject pour.prick-do inject(7)
4 han.nyuu-suru CHI bring in carry.put into-do Matsumoto (1997: 175)
4 kumu simple draw, ladle ladle(3)
4 shiireru simple stock stock(with)(7)
4 shimau simple stow away pack(7), house(10), stash(1)
4 sosogu simple pour into funnel(3), pour(5)
4 sukuu simple scoop ladle(3), scoop(3), shovel(3)
4 suu simple suck siphon(3)
4 tsugu simple pour into pour(5)
5 ateru simple put on diaper(9), patch(9), dab(with)(7)
5 chaku.sou-suru CHI wear wear.decorate-do JSD p. 1042
5 chaku.you-suru CHI wear wear.use-do JSD p. 1042
5 chiri-bameru frozen encrust scatter-fit into encrust(8), inlay(8), stud(8)
5 chou.mi-suru CHI season, flavour adjust.taste-do season(8)

454
CLS Scls Japanese V-type English Direct Translation English Levin
5 hameru simple fit on/in mount(1), panel(9), gag(9), frame(9), glove(9)
5 haru simple stick on/in stick(on, in)(7), cover(8), mount(1), plaster(7), tile(8), label(9), plank(9), board(9),
plank(9), paper(9), wallpaper(9), veneer(9), poster(9), tile(9), parquet(9)
5 haru simple put up string(9)
5 haru simple fill JSD p. 1048
5 hasamu simple put between interleave(8), lard(8)
5 hikkaburu simple wear,put on JSD p. 54
5 itadaku simple wear JSD p. 54
5 kabuseru simple cover, put over lay(2), coat(8), cover(8), case(10), cap(9), silver(9), cap(9), crown(9), sugar(9),
garland(8)
snare(10), trap(10)
5 kakeru simple ensnare
5 kakeru simple put on, cover wrap(around)(7), yoke(9), muzzle(9)
5 kakeru simple put up sling(1)
5 kamuru simple wear JSD p. 54
5 kiseru simple dress, plate robe(8), cloak(9), mantle(9), robe(9), tin(10)
5 kobosu simple spill slop(5), spill(5)
5 kuberu simple put on fire fuel(9)
5 kuwaeru simple add drug(9)
5 maku simple sow set(1), seed(7), sow(7), stock(with)(7)
5 matou simple wear, drape around wrap(around)(7), drape(around)(7)
5 megurasu simple surround fence(9)
5 mori-ageru frozen serve, heap up mound(7)
5 moru simple heap JSD p. 395
5 noseru simple put on, place perch(2), rest(2), load(7), stand(2), shelve(10)
5 okkabuseru simple wear, cover JSD p. 54
5 sashi-hasamu dev1 interleave interleave(8)
5 sashi-kakeru dev1 hold over JSD p. 55
5 shiku simple spread on/over lay(2), pave(8), gravel(9), brick(9), sod(9), carpet(9)
5 soeru simple add, garnish season(8), smother(8), garnish(8)
5 sou.chaku-suru CHI wear decorate.wear-do JSD p. 1042
5 tataeru simple fill JSD p. 1048

455
CLS Scls Japanese V-type English Direct Translation English Levin
5 ten.koku-suru CHI stipple spot.engrave-do stipple(8)
5 ten.sai-suru CHI stipple spot.colour-do stipple(8)
5 tsukeru simple wear cover(8), cloak(9)
5 tsumi-kasaneru pair pile up pile-pile up pile(on)(7), stack(?)(7)
5 tsumu simple pile on stow(1), stack?(7), pile(with)(7)
5 ueru simple plant in set(1), bed(10), plant(7), bed(10), plant(7)
5 zou.gan-suru CHI inlay elephant.eye-do inlay(8)
6 1 haku simple apply, brush JSD p. 415
6 1 hiku simple apply oil(9), wax(9)
6 1 sasu simple apply oil(9), grease(9), rouge(9)
6 1 tsukeru simple apply spread(7), string(7), fleck(8), splotch(8), trim(8), frame(8), trim(8), smudge(7),
soil(8), spot(8), stain(8), speckle(8), line(8), harness(9), rouge(9), buttonhole(9),
bait(9), feather(9), label(9), halter(9), ticket(9), frame(9), sulphur(9), string(10),
6 2 abise-kakeru pair shower shower-splash pound(3)
6 2 abiseru simple shower, pour into pour(5), spray(with)(7), bombard(8), shower(on)(7)
6 2 baramaku simple scatter scatter(7), strew(7)
6 2 bukkakeru simple dump, pour, dash douse(8)
6 2 chirakasu simple scatter litter(8)
6 2 chirasu simple sprinkle sprinkle(7)
6 2 haku simple vomit, spew spew(5)
6 2 hane-kakeru pair splash splash-shower spatter(7), splash(7), cover(8)
6 2 hanekasu simple splash slosh(5)
6 2 kakeru simple shower, spread smother(8), water(9), sand(9), spread(7), cover(8), veil(8), veil(9), shawl(9), drape(7)
6 2 mabusu simple sprinkle, dust bread(9)
6 2 maki-chirasu pair scatter scatter-scatter scatter(7), sprinkle(7), strew(7), bestrew(8), cover(8)
6 2 maku simple scatter, sprinkle A scatter F on/over G gravel(9), salt(9), sand(9), seed(9), water(9)
6 2 tori-chirakasu dev1 scatter clutter(8), litter(8)
7 1 burasageru simple dangle dangle(2)
7 1 kakeru simple hang hang(2), suspend(2), hang(on)(7)
7 1 sageru simple hang hang(2), hang(on)(7)

456
CLS Scls Japanese V-type English Direct Translation English Levin
7 1 tarasu simple hang, drip hang(2), dribble(5), drip(5)
7 1 tsuri-sageru pair hang hang-hang Matsumoto (1997: 173-4)
7 1 tsurusu simple hang sling(1), hang(from)(2), suspend(2), hang(with)(7)
7 2 tateru simple stand set(1), stand(2)
7 2 yokotaeru simple lay JSD p. 101
8 noseru simple load, have on board load(7), load(with)(8)
8 tou.sai-suru CHI mount load.load-do mount(1)
8 tsumu simple load load(7), shovel(3)
9 huchidoru simple fringe edge(8), face(8)
9 husagu simple block wad(3), jam(with)(7), block(8), choke(8), clog(8), plug(8), stop up(8), board(9),
brick(9)
block(8)
9 huu.sa-suru CHI block block.shut-do
9 kakomu simple surround encircle(8), surround(8), brick(9)
9 kakou simple enclose surround(8), board(9)
9 oou simple cover, hide drape(7), spread(7), encrust(8), sheathe(10), veil(8), festoon(8), mask(8), veil(9),
forest(9), blanket(9), pitch(9), brick(9), turf(9), sod(9), roof(9), cover(8),
blindfold(9), bathe(8), cloak(8), shroud(8), suffuse(8), cloak(9), mantle(9),
9 sekitomeru simple dam smother(8)
dam(8)
9 tori-kakomu dev1 enclose encircle(8), ring(8)
9 tori-maku dev1 surround ring(8), surround(8), wreathe(8)
10 hei.soku-suru CHI block shut.block-do block(8)
10 ho.sou-suru CHI pave spread.decorate-do pave(8), asphalt(9)
10 huku simple cover, thatch tile(8), tile(9), shingle(9), thatch(9), slate(9), roof(9)
10 hyou.sou-suru CHI mount face.decorate-do cover(8)
10 jun.shoku-suru CHI embellish decorate.color-do embellish(8)
10 kegasu dirty, soil GJD p. 660
10 kyou.ka-suru CHI reinforce, intensify strong.change-do enrich(8)
10 nurasu simple wet squirt(with)(7), slop(5)
10 o.sen-suru CHI contaminate, pollute dirty.dye-do pollute(8), contaminate(8), infect(8)
10 someru simple dye imbue(8), infect(8)
10 sou.shoku-suru CHI decorate decorate.decorate-do adorn(8), deck(8), embellish(8)

457
CLS Scls Japanese V-type English Direct Translation English Levin
10 to.hu-suru paint smear.spread-do JSD p. 415
10 to.matsu-suru paint smear.spray-do JSD p. 415
10 to.sou-suru CHI paint, coat smear.decorate-do JSD p. 415
10 yogosu simple dirty slop(5), smear(7), smudge(7), blot(8), dirty(8), pollute(8), taint(8), soil(8), spot(8),
spatter(7), ink(9)
11 kazari-tateru dev2 decorate a lot decorate-a lot adorn(8), deck(8)
11 kazaru simple decorate, display hang(2), drape(7), hang(with)(7), adorn(8), deck(8), decorate(8), embellish(8),
emblazon(8), garnish(8), lard(8), ornament(8), trim(8), ornament(9), panel(9),
veneer(9), garland(8), wreathe(8), festoon(8), intersperse(8), wreathe(9), garland(9)
11 mitasu simple fill fill(8), inundate(8), line(8), replenish(8)
11 nuri-tateru dev2 plaster plaster-a lot plaster(7)
11 nuru simple smear, apply lay(2), slather(7), swab(with)(7), slather(7), dab(on, over)(7), wash?(with)(7),
anoint(8), poison(9), ink(9), salve(9), pitch(9), rosin(9), wax(9), tar(9), coat(8),
cover(8), stucco(9), whitewash(9), plaster(9), brush(7), oil(9), grease(9), butter(9), sp
11 tsumeru simple pack, fill pack(7), crowd(7), cram(7), choke(8), box(10), crate(10), caulk(9), fill(8), plug(8)
12 kakusu simple cover, hide stash(1), mantle(9), veneer(9), shelter(10), mask(8)
12 kurumu simple wrap roll(6), bind(8), wrap(in)(7), swaddle(8), tuck(3), blanket(9)
12 maku simple wrap, tuck curl(6), roll(6), wind(6), spool(10), wrap(around)(7), wreathe(8), wreathe(9),
swaddle(8), swathe(8)
12 tsutsumu simple wrap, veil wrap(in)(7), bind(8), cover(8), case(10), sheathe(10), wind(6), swathe(8), shroud(8),
smother(8), mantle(9), wreathe(8)
13 fairu-suru ENG file file-do file(10)
13 hou.boku-suru CHI graze release.pasture-do pasture(10)
13 nou.kan-suru CHI lay in a coffin keep.coffin-do Matsumoto (1997: 176)
13 nyuu.kan-suru CHI lay in a coffin put into.coffin-do Matsumoto (1997: 176)
13 tou.goku-suru CHI jail throw.prison-do jail(10)
13 tou.kan-suru CHI drop into mailbox throw.mailbox-do Matsumoto (1997: 176)
14 hou.sui-suru CHI spray water release.water-do Matsumoto (1997: 176)
14 mizu.maki-suru CHI water water.scatter-do JSD p. 413
14 nyuu.ka-suru CHI receive goods put into.load-do Matsumoto (1997: 176)
14 nyuu.kin-suru CHI deposit money put into.money-do Matsumoto (1997: 176)

458
CLS Scls Japanese V-type English Direct Translation English Levin
14 san.sui-suru CHI water scatter.water-do Matsumoto (1997: 176)
14 shoku.hi-suru CHI plant skin plant.skin-do GJD p. 1071
14 shoku.ju-suru CHI plant a tree plant.tree-do Matsumoto (1997: 176)
14 shoku.min-suru CHI colonise plant.people-do settle(in)(7)
14 shoku.mou-suru CHI plant hair plant.hair-do GJD p. 1072
14 shuk.ka-suru CHI ship take out.load-do Matsumoto (1997: 176)
14 shup.pin-suru CHI send to exhibit take out.goods-do Matsumoto (1997: 176)
14 uchi.mizu-suru CHI water dash.water-do JSD p. 413
15 chaku.shoku-suru CHI colour, paint apply.color-do Matsumoto (1997: 176)
15 juu.den-suru CHI charge battery fill.electricity-do Matsumoto (1997: 176)
15 shoku.rin-suru CHI afforest plant.forest-do Matsumoto (1997: 176)
16 chuu.nyuu-suru CHI pour into pour.put into-do pump(into)(7), inject(7)
16 daki-ageru means lift hold-raise lift(4)
16 haki-dasu means vomit out vomit-take out spew(5)
16 hakobi-komu means carry into carry-put into Kojien p. 2053
16 hame-komu means set/fit into fit-put into set(1), inlay(8)
16 hane-kaesu means slosh splash-return slosh(5)
16 hasami-komu means put between put between-put into JSD p. 1048
16 hiki-ageru means pull up pull-raise lift(4), raise(4), land(10), beach(10)
16 hiki-ireru means pull into pull-put into Matsumoto (1997: 173-4)
16 hiki-komu means draw into pull-put into channel(3)
16 hiki-orosu means pull down pull-lower Matsumoto (1997: 173-4)
16 hiki-otosu means pull down pull-drop Matsumoto (1997: 173-4)
16 hiki-yoseru means pull into pull-pull to Matsumoto (1997: 173-4)
16 hou.ryuu-suru CHI stock, discharge release.flow to-do stock(with)(7)
16 houri-komu means throw into throw-put into shovel(3)
16 huki-dasu means spew out spew-take out spew(5)
16 huki-komu means inspire with blow-put into stuff(7), imbue(8), impregnate(8)
16 kakae-ageru means lift hold-raise *
16 kaki-ageru means rake up rake-raise fork(10)
16 keri-ageru means kick up kick-raise Matsumoto (1997: 173-4)
16 keri-ireru means kick into kick-put into Matsumoto (1997: 173-4)

459
CLS Scls Japanese V-type English Direct Translation English Levin
16 kurumi-komu means tuck in wrap-put into tuck(3)
16 kyuu.nyuu-suru CHI inhale suck.put into-do Matsumoto (1997: 175)
16 maki-ageru means roll up roll-raise hoist(4)
16 maki-komu means roll into roll-put into entangle(8)
16 maneki-ireru means invite to invite-put into Matsumoto (1997: 173-4)
16 mochi-ageru means lift take hold of-raise lift(4), fork(10), raise(4)
16 mochi-kaeru means take back take hold of-return Kojien p. 2539
16 mori-komu means heap into, serve heap-put into JSD p. 395
16 nagashi-komu means pour into pour-put into Matsumoto (1997: 173-4)
16 nage-ageru means throw upto throw-raise Matsumoto (1997: 173-4)
16 nage-ireru means throw into throw-put into Matsumoto (1997: 173-4)
16 nage-kaesu means throw back throw-return Matsumoto (1997: 173-4)
16 nage-komu means throw into throw-put into Matsumoto (1997: 173-4)
16 nori-ageru means beach put-raise beach(10)
16 nui-komu means sew into sew-put into sew(in, into)(7)
16 oi-ageru means chase up chase-raise tree(10)
16 okuri-kaesu means send back send-return Matsumoto (1997: 173-4)
16 okuri-komu means sent into send-put into funnel(3)
16 ori-komu means weave into weave-put into interweave(8)
16 oshi-ageru means push up push-raise Matsumoto (1997: 173-4)
16 oshi-dasu means push out push-take out push(3)
16 oshie-komu means instill into teach-put into pump(into)(7)
16 oshi-ireru means push into push-put into Matsumoto (1997: 173-4)
16 oshi-komeru means push into push-put into squash(3)
16 oshi-komu means cram into push-put into crowd(7), tuck(3), squeeze(3), cram(7), jam(into)(7), wedge(3)
16 oshi-yaru means push aside push-put away push(3)
16 otoshi-ireru means drop into drop-put into drop(4)
16 sashi-komu means insert prick-put into Matsumoto (1997: 173-4)
16 sasoi-komu means invite to invite-put into Matsumoto (1997: 173-4)
16 shibori-dasu means squeeze out squeeze-take out wring(3)
16 shimai-komu means stow in put back-put into stow(1)
16 sosogi-ireru means pour into pour-put into *

460
CLS Scls Japanese V-type English Direct Translation English Levin
16 sui-komu means suck into suck-put into Matsumoto (1997: 173-4)
16 sukui-ageru means scoop up scoop-raise scoop(3), spoon(3)
16 sukui-dasu means scoop out scoop-take out scoop(3)
16 suri-komu means rub into rub-put into rub(on, into)(7), anoint(8)
16 tataki-komu means hammer into hit-put into bang(3), hammer(3)
16 tataki-otosu means hit down hit-drop Matsumoto (1997: 173-4)
16 toji-komu means file in file-put into file(10)
16 tori-ageru means lift take-put into Matsumoto (1997: 173-4)
16 tori-ireru means put into take-put into Matsumoto (1997: 173-4)
16 tori-komu means put into take-put into Matsumoto (1997: 173-4)
16 tori-orosu means take down take-drop Matsumoto (1997: 173-4)
16 tou.ka-suru CHI throw down throw.lower-do Matsumoto (1997: 175)
16 tou.nyuu-suru CHI throw into throw-put into-do Matsumoto (1997: 175)
16 tsugi-komu means pour into pour-put into pour(5)
16 tsumami-ageru means pick up pick-raise Matsumoto (1997: 173-4)
16 tsume-komu means cram into cram-put into stow(1), ram(3), squeeze(3), tuck(3), pack(7), pump(into)(7), stuff(7), pad(8),
jam(into)(7), load(with)(7), crowd(7), cram(7)
16 tsumi-ageru means pile up pile-raise pile(with)(7), stack(with)(7)
16 tsumi-komu means load into load-put into load(7)
16 tsure-komu means bring into bring-put into Matsumoto (1997: 173-4)
16 tsuri-ageru means lift by hanging hang-raise hoist(4)
16 tsuri-ageru means lift by fishing fish-raise land(10)
16 tsutsumi-komu means wrap in wrap-put into roll(6)
16 uchi-ageru means hit up hit-raise Matsumoto (1997: 173-4)
16 uchi-kaesu means hit back hit-return Matsumoto (1997: 173-4)
16 uchi-komu means drive into hit-put into ram(3), hammer(3)
16 uchi-komu means shoot into shoot-put into lodge(1)
16 uchi-otosu means shoot down shoot-drop Kojien p. 231
17 hari-tsukeru means stick into stick-attach stick(on, in)(7), plaster(7)
17 huki-kakeru means spray spray-shower spray(7), squirt(with)(7)
17 huri-kakeru means sprinkle shake-shower shake(3), dust(7), sprinkle(7), flour(9), pepper(9), powder(9), sugar(9), salt(9)

461
CLS Scls Japanese V-type English Direct Translation English Levin
17 huri-maku means sprinkle spray-scatter perfume(9)
17 kazari-tsukeru means decorate to decorate-attach Matsumoto (1997: 173-4)
17 kosuri-tsukeru means rub into rub-apply rub(on, into)(7)
17 maki-tsukeru means wind on/to wind-attach coil(6), curl(6), loop(6), twist(6), wind(6), wreathe(9)
17 mori-tsukeru means dish up heap-apply JSD p. 395
17 motase-kakeru means lean against lean-put on lean(2)
17 nade-tsukeru means plaster stroke-attach plaster(9)
17 nasuri-tsukeru means rub into rub-apply rub(with)(7)
17 nui-tsukeru means sew onto sew-attach sew(on, onto)(7)
17 nuri-tsukeru means dab on smear-apply dab(on)(7), daub(7), smear(7), coat(8)
17 oi-tsumeru means chase into chase-pack tree(10)
17 ooi-kabuseru means cover cover-put over GJD p. 264
17 oshi-nagasu means sweep away push-pour sweep(3)
17 oshi-tsukeru means force on push-attach push(3)
17 tate-kakeru means stand against stand-put on lean(2), stand(2)
17 tsuki-tateru means lodge on stab-stand lodge(1)
17 uchi-tsukeru means pound into hit-attach pound(3)
17 ue-tsukeru means implant plant-attach impregnate(8)
17 yaki-tsukeru means attach by burning burn-attach Matsumoto (1997: 173-4)
18 ooi-kakusu means cover (up) cover-hide wrap(in)(7), shroud(8)
18 ooi-tsutsumu means cover cover-wrap wrap(in)(7)
20 1 hakobi-saru means carry away carry-go away sweep(3)
20 1 hakobu simple carry channel(3)
20 1 katsugu simple carry shoulder(10)
20 1 ninau simple carry shoulder(10)
20 1 okuri-tsukeru dev2 send send impolitely Kojien p. 347
20 1 okuru simple give endow(8)
20 1 ou simple carry shoulder(10), shoulder(10)
20 1 seou simple carry sling(1)
20 1 utsusu simple transfer siphon(3), infect(8)
20 2 ataeru simple give, supply vest?(7), shower(with)(7), bed(10), berth(10), salt(9)

462
CLS Scls Japanese V-type English Direct Translation English Levin
20 2 ategau simple give shelter(10)
20 2 azukeru simple entrust lodge(1), lodge(10), bank(10)
20 2 ho.jyuu-suru CHI replenish supplement.fill-do replenish(8)
20 2 hu.yo-suru CHI grant, give give.give-do vest?(7), endow(8)
20 2 ki.hu-suru CHI contribute give.give-do endow(8)
20 2 ki.sou-suru CHI denote, present give.give-do endow(8)
20 2 kyou.kyuu-suru CHI provide offer.supply-do stock(with)(7)
20 2 kyuu.sui-suru CHI supply with water supply.water-do Matsumoto (1997: 176)
20 2 kyuu.yu-suru CHI supply with oil supply.oil-do Matsumoto (1997: 176)
20 2 sazukeru simple give, grant endow(8)
20 2 tei.kyou-suru CHI offer offer.offer-do billet(10)
20 2 yo.kin-suru CHI deposit entrust.money-do bank(10) also Matsumoto (1997: 176)
20 3 ami-awaseru means combine interweaving weave-combine interweave(8)
20 3 kumi-awaseru means combine put together-combine interlace(8)
20 3 mazeru simple mix interlard(8), drug(9)
20 3 ori-mazeru means mix by interweaving weave-mix interlace(8), interweave(8)
20 3 tsugu simple connect putty(9)
20 3 tsunagu simple connect, tie yoke(9), festoon(9), harness(9), leash(9)
20 4 butsukeru simple hit ram(3)
20 4 sasu simple stab stick(7), fork(10), skewer(10), spit(10), prick?(7)
20 4 tataku simple pat dab(with)(7)
20 4 toosu simple thread string(10)
20 4 tsuki-ateru means ram prick-hit ram(3)
20 4 tsuki-sasu pair stick prick-prick stick(7)
20 4 utsu simple hammer shoe(9)
20 5 nage-suteru means throw away throw-throw away dump(3)
20 5 suteru simple throw away dump(3)
20 6 horu simple dig groove(9)
20 6 huku simple wipe wipe(3)
20 6 kosuru simple rub rub(7), rosin(9)
20 6 migaku simple polish polish(9), rub(with)(7), wax(9), sand(9)
20 6 nuguu simple wipe wipe(3)
20 6 shiboru simple wring squish(3)

463
APPENDIX E. Levin’s classification of verbs of removing
(Levin 1993: 122-132)

10.1 Remove Verbs (dismiss, remove, eliminate)

ex. Doug removed the smudges from the tabletop.

(i) they relate to the removal of an entity from a location

(ii) they express one of their arguments (location) with a prepositional
phrase headed by from
(iii) they can also take from followed by one of a variety of locative
prepositions
ex. Doug removed the smudges from around the sink.
(iv) most of them cannot take the prepositions out of or off of
ex. *Doug removed the smudges out of the drawer.
(v) each of them imposes particular restrictions on the set of possible direct
objects
(vi) they do not allow the locative alternation
ex. a. Doug removed the smudges from the tabletop.
b. *Doug removed the tabletop of smudges.

10.2 Banish Verbs (banish, deport, recall)

ex. The king banished the general from the army.

(i) they relate to the removal of an entity, typically a person, from a

location
(ii) The location argument is expressed in a prepositional phrase headed by
the preposition from.
(iii) they do not allow any of the other prepositions such as out, off, or a
combination of from and another preposition
(iv) they allow to phrases as well as from phrases, though not
simultaneously
ex. *The king banished the general from the palace to a mountain
fortress.
(v) they do not allow the locative alternation

10.3 Clear Verbs (clear, clean, empty)

ex. Doug cleared dishes from the table.

(i) they relate to the removal of a substance from a location but in some of
their uses they are better characterised as verbs of change of state
(ii) they allow the (transitive) locative alternation
ex. a. Doug cleared dishes from the table. (locative variant)
b. Doug cleared the table of dishes. (of variant)
(unlike spray/load verbs, they take the preposition of)
(iii) Most of them (except clean) show the causative/inchoative
alternation
ex. a. The strong winds cleared the skies.
464
 b. The skies cleared.
(iii) they allow the (intransitive) locative alternation
ex. a. The sky cleared (?of clouds).
b. Clouds cleared from the sky.
(iv) the state that each of these verbs lexicalises is a state that can hold of a
“location” as a result of removing something from that location
(v) when they refer to possessional deprivation, they take from in the
locative variant, but when they describe the removal of an entity from a
location, they may take a variety of locative prepositions
(vi) the location argument receives the “holistic” interpretation

10.4 Wipe Verbs

10.4.1 Manner Subclass (scrape, wipe, wash)

ex. Brian wiped the fingerprints from the counter.

Brian wiped the fingerprints from inside/outside/under the cupboard.

(i) they relate to removing things from surfaces or containers

(ii) they participate in the locative alternation and show the
“holistic/partitive” effect but they do not allow an of phrase when the
location is the direct object
ex. a. Brian wiped the fingerprints from the counter.
b. Brian wiped the counter (*of fingerprints).
(iii) they lexicalise a manner or means of removal
(iv) most of them in their basic meanings are probably not verbs of
removing. Since some of the manner or means that are part of the
meanings of these verbs are specifically associated with removing
things from surfaces or containers, these verbs show properties of verbs
of removing.
(v) verbs whose meaning involves some manners or means that may be
associated with putting things on surfaces or in containers as well as
with removing things from surfaces or containers can be used both as
verbs of putting and verbs of removing
(vi) some of them allow the unspecified object alternation
ex. a. Brian was wiping the counter.
b. Brian was wiping.

10.4.2 Instrument Subclass (brush, rake, shovel)

ex. Carla shoveled the snow from the walk.

Carla shoveled the snow from under/near/among the bushes.

(i) they are zero-related to a noun that is the name of an instrument. These
verbs in their most basic meaning probably refer to using the instrument
they take their name from in a conventional way. Since many of these
instruments are used for removing things from surfaces or containers,
they show properties of verbs of removing.
(ii) they relate to removing things from surfaces or containers
(iii) they participate in the locative alternation and show the
“holistic/partitive” effect but they do not allow an of phrase when the

465
 location is the direct object
ex. a. Carla shoveled the snow from the walk.
b. Carla shoveled the walk (*of snow).
When the location is the direct object, it receives the “holistic”
interpretation.
(iii) some of them can be used as verbs of putting
(iv) some of them allow the unspecified object alternation
ex. a. Carla was shoveling the walk.
b. Carla was shoveling.

10.5 Verbs of Possessional Deprivation: Steal Verbs (abduct, pilfer, steal)

ex. The thief stole the painting from the museum.

(i) they primarily describe the removal of something from someone’s

possession; the previous possessor or a location associated with this
possessor is expressed in a from prepositional phrase.
(ii) they can also take benefactive for phrases to indicate the person on
whose behalf the removal was done, but they do not participate in the
benefactive alternation.
ex. a. The thief stole the painting for Mr. Smith.
b. *The thief stole Mr. Smith the painting.
(iii) they are found in a syntactic frame that resembles the location variant of
the locative alternation, but they do not participate in the locative
alternation.
ex. a. The thief stole the painting from the museum.
b. *The thief stole the museum of the painting.
(iv) some of these verbs can also be used as verbs of obtaining because in
many situations in which someone obtains something, someone else
loses possession of that thing.

10.6 Verbs of Possessional Deprivation: Cheat Verbs (cheat, cure, deprive)

ex. The doctor cured Pat of pneumonia.

(i) they typically describe depriving someone/something of an inalienable

possession (in a broad sense).
(ii) they are found in one of the two syntactic frames shown by the form of
the locative alternation (of variant) but it does not participate in the
locative alternation. The direct object is said to receive the “holistic”
interpretation.
ex. a. *The doctor cured pneumonia from Pat.
b. The doctor cured Pat of pneumonia.
(iii) a few of them allow the preposition to to alternate with out of
ex. The swindler cheated Pat of her fortune.
The swindler cheated Pat out of her fortune.

10.7 Pit Verbs (bark, pit, stone)

ex. The cook boned the fish.

466
 (i.e., The cook removed the bones from the fish.)

(i) They all have zero-related nominals. These nominals could be found as
X as in “remove X from (something).” Moreover, they are considered to
be an inalienably possessed part of an animal or plant.
(ii) The direct object of these verbs receives the “holistic” interpretation
associated with the with/of variant of the locative alternation: in the
above sentence, it is interpreted that all of the bones have been removed.
(iii) They do not allow a cognate of phrase.
ex. a. *The cook boned the fish of bones.
b. *The cook boned the fish of its backbone.

10.8 Debone Verbs (debark, debone, degrease)

ex. The cook deboned the fish.

(i.e., The cook removed the bones from the fish.)

(i) they all have related nominals which form the verbs with the prefix de-.
These verbs are paraphrased as “remove X from (something),” where X
is the noun related to the verb.

(ii) the nouns that these verbs are based on might for the most part be
considered to be an inalienably possessed part of an animal or plant (in
a broad sense).

(iii) the direct object of the verbs receives the “holistic” interpretation
associated with the with/of variant of the locative alternation as pit
verbs.

(iv) they do not allow a cognate of phrase.

ex. *The cook deboned the fish of all its bones.

10.9 Mine Verbs (mine, quarry)

ex. The men mined the gold.

(i.e., The men removed the gold from the mine.)

(i) they have zero-related nominals; the nouns name locations that one
typically removes something from.
(ii) a cognate source phrase may be found if it expresses further
information about the source.
ex. The men were able to mine more gold from the abandoned mine.

467
APPENDIX F. List of classes of verbs of removing in Japanese

468
CLS Scls Type Name No FA-type linkings GA-type linkings Goal Other aspect prominent syntactic or semantic relations bet. F
A-linking G-linking FA-linking RA-linking GA-linking s.pattern characteristics and G
1 1 tori-nozoku 'remove' 43 OK depends OK * * * punctual/ (1) general removing verbs, (2) (extended)
(mostly) extended most of them allow FA-linking locational

1 2 hiku 'subtract' 6 OK * * * * * punctual/ (1) mostly, they are verbs of whole-part or

extended subtraction, (2) does not allow extended
G-linking locationl
2 1 dasu 'put out' 4 OK ? * * * OK path pattern punctual (1) are productive as V2 locational
components, (2) spatial transfer
with a specified direction

2 2 nokeru 'put aside' 2 OK ? ?? * * OK path pattern punctual (1) can be V2 components, (2) locational
spatial transfer with a specified
direction
3 katazukeru 'clear' 1 OK OK OK * OK OK path pattern extended (1) manifest the locative locational
alternation
4 haku 'sweep' 17 ?? OK OK * OK * extended (1) denote means/manner of locational
activity of removing impurities, (impurities)
(2) compounded as V1

5 muku 'peel' 22 ?? OK OK (some) * OK * extended (1) refer to activity of whole-part

separation, (2) compounded as
V1
6 hagasu 'peel' 4 OK OK OK * * * extended (1) similar to type 1, but entails locational
a certain manner/means of (attachment)
removing

469
CLS Scls Type Name No FA-type linkings GA-type linkings Goal Other aspect prominent syntactic or semantic relations bet. F
A-linking G-linking FA-linking RA-linking GA-linking s.pattern characteristics and G
7 1 arau 'wash' 5* * * * OK * extended (1) denote activity of cleaning, locational
(2) can be used as V1 compound (impurities)

7 2 kiyomeru 'cleanse' 5* * * * OK * punctual (1) used as change of state locational

(extended) verbs, (2) compounded as V2 (impurities)

8 1 nusumu 'steal' 37 OK OK OK * *? * punctual (1) verbs of possessional (extended)

deprivation possessional
8 2 dak.kan-suru 5 OK * OK * * * punctual (1) verbs of possessional possessional
'recapture' regaining, (2) do not allow G-
linking
9 tsui.hou-suru 'expel' 11 OK OK OK * * OK path pattern punctual (1) include a lot of frozen locational
compounds, (2) denote causing
animate's change of location
(with/without its willingness)

10 ha.mon-suru 8 OK ?? * * * * punctual (1) denote more abstract expel (extended)

'excommunicate' of people from social locational,
organization or social status etc. identifica-tional

11 kai.hou-suru 2 OK * * * * * punctual (1) denote release of somebody extended

'liberate' from an abstract notion locational

12 toku 'relieve' 5* OK * OK * * punctual (1) take RA-linking, (2) denote extended

release from abstract notion possessional

470
CLS Scls Type Name No FA-type linkings GA-type linkings Goal Other aspect prominent syntactic or semantic relations bet. F
A-linking G-linking FA-linking RA-linking GA-linking s.pattern characteristics and G
13 kai.nin-suru 'dismiss' 5* OK OK ?? ?? * punctual (1) most of them are originally identifica-tional
'figure' incorporation type, (2)
denote "dismissal"

14 sen.patsu-suru 'wash 2* * * * OK * intransitive extended (1) has intransitive version, (2) locational
hair' ground incorporation (impurities)
15 hai.sui-suru 'drain 2 OK OK OK * * OK path, *GO- extended (1) figure incorporation locational
water' linking
16 jo.setsu-suru 'remove 23 OK OK OK * OK * *GO-linking extended (1) figure incorporation with locational
snow' alternative linkings
17 nuki-dasu 'pull out' 58 OK ? * * * OK path pattern punctual/ (1) means compound with (extended)
(some) extended specified direction locational
18 arai-otosu 'wash- 39 OK OK OK (some) * * * punctual/ (1) means compound of FA-type varies
remove' extended linking
19 haki-kiyomeru 2* * * * OK * punctual (1) means compound of GA- locational
'sweep-cleanse' type linking (impurities)

471
APPENDIX G. List of verbs of removing in Japanese

The convention follows that of Appendix D for verbs of putting.

(1) The major topics in the list:

[CLS] Class of verbs (of putting)

[Scls] Subclass of verbs (of putting) (or type of events)
[Japanese] Japanese verbs
[Vtype] Type of Verb
[English] English translation equivalent
[Direct Translation] Direct and detailed translation
[English Levin] English Corresponding Verbs in Levin (1993)

The following is added information that is needed in order to read the list.

[CLS] (and [Scls]): class of verbs

As with verbs of putting, I also leave in the list verbs of other classes that are

semantically close to verbs of removing.

Type 20-1—verbs of sending and carrying

Type 20-2—verbs of possessional transfer
Type 20-3—verbs of separating
Type 20-4—verbs of throwing
Type 20-5—verbs of catching
Type 20-6—verbs of mental request

[Vtype]: Type of verb

There is one type of verbs which are found in verbs of removing.

Japansese VN suru—JAP

472
CLS Scls Japanese V-type English Direct Translation English Levin
1 1 bas.sui-suru CHI extract pull out.element-do extract(5), cull(6)
1 1 boku.metsu-suru CHI exterminate hit/knock.destroy-do eradicate(1)
1 1 bun.ri-suru CHI separate separate.separate-do abstract(1), separate(1)
1 1 chuu.shutsu-suru CHI extract, sample pull out.take out-do distill(4.1) also Matsumoto (1997: 175)
1 1 habuku simple omit JSD p. 1084
1 1 hai.jo-suru CHI remove, exclude remove.remove-do clear(3)
1 1 hai.seki-suru CHI reject, boycott push aside.push aside-do ostracize(1)
1 1 haisuru simple remove JSD p. 145
1 1 hanasu simple separate disengage(1)
1 1 harau simple get rid of, remove dust(4.1), whisk(4.1)
1 1 hazusu simple remove, take off disengage(1)
1 1 hedateru simple separate separate(1), sever(1)
1 1 hiki-hanasu frozen separate pull-separate separate(1), sever(1), wean(6)
1 1 hus.shoku-suru CHI wipe away sweep.wipe-do JSD p. 147
1 1 is.sou-suru CHI clear away, root out one.sweep-do extirpate(1), sweep(4.1)
1 1 jo.gai-suru CHI exclude, omit remove.exclude-do omit(1)
1 1 jo.kyo-suru CHI remove, eliminate remove.remove-do eliminate(1), rid(6)
1 1 kai.jo-suru CHI disarm, cancel relieve.remove-do disarm(6)
1 1 kai.metsu-suru CHI annihilate destroy.destroy-do JSD p. 1084
1 1 kaki-otosu frozen omit write-remove omit(1)
1 1 kesu simple erase delete(1), eradicate(1), erase(4.1), soak(4.1)
1 1 kon.zetsu-suru CHI eradicate root.exterminate-do eradicate(1), extirpate(1), uproot(1)
1 1 ku.jo-suru CHI get rid of drive away.remove-do worm(7), delouse(8)
1 1 mas.satsu-suru CHI eliminate erase.kill-do JSD p. 1082
1 1 mas.shou-suru CHI delete erase.erase-do expunge(4.1)
1 1 nozoku simple exclude, remove eliminate(1), leach(4.1), purge(4.1), weed(4.1), cure(6), divest(6), purify(6), hull(7),
pip(7), string(7), tassel(7), weed(7), debug(8), defog(8), degas(8), delouse(8),
1 1 nuku simple remove descale(8)
soak(4.1)
1 1 otosu simple remove soak(4.1)
1 1 saku.jo-suru CHI delete delete.remove-do delete(1), eliminate(1), excise(1), expunge(4.1)
1 1 shou.kyaku-suru CHI eliminate erase.drive away-do JSD p. 1082

473
CLS Scls Japanese V-type English Direct Translation English Levin
1 1 shou.kyo-suru CHI eliminate erase.remove-do JSD p. 1082
1 1 sou.metsu-suru CHI exterminate sweep.destroy-do JSD p. 147
1 1 sou.tou-suru CHI mop up, clear sweep.attack-do JSD p. 146
1 1 suteru simple abandon, throw away dismiss(1), divest(6)
1 1 tek.kyo-suru CHI remove throw away.remove-do JSD p. 146
1 1 tep.pai-suru CHI abolish throw away.remove-do JSD p. 145
1 1 topparau simple remove JSD p. 145
1 1 tori-harau dev1 clear away take-remove remove(1)
1 1 tori-kesu dev1 cancel, revoke take-delete void(6)
1 1 tori-nozoku dev1 remove take-remove dislodge(1), lop(1), remove(1), banish(2), clear(3), distill(4.1), prune(4.1), scrub(4.1),
skim(4.1), smooth(4.1), strain(4.1), strip(4.1), sweep(4.1), weed(4.1), filter(4.2),
shovel(4.2), exorcise(5), cure(6), free(6), purify(6), relieve(6), rid(6), unburden(6),
pit(7), pulp(7), seed(7), stem(7), worm(7), debone(8), debug(8)
1 1 tori-saru pair remove remove-remove remove(1), strip(6), pith(7), dehorn(8)
1 1 toru simple remove, take, steal eradicate(1), dust(4.1), weed(4.1), iron(4.2), rake(4.2), snatch(5), bleed(6),
swindle(6), bone(7), gill(7), gut(7), head(7), hull(7), husk(7), louse(7), rind(7),
shell(7), string(7), stone(7), weed(7), defrost(8)
1 1 zetsu.metsu-suru CHI exterminate all.destroy-do uproot(1)
1 2 genjiru simple reduce substract(1)
1 2 hiku simple subtract draw(1), substract(1)
1 2 kou.jo-suru CHI subtract subtract.remove-do subtract(1)
1 2 sashi-hiku dev1 deduct prick-subtract JSD p. 1081
1 2 sorasu simple avert withdraw(1)
1 2 yori-suguru pair select select-select winnow(4.1)
2 1 dasu simple take out empty(3), void(6)
2 1 kaesu simple return JSD p. 1018
2 1 modosu simple return JSD p. 1018
2 1 orosu simple bring down, unload discharge(1), unburden(6)
2 2 dokeru simple put aside/away JSD p. 145
2 2 nokeru simple put aside/away JSD p. 145

474
CLS Scls Japanese V-type English Direct Translation English Levin
3 katazukeru simple clear, put away clear(3), free(6)
4 kosuru simple rub rub(4.1), scrape(4.1)
4 haku simple sweep sweep(4.1)
4 hataku simple dust, beat whisk(4.1)
4 huku simple wipe rub(4.1), swab(4.1), wipe(4.1), mop(4.2), mop(4.2), sponge(4.2), towel(4.2)
4 nameru simple lick lick(4.1)
4 nuguu simple wipe mop(4.2)
4 tokasu simple comb comb(4.2)
5 bas.sai-suru CHI deforest, cut down cut down.take/pick-do deforest(8)
5 sukuu simple scoop shovel(4.2)
5 chigiru simple tear JSD p. 51
5 eguru simple scoop, gouge JSD p. 52
5 hiru simple winnow winnow(4.1)
5 horu simple dig mine(9)
5 karu simple reap, cut shear(4.2), fleece(6)
5 kezuru simple chip, curtail, sharpen shave(4.1)
5 kiru simple cut strain(4.1), shear(4.2), head(7), poll(7), tail(7)
5 kizamu simple cut JSD p. 51
5 kosu simple filter strain(4.1), filter(4.2)
5 kurinuku simple gouge out JSD p. 52
5 kuru simple gouge JSD p. 52
5 muku simple peel bark(7), husk(7), peel(7), pod(7), rind(7), scale(7), shuck(7), skin(7)
5 ro.ka-suru CHI filter filter.pass-do filter(4.2)
5 sai.kutsu-suru CHI mine take/pick.dig-do mine(9)
5 setsu.dan-suru CHI cut, disconnect cut.sever-do sever(1)
5 shiboru simple squeeze, wring wring(4.1), milk(6), milk(7)
5 sogu simple chip off JSD p. 52
5 soru simple shave shave(4.1)
5 suu simple suck suck(4.1)
5 tatsu simple cut JSD p. 51

475
CLS Scls Japanese V-type English Direct Translation English Levin
6 hagasu simple peel strip(4.1)
6 hagu simple strip, peel, skin, bark divest(6), bark(7), scalp(7), skin(7)
6 kiru simple drain off drain(3)
6 nuku simple pull out bleed(6), beard(7), bone(7), core(7), gut(7), milk(7), degas(8)
7 1 arau simple wash, clean scrub(4.1), hose(4.2), sponge(4.2)
7 1 migaku simple polish buff(4.1), polish(4.1), rub(4.1), scrub(4.1), file(4.2), sandpaper(4.2)
7 1 sei.sou-suru CHI clean cleanse.sweep-do JSD p. 413
7 1 sou.ji-suru CHI clean sweep.remove-do sweep(4.1), clean(3), mop(4.2), vaccuum(4.2)
7 1 togu simple sharpen, polish buff(4.1), file(4.2)
7 2 jou.ka-suru CHI purify, clean up purify.change-do cleanse(6)
7 2 jou.ryuu-suru CHI distill steam.take elements-do distill(4.1)
7 2 jun.ka-suru CHI purify purity.change-do JSD. 1064
7 2 kiyomeru simple purify, cleanse purge(4.1), purge(4.1), cleanse(6)
7 2 tan.sui.ka-suru CHI desalt non-salt-water-change-do desalt(8)
8 1 bos.shuu-suru CHI confiscate confiscate.put away-do confiscate(5), impound(5), seize(5)
8 1 bundoru simple pilfer JSD p. 120
8 1 chaku.huku-suru CHI embezzle wear.dress-do embezzle(5), grab(5)
8 1 das.shu-suru CHI capture steal.take-do GJD p. 1328
8 1 datsu.ryaku-suru CHI plunder steal.snatch-do GJD p. 1330
8 1 gomakasu simple pocket, cheat wangle(5), weasel(5), cheat(6)
8 1 gou.datsu-suru CHI rob, plunder strong.steal-do pirate(5), snatch(5), rob(6)
8 1 haku.datsu-suru CHI deprive, forfeit strip.steal-do divest(6), strip(6)
8 1 hit-takuru frozen snatch pull-tuck in wrench(1), grab(5), nab(5), snatch(5)
8 1 humi-taosu frozen bilk step on-knock down bilk(6)
8 1 hundakuru simple snatch relieve(6)
8 1 hyou.setsu-suru CHI plagiarise snatch.steal-do lift(5), plagiarize(5)
8 1 in.you-suru CHI quote pull.use-do extract(5)
8 1 kadowakasu simple abduct, kidnap abduct(5)
8 1 kapparau simple snatch liberate(5), swipe(5)
8 1 kasumeru simple pilfer JSD p. 120
8 1 kusuneru simple pilfer filch(5), pilfer(5)
8 1 maki-ageru frozen cheat roll-raise pinch(5), bleed(6), cheat(6), fleece(6), milk(6), con(6), gull(6)

476
CLS Scls Japanese V-type English Direct Translation English Levin
8 1 man.biki-suru JAP steal ten thousand.pulling-do JSD p. 611
8 1 mitsu.yu-suru CHI smuggle secret.carry-do smuggle(5)
8 1 nugasu simple strip of clothes divest(6)
8 1 nusumu simple steal, plagiarise abstract(1), cop(5), flog(5), liberate(5), pinch(5), purloin(5), rustle(5), sneak(5),
steal(5), swipe(5), take(5), thieve(5), thieve(5), plunder(6), relieve(6), rob(6), rob(6)
8 1 ou.ryou-suru CHI embezzle wicked.get-do embezzle(5), plunder(6)
8 1 ou.shuu-suru CHI confiscate push.put away-do confiscate(5), impound(5), seize(5)
8 1 ryaku.datsu-suru CHI plunder snatch.steal-do pirate(5), despoil(6), plunder(6), ransack(6)
8 1 sa.shu-suru CHI swindle deceive.take-do defraud(6)
8 1 sarau simple carry off sweep(4.1), kidnap(5)
8 1 sashi-osaeru frozen distrain, seize prick-push seize(5)
8 1 seshimeru simple swindle wangle(5)
8 1 set.tou-suru steal steal.steal-do JSD p. 611
8 1 shik.kei-suru CHI steal make mistake.respect-do filch(5)
8 1 shuu.datsu-suru steal put away.steal-do JSD p. 119
8 1 tori-ageru frozen deprive, confiscate take-raise withdraw(5), wrest(5), disarm(6), dispossess(6)
8 1 tou.you-suru CHI plagiarise steal.use-do lift(5), plagiarise(5), steal(5)
8 1 ubau simple take away bereave(6), denude(6), deprive(6), drain(6), ease(6), rob(6), strip(6)
8 1 yoko.dori-suru JAP embezzle wicked.taking-do JSD p. 120
8 1 yuu.kai-suru CHI kidnap deceive.abduct-do abduct(5), kidnap(5)
8 2 chou.hatsu-suru CHI requisition, commandeer collect.take out-do confiscate(5)
8 2 dak.kai-suru CHI recapture steal.return-do recover(5), rescue(5)
8 2 dak.kan-suru CHI recapture steal.return-do JSD p. 121
8 2 kai.huku-suru CHI regain, recover return.return-do recover(5), regain(5)
8 2 kai.shuu-suru CHI recover, withdraw return.put away-do retrieve(5), withdraw(5)
9 hanasu simple set free, release JSD p. 83
9 oi-dasu frozen drive out chase-take out eject(1), extrude(1), expel(2), flush(4.1), exorcise(5), dispossess(6)
9 oi-harau frozen drive away chase-remove dislodge(1), shoo(1), sweep(4.1), exorcise(5)
9 oi-tateru frozen chase chase-stand evict(1), uproot(1)
9 tsui.hou-suru CHI expel, banish chase.release-do ostracize(1), banish(2), deport(2), purge(6)

477
CLS Scls Japanese V-type English Direct Translation English Levin
9 ke-chirasu frozen drive away kick-scatter JSD p. 145
9 oi-chirasu frozen drive away chase-scatter JSD p. 145
9 oi-otosu frozen drive away chase-drop JSD p. 145
9 opparau simple drive away JSD p. 145
9 shou.kan-suru CHI recall call.return-do recall(2), redeem(5)
9 uchi-harau frozen drive away hit-remove JSD p. 145
10 geki.tai-suru CHI repulse, reject attack.drive back-do dislodge(1)
10 hou.chiku-suru CHI throw out release.drive away-do JSD p. 146
10 jo.mei-suru CHI oust, expel remove.name-do excommunicate(1), expel(2) also Matsumoto (1997: 177)
10 jo.seki-suru CHI strike off, expel remove.membership-do GJD p. 1075
10 ku.chiku-suru CHI expel, drive out drive away.drive away-do JSD p. 147
10 shirizokeru simple oust, remove dismiss(1)
10 ha.mon-suru CHI excommunicate break.collegue-do excommunicate(1)
10 shaku.hou-suru CHI release release.release-do discharge(1), liberate(5), free(6)
11 hou.men-suru CHI release, acquit release.exempt-do absolve(6), acquit(6)
11 kai.hou-suru CHI release, liberate relieve.release-do discharge(1), disengage(1), emancipate(5), liberate(5), acquit(6), disabuse(6),
disencumber(6), exonerate(6), free(6), rid(6)
12 men.jo-suru CHI exempt exempt.remove-do discharge(1), absolve(6), exonerate(6)
12 menzuru simple exempt JSD p. 591
12 sha.men-suru CHI pardon pardon.exempt-do pardon(6)
12 toku simple relieve disengage(1)
12 toku.sha-suru CHI pardon special.pardon-do pardon(6)
13 hi.men-suru CHI dismiss, relieve stop.exempt-do JSD p. 590
13 kai.ko-suru CHI dismiss, discharge relieve.employ-do discharge(1), dismiss(1), remove(2)
13 kai.nin-suru CHI dismiss relieve.duty-do separate(1), recall(2), remove(2)
13 kai.shoku-suru CHI dismiss relieve.job-do JSD p. 590
13 men.shoku-suru CHI dismiss exempt.job-do dismiss(1), expel(2), remove(2)
14 sen.gan-suru CHI wash face wash.face-do JSD p. 412
14 sen.patsu-suru CHI wash hair wash.hair-do JSD p. 412
15 hai.ki-suru CHI exhaust gas, ventilate put away/out.air-do JSD p. 1048
15 hai.sui-suru CHI drain put away/out.water-do Matsumoto (1997: 177)
16 bas.shi-suru CHI remove stitches pull out.thread-do Matsumoto (1997: 177)

478
CLS Scls Japanese V-type English Direct Translation English Levin
16 dak.koku-suru CHI thresh remove.grain-do GJD p. 1328
16 das.shi-suru CHI remove fat remove.fat-do GJD p. 1328
16 das.shoku-suru CHI decolorise remove.color-do Matsumoto (1997: 177)
16 das.shuu-suru CHI deodorise remove.smell-do GJD p. 1328
16 das.sui-suru CHI dehydrate remove.water-do Matsumoto (1997: 177)
16 datsu.en-suru CHI desalt remove.salt-do desalt(8)
16 jo.mou-suru CHI depilate remove.hair-do GJD p. 1079
16 jo.setsu-suru CHI clear of snow remove.snow-do Matsumoto (1997: 177)
16 jo.shitsu-suru CHI dehumidify remove.dampness-do Matsumoto (1997: 177)
16 jo.sou-suru CHI weed remove.weed-do Matsumoto (1997: 177)
16 jo.sou-suru CHI defrost remove.frost-do JSD p. 144
16 sai.seki-suru CHI quarry pick/take.stone-do quarry(9)
17 tori-kaesu means get back take-return regain(5)
17 arai-dasu means wash out, identify wash-take out GJD p. 72
17 arai-nagasu means wash away wash-put away wash(4.1)
17 eguri-dasu means scoop out scoop-take out Sakabikijiten p. 345
17 erabi-dasu means pick out select-take out Matsumoto (1997: 173-4)
17 hai.shutsu-suru CHI discharge put out-take out eject(1), expel(1) also Matsumoto (1997: 175)
17 hajiki-dasu means eject eject-take out eject(1)
17 haki-dasu means spew, disgorge spew-take out disgorge(1), expel(1)
17 hakobi-dasu means carry out carry-take out Matsumoto (1997: 173-4)
17 hakobi-saru means carry away carry-go away sweep(4.1), whisk(4.1), snatch(5)
17 hiki-dasu means draw, withdraw draw-take out draw(1), withdraw(1), extract(5), extort(5), winkle(5), milk(6)
17 hik-komeru means withdraw draw-take in withdraw(1)
17 hineri-dasu means squeeze out squeeze-take out squeeze(4.1)
17 hojikuri-dasu means dig out dig-take out Sakabikijiten p. 345
17 hori-dasu means dig out dig-take out extract(1)
17 hori-okosu means dig up dig-raise plow(4.2)
17 hou.shutsu-suru CHI spout, release release.take out-do Matsumoto (1997: 175)
17 houri-dasu means throw out throw-take out Sakabikijiten p. 345
17 hun.shutsu-suru CHI spout spout.take out-do eject(1)
17 kai-dasu means bail out bail.take out-do Sakabikijiten p. 344

479
CLS Scls Japanese V-type English Direct Translation English Levin
17 kai-modosu means redeem buy-return redeem(5)
17 kaki-dasu means rake out rake-take out Sakabikijiten p. 344
17 kiki-dasu means get out of listen-take out pry(1)
17 kiri-dasu means quarry cut-take out quarry(9), JSD p. 51
17 koshi-dasu means filter out of filter-take out leach(4.1)
17 kumi-dasu means bail out bail-take out draw(1), bail(4.1)
17 mochi-dasu means take out take hold of-take out smuggle(5), sneak(5)
17 mochi-kaeru means take back take hold of-return GJD p. 2165
17 mochi-saru means take away take hold of-go away remove(1), whisk(4.1), take(5)
17 nagashi-dasu means flow out flow-take out drain(3)
17 nagashi-saru means flow away flow-go away scour(4.1)
17 nege-dasu means throw out throw-take out Matsumoto (1997: 173-4)
17 nugui-saru means wipe off wipe-go away wipe(4.1)
17 nuki-dasu means pull out, select from pull out-take out draw(1), distill(4.1)
17 nusumi-dasu means steal out steal-take out Matsumoto (1997: 173-4)
17 obiki-dasu means lure away lure-take out Matsumoto (1997: 173-4)
17 okuri-kaesu means send back send-return Matsumoto (1997: 173-4)
17 oshi-dasu means push out push-take out extrude(1)
17 oshi-nagasu means float down push-pour away sweep(4.1)
17 saguri-dasu means spy out spy-take out draw(1)
17 sasoi-dasu means ask out invite-take out Matsumoto (1997: 173-4)
17 shibori-dasu means squeeze out squeeze-take out extract(1), extrude(1), squeeze(4.1), wring(4.1)
17 shime-dasu means lock out lock-take out Sakabikijiten p. 345
17 sui-ageru means siphon up, pump suck-raise siphon(4.2)
17 sui-dasu means suck out suck-take out Sakabikijiten p. 344
17 tataki-dasu means kick out hit-take out flog(5)
17 tataki-otosu means knock down hit-drop Koujien p. 1587
17 teki.shutsu-suru CHI pick out, extract pick out.take out-do extirpate(1) also Matsumoto (1997: 175)
17 tori-dasu means take out, extract, pick out take-take out abstract(1), draw(1), extract(1), empty(3)
17 tori-modosu means get back take-return evict(1), reclaim(5), recover(5), regain(5), repossess(5), retrieve(5)
17 tsumami-dasu means throw out pick-take out extrude(1)

480
CLS Scls Japanese V-type English Direct Translation English Levin
17 tsumi-dasu means ship out load-take out Sakabikijiten p. 345
17 tsure-dasu means take out take-take out Matsumoto (1997: 173-4)
17 tsure-modosu means take back take-return Matsumoto (1997: 173-4)
17 tsure-saru means take away take-go away remove(1), whisk(4.1)
17 uchi-otosu means shoot down shoot-drop GJD p. 191
17 yobi-dasu means tell to come call-take out Matsumoto (1997: 173-4)
17 yobi-modosu means recall call-return recall(2)
18 aogi-nozoku means remove by winnowing winnow-remove winnow(4.1)
18 arai-otosu means wash off wash-remove rinse(4.1), wash(4.1), sponge(4.2)
18 chigiri-toru means tear off tear-remove Matsumoto (1997: 173-4)
18 damashi-toru means defraud deceive-take extort(5), cheat(6), con(6), defraud(6), fleece(6), mulct(6)
18 eguri-toru means scoop out scoop-remove winkle(1)
18 hagashi-toru means peel off peel-remove scratch(4.1)
18 hagi-toru means strip off strip-remove strip(4.1), denude(6)
18 harai-nokeru means brush off brush-put away banish(2), brush(4.2)
18 harai-otosu means dust off brush-remove dab(4.1), brush(4.2)
18 hiki-nuku means pull out pull-pull out extract(1), uproot(1), pluck(4.1)
18 huki-toru means wipe off wipe-remove swab(4.1), wipe(4.1), mop(4.2)
18 huri-otosu means shake off shake-remove Matsumoto (1997: 173-4)
18 hurui-otosu means shake off brandish-remove winnow(4.1)
18 kaki-harau means rake off rake-remove rake(4.2)
18 kari-toru means cut down, reap reap-remove trim(4.1)
18 kezuri-otosu means shave off file-remove file(4.2)
18 kezuri-toru means shave off file-remove shave(4.1)
18 kiri-nuku means cut out cut-pull out JSD p. 51
18 kiri-otosu means cut off cut-remove lop(1), trim(4.1)
18 kiri-toru means cut out of cut-remove excise(1), prune(4.1), trim(4.1), gill(7), top(7), tail(7), pinion(7)
18 kosuri-otosu means scrape off rub-remove scrape(4.1)
18 kosuri-toru means scrape off rub-remove scour(4.1), scrape(4.1), scratch(4.1), scrub(4.1)
18 mogi-toru means pluck off pluck-remove wrench(1), pluck(4.1), wrest(5)
18 mushiri-toru means pluck off pluck-remove pluck(4.1)
18 nugui-toru means wipe off wipe-remove sponge(4.2)

481
CLS Scls Japanese V-type English Direct Translation English Levin
18 nuki-toru means pull out, extract pull out-take abstract(1)
18 nusumi-toru means rob of rob-take purloin(5)
18 setsu.jo-suru CHI excise cut.remove-do extirpate(1)
18 shibori-toru means squeeze out squeeze-remove squeeze(4.1), wring(4.1), siphon(4.2), bleed(6), milk(6), sap(6)
18 sori-otoru means shave off shave-remove shave(4.1)
18 sui-toru means suck up suck-take suck(4.1), sponge(4.2)
18 sukui-toru means scoop up scoop-take skim(4.1)
18 susugi-otosu means rinse off rinse-remove rinse(4.1)
18 toki-hanasu means release relieve-release JSD p. 83
18 tsukami-toru means snatch grab-take seize(5)
18 tsumi-toru means remove by picking pick-remove pluck(4.1), trim(4.1)
18 ubai-toru means deprive steal-remove seize(5), rifle(6)
18 yori-nuku means select select-pull out cull(1)
18 yusuri-toru means extort from extort-take extort(5), mulct(6)
19 arai-kiyomeru means cleanse by washing wash-cleanse wash(4.1)
19 haki-kiyomeru means cleanse by sweeping sweep-cleanse Kageyama (1993: 105)
20 1 utsusu simple transfer siphon(4.2)
20 2 hiki-watasu dev1 deliver pull-hand extradite(2)
20 2 kaku.toku-suru CHI acquire acquire.get-do cop(5)
20 2 ukeru simple inherit, obtain reap(1)
20 3 aogi-wakeru means separate by winnowing winnow-separate winnow(4.1)
20 3 bun.katsu-suru CHI divide separate.divide-do partition(1)
20 3 bun.pai-suru CHI distribute separate.distribute-do partition(1)
20 3 ku.bun-suru CHI divide partition.divide-dp parition(1)
20 3 shikiru simple partition partition(1), separate(1)
20 3 wakeru simple divide separate(1)
20 4 has.sha-suru CHI discharge, fire emit.shoot-do expel(1)
20 5 hit-tsukamaeru dev1 seize pull-seize nab(5)
20 5 hit-tsukamu dev1 grab pull-grab grab(5), nab(5)
20 5 tai.ho-suru CHI arrest reach.catch-do capture(5), nab(5), pinch(5), seize(5)
20 5 toraeru simple catch capture(5), seize(5)
20 5 tsukamaeru simple catch cop(5)

482
CLS Scls Japanese V-type English Direct Translation English Levin
20 5 tsukamu simple grasp seize(5)
20 6 kyou.you-suru CHI exact strong.request-do extort(5)
20 6 nedaru simple beg cadge(5), cop(5)
20 6 sebiru simple pester cadge(5), sponge(5)
20 6 takaru simple bum off sponge(5)
20 6 you.kyuu-suru CHI demand of/from need.require-do reclaim(5)

483
APPENDIX H. List of classes of verbs of putting and removing in
Japanese

484
CLS Scls Type Name No FA-type linkings GA-type linkings GO-linking Source or Goal Other syntactic
L-linking A-linking G-linking FA-linking I-linking RA-linking GA-linking Source (P) Goal (R) patterns
(for P) (for R) (for R) (P and R) (for P) (for R) (P and R)
Verbs with specified direction
P2 1 ireru 'put into' 5 OK * * * * OK path pattern
R2 1 dasu 'take out' 4 OK ? * * * * OK path pattern
P2 2 ageru 'raise' 3 OK OK * * * OK path pattern
R2 2 nokeru 'put aside' 2 OK ? ?? * * * OK path pattern
Ground incorporation
P13 hou.boku-suru 'graze' 6 OK OK * * * *

R14 sen.patsu-suru 'wash 2 * * * * OK * *

hair'
Figure incorporation 1
P14 san.sui-suru 'scatter 12 OK OK (some) * * OK OK (some) path pattern
water'
R15 hai.sui-suru 2 OK OK OK * * * OK path pattern
'drain water'
Figure incorporation 2
P15 shoku.rin-suru 3 OK * OK OK OK * path pattern
'afforest'
R16 jo.setsu-suru 'remove 23 OK OK OK * OK * *
snow'

485
CLS Scls Type No FA-type linkings GA-type linkings GO-linking Source or Goal Other syntactic
L-linking A-linking G-linking FA-linking I-linking RA-linking GA-linking Source (P) Goal (R) patterns
(for P) (for R) (for R) (P and R) (for P) (for R) (P and R)
Compounds 1
P16 hame-komu 'put 80 OK * * * * OK (some) path pattern
in/on-put.into'
R17 nuki-dasu 'pull 58 OK ? * * * * OK (some) path pattern
out'
Compounds 2
P17 nuri-tsukeru 'smear- 22 OK * * * * *
apply'
R18 arai-otosu 39 OK OK OK (some) * * * *
'wash-remove'
Compounds 3
P18 ooi-kakusu 'cover- 2* * OK OK * *
hide'
R19 haki-kiyomeru 'sweep- 2 * * * * OK * *
cleanse'

486
CLS Scls Type No FA-type linkings GA-type linkings GO-linking Source or Goal Other syntactic
L-linking A-linking G-linking FA-linking I-linking RA-linking GA-linking Source (P) Goal (R) patterns
(for P) (for R) (for R) (P and R) (for P) (for R) (P and R)
FA-type linkings
P1 oku 'put' 15 OK * * * * *
P3 hitasu 'soak' 18 OK * * * * *
P4 sosogu 'pour into' 9 OK * * * * ?? (some path pattern
ok)
P5 kabuseru 'put on' 38 OK * * * * *
P6 1 tsukeru 'apply' 4 OK * * * * *
P6 2 maku 'scatter' 14 OK OK * * * OK path (mostly)
(mostly) (mostly)
P7 1 tsurusu 'hang' 6 OK OK * * * OK
P7 2 tateru 'stand' 2 OK OK * * * *
P8 tsumu 'load' 3 OK * * * * * GS with te-iru,
??path
R1 1 tori-nozoku 'remove' 43 OK depends OK * * * *
(mostly)
R1 2 hiku 'subtract' 6 OK * * * * * *
R6 hagasu 'peel' 4 OK OK OK * * * *
R8 1 nusumu 'steal' 37 OK OK OK * *? * *
R8 2 dak.kan-suru 5 OK * OK * * * *
'recapture'
R9 tsui.hou-suru 'expel' 11 OK OK OK * * * OK path pattern

R10 ha.mon-suru 8 OK ?? * * * * *
'excommunicate'
R11 kai.hou-suru 'liberate' 2 OK * * * * * *

R13 kai.nin-suru 'dismiss' 5 * OK OK ?? ?? * *

487
CLS Scls Type No FA-type linkings GA-type linkings GO-linking Source or Goal Other syntactic
L-linking A-linking G-linking FA-linking I-linking RA-linking GA-linking Source (P) Goal (R) patterns
(for P) (for R) (for R) (P and R) (for P) (for R) (P and R)
GA-type linkings
P9 oou 'cover' 9* * OK OK (some) * * IS-pattern
P10 yogosu 'dirty' 15 * * OK OK * *
R7 1 arau 'wash' 5 * * * * OK * *
R7 2 kiyomeru 'cleanse' 5 * * * * OK * *

FA-type linkings and GA-type linkings

P11 kazaru 'decorate' 6 OK * OK OK (some) * *
P12 tsutsumu 'wrap' 4 OK * OK (RI) * * *
R3 katazukeru 'clear' 1 OK OK OK * OK * OK path pattern
R4 haku 'sweep' 17 ?? OK OK * OK * *
R5 muku 'peel' 22 ?? OK OK (some) * OK * *
R12 toku 'relieve' 5 * OK * OK * * *

488