The Meaning of Space in Sign Language

Sign Languages
and Deaf Communities 4

Editors
Annika Herrmann
Markus Steinbach
Ulrike Zeshan

Editorial board
Carlo Geraci
Rachel McKee
Victoria Nyst
Marianne Rossi Stumpf
Felix Sze
Sandra Wood

De Gruyter Mouton · Ishara Press

The Meaning of Space
in Sign Language
Reference, Specificity and Structure
in Catalan Sign Language Discourse

By
Gemma Barberà Altimira

De Gruyter Mouton · Ishara Press

ISBN 978-1-61451-866-2
e-ISBN (PDF) 978-1-61451-881-5
e-ISBN (EPUB) 978-1-5015-0055-8
ISSN 2192-516X
e-ISSN 2192-5178

Library of Congress Cataloging-in-Publication Data

A CIP catalog record for this book has been applied for at the Library of Congress.

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available on the Internet at http://dnb.dnb.de.

” 2015 Walter de Gruyter Inc., Boston/Berlin and Ishara Press, Preston, UK

Printing and binding: CPI books GmbH, Leck
 Printed on acid-free paper
Printed in Germany
www.degruyter.com
Acknowledgements

Cloc els ulls i sé

que no hi ha res més bonic,
quan aquest gest és sincer,
que dar, cada ú, un xic.
(Eulàlia Ribera Llonc, 2011)

This book is a stimulating project that goes beyond what it is contained within
the following pages. There have been inspiring people and appealing situa-
tions that have influenced the path I have taken and shaped this adventure.
They are also in part responsible that I have come so far as to transform an
almost naive intuition into a scientific work. I would like to take the opportu-
nity to acknowledge them now. Josep Quer, my mentor, has been on this path
since the very beginning. I would like to express my most sincere gratitude
for his guidance, constant support and all the opportunities he has all the way
provided. As I have previously said, this is only (part of) the beginning. Un
milió de gràcies!!
I am most grateful to the Catalan Sign Language deaf signers who contrib-
uted to the creation of the small corpus with their natural signing. Special
thanks go to Josep Boronat, Albert R. Casellas, Pepita Cedillo, Encarna
Muñoz, Frank Vidiella, Santiago Frigola, and Delfina Aliaga. ¡Muchísimas
gracias por todo, amigos! Also, I am much indebted to the Catalan Deaf
Community to welcoming me to their visual world, sharing with me their
wonderful language and instructing me how to express irony in sign language,
make fun and argue (well, maybe not that much on the arguing side...).
Short trips within this journey drove me to meet very interesting people
here and elsewhere who contributed in some way or another to the shaping
of the ideas herein. I am very grateful to have travelled on the same journey
as Celia Alba, Delfina Aliaga, Stefan Bott, Elena Castroviejo, Brendan
Costello, Onno Crasborn, Kathryn Davidson, Javi Fernández, Santi Frigola,
Berit Gehrke, Carlo Geraci, Annika Herrmann, Vadim Kimmelman, Els van
der Kooij, Diane Lillo-Martin, Guillem Massó, Laia Mayol, Marta Mosella,
Louise McNally, Ellen Ormel, Roland Pfau, Lali Ribera, Joana Rosselló,
Markus Steinbach, Alexandra Spalek, Henriette de Swart, Enric Vallduví,
Saúl Villameriel, Martine Zwets, and Inge Zwitserlood. I would also like
to thank my students throughout these years for having taught me so much
about teaching.
Last but not least, my beloved family deserves an exceptional mention.
My parents, Jaume and Teresa, as well as Cesc, my fellow traveller, have
supported me on every single step. Moltíssimes gràcies per ser-hi sempre.
Contents

Acknowledgements ..................................................................................... v

List of figures ............................................................................................xiii

List of tables............................................................................................... xv

Abbreviations ...........................................................................................xvi

Notational conventions ..........................................................................xviii

1. Introduction ............................................................................................ 1
1.1. Objectives and goals ........................................................................ 1
1.2. Sign language research .................................................................... 2
1.3. Catalan Sign Language .................................................................... 3
1.4. Methodology .................................................................................... 4
1.4.1. Sign language corpora .......................................................... 4
1.4.2. Small-scale LSC corpus ........................................................ 6
1.4.3. Annotation conventions ........................................................ 8
1.5. Organisation of this book ............................................................... 10
2. Space in sign languages: background ................................................ 13
2.1. Introduction .................................................................................... 13
2.2. Signing space ................................................................................. 14
2.3. Modality effects ............................................................................. 18
2.3.1. Modality and space ............................................................. 19
2.3.2. Modality and gesture .......................................................... 21
2.4. Spatial functions ............................................................................ 22
2.4.1. Descriptive localisation....................................................... 23
2.4.2. Non-descriptive localisation ............................................... 25
2.4.3. One function or two? .......................................................... 26
viii Contents

2.5. Previous accounts .......................................................................... 28

2.5.1. Spatial mapping view.......................................................... 29
2.5.1.1. Locus and real space ............................................. 29
2.5.1.2. Limitations ............................................................ 31
2.5.2. R-locus view ....................................................................... 35
2.5.2.1. Referential locations and linguistic space ............. 36
2.5.2.2. Advantages ............................................................ 37
2.6. The pointing hodgepodge .............................................................. 38
2.6.1. The morphosyntax of pointing signs................................... 38
2.6.2. The semantics of pointing signs.......................................... 39
2.6.3. Sign language pronouns ...................................................... 41
2.7. Acquisition and emergence of new sign languages ....................... 44
2.8. Proposal ......................................................................................... 46
2.9. Summary ........................................................................................ 47
3. A morpheme on spatial planes ............................................................ 49
3.1. Introduction .................................................................................... 49
3.2. The spatial morpheme .................................................................... 50
3.3. Localisation mechanisms ............................................................... 52
3.3.1. Index signs .......................................................................... 55
3.3.2. Spatial modification ............................................................ 57
3.3.3. Verb agreement ................................................................... 57
3.3.4. Nonmanual mechanisms ..................................................... 59
3.4. Non-descriptive use of spatial planes ............................................ 60
3.4.1. Horizontal ........................................................................... 62
3.4.1.1. Kinds of spatial entities......................................... 64
3.4.1.2. Contrastive topics.................................................. 68
 Contents ix

3.4.2. Frontal ................................................................................. 69

3.4.2.1. Hierarchical relations ............................................ 70
3.4.2.2. Locatives ............................................................... 72
3.4.2.3. Specificity ............................................................. 74
3.4.2.4. Absence in the physical context ............................ 78
3.4.3. Midsaggital ......................................................................... 79
3.5. Features on spatial planes .............................................................. 80
3.6. Body-anchored locations ............................................................... 83
3.7. Summary ........................................................................................ 86
4. Spatial locations and discourse referents ........................................... 87
4.1. Introduction .................................................................................... 87
4.2. Dynamic semantics ........................................................................ 87
4.2.1. Discourse and discourse model........................................... 88
4.2.2. Discourse representation theories ....................................... 89
4.2.2.1. Donkey anaphora in DRT ..................................... 93
4.2.2.2. Accessibility .......................................................... 94
4.2.3. Discourse referents ............................................................. 95
4.2.3.1. S-Topic .................................................................. 98
4.2.3.2. Referential status ................................................. 100
4.2.4. Desiderata for a DRT application to sign language .......... 101
4.3. Locations and discourse referents ................................................ 102
4.3.1. Locationsas variables ........................................................ 103
4.3.2. Identity features ................................................................ 104
4.4. Scope of discourse referents’ quantifiers ..................................... 106
4.4.1. (Non-)argumental NPs ...................................................... 109
4.4.2. Donkey sentences ............................................................. 112
x Contents

4.4.3. Distributivity and quantification ....................................... 114

4.4.4. Genericity.......................................................................... 118
4.4.5. Kinds ................................................................................. 119
4.5. Summary ...................................................................................... 121
5. Deixis and familiarity ........................................................................ 123
5.1. Introduction .................................................................................. 123
5.2. Definiteness: background ............................................................. 124
5.2.1. Uniqueness ........................................................................ 124
5.2.2. Familiarity ......................................................................... 125
5.2.2.1. Weak/strong familiarity....................................... 127
5.2.2.2. Discourse/addressee familiarity .......................... 129
5.2.3. Deixis ................................................................................ 131
5.2.4. Definiteness in sign language ........................................... 134
5.3. Anaphoric deixis in LSC .............................................................. 136
5.4. Familiarity .................................................................................... 141
5.4.1. Assertion of existence ....................................................... 142
5.4.2. Presupposition of existence .............................................. 144
5.5. Indefiniteness marking ................................................................. 146
5.5.1. Indefinite determiners ....................................................... 146
5.5.2. Nonmanual indefiniteness marking .................................. 146
5.6. Summary ...................................................................................... 147
6. Specificity ............................................................................................ 149
6.1. Introduction .................................................................................. 149
6.2. Specificity: background ............................................................... 150
6.2.1. Scope ................................................................................. 151
6.2.2. Partitivity .......................................................................... 153
6.2.3. Identifiability ..................................................................... 154
6.2.4. Specificity in sign language .............................................. 156
 Contents xi

6.3. Specificity in LSC ........................................................................ 157

6.3.1. Scope ................................................................................. 159
6.3.2. Partitivity .......................................................................... 162
6.3.3. Identifiability ..................................................................... 164
6.3.4. Narrow scope marking (or what this chapter is not about)168
6.4. Localisation pattern...................................................................... 172
6.4.1. Compositional analysis of the data ................................... 173
6.4.1.1. Direction towards the frontal plane..................... 173
6.4.1.2. Amount of mechanisms....................................... 175
6.4.1.3. Eye gaze duration ................................................ 176
6.4.1.4. Simultaneity and coincidence in direction .......... 177
6.4.2. Spatially modified categories ............................................ 179
6.4.3. Dual nature of localisation ................................................ 183
6.4.3.1. Non-specific partitives ........................................ 183
6.4.3.2. Modal subordination ........................................... 186
6.5. Existence in the model ................................................................. 190
6.6. Summary ...................................................................................... 192
7. Discourse structure and prominence................................................ 195
7.1. Introduction .................................................................................. 195
7.2. Background .................................................................................. 196
7.2.1. Discourse anaphora and underspecification ...................... 196
7.2.2. Prominence ....................................................................... 198
7.2.3. Noteworthiness ................................................................. 201
7.3. Sign language pronominal issues ................................................. 203
7.3.1. Infinity and unambiguity ................................................... 203
7.3.2. Pronouns versus DRs ........................................................ 204
7.4. Prominence .................................................................................. 209
7.4.1. Global discourse structure ................................................ 209
xii Contents

7.4.2. Topicality .......................................................................... 210

7.4.3. Topical variables ............................................................... 211
7.5. Underspecification ....................................................................... 216
7.5.1. Informativity ..................................................................... 218
7.5.2. Rigidity ............................................................................. 221
7.6. Summary ...................................................................................... 224
8. Final remarks ..................................................................................... 225

Appendix .................................................................................................. 230

Notes ......................................................................................................... 231

References ................................................................................................ 238

Index ......................................................................................................... 266

List of figures
Figure 1 ELAN screenshot 8
Figure 2 ELAN linguistic tiers annotated 8
Figure 3 Signing space 14
Figure 4 LSC minimal pairs distinguished
the location parameter 15
Figure 5 Regular agreement verb in LSC 16
Figure 6 First and futher mentions of a referent localised in space 17
Figure 7 Descriptive localisation 22
Figure 8 First and further mention of a localised
discourse referent 23
Figure 9 Book on the table, represented with an entity classifier 24
Figure 10 Spatial planes 26
Figure 11 Mixed position of the members of two deaf clubs 34
Figure 12 First and further mention of a localised
discourse referent 53
Figure 13 Localisation mechanisms 55
Figure 14 Index sign 55
Figure 15 PERSON sign 56
Figure 16 Spatial modification of signs 57
Figure 17 Agreement verbs 58
Figure 18 Nonmanual mechanisms of localisation 60
Figure 19 Spatial planes 61
Figure 20 Horizontal plane 62
Figure 21 Reference to present objects 64
Figure 22 Index signs directed to non-present entities 65
Figure 23 Coreferential index referring to a non-entity in LSC 67
Figure 24 Discourse referents with a semantic affinity 69
Figure 25 Frontal plane 70
Figure 26 Expression of hierarchical relations 71
Figure 27 Locative signs directed towards the upper frontal plane 73
Figure 28 Non-iconic convention 74
xiv  List of figures

Figure 29 Specificity marking on the frontal plane 75

Figure 30 Midsaggital plane 79
Figure 31 Body-anchoredlocation 84
Figure 32 Citation form vs. localisation 85
Figure 33 Localisation of discourse referents 105
Figure 34 Scope representation in DRT 108
Figure 35 Donkey sentence in LSC 113
Figure 36 Sign to denote genericity 119
Figure 37 Definiteness marking on the frontal plane in
ASL and HKSL 135
Figure 38 Mixed position of the members of two deaf clubs 137
Figure 39 (In)definiteness marking in LSC signing space 145
Figure 40 Indefiniteness nonmanual marking 147
Figure 41 Upper and lower features of frontal plane 158
Figure 42 Definiteness and specificity marking on
LSC frontal plane 159
Figure 43 Localisation of a wide scope DR 160
Figure 44 Localisation of a narrow scope DR 161
Figure 45 Quantified NP for a specific DR 163
Figure 46 Quantified NP for a non-specific DR 164
Figure 47 Localisation of an identifiable DR 165
Figure 48 Introduction of a non-identifiable DR 167
Figure 49 Strong localisation and use of lower frontal plane 174
Figure 50 Weak localisation and use of upper frontal plane 174
Figure 51 Partitive construction with specific determiner 185
Figure 52 Partitive construction with non-specific determiner 185
Figure 53 Resumptive pronoun within an intensional context 187
Figure 54 Darting eye gaze 189
Figure 55 Contrastive locations 219
Figure 56 Same location for two DRs 220
Figure 57 Two lateral directions for the same DR 222
Figure 58 Two lateral directions for the same DR II 223
List of tables
Table 1 Distribution: types of data and signers 8
Table 2 Terminology and treatments of s-topic 99
Table 3 Equivalence of information-status w.r.t. definiteness 130
Table 4 Indefinites localised on the lower/upper frontal plane 175
Table 5 Number of mechanisms used simultaneously
in indefinites and the corresponding interpretation 176
Table 6 Eye gaze duration 177
Table 7 Instances of simultaneous co-occurrence with
manual and nonmanual component 177
Table 8 Features of pronominal person distinction 206
Table 9 English sentence with co-speech gesture 207
Table 10 English sentence 208
Table 11 LSC sentence 208
Abbreviations

Sign language acronyms

ABSL Al-Sayyid Bedouin Sign Language

AdaSL Adamorobe Sign Language (Ghana)
ASL American Sign Language
Auslan Australian Sign Language
BSL British Sign Language
DSL Danish Sign Language
DGS Deutsche Gebärdensprache, German Sign Language
HKSL Hong Kong Sign Language
HZJ Hrvatski znakovni jezik, Croatian Sign Language
IPSL Indo-Pakistani Sign Language
ISL Irish Sign Language
ISL Israeli Sign Language
ISN Nicaraguan Sign Language
KK Kata Kolok (Bali)
LIS Lingua dei Segni Italiana, Italian Sign Language
LSC Llengua de signes catalana, Catalan Sign Language
LSE Lengua de signos española, Spanish Sign Language
LSF Langue des Signes Française, French Sign Language
LSFB Langue des signes de la Belgique Francophone,
Southern Belgium Sign Language
LSQ Langue des Signes Québécoise, Quebec Sign Language
NGT Nederlandse Gebarentaal, Sign Language of the Netherlands
RSL Russian Sign Language
 Abbreviations xvii

SSL Swedish Sign Language

TSL Taiwanese Sign Language

Other abbreviations

DR Discourse referent
DRT Discourse Representation Theory
DRS Discourse Representation Structure
NP Noun Phrase
SL Sign language
Notational conventions

General

table Lexical sign

person+++ Reduplication of sign
a-n-n-a Finger spelled word
yes/true Two words needed to denote the precise meaning
ix1 First person pronoun
ix2 Second person pronoun
ix3 Third (person) pronoun / determiner
ix3pl Plural third (person) pronoun / determiner
ix3i Coreferential index
2-ix2 Second person pronoun with number incorporation
(“the two of us”)
1-give-2 Agreement verb inflected for first and second person
(“I explain to you”)
cle Entity classifier
clh Handling classifier
cle.1-duck-walks Entity classifier with 1 handshape and denoting a duck
walking
clh.take-book Handling classifier denoting a book that is taken
disappear(h) Final hold of the sign

Location and direction in signing space

ix3ip Sign localised towards the ipsilateral part

ix3cl Sign localised towards the contralateral part
ix3l Sign localised towards the lower part
ix3u Sign localised towards the upper part
 Notational conventions xix

Possible combinations

ip-l Localisation towards the ipsilateral and lower part

ip-u Localisation towards the ipsilateral and upper part
cl-l Localisation towards the contralateral-lower part
cl-u Localisation towards the contralateral-upper part

Nonmanual marking

eg:cl
friend Eye gaze with scope over the sign, directed to the
contralateral part
bl:ip
friend Body lean directed to the ipsilateral part
br
friend Raised eyebrows with scope over the sign
br
friend Furrowed eyebrows with scope over the sign
rs
ix1 walk Role shift scope
Chapter 1
Introduction

What about spoken language? I am still learning it, very slowly. I do not in-
tegrate it as naturally as LSC. Spoken language messages are opaque to me,
shady. I am lucky to be very fast acquiring a clean, transparent and crystal-
clear language to express my feelings, to access knowledge, to discover the
world around through conversations with deaf colleagues, to live and defi-
nitely to have a place in society.
Háblame a los ojos, Pepita Cedillo (2004: 43)

1.1. Objectives and goals

When we look at a signed conversation for the first time the most striking
difference that we may find is that sign languages (SLs) use space for the
representation of meaning. While spoken languages use the audio-vocal
modality, SLs use the visual-spatial modality. As a consequence of this
modality, signing space, which is the three dimensional space in front of
the signer’s body, is thoroughly used. Linguistic expressions in SLsrely on
signing space and the different components of the grammar show dependence
on it. In fact, signing space plays a role at the phonological, morphosyntactic
and discourse level of all SLs studied to date. However, the interpretation
of the use of signing space is not free of controversy and there are opposing
views considering the status of locations with respect to signs that use space,
such as pronouns, agreement verbs and classifiers. For instance, it is not clear
how spatial locations are associated with meaning, or whether they belong to
the grammar of the language or rather to the gesture domain. This book aims
at clearly developing a description and analysis of how spatial locations are
integrated in the discourse grammar of Catalan Sign Language (henceforth,
LSC) concerning the dynamic nature of discourse and taking into account
dynamic semantic theories.
The main goals of this book are three-fold:

G1. To show that spatial locations are integrated into the grammar of
LSC and, even more, that they denote specificity. The incorporation of
spatial locations into the grammar of SLs is a controversial issue. The present
book shows how they are associated with meaning and the role they play in
specificity marking.
2 Introduction

G2. To analyse how spatial locations are set, given the dynamic nature
of discourse. The establishment of spatial locations has been mainly studied
within the scope of clauses, but their discourse behaviour has not been
considered. Using a small-scale LSC corpus, it is shown that spatial locations
consist in abstract points established in space independently of the direction
towards spatial planes manual signs may take, which are categorically inter-
preted within the linguistic system.

G3. To apply a dynamic semantic theory, such as classical Discourse

Representation Theory (Kamp and Reyle 1993), to a visual-spatial
language like LSC. Due to its face-to-face interaction, LSC uses signing
space as well as deictic pronominal elements in the development of discourse.
Here deictic uses and signing space are incorporated to the semantic
representation.

1.2. Sign language research

About 60 years ago it was proven that natural SLs are not mere panto-
mime and that they are provided with grammatical structure (Stokoe 1965;
Teervort 1953). Although the depth of knowledge is not comparable to
that of spoken languages, since the 1950’s research in SLs has advanced
and reached different levels of linguistic analysis (see Brentari 2010;
Pfau, Steinbach and Woll 2012, and Sandler and Lillo-Martin 2006 for an
overview). The areas which have received more attention are phonology,
including phonetics and prosody (Brentari 1998; Crasborn 2001; van der
Kooij 2002; Liddell and Johnson 1989; Sandler 1989 among others), as well
as morphosyntax, studied from a theoretical point of view (Aarons 1994;
Aronoff et al. 2004; Bahan 1996; Benedicto and Brentari 2004; Branchini
and Donati 2009; Cecchetto, Geraci, and Zucchi 2007; Fischer 1975;
Friedman 1976; Kegl 1986; Liddell 1990; Lillo-Martin 1986; Meir 2002;
Neidle et al. 2000; Padden 1988; Pfau 2002; Pizzuto et al. 1990; Rathmann
and Mathur 2008; Quer 2004; Schembri 2003; Suppalla 1986; Steinbach and
Pfau 2007; Wilbur 1997; Zeshan 2004; Zwitserlood 2003, only to indicate a
very short representative list of references).
Discourse analysis is an area where research has started to reach a basic
level of understanding (Baker 1977; Coates and Sutton-Spence 2001;
Metzger 1995; Metzger and Bahan 2001; Meurant 2004, 2006, 2007, 2008;
Morgan 1996, 1999; Nilsson 2004, 2007; Wilson 1996; Winston 1995,
among others), and more research is still needed.
 Catalan Sign Language 3

However, formal semantics and pragmatics is the interface that is still at a

very incipient state (however, see Caponigro and Davidson 2011; Cecchetto
and Zucchi 2006; Davidson 2012, 2013, 2014; Kuhn 2014; Quer 2005a,
2005b, 2011a, 2011b, 2012; Schlenker 2011a, 2011b, 2011c; Schlenker and
Lamberton 2012; Schlenker, Lamberton and Santoro 2013; Wilbur 2011;
Zucchi 2004, 2011, 2012). Hence, this book aims at contributing to the
analysis of SL, and more specifically to LSC discourse from a semantic-
pragmatic perspective.

1.3. Catalan Sign Language

Catalan Sign Language (LSC, llengua de signes catalana)1 is the natural

language used by the signing deaf and deaf-blind community in Catalonia,
the northeastern autonomy in Spain. Together with Spanish Sign Language
(LSE), they are the two sign languages recognised in Spain. Interestingly,
although in the hearing community the bilingualism between spoken Spanish
and spoken Catalan is present, no bilingualism between LSC and LSE exists
in Catalonia. Signers living in Catalonia use only LSC, apart from spoken/
written Catalan and Spanish. Hence the two sign languages do not co-exist
in Catalonia.
Although there is no official count of deaf people, the Catalan Federation
for the Deaf estimates that there are around 25,000 signers around the
Catalan territory, including both deaf and hearing people. Both LSE and LSC
were legally recognised by a law that was passed in 2007 in the Spanish
Parliament (Ley 27/2007, October 23rd 2007). The Catalan Autonomy Law
of 2006 already includes the right to use LSC and on May 26th 2010 a bill
was approved by the Catalan Parliament to regulate the LSC use in the
areas of public life (Llei 17/2010, June 3rd 2010) (see Quer, Mazzoni, and
Sapountzaki 2010).
Teaching and learning materials about LSC are still limited, but a
small amount of materials is available to study and practice the language
(Codorniu, Segimon, and Fernández-Viader 2005; Domad 2002a, 2002b;
García and Codorniu 2007; Illescat 2002a, 2002b; Segimon and Fernández-
Viader 2000). Some dictionaries have also been designed (Ferrerons 2011;
Illescat 2004; Martín and Alvarado 2004; Perelló and Frigola 1998), as well
as a basic descriptive grammar with its corresponding exercises (Quer et
al. 2005; Frigola et al. 2011).
LSC does not have a standardised variety yet, although an indirect plan-
ning is taking place with the use of Webvisual,2 an important media on
4 Introduction

internet, and with LSC being taught at different institutions.3 The dialect
from the capital, Barcelona, is the most standardised one, since it is where
the Federation for the Deaf is located and also where all the largest popu-
lation of deaf people lives, with the corresponding deaf schools and deaf
clubs.
Although LSC is still an understudied language, some published works
and master thesis are already available in different linguistic areas, namely
phonology (Bosch i Baliarda 2005; Massone, Bosch i Baliarda, and Fernández-
Viader 2003), morphosyntax, including lexicalization (Jarque et al. 2012),
word order (Jarque et al. 2007), agreement (Morales et al. 2005; Quadros
and Quer 2008; Quer 2009, 2011b), negation (Pfau and Quer 2007; Quer and
Boldú 2006), possessives (Quer and GRIN 2008), classifier constructions
(Barberà and Quer, in press; Benedicto, Cvejanov, and Quer 2007; Benedicto,
Cvejanov, and Quer 2008; Jarque 2011), Wh-questions (Alba 2010); relative
clauses (Mosella 2012); metaphor (Jarque 2005), role shift (Frigola and Quer
2006; Quer 2005b, 2011a); discourse cohesion (Barberà 2007); machine
translation (Massó and Badia 2010); lexical access in production (Baus et al.
2008); lexicography (Barberà and Ribera 2010; Ribera 2007), and sociolin-
guistics and language planning (Gras 2006; Jarque 2012; Morales-López et
al. 2002, Quer 2010b), among others.

1.4. Methodology

1.4.1. Sign language corpora

The main aim of this book is to analyse what it takes to have a spatial loca-
tion established in LSC discourse with respect to the referential domain. The
referential meaning of nouns in natural languages, and in LSC in partic-
ular, is an intricate topic that is extremely hard to investigate on the basis of
elicitation only. In the end, it is connected discourse in the language under
investigation that provides the most important clues for analysis of these
grammatical domains (Dimmendaal 2001). In this section the peculiarities of
sign language corpora and the methodology for the annotation of the small-
scale LSC corpus are presented.
A corpus is a representative collection of language samples in a machine-
readable form that can be used to study the type and frequency of linguistic
units (McEnery and Wilson 2001). For SLs, corpora are collections of video
that are annotated, i.e. they contain written material that is added to and
time-aligned with the primary sign language digital video data (Schembri
 Methodology 5

and Crasborn 2010). It represents a description and it is very useful for

analysis of the data. The beginnings of standardised annotation conventions
for SL corpora can be found in the ECHO project (Crasborn et al. 2007).
This project consisted in the recording of some Aesop’s fables in three sign
languages, namely Sign Language of the Netherlands (NGT), British Sign
Language (BSL) and Swedish Sign Language (SSL).4 The three university
groups responsible for the project had different research interests, but a basic
annotation layer was established which has been used for the annotation of
LSC data.
SL corpora have the main advantage of improving peer review of descrip-
tions of SLs and make possible a corpus-based approach to SL analysis.
Corpora are important for hypotheses testing at all grammatical components
(from phonology, morphology, lexis, syntax and discourse). Importantly,
when documenting SLs there are different reasons why corpora are neces-
sary tools. First, SLs are young languages of minority communities and lack
written forms and developed standards of correctness that often accompany
literacy (Johnston 2010). Second, they have discontinuous transmission and
also have few native speakers (see Costello, Fernández, and Landa 2008).
And third, the traditional annotation of SL examples using glosses remains
often inaccessible for some researchers. Within the last ten years, more
corpus projects have been set up for different SLs. Among the most known
ones, there are the NGT corpus,5 the BSL corpus,6 the AUSLAN (Australian
SL) corpus,7 the DGS (German SL) Corpus,8 and the ISL (Irish SL) Corpus.9
Different softwares allowing precise time-alignment of annotations
with the corresponding video sources on multiple user-specifiable tiers
have been designed. Among the most used ones are ELAN,10 Sign Stream
(Neidle 2007),11 Anvil,12 and Colin.13 ELAN is a digital video annotation
software developed at the Max Planck Institute and it is the software mostly
used for SL annotation. One of the main advantages is that it has customis-
able, study-specific tiers that can always be added at any time of the annota-
tion process.
Data gathering is an extremely valuable part of a fieldworker’s repertoire
since it is by collecting spontaneous or semi-spontaneous data that linguists
are exposed to phenomena that are outside the boundaries of their imagina-
tion. A body of textual material enables to study how discourse is connected,
for example. However, corpora work entails two drawbacks. First, corpus
data cannot be used exclusively when the aim is to describe and analyse
a natural language, since not everything that is in the language is always
contained in the corpus. The observation of the restricted set of data can
be a limitation once we want to obtain, for instance, negative evidence
6 Introduction

(Matthewson 2004). This is why linguistic studies cannot only be based

on corpora data, because the language under study needs to be tested and
this cannot be done alone with the data gathered in the corpus. Therefore,
elicitations and grammaticality/felicity judgments are also very important.
A second drawback is that although annotations should be as a theoretical
as possible, it is quite unlikely to have a theory-neutral annotation system
(Dryer 2006). To avoid problems related to this issue, Zwitserlood, Özyürek,
and Perniss (2008) propose a two-level annotation. Since some structures
that researchers are interested in still lack a thorough study, it is better not to
provide interpretation during the annotation process. They propose to make
a distinction between annotations on a mere descriptive level and annota-
tions on an analytic level. Annotations on the descriptive level describe
signs in terms of their phonological/phonetic form only, while annotations
on the analytic level tiers provide an interpretation and/or analysis. Analytic
annotations can be based on descriptive annotations, but they can also be
independently re-analysed if necessary. Importantly, any annotation-related
project would also possibly require studies into intra-annotator and inter-
annotator reliability, as well as the creation of computational tools that can
increase the reliability of annotators’ work (Schembri and Crasborn 2010).
For this goal, the online Kappa calculator calculates inter-annotator agree-
ment amongst several annotators and it can be a reliable tool to be used in
future practices.14
As for the annotation of signs localised in signing space, different groups
have developed complex spatial grids in their own annotation guidelines:
Zwitserlood, Özyürek, and Perniss (2008) using ELAN; specific signing
space font used in iLex;15 Lenseigne and Dalle (2006) propose a computa-
tional representation of the French Sign Language (LSF) spatial organisation
both for automatic interpretation and generation; Nonhebel, Crasborn and
Kooij (2004) establish the general guidelines for spatial annotations used in
the rest of the corpus cited previously.

1.4.2. Small-scale LSC corpus

By the time this book was written, LSC did not have a corpus yet. With
the aim of studying how definiteness and specificity are encoded in spatial
locations, a small-scale LSC corpus was built for the purposes of this work.
The small-scale LSC corpus consists of three types of data, namely semi-
spontaneous, videos recorded for other purposes, and elicited data. The natu-
ralistic data consists of recorded LSC conversations. Videos recorded for
 Methodology 7

other purposes include news presentation and materials to learn LSC. This
data was used at a preliminary stage in order to have a general sense of how
LSC spatial locations are used in different language situations and what they
encode for. This provided a picture within which I would frame specific data
questions and intuitions.
As mentioned before, corpus data cannot be the only reliable piece of
data on which hypotheses are based, since not everything that is gram-
matical is found in the corpus. Therefore, elicited data is crucial in order
to know whether a specific construction can be used or rather rejected in
some contexts. In the elicitation tasks, the use of translations from spoken
language (i.e. Catalan or Spanish) as well as the use of glosses were avoided
as much as possible in order to keep the output away from the influence of
the spoken modality and from sign supported speech variants. In SL research,
when presenting contexts it is much better not to use a metalanguage (i.e. a
language different from the language object of study) (Neidle et al. 2000).
Instead, drawings avoid any interference from the spoken language in the
surrounding community. My elicitation materials consisted of drawings as
well as signed contexts that provided the informants with stimuli to provide
the target discourse. I also asked for felicity judgments about the data, which
are comments that native signers are qualified to give by virtue of knowing
the language. Recorded fragments of discourse were shown to native inform-
ants and they had to judge the felicity of those constructions. These felicity
judgements were based on the intuitions of two native deaf signers. Moreover,
some comments signers gave were taken into account, although not included
as conclusions for the work but only as aside comments. Research cannot be
restricted to informants’ intuitions and it is important to note that when doing
fieldwork research comments should be viewed as clues, and not as results
by the researchers (Matthewson 2004). Researchers then have to decide and
determine whether the clues are relevant for the analysis. Data collection
sessions were conducted in LSC with the informants and myself. I am not a
native signer, but after working and being very actively involved in the Deaf
Community for many years my signing is very fluent.
The small-scale LSC corpus used in this book includes data from seven
native deaf signers (three women and four men), aged between 41 and 62
years old and living in the area of Barcelona. The corpus comprises so far
about 5,108 signs. It is a composite of genres, such as news, interviews,
documentaries, tales, as well as different discourse modes, namely narra-
tive, explicative, and dialogue (Smith 2003). The distribution across types
of data and the signers that participated in each one is illustrated in the table
below.16
8 Introduction

Table 1. Distribution: types of data and signers

Types of data Signers
Semi-spontaneous 2, 4, 6
Recorded for other purposes 1, 2, 3, 6, 7
Elicited 1, 3, 5, 6, 7

1.4.3. Annotation conventions

The software used for the annotation is ELAN. When elicited data has been
recorded, two cameras were used and the recording was synchronised in the
ELAN annotation file.

Figure 1. ELAN screenshot

The annotation conventions are based on the guidelines established in

Nonhebel, Crasborn and Kooij (2004). The ten linguistic tiers that the anno-
tation of the corpus comprises are shown in Figure 2 and defined below.
They are listed in the Annotation conventions and described below.

Figure 2. ELAN linguistic tiers annotated

 Methodology 9

The description of each linguistic tier is presented in what follows:

(i) The gloss is an approximate and consistent translation of the sign. The
glosses for the active and non-active hands have been annotated. This
was not meant to be a big corpus, but only a representative part of real
LSC use. If this had been a large-scale LSC corpus, we would have
used ID-glosses for every sign. Lexical signs need to be identified with
a gloss, which uniquely identifies the sign. This is referred in the Aus-
tralian SL (Auslan) Corpus as ID-gloss, which is the spoken word that
is used to label a sign all the time within the corpus, regardless of what
a particular sign may mean in a specific context (Johnston 2008). With
this consistent annotation it is possible to use the corpus productively
and convert it into a machine-readable format.
(ii) Direction and location of signs towards signing space have also been
considered distinguishing among ipsilateral, contralateral and centre,
as well as upper and lower.
(iii) Our annotation of co-reference chains consists in giving the same en-
tity number to all coreferring mentions. To this end, in the co-ref tier in
each ELAN file, an index number is assigned to each discourse refer-
ent introduced. To keep track of the establishment of every discourse
referent introduced, we also assign a 0 next to the index, which indi-
cates the first mention of the referent. This annotation convention is
based on the designing annotation guidelines of coreferential chains
in spoken language corpus, work that started during the last part of the
decade of the 90’s. The guidelines of spoken language corpus (MATE/
GNOME, Poesio 2004; MUC, Hirschman 1997; ANCORA, CLiC-UB
2008) have inspired our LSC annotated corpus.
(iv) The referring term used is also annotated, and distinguished either as
a noun, a pronoun, or a Noun Phrase (NP). Verb agreement and classi-
fier constructions have also been annotated.
(v) The segmentation of utterances has been done according to units of
information that contain a predicate, have a semantic interpretation and
are delimited by major prosodic boundaries. Utterance boundaries are
a major concern both for spoken languages (Himmelmann 2006) and
for signed languages (Nicodemus 2009). Specifically considering SLs
there are two specific problems in delimiting sentences and clauses:
the difficulty of determining what is considered a predicate and the
availability of simultaneous constructions (Crasborn 2007). Discourse
units in LSC have been first singled out, with the help of identifying
the topic markers, which are markers of discourse units (Asher and
10 Introduction

Lascarides 2003). As markings of the intonational phrase in LSC,

the traditional markings have been taken into consideration, namely
change in head or body position, and a change in all aspects of facial
expression (Nespor and Sandler 1999).
(vi) The scope of role shift for the different characters in the discourse has
been also annotated.
(vii) Nonmanual elements have also been marked. As for brows, raised and
furrowed brows have been distinguished. As for eye gaze, the direction
in signing space (similar to the direction and location tier of manual
signs) has been annotated.
(viii) Comments on the annotation are important and useful for later analy-
sis.

Concerning the distinction between the two levels of annotation as mentioned

in the previous section, the linguistic tiers (i), (ii), (vi), and (vii) form the
descriptive level and (iii), (iv), (v), and (viii) form the analytic level of
annotation.
Many examples in this book are graphically represented with stills. The
stills correspond to different instances of signs in a fragment of discourse.
Below each still, the gloss of the utterance and a free translation are given.
Because the important signs localised in space are found in long stretches
of discourse, I have opted for only illustrating instances of localised and/or
relevant signs. In between each mention the signer keeps signing, but these
stills have been omitted from the figures in the interest of space. The signing
in between is indicated by a larger separation between two stills.

1.5. Organisation of this book

This book is concerned with the semantics and pragmatics of signing space
in Catalan Sign Language (LSC). Different phenomena are presented related
to definiteness, specificity and discourse structure. Hence the theoretical
background concerns different linguistic phenomena. This is the reason
why every chapter has a first section X.1 where the theoretical ingredients
needed for the presentation of the hypothesis are presented. The chapter then
develops the main analysis and findings in LSC of the concerned topic. The
rest of the book is organised as follows.
Chapter 2 provides background assumptions, analyses and views on
signing space found in the literature. The two main views concerning the
analysis of space, namely the spatial mapping and the r-locus view, are
 Organisation of this book 11

presented. This book clearly favours the r-locus view and presents new and
fresh arguments from LSC discourse data. The modality effects and the
different analyses of index signs directed to space are also presented.
Chapter 3 is devoted to the use of space of non-descriptive locations,
which are categorically defined as taking place in the different areas within
the three spatial planes projected with respect to the body of the signer. A
description of the uses of the three spatial planes and the features is presented.
Signs directed to the different parts of spatial planes contribute to the estab-
lishment of a grammatical morpheme that consists in an abstract point in
space (p), which is categorically interpreted within the linguistic system. In
LSC, (p) can be abstractly established in different parts of the three spatial
planes. Yet, only the two directions of the frontal plane, namely upper and
lower, are grammatically relevant and distinctively interpreted. As for the
horizontal plane, the features [ipsilateral], [contralateral], and [centre] do not
encode a specific and grammatical meaning by themselves, and also entities
of different nature are localised in each part. Finally, the two features in the
midsaggital plane, namely [proximal] and [distal], are not relevant for LSC
discourse. This chapter supports goal 1 and goal 2 (see §1.0).
Chapter 4 shows that (p) undertakes a semantic function: that of being
the overt manifestation of discourse referents. Under the specific Discourse
representation Theory (Kamp and Reyle 1993) formalisation, the discourse
referent established in space corresponds to a variable established in the
main universe of discourse. Hence, the establishment of (p) correlates with
discourse referents which are attached to a quantifier that has wide scope.
In contrast, variables attached to narrow scope quantifiers, such as donkey
sentences, quantified noun phrases, genericity and reference to kinds, lack a
spatial location establishment. This chapter offers new evidence in favour of
the r-locus view, as well as supports goal 1 and goal 3 (see §1.0).
Chapter 5 provides evidence that definiteness is not formally encoded
in LSC spatial locations. The distinction to show definite and indefiniteness
marking established in signing space is implemented here with respect to the
status of the DR in the model, based on whether the DR is presupposed or
asserted. It is shown that in LSC both possibilities establish (p). The chapter
also focuses on how information is incorporated into the model, and it is
claimed that both deictic uses, which do not have an explicit linguistic ante-
cedent, and Discourse Referents (DRs) with explicit antecedents appearing
in the previous discourse may introduce variables to the model. In the first
case, a default variable is present in the semantic representation, which
makes all references anaphoric to the model. This chapter supports goal 1
and goal 3 (see §1.0).
12 Introduction

Chapter 6 turns to specificity marking and proposes that spatial locations

encode specificity. The frontal plane is grammatically relevant for specificity
marking: lower spatial locations correlate with specific discourse referents,
whereas upper spatial locations correlate with non-specific ones. In LSC
two kinds of localisation on the frontal plane are found, namely a strong
and a weak localisation. Strong localisation is instantiated by the feature (p),
while weak localisation is instantiated by the marked feature (p)[up]. The
formalisation offered to explain this distinction is framed within the distinc-
tion between main and subordinate variables in a Discourse Representation
Structure (DRS). This chapter provides support for goal 1 and goal 3
(see §1.0).
Chapter 7 addresses the fact that lower spatial locations correspond to
discourse prominence, defined as backward looking properties as well as
forward looking properties. It is shown that independently of the scope of
the quantifier attached to the variable, narrow scope variables that are linked
to the prominent DR at a specific point in a discourse behave like wide scope
ones and establish a lower spatial location. It is also shown that (p) is an
abstract point in space which does not correspond to an exact point nor it
is related to a specific direction on spatial planes. In connected discourse,
locations associated with the most prominent DR can be shifted in space,
showing that the exact direction on planes is irrelevant for the nature of (p).
What it is relevant is that the spatial location (p) is associated with a DR from
the model and it is categorically interpreted independently of the direction
towards the horizontal plane. This chapter provides support to goal 2 and
goal 3 (see §1.0).
Finally, Chapter 8 provides a summary of the main findings and contri-
butions of this book, as well as interesting issues raised along these lines that
should be accounted for in future research.
Chapter 2
Space in sign languages: background

– I am still in the acquaintance phase.

– Acquaintance? How does that work?
– Basically, you read thousands of pages to learn the most necessary aspects
and to get to the essence of a subject, to its emotional truth, and then you
unlearn everything and start from scratch.
El joc de l’àngel, Carles Ruiz Zafón (2008: 300)

2.1. Introduction

Linguistic expressions in sign languages (SLs) depend on the use of space.

The different components of the grammar show dependence on it, from
phonology, to morphosyntax and discourse. This has been considered
to be unique to the visual-spatial modality and in some works it has even
been connected to co-speech gesture properties. This contrasts with a main
opposing view that considers that modality is not reflected in the structure
of the grammar and, besides the use of space, SLs and spoken languages do
share the same basic properties. According to this opposing view, SLs mainly
differ from spoken languages in that the referential indices are expressed
overtly.
This book claims that the use of space does not make the grammar of
SLs especially different. In order to show that, the present chapter is devoted
to presenting a broad revision of signing space, which is the main theme of
the present proposal. Here, a state of the art of the accounts, analyses and
views on the use of signing space are offered. After defining “signing space”
in §2.1, §2.2 explains the effects that the difference in modality has on the
language. §2.3 presents the two spatial functions associated with the use of
space. §2.4 compares the two main views concerning the analysis of space
and §2.5 presents the different syntactic and semantic analyses of pointing
signs. §2.6 provides evidence for the linguistic status coming from acquisi-
tion and studies about emergence of new SLs. §2.7 presents the proposal
underlying this book. Finally, §2.8 summarises the main claims presented in
the chapter.
14 Space in sign languages: background

2.2. Signing space

The actual space where the articulations of signs take place is called “signing
space”. It is generally considered being constrained to the horizontal and the
frontal plane in front of the signer’s torso (Figure 3). Pointing signs directed
to the back of the signer are also possible but the articulation of the sign does
not reach further than the back of the signer’s body.17 The body of the signer
itself is also used as a possible location for the articulation of the signs. As
argued by Klima and Bellugi (1979), it is important to note that this space
is not only used for articulatory reasons where the hands and the arms can
move (like the tongue is accommodated in the mouth for spoken languages)
but, more importantly, it carries linguistic meaning.

Figure 3. Signing space

As previously mentioned, the different components of the grammar rely on

signing space: starting from phonology, passing through morphosyntax and
reaching the discourse level. At the phonological level, space is used contras-
tively in the place of articulation parameter of signs. After Stokoe’s (1960)
study, signs are considered to be a bundle of manual and nonmanual compo-
nents. Nonmanual markers are movements linguistically significant, which
are done with the body, the head and parts of the face. Manual components
comprise the characterisation of the shape and position of fingers (i.e. the
handshape), the movement of the hand and/or fingers, the orientation of the
palm of the hand, and the place of articulation or location, where the articula-
tion of the sign takes place. Hence location is one of the minimal parameters
into which a lexical sign can be decomposed and it is considered to be a
major phonological category in SLs (Sandler 1989).
 Signing space 15

In Catalan Sign Language (LSC) the minimal pair of signs formed by

remember and difficult is distinguished by the place of articulation.
While the sign for remember is an index finger with contact on the forehead
and moving forward, the same manual articulation located on the chin results
in the sign for difficult, as it can be observed in the following figure.18

a. remember

b. difficult
Figure 4. LSC minimal pairs distinguished by the location parameter

At the morphosyntactic level, signs are modulated in space for grammatical

purposes to express number, person, and also the arguments of the verb.
Padden (1988) classifies ASL verbs in terms of agreement features. Three
types are distinguished, namely plain, agreement and spatial verbs. Plain
verbs do not inflect. Agreement and spatial verbs use space to express agree-
ment. The main difference between the two, following Padden, is that agree-
ment verbs inflect for person and number, and that the locations in space
indicate subject and object. In contrast, spatial verbs make reference to loca-
tions; that is, to the initial and final location of the entity being moved or to
the location where an entity is.
Both agreement and spatial verbs consist of a location-movement-location
sequence achieved by means of a path movement (Sandler 1989). Sometimes
it is also marked with the orientation of the palm of the hand. For regular
16 Space in sign languages: background

agreement verbs involving movement, the trajectory goes from the location
associated with the subject towards the location associated with the object
(Janis 1992, 1995; Mathur 2000; Zwitserlood and van Gijn 2006, among
others). Verb agreement issues are treated more deeply in §3.2.3. As shown
in Figure 5, the LSC regular agreement verb give is inflected for subject and
object. In the first still the movement is articulated from first to third person,
and in the second still it is articulated from third to first person.

a. 1-give-3 ‘I give (it) to him.’

b. 3-give-1 ‘He gives (it) to me.’

Figure 5. Regular agreement verb in LSC

Concerning the discourse level, it has been repeatedly noted in the litera-
ture that spatial locations are associated with discourse referents: a discourse
referent is assigned to a certain location on the horizontal plane and may be
referred back to later in the discourse (Klima and Bellugi 1979). In fact, it
is when we enter the syntax-discourse interface that signing space is greatly
used. Such spatial location associated with an entity is called “referential
locus” or “r-locus” (Lillo-Martin and Klima 1990). In isolated sentences,
entities tend not to be localised in signing space.19 To a great extent, it is
in the realm of connected discourse when the complexity and intricacies
 Signing space 17

of signing space are more evident. The following figure is an example of

different instances of reference to an entity localised in space across sentence
boundaries.

Figure 6. First and futher mentions of a referent localised in space

The set of linguistic mechanisms used to both establish an individual in

space and to refer back to it have been described for specific SLs, namely
ASL (Winston 1995), BSL (Morgan 1996), and LSC (Barberà 2007). The
most commonly used mechanisms are index signs,20 agreeing verbs, weak
hand in dominance reversals sequences, body lean, and classifiers construc-
tions.21 These mechanisms, as described for LSC in Barberà (2007), not
only establish an individual in signing space but they also contribute to
keeping the referent active in the discourse and constitute the strategies for
reference-tracking.
Several studies have explored how SL signers understand and maintain
the association between spatial locations and discourse referents. It has been
proven by psycolinguistic experiments that ASL pronouns activate their ante-
cedent and suppress non-antecedents in memory, just as has been found for
spoken languages (Emmorey 1997). In addition, ASL agreeing verbs license
phonologically null pronouns (Lillo-Martin 1986). Like in some romance
languages, subjects and objects in clauses with agreeing verbs can appear
as null elements due to the rich verbal morphological marking. Indeed, null
pronouns also activate their antecedent to the same extent as overt pronouns,
similarly to what has been found for spoken languages (Emmorey and Lillo-
Martin 1995). These SL comprehension studies prove that although the
association between spatial locations and space is a typical characteristic of
signed languages, the interpretation and resolution of coreferential relations
do not differ across language modalities (Emmorey 2007).
The definition of natural languages as discrete systems has been
applied with difficulty to SLs when it comes to the analysis of space. The
18 Space in sign languages: background

boundless three-dimensional extent in front of the body of the signer is, at

first sight, difficult to be considered a discrete system. It has been claimed,
for instance, that the distinctive locations in signing space are evident at
the syntax-discourse interface rather than at the phonological-lexical level.
Kooij (2002) analyses the NGT frequencies of different locations in the
SignPhon database (Crasborn et al. 2002) and provides an overview of the
distribution of the different lexical distinctive locations, namely neutral
space, head, neck, trunk, arm, and weak hand. Kooij concludes that at the
lexical level there are no phonological specifications in neutral space, except
for the horizontal plane, which is used as a major location. This implies,
according to her, that locations in neutral space with respect to which the
hand can move are determined and established according to some referent
found in the discourse.
In fact, the analysis of space in SLs is not free of controversy and it has
been, and still is, a matter of debate among SL researchers. Some researchers
have found similarities in the use of space in co-speech gesture. Since the
gestural component of spoken language conversations can externally be seen
as very similar to co-speech gesture, some researchers have argued that SL
signs that necessarily use locations in space (such as pronouns and agree-
ment verbs) are composed of a linguistic and a gestural part (Liddell 1990;
Meier 1990, among many others, but see §2.4.1 for a revision of the relevant
literature). For them, this implies that signs directed to space are analysed as
a gradient continuum and very rarely as categorical elements.

2.3. Modality effects

An important contribution to SL research specifically but also to (neuro-)

linguistics in general was Poizner et al. (1987)’s book. This pioneering work
provided evidence that language is not limited to hearing and speech. By
studying deaf subjects with brain lesions, they showed that ASL has the same
properties and relies on the same brain areas as spoken languages. Poizner
and his colleagues state that signing space has linguistic meaning:
“Overall the ASL system of spatialized syntax is similar in function to gram-
matical devices found in spoken languages of the world. However, in its form
– marking connections among spatial points – spatially organised syntax in
ASL bears the clear imprint of the mode in which the language evolved.”
(Poizner et al. 1987: 18)
There is in fact a general assumption in the SL literature that the strong candi-
date for a modality effect in the language is the use of space for indicating
 Modality effects 19

reference in pronouns and verbs. But it is still an open question whether

the use of space is a total reflection of the use of a different modality. By
modality it is meant how the language is perceived and produced. Spoken
languages are considered to use primarily the audio-vocal modality, although
to a lesser extent the visual-gestural modality is also used during co-speech
gesture. SLs are natural visual-gestural languages and this modality interacts
with space, as well as with gesture. These interactions are the main concern
of the next two subsections.

2.3.1. Modality and space

Bringing this debate into the generativist framework, Lillo-Martin (2002)

convincingly states that the modality of the language mainly affects the
phonology module since this is the component of the grammar that inter-
acts with articulation and perception of the language. According to this
work, phonology and syntax are independent modules, and modality in SL
only affects the former. She also argues that abstract syntactic principles of
universal grammar apply to both SLs and spoken languages. Just as universal
grammar allows parametric variation between languages, SLs may vary from
spoken languages (Lillo-Martin 2002: 243).
As far as agreement is concerned, in the general literature verbs have been
described as lexically specified for certain components, such as handshape
and skeletal structure. The template that has been described and applied to
many SL verbs is location-movement-location (Sandler 1989), as seen in
§2.1. This information relates to the morphological process of agreement.
But the problem that arises is how these locations are specified in the finite
lexicon of the language since they are left unspecified. Locations must be
thus filled in. The general assumption is that verbs agree with subject/object
in person and number. This view is clearly stated by Neidle et al. (2000)
and Neidle and Nash (2012), which analyse spatial locations as consti-
tuting an overt instantiation of phi features, and more specifically person
features. This is coherent with the argument that the matching of features
among syntactic elements is of essentially the same nature as in other agree-
ment systems. However, from an opposed perspective, Liddell (1990, 2000,
2003) and Meier (1990), among others, point to some modality differences
between spoken and signed languages, the most important one being the
spatial locations established in space in signed languages due to the visual-
gestural modality. According to this view, these locations do not constitute a
finite set of discrete elements since verbs and pronouns can be directed to an
infinite number of possible locations in signing space.
20 Space in sign languages: background

Another aspect related to modality is the interlinguistic variation. It has

been noted that the variation observed in SLs seems much more limited
than the variation found in spoken languages (Newport and Suppalla 2000).
In a detailed study of the effects of modality upon linguistic structure,
Meier (2002) argues that the differences in the properties of articulators, in
the perceptual system and in the youth of SLs are possible sources of modality
effects on linguistic structure that underlie the limited variation. According
to the author, the possible outcomes of modality effects range among differ-
ences between languages of the two modalities in statistical tendencies. That
is to say, one modality has more instances of a certain linguistic feature
than the other. Also one modality has some preferred typological proper-
ties in comparison to the other (e.g. SLs generally opt for nonconcatena-
tive morphology, as convincingly argued in Mathur and Rathmann (2010)).
A relative structural uniformity of SLs vs. a relative diversity of spoken
languages could be a product of the visual-gestural modality, but it has been
noted that this apparent uniformity may be a consequence of the youth of
such languages and the scarcity of cross-linguistic studies (Meier 2002;
Newport and Suppalla 2000).
Interestingly, more and more studies are focusing on the comparison of
structures among different SLs showing that, once grammatical aspects are
closely examined, SLs may be more different than previously thought (see
Perniss, Pfau, and Steinbach 2007). Moreover, the study on village SLs,
i.e. SLs in small communities with a high incidence of hereditary deafness
(Nyst 2012), enhances the idea that there is more linguistic diversity across
SLs than previously assumed. In fact, data from Kata Kolok in Bali and
AdaSL (Adamorobe SL) in Ghana adds appreciably to this diversity. For
instance, these languages considerably differ between them, and also differ
from the features previously thought to be modality-specific, such as the use
of spatial grammar. AdaSL uses directional verbs, but does not use entity
classifiers (Nyst 2007). This contrasts with Kata Kolok, which does not use
directional verbs, but does use entity classifiers (Marsaja 2008; de Vos 2012).
In fact, Kata Kolok pointing signs may not be directed to abstract areas in
space and spatially verb inflection is virtually absent from the language
(de Vos 2012). It also seems that the small-scale setting in deaf villages
enables these languages to have structures that are less frequently attested or
even not found at all in other Western SLs, such as absolute pointing instead
of using locative names (Zeshan 2010). Research on village SLs is thus of
great importance to contrast what was thought to be modality-specific with
other structures found in languages which actually use the same modality.
 Modality effects 21

2.3.2. Modality and gesture

The visual-gestural modality is not only used by SLs, but also by co-speech
gesture co-occurring with spoken languages. This is the reason why some
authors have put into question the grammatical use that SLs make of it,
and they have compared it to gesture (see Liddell 2003 for an overview).
However, there is a non-symmetric use of both modalities since the gesture
accompanying speech is highly under specified and dependent on it
(Lascarides and Stone 2009). As argued in Barberà and Zwets (2013), the
audio-vocal modality in spoken languages is to be considered the dominant
one, and the visual-gestural modality relies heavily on it for its interpreta-
tion. That is, a listener gets the cues for interpreting the pointing gesture
from the dominant modality that co-occurs with it, in combination with
the physical environment surrounding the speech participants. Since the
verbal element and the pointing gesture are performed in different modali-
ties they can still occur simultaneously. In contrast, in SL the addressee
interprets the pointing sign expressed only in the visual-gestural modality
by considering both the linguistic and the physical context. In fact, experi-
mental studies show that when using gesture alone, hearing adults placed
gestures for particular entities in non-neutral locations and then used
these locations to refer back to them. But when using gesture and speech
together, hearing adults also produced gestures in non-neutral locations
but used the locations coreferentially far less often (Chee So et al. 2005).
These results lead to the hypothesis that when the visual-gestural modality
is totally responsible for communicating the message, space is exploited
for co-reference.
According to what has been said in the previous paragraphs, modality not
only determines the extra use of space that SL makes compared to the spoken
languages counterpart, but also the gestural domain. SLs have additional
possibilities of developing grammatical markers directly from gestures. This
unusual source is evidently due to the particular way signing is produced and
perceived, i.e. a modality which exploits the visual-manual medium, also
exploited by the gestures that accompany speech. However, the grammati-
calisation patterns are considered to be parallel to the spoken languages ones.
The interested reader is referred to Pfau and Steinbach (2006) for convincing
arguments based on different SLs showing that the typical paths taken by
lexical items as they are transformed into grammatical elements are the same
in both signed and spoken languages.
22 Space in sign languages: background

2.4. Spatial functions

Since the beginnings of SL linguistics research, it has been argued that space
undertakes two functions, namely a syntactic and a topographic one (Poizner
et al. 1987). The syntactic function is an abstract use of space in which enti-
ties are localised arbitrarily to identify the arguments of the verb. Entities are
assigned a specific location, which is movable as it can be shifted without
affecting the truth conditions of the sentence. The topographic function, in
contrast, is used to express spatial relations among objects and it is repre-
sented by meaningful locations that exploit the iconic properties of the visual-
spatial modality. Topographic locations are meaningful by themselves, so a
small change in the location affects its truth conditions. In this latter case,
space is used to represent spatial arrangements via signed descriptions, and
thus the actual spatial relations of signs are significant. Following Quer et al.
(2005) and in order to avoid the implication that the topographic use of space
is deployed without syntax, I call the localisations occurring in the syntactic
use of space “non-descriptive”, and the ones that occur in the topographic
use of space, “descriptive” localisations. In descriptive localisations the
relations among spatial locations become significant because they represent
actual spatial relations topographically. The descriptive location in Figure 7a
represents a bike leaning against a tree; and in Figure 7b, a person seated on
a tree. In both cases the location of the manual articulators is meaningful.

a. Bike leaning against a tree b. Person seated on a tree

Figure 7. Descriptive localisation

This contrasts with non-descriptive localisations, which are arbitrarily and

abstractly established for the syntax and discourse of the language. As the
 Spatial functions 23

following figure shows, the signer has arbitrarily localised an entity on his
contralateral side for the discourse referent “son” (Figure 8a). Afterwards,
the signer co-referentially picks up the non-descriptive location previously
established (Figure 7b).

a. son ix3 b. ix3

Figure 8. First and further mention of a localised discourse referent

Comparing both figures shows that LSC descriptions of spatial layout

(i.e. descriptive localisation) use the same horizontal plane of signing space
as do SL nominals, pronominal reference and verb agreement devices
(i.e. elements within a non-descriptive localisation). Both functions make
use of locations in space, the difference being in its significance: locations
in signed descriptions are meaningful because they represent actual spatial
topographical relations, whereas abstract locations are not meaningful by
themselves because they are established for syntactic and discourse purposes.

2.4.1. Descriptive localisation

Descriptive use of space is circumscribed to the expression of spatial infor-

mation, such as the position of an object or a relative positioning of an object
with respect to another one. Most spoken languages encode spatial relations
with affixes and prepositions, as in (1).

(1) a. El llibre és sobre la taula. [Catalan]

b. The book is on the table.

Descriptive localisation has largely received a great attention in the litera-

ture, especially the relationship between the properties of space and the
24 Space in sign languages: background

perspective the signer takes while providing spatial information (Emmorey

and Falgier 1999; Emmorey 2001; Emmorey 2002a, 2002b; Emmorey and
Tversky 2002 for ASL; and Perniss 2007a, 2007b; Perniss and Özyürek 2008
for DGS, among others). Spatial information in SLs is mainly conveyed by
classifier constructions and by the placement of the hands in certain locations
in signing space and also with respect to the body of the signer (Suppalla
1986; Engberg-Pedersen 1993). When an LSC signer wants to convey that the
book is on the table, an entity classifier22 will be used to refer to the book and
it will be localised above a flat surface representing the table, as shown below.

Figure 9. Book on the table, represented with an entity classifier.

If more than one referent is represented in space, first the backgrounded

entity is introduced (the so-called “ground” in the literature on language and
space, e.g. Talmy 1985), and then the smaller entity, which is in the focus of
attention (the so-called “figure”). The particular position of one handshape
with respect to the other expresses the spatial relation between referents
(Arık 2009; Chang et al. 2005; Miller 1994; Morgan and Woll 2007; Perniss
2007a; Supalla 1986). This is achieved by virtue of the simultaneous use that
SLs make of the two manual articulators.
It has been argued that there is an isomorphic mapping between the loca-
tion of the hands in space and the location of the objects described (Emmorey
and Herzig 2003). The spatial mapping can be set from the signer’s point of
view (i.e. the character perspective) or from the addressee’s point of view
(i.e. the observer perspective).23 Spatial scenes are commonly represented
according to the signer’s point of view and in fact ASL signers comprehend
much better spatial descriptions when they are represented from the signer’s
perspective rather than the addressee’s one, despite the mental rotation these
descriptions entail (Emmorey et al. 1998). Finally, when both signer and
addressee jointly view an environment, ASL signers use a “shared space”
 Spatial functions 25

(Emmorey and Tversky 2002). That is, the physically observed setting maps
the linguistic represented setting in signing space to both the signer’s and
addressee’s view. In the shared space use there is no true signer’s or address-
ee’s point of view, nor mental rotation required. Both interlocutors refer to
the same locations, regardless of their actual location. According to some
works, non-descriptive localisations consist in conceptualisations of spatial
relations of objects, which are conceptualised under some frame of reference
(Levinson and Wilkins 2006). As the works by Arık (2010, 2011) and Perniss
(2007a, 2007b) show, different sign languages make use of different perspec-
tives and frames of reference.
Before finishing this section, it should be noted that some village SLs
have been described as making extensive use of descriptive localisations.
Kata Kolok SL predominantly employs topographical space (Marsaja 2008;
de Vos 2012). As previously mentioned, signers use real-world locations
instead of establishing abstract locations despite the ambiguities (e.g. the
sign for a place may be localised differently depending on where the signer is
in relation to the referent). Kata Kolok uses thus an absolute frame of refer-
ence (Levinson 1996), which is very rarely used in western SLs.

2.4.2. Non-descriptive localisation

The main difference between descriptive and non-descriptive localisations

from an interpretation and analytical level is that the former conceptualise
the position of the object in signing space and the latter establish a formal
and abstract relationship between the object and the corresponding associ-
ated spatial location. Moreover, they also present another difference at the
form level: while descriptive uses of space exploit richer and freer sets of
locations in the three-dimensional space, non-descriptive ones are composed
of spatial planes and fixed trajectories within each plane. Importantly, non-
descriptive locations are not restricted to the horizontal plane in front of the
signer as originally argued by Klima and Bellugi (1979). Some authors note
that nouns can also be assigned locations vertically above or below the hori-
zontal plane in certain circumstances (Davidson and Gagne 2014; Fischer and
Gough 1978; Morales-López et al. 2005; Schlenker and Lamberton 2012;
Schlenker, Lamberton and Santoro 2013; Shepard-Kegl 1985). Liddell and
Johnson (1989), Sandler (1989), and Brentari (1998) provide a detailed
phonological analysis of locations and especially a thorough description
of body locations for the production of signs. However, when it comes to
analyse space as used in actual discourse, the description is not as detailed
26 Space in sign languages: background

as it is at the lexical level. The phonological works just cited agree that the
syntax-discourse interface determines the position of the sign in space and
they postulate some major spatial areas interfacing with the discourse domain.
Place of articulation, and thus also localisation, may be divided as occur-
ring on three different planes projected with respect to the body of the signer
(Brentari 1998: 120). First, the horizontal or transverse plane stands perpen-
dicularly to the body of the signer and it is the default plane where the majority
of the signs are localised (Figure 10a). Second, the frontal or vertical plane
is defined by all those points encountered on the plane in parallel to the body
(Figure 10b). Finally, the midsagittal plane is vertically perpendicular to the
body of the signer (Figure 10c).

a. Horizontal b. Frontal c. Midsaggital

Figure 10. Spatial planes

Non-descriptive locations are constrained within these three planes. In §3.3

the major distinctions made by the works cited above and the three planes are
applied and extended to the LSC discourse data.

2.4.3. One function or two?

Whether there is a clear-cut distinction between a descriptive and non-

descriptive function is still a matter of debate among linguists. Some
researchers, in particular Liddell (1990, 1995, 1998, 2003), van Hoek (1992,
1996), Engberg-Pedersen (1993, 2003) and Perniss (2012) propose a strongly
integrated view of the double function of spatial locations, and have argued
against maintaining a distinction between them. According to them, both
descriptive and non-descriptive locations are projections of mental represen-
tations and the frontier between the two functions is blurred. For example,
a signer could use a classifier predicate to establish a referent at a certain
 Spatial functions 27

descriptive location in signing space, e.g. a man on a tree, as represented in

Figure 7b above. Subsequently, the signer could direct a verb sign, e.g. tell,
to the same location, specifying the man as the grammatical object of the
predicate (see Liddell 1990: 318 and Perniss 2012, for a similar example).
The man is still conceived of as seated above the tree at the time he is told
something. In this example, the location associated with the man is both
functioning referentially and topographically.
Psycholinguistic studies have been undertaken to motivate this difference
from an experimental and testable perspective. Emmorey et al. (1995) present
some empirical evidence that these two functions can be dissociated with
brain injury, during on-line processing and memory encoding. Their results
suggest that the topographic use of space may be tightly linked to spatial
cognitive abilities attributed to the right hemisphere. Although they empha-
size the distinctness of these two spatial functions, they also note that they
are not mutually exclusive, noting that it is an issue of how a location func-
tions within signing space, and not of two distinct types of signing space, as
previously suggested by Poizner et al. (1987). Depending on how it is used,
the same location may function both non-descriptively and descriptively.
Emmorey et al. (1995) show that the two spatial functions are processed
differently. When incongruent spatial information is given, processing of
descriptive locations is more affected than processing non-descriptive ones.
Since topographic space uses space as a map, the information provided must
be more explicit in encoding the interrelation between locations and objects.
In contrast, the setting for syntactic space does not convey spatial information
about their referents, and subjects can easily understand sentences where the
spatial setting is completely arbitrary and incongruent. More recent studies
provide new evidence for these spatial distinctions. In a PET study, Emmorey
et al. (2002) propose that naming spatial relations with classifier construc-
tions in ASL (hence when signing space is used descriptively) involve right
hemisphere processing. However, in a related study, MacSweeney et al.
(2002) used fMRI to investigate the neural areas when deaf and hearing users
of British SL comprehend sentences that used space descriptively compared
to sentences that used it non-descriptively. Interestingly, their results did not
show more right hemisphere activation for processing descriptive locations
compared to non-descriptive ones. These differences in the results may be
most probably motivated by the difference in the demands of the task. While
in the ASL study subjects were asked to translate the spatial relation between
the hands in signing space into another representation, in the BSL study the
signers were asked to press a button when they detected a semantic anomaly.
More research on similar tasks is needed in order to confirm these findings.
28 Space in sign languages: background

Also further work is needed on contexts where the descriptive and non-
descriptive uses of space are simultaneously inserted in the same context. In
the psycholinguistic studies described above the majority of tests presented
sentences dealing with one or the other function, but there were no strings
of sentences where the two functions were fused. This overlap in the same
fragment of discourse would allow us to make a straightforward distinction
between the two of them and also see whether one function is more predomi-
nant than the other when they co-occur.
Finally, it is worth mentioning that it has been described that some
SLs represent temporal reference in space. Brennan (1983) and Engberg-
Pedersen (1993) first identified several distinct time lines, which are spatial
constructs that represent distinct types of temporal information. These time
lines extend on the horizontal plane and can be divided into: basic, anaphoric,
sequential, and mixed. This book focuses only on referentiality associated
with spatial locations and on the referring back function. On temporal refer-
ence, the interested reader is referred to Brennan (1983) for a description on
BSL, Engberg-Pedersen (1993) for DSL, Emmorey (2001, 2002) for ASL,
and Cabeza and Fernández-Soneira (2004) and Herrero (2009) for LSE. Let
us turn now to the two main and opposing views concerning signing space.

2.5. Previous accounts

At the beginning of the 90’s many researchers focused their attention

primarily on the analysis of space. One of the main concerns was, and still is,
how to integrate the infinite directions in space of localised signs into a finite
system like grammar. In SLs second and third person pronouns are always
dependent on the location that the denoted referent occupies and according
to some researchers their phonological form does not have linguistic distinc-
tions. The multiple or even infinite number of locations where index signs
may be directed to turns the integration of these infinite locations to a discrete
linguistic system into an impossible task. This infinity issue splits the view
that researchers have in space into two main analyses. On the one hand,
what I call the “spatial mapping view” considers that spatial locations are not
phonologically specifiable. They are a projection of a mental representation
and thus they can occupy different non-discrete places on signing space. On
the other hand, the “r-locus view” associates spatial locations with referen-
tial indices. Locations represent overt referential indices that correspond to
the referent the nominal denotes. As such, no listability issue is found in the
phonological form since locations are derived from the discourse. The main
 Previous accounts 29

difference that the r-locus view argues for is that while referential indices are
covert in spoken languages, they are overt in SLs. Below the main claims of
each view are detailed, as well as the problems and the advantages of each
one. As it will become clear, this book follows the r-locus view.

2.5.1. Spatial mapping view

Space has interested SL researchers coming from different perspectives and

theoretical frameworks. From the cognitive linguistics area, a large amount
of linguists have devoted time to describe how locations in space function.
The basis of their analysis is that SLs imply a spatial mapping, i.e. “the
process used by the signer to reflect mental representations in physical space
for reference and subsequent co-reference in discourse” (Winston 1995: 87).
The spatial mapping view is followed by Engberg-Pedersen (1993), van
Hoek (1992, 1992), Liddell (1990, 1994, 1995, 2000, 2003), Mandel (1977),
Rinfret (2009), and Winston (1995), among many other authors. Because
of the great influence of Liddell’s work on the spatial mapping view, the
following subsections review the main ideas developed in Liddell (1990,
1994, 1995, 2000, 2003). I first start with the definition of “locus” according
to this view and the relationship it has with real space according to the spatial
mapping account. Afterwards, the limitations of this view are also presented.

2.5.1.1. Locus and real space

The spatial mapping view defines “locus” as a projection of the referent

into space in the absence of the entities in the situational context (Engberg-
Pedersen 1993: 53). The majority of authors within this account follow
Liddell, who bases his approach on the mental space theory (Fauconnier
1985). According to this theory, entities are grounded in space. This means
that entities are conceived of as present in the immediate environment.
Liddell and followers of the spatial mapping view claim that signers concep-
tualise the referents they talk about in space. Hence they “visualise” the posi-
tion they occupy, as well as their height. According to him, agreement verbs
are directed towards the areas in space the referents of the arguments are
associated with, but there is also a “further agreement” with the concep-
tualisation of the height of the referent. In order to show this, he bases his
argumentation on some lexical verbs. An instance is the verb ask, which is
inflected from first to second person and which is articulated at mid height
30 Space in sign languages: background

to a referent that is higher than the signer. This example, following Liddell,
is ungrammatical as there has to be a correlation between the height of the
referent (either present or not, as what matters is the conceptualisation) and
the height of the chin. He says that “since the signer must conceptualise the
location of body parts of the referent imagined to be present, there is a sense
in which an invisible body is present. The signer must conceptualise such a
body in order to properly direct agreement verbs” (Liddell 1990: 184). Thus,
the verb should be directed to the supposed chin of the imagined referent.
Locations consist then of place-holders which stand for the conceptu-
alisation of a referent. This conceptualisation is expressed by signs directed
to space (such as pronouns and verb agreement). The height features are
found in the lexical properties of the verb. In fact, locations express topo-
graphic localisation, since they always situate a referent in space as if it were
present. He disagrees with the traditional view (Friedman 1975) according
to which the relationship between a locus and a referent is that of referential
equality (i.e. referent-a = locus-a), and he proposes a location fixing rela-
tion. According to him, every use of space is topographic and the locus itself
expresses nothing but that referent x is at locus y. According to him, “estab-
lishing an index serves as a way of saying where the referent is, not what
point is referentially equivalent to the referent. Evidence for this conclu-
sion comes from the fact that agreement verbs were not directed toward the
locus at which the index was established, but directed to points in space
whose height was a function of the lexical properties of the verb rather than
a function of the height of the locus” (Liddell 1990: 186). In fact, Liddell
(1990: 185) establishes the lexical properties that some ASL verbs have in
this respect. By means of some examples, he argues that the verb say-no-to
is directed to the nose, ask is directed to the chin, remind is directed to
the shoulder and give is articulated at the height of the chest. Mental space
conceptualisations allow Liddell to explain the infinity and multiplicity of
locations in space where signs can be directed to, which was already claimed
since De Matteo (1977) and Mandel (1977). However, as it will become
obvious in what follows this is not a satisfactory solution.
Liddell (1995) also discusses the behaviour of pronouns and verb agree-
ment24 in relation with possible spaces he himself defines, namely real space,
surrogate space and token space. Following Liddell, real space consists in
the person’s mental representation of what is real in the current environment.
Surrogates are invisible entities (person or objects) that signers conceptualise
as if they were present. They may take first, second and third person roles.
This kind of reference coincides with what is generally known as role shift.25
Token space is the situation where the signer places an invisible entity in
 Previous accounts 31

space. Tokens are not normal sized as surrogates and they are limited to third
person referents. Liddell (1995) argues that grammatical reference, when
surrogate and token space are used, is the same as with real space because the
signer can imagine surrogates and tokens as being in an unlimited number of
locations and therefore can treat them as physically present. Thus reference
in surrogate and token space are like they are in real space. Therefore, in his
opinion, they are deictic and not anaphoric.

2.5.1.2. Limitations

Although Liddell’s account provides an explanation to the infinity issue, it

also faces many problems that a theory of language should not be forced
to address. In what follows I present some arguments against the spatial
mapping view concerning a distinction Liddell’s account makes between
pronouns and agreement verbs, as well as the lack of consistency of use of
agreement with respect to the lexical properties. I will support my arguments
by considering the auxiliary agreement sign and by incorporating abstract
referents into the picture.
Liddell’s conceptualisation of referents in space presented in the previous
section is apparently only present in the use of some agreement verbs and it
is not present in the behaviour of pronouns. However, since spatial locations
stand for the arguments of the predicate it is widely accepted that SLs verbs
and pronouns behave similarly. Hence, if such conceptualisation is at work,
why should there be a distinction between the features that verbs incorporate
and pronouns? We could think that the lexical properties of verbs should be
also represented in the direction of pronouns if they are directed towards
surrogates, and that both an agreement verb and a pronominal form referred
to a taller referent should have the same direction in signing space. But this
is not the case and pronouns do not show this behaviour of representing
different heights of conceptualisations of referents. Notwithstanding, let us
suppose that we accept Liddell’s conceptualisation of referents, as well as
the lexical information incorporated in some verbs. In such a situation, it
is pertinent to ask whether the lexical specifications of the verbs are cogni-
tively coherent, especially if they are treated under such a cognitive frame-
work as mental space theory. According to Liddell (1990), say-no-to and
ask are directed to the nose and to the chin, respectively. However, since the
actions of saying and asking in speech are articulated in the mouth, wouldn’t
it be more cognitively coherent to direct the verbs that represent each action
towards the mouth? And even more importantly, since in SLs the action of
32 Space in sign languages: background

saying is connected to the hands, why is the verb not directed to the hands
of the addressee? Last but not least, give is expressed at the height of the
chest. But why is it not articulated at the height of the hands? The lexical
properties attributed to verbs are not motivated and they do not seem to be
coherent with the actual action they represent, especially under an account
that deals precisely with conceptualisations of referents. Also, Liddell does
not explain why these lexical properties are only present in some verbs but
not in others.
If, following Liddell, verbs incorporated lexical properties of these kinds
in languages with auxiliary agreement this distinction should be also evident.
So far, studies done on agreement auxiliary signs (i.e. signs which generally
co-occur with plain verbs to mark agreement) have claimed that they behave
very similar to agreement verbs, i.e. with a movement from subject to object
(Steinbach and Pfau 2007). Nevertheless, to the best of my knowledge, no
work has noted so far that the auxiliary sign also represents the conceptuali-
sation of the referent and that it can be directed to different heights according
to the height attributed to the referent. For instance in LSC the auxiliary sign
directed towards first person is at the chest, and for second and third person is
directed towards the head (Josep Quer, p.c., August 2010). Thus no specific
conceptualisation seems to be at play here.
Finally, if referents are conceptualised in space, it is difficult to explain
what happens with abstract referents, which do not have a specific height
to direct signs to. So if, according to Liddell (1990), in ASL the verb ask
is expressed at the chin level, when we use it with an abstract referent it
may be difficult to determine how to localise it. Imagine, for instance, that
someone is saying that he is very curious about deaf traditions, culture,
habits, etc. If he utters the sentence “I may ask all these questions to the
Deaf Community”, to which height is he supposed to direct the verb? What
are the lexical properties of a sign going to be for an abstract referent, such
as “Deaf Community”? This would predict that we need two lexical entries,
each related to the denoted referent (i.e. ask_concrete and ask_abstract),
and not a single entry with specific lexical properties. Again, this is not a
satisfactory solution. However, in Liddell (1995) this problem is solved with
the distinction made between surrogates and tokens. Although he does not
make this point precise, I assume that abstract nouns may be only repre-
sented by tokens and thus no conceptualisation in space needs to be made.
Yet, this issue is not presented in his influential 1990 paper.
Another important aspect in the spatial mapping view is the iconicity
attributed to index signs directed to space. It is a widespread belief that
pointings are iconic since their meaning depends on the visual connection
 Previous accounts 33

between the pointing gesture and its target (Cormier 2007; Liddell 2003;
Mandel 1977). However, pointing signs are very often directed to an object
in space not to denote that object but rather to refer to an idea or an entity
related to that object. This indirect reference instances are very frequent in
SLs, and in these cases there is not a relation of contiguity between the index
sign and the object pointed at, as widely claimed. In indirect reference, for
instance, the orientation of the pointing is always directed towards an object
that has a strong contextual link with the actual referent that is to be inter-
preted by the addressee, but the interpretation is not only derived from conti-
guity. We can easily imagine a situation in which a signer directs a pointing
sign towards a book present in the physical environment, which is about a
deaf school that existed in the 60’s in southern Catalonia, while uttering the
following sentence.

(2) ix3 interesting.

Depending on the context, the pointing directed to the book can have different
meanings, listed in (3) from more iconic to more indirect.

(3) a. The copy of the book I am presently pointing to is interesting.

b. The deaf school in Catalonia (the book is about) is an interesting
topic.
c. The problems deaf kids had during the 60’s (which are reported
about in the book) are an interesting topic.

For the interpretation of indirect reference, context becomes of great impor-

tance, and physical contiguity is not only insufficient, but it can also be
misleading, as shown in (3). To get the interpretation in (3a), physical conti-
guity is decisive. But for the interpretation of (3b) and (3c) physical conti-
guity can be rather misleading.
If points in space were precisely what matters, the concrete area where the
pointing sign is directed to would be difficult to determine. Since a pointing
sign is an index handshape with the fingertip oriented somewhere, to exactly
delimit where the exact and relevant point is becomes an impossible task.
The relevant point could be 5 cm away from the fingertip, but it could also be
20 cm away, or even 3 metres. Thus, to precisely determine its physical end
point becomes an arduous task, even an impossible one. Hence, it is rather
a matter of direction of the pointing sign together with linguistic contextual
clues (Barberà and Fernández 2009). But according to Mandel, Liddell and
34 Space in sign languages: background

Cormier the exact direction is crucial for the understanding of an index sign
and the significant difference between English and ASL pronouns does not
lie in their ability to point conceptually toward their referents, since both do
that. Instead, according to Liddell (2003: 68):
“The significant difference is that during the production of the ASL pronoun
the hand also physically points toward the present referent. Its significance
can only be determined by following the directionality to see what it leads to.”
Pointing signs do not point to present referents physically, but conceptually.
If they pointed to present entities in the surrounding space it would be very
difficult for the addressee to disambiguate whether the signer is pointing to,
for instance, the man the pronoun is referring to, to his shirt, to the stain on
his shirt or even to the dog that crosses the room and it is precisely in front
of the man while the signer is pointing (Barberà and Fernández 2009). In
order to test Liddell’s (2003) hypothesis, I will report on an example that I
used with our deaf informants. The setting is the following: let us imagine
a meeting between the president of the Catalan Federation of the Deaf and
seven members of two deaf clubs in Barcelona. Three members belong to
club x, and the other four belong to club y. Members are seated separately
from their comrades, and hence the positions they occupy at the table are
mixed.

President
Figure 11. Mixed position of the members of two deaf clubs

After one hour of discussion, it is time to establish the duties each club will
have to undertake. The president can direct an incorporated pronoun and
say “you-three will do this, and you-four will do that”. Even if the members
are mixed and the three members of club x do not sit next to each other,
the pronoun may be directed to the area where most members of club x
are sitting -even if in between there is a member of club y. The linguistic
context may guide the interpretation of the pronominal form and hence the
 Previous accounts 35

most salient referent (members of club x for “you-three”, and members of

club y for “you-four”) will be picked irrespectively of the exact direction of
the pointing sign. Hence the exact position of the referent denoted by the
pronoun is irrelevant to get the felicitous interpretation. Rather, the linguistic
context is the needed clue. If pronouns were to be understood as the exact
direction of the index sign, the reference of the incorporated pronouns would
be as “you three that I am actually pointing at” independently of the club
they belong to. However, this is not the case in LSC and what is important
is that the reference of the pronoun is derived from the combination of the
direction of the manual sign and the linguistic context, but not considering
the exact direction alone. I agree with Liddell (2003) that the amplitude of
the movement of the pronominal form will be in accordance with the setting
of the objects in reality denoted: a smaller movement means that objects are
closer to each other, and a bigger movement denotes objects that are sepa-
rated. Nevertheless, the combination of amplitude of movement, direction,
and linguistic context are required for the disambiguation of under specified
forms directed to space.
Liddell’s work and specifically the mental space theory do not make any
distinction between the actual world and the discourse. Referents exist in the
same domain, which is the fusion of the two domains (i.e. real world and
discourse). However, a clear distinction must be made between what exists
in the real world and what exists in the discourse. If such a distinction is not
made, what happens with reference to entities that do not exist or also with
quantified expressions? According to the spatial mapping view, the repre-
sentation of space consists in an abstract mental representation in contexts
where the referent is not physically present and with a direct mapping with
the referent in contexts where it is present. The main and crucial difference
between the spatial mapping view and the r-locus view is that the former
considers locations to be projections of mental representations that are not
part of the grammar. In contrast, the latter analyses locations in space as inte-
grated into the grammar of the language, as will be shown below.

2.5.2. R-locus view

When establishing a referent, a signer is referring to the discourse model

itself. With this establishment, a signer in effect says no more than “when I
refer to this spatial location, I will mean such an individual”. The signer is
in fact defining a formal relationship between a referent and a location, for
further use later in the discourse. This is the main claim that has inspired the
36 Space in sign languages: background

r-locus view, which has been pursued by many authors (Aronoff et al. 2000;
Bahan 1996; Cormier, Wechsler, and Meier 1998; Friedman 1975; Janis
1992; Kegl [1976] 2003; Lacy [1974] 2003; Lillo-Martin and Klima 1990;
Lillo-Martin and Meier 2011, using data from ASL; Schlenker 2011a, 2011b,
using data from ASL and French SL (LSF); Meir 1998 using data from
Israeli SL (ISL), among many others). The r-locus view claims that although
it is generally assumed in the syntactic literature that NPs are considered to
contain referential features that are abstract, SLs show the overt morpho-
logical expression of referential distinctions through association of distinct
referents with specific spatial locations. Hence locations in space, the
so-called loci, are analysed as the overt manifestation of referential indices.

2.5.2.1. Referential locations and linguistic space

This approach argues that locations are identified with referential indices
(r-indices). R-indices are variables in the linguistic system, whose content
comes from discourse and that are overtly represented in the linguistic system
of SLs. Variables can be one among an infinite number of possible referents,
since a discourse model can contain an infinite number of possible referents.
The basic properties that characterise the r-locus view support the anal-
ysis that locations are more similar to indices than to pronouns (Lacy [1974]
2003). They are briefly presented below (see Janis 1992; Meir 1998 for
detailed comments on each claim):

(i) Infiniteness: There are an infinite number of possible locations where

index signs can be directed to. Hence they resemble more indices, since
pronouns are a closed class with a restricted number of members.
(ii) Discourse determinacy: The spatial location that is associated with a
referent is discourse-determined, rather than lexically specified. That
is, it is considered that there is nothing in the lexical specifications that
will determine to which location an index sign will be directed.
(iii) Non-ambiguity: In a fragment of discourse, referents are associated
with a specific spatial location. Hence an index sign directed to space
does not show the ambiguity that spoken language pronouns show,
since they are associated with a unique referent in a given stretch of
discourse.

The infinity issue is thus transferred from pronouns to referents and this is
a legitimate move since referents are constrained by the discourse model,
 Previous accounts 37

which can only be limited by perceptual and memory limitations, but not by
purely linguistic reasons. Hence we must accept that the number of referents
in a discourse model can be infinite. Even if these characteristics can be
applied to indices, in §7.2 it will be shown that these criteria do not exactly fit
LSC data and that, in fact, the same properties that apply to spoken languages
pronouns also apply to LSC pronouns.
The fact that index signs may be directed to the 3-dimensional extension
that signing space is and that the value for location in space is difficult to
categorise has led some authors to argue for a phonemic/phonetic distinction
in the direction that index signs may take (Janis 1992; Kooij 2002: 165).
While the phonology of index signs is considered to be an abstract point in
space, their phonetics is the actual direction (and thus the broader dimen-
sion) that signs pointing to space can get. The different directions that an
index sign towards space may have are in fact considered a gradient prop-
erty, which can be compared to the opening of vowels in spoken languages
(Rathmann and Mathur 2002; Russell and Janzen 2008).

2.5.2.2. Advantages

One of the main differences between the r-locus view and the spatial mapping
view is the conception they have of space. The r-locus view considers that
space in front of the signer and around the signed conversation is always a
linguistic construct that is only built on the basis of discourse. Linguistic
space is constructed as long as a sign conversation or monologue takes place.
Without a conversation and without the use of referring expressions directed
to it, signing space does not exist. It is in fact made evident by means of
signs directed to it. This linguistic construct has to be differentiated from
real space, where objects in reality stand and which can be perceived by our
senses. As presented in Berenz and Ferreira-Brito (1990) and Herrero (2009),
real space is the three-dimensional extent that is unlimited and continuous. In
contrast, linguistic space, which is part of real space, is limited and discrete.
In fact, the distinction between the linguistic space and the real space
can be proven by the fact that when talking about objects in the real world
they always have to be introduced into the linguistic context. Even if we
refer to objects that are present in the immediate surrounding context we
have to refer to them by pointing at the direction of the area where they are
(but of course, not at the precise point or place) in order to introduce them
into discourse. What matters most for the construction of linguistic space
is that sign interlocutors share the same coordinates in which discourse is
38 Space in sign languages: background

built. Index signs directed to objects present in the physical environment

cannot denote anything outside the linguistic context. Hence, the only refer-
ence made is towards discourse and not towards real space around the signer.
Signing space is a construction space where the conversation takes place,
and thus has to be built by the conversation participants. As previously seen
in (2), where the issue of indirect reference was raised, an index directed to
an object can only get its meaning from the linguistic context.
This book follows the r-locus view, which strongly separates real space from
linguistic space, and it considers that spatial locations established along the
discourse are the overt expression of discourse referents. A formalisation will be
given (Chapter 4), which contributes and adds new evidence to the r-locus view.

2.6. The pointing hodgepodge

Pointing signs (or index signs) have been an important focus of research since
late 70’s and different accounts have been proposed. So far, studies have
shown that although pointing signs are formally very similar, they do not form
a unique category in SLs and they may have different morphosyntactic func-
tions according to the different proposals. Yet, this hodgepodge is known under
the same label in the literature, namely “pointing”. Pointings are difficult to
analyse because of their varied distribution and functionality (Pfau 2011). They
can indicate a location (“the book over there”), they can be used predicatively
(“the book is over there”). When co-occurring with a noun they can act as a
definite determiner (“the book”) or as a demonstrative (“this/that book”). Their
distribution is quite free since they can appear both pre-nominally and post-
nominally. And they can also be used anaphorically as pronouns. Moreover,
pointing signs have been related with definiteness, determinacy and adver-
bials. Below a summary of the most relevant accounts is offered.

2.6.1. The morphosyntax of pointing signs

Concerning the syntax of pointing signs, Bahan et al. (1995) and MacLaughlin
(1997) observe systematic differences between the syntactic positions index
signs occupy. Prenominal pointings correlate with definiteness and can
express plurality, while postnominal pointings can be used both for definite
and indefinite entities, but they are not marked for plural. Postnominals that
are marked for plural are analysed by MacLaughlin as adverbials. Following
this account, Neidle et al. (2000) argue for an analysis of index signs as
 The pointing hodgepodge 39

definite and indefinite determiners when occurring in prenominal position,

and for a locative adverbial analysis when occurring in a postnominal posi-
tion. The authors believe that what we see in (4) is a construction parallel to
the construction found in Norwegian and French shown in (5).

(4) john see [man ix]. (ASL)

‘John saw a man (there).’

(5) a. den mannen der (Norwegian)

b. cet homme-là (French)
‘that man (there)’
(Bahan et al. 1995)

Another analysis establishes that pointing signs in SLs, and more specifi-
cally in SL of The Netherlands (NGT), can occupy a clause final position.
When this happens, this is considered to be a right dislocation with a pronoun
copy at the end of the sentence (Bos 1995). However, Crasborn et al. (2009)
analyse this NGT index (pointing) sign at the end as an agreement marker
with the topic of the sentence.
As for the morphology, a point of view that makes the grammatical
status of pointing signs very weak and relates them more closely to gesture
is the one of Liddell (1990, 2003). According to him a pointing sign is not
symbolic. It is indexic and its significance depends on what the pointing is
directed toward, as explained in §2.4.1. According to this account, pointing
signs are composed of two morphemes: the root, which includes handshape
and movement, and a spatial morpheme. The spatial morpheme is fully
dependent on the actual position that a present referent occupies. Since there
can be an infinite number of possible positions for referents, and thus an infi-
nite number of spatial morphemes, this led Liddell (but also Meier and some
others, as we have seen in §2.1) to argue that pointing signs are a combina-
tion of linguistic and gestural morphemes. The gestural part comes from the
infinite possibilities that spatial morphemes can have since they cannot be
integrated in a finite system. This debate is linked up to personal pronouns,
which are further treated in §2.5.3.

2.6.2. The semantics of pointing signs

Besides the works where pointing signs have been considered to be locatives
(Emmorey 2002a; Padden 1988; Shepard-Kegl 1985), they have mainly
40 Space in sign languages: background

been treated as determiners. Wilbur (1979) is, to the best of my knowl-

edge, the first work to hypothesize that the definite/indefinite distinction in
ASL may be due to the contrast between the existence and the lack of a
surface determiner. Other works have followed this definiteness hypothesis
considering that pointings directed to space are used to express definite NPs
(Ahlgren and Bergman 1994, for Swedish SL; Bahan 1996; MacLaughlin
1997; Wilbur 2008 for ASL; Tang and Sze 2002, for Hong Kong SL). In fact,
according to Tang and Sze (2002), in HKSL definite reference is marked
with eye gaze directed to a spatial location, while indefinite reference is
marked with eye gaze directed to the addressee. This analysis contrasts with
Neidle et al. (2002), who claim that head tilt and eye gaze are abstract agree-
ment features in ASL. Going beyond the definiteness debate, Bahan (1996)
and MacLaughlin (1997) state that non-specific referents are not associated
with any fixed location in the signing space. In these cases, a special loca-
tion in front of the signer’s body “and with a little higher than normal end
point associated with object spatial location, is used to express agreement
with a non-specific object” (Bahan 1996: 105). Hence, according to these
authors definiteness is marked in ASL with an index sign directed to space,
and indefiniteness is marked with an upward sign. However, as we will see
in Chapter 5 definiteness distinctions concerning signs directed to space are
not found in LSC.
Other works conclude that pointing signs in ASL and Italian Sign
Language (LIS) function as a type of determiner used to specify the noun
(Bertone 2007, 2009, for LIS; Zimmer and Patschke 1990, for ASL). Zimmer
and Patschke (1990) found many instances in which a noun being mentioned
for the first time does occur with a determiner; hence they avoid relating it
with definiteness. They also did not find instances of pointing co-occurring
with generic nouns nor with abstract nouns (according to them, in ASL they
happen to be ungrammatical with signs such as concept or theory). They
did not identify a specific marking for the distinction between definite and
indefinite. The authors explicitly say that their conclusion is determined by
informants’ comments about pointing serving to describe only specific enti-
ties and they do not provide any further description or analysis of specificity
marking. Chapter 6 is devoted to the description and analysis of specificity
marking in LSC and it will be shown that specificity distinctions are clearly
marked in LSC signing space.
A different view is that space is used to express topicality (Engberg-
Pedersen 1993; Winston 1995). To refer to topicality Engberg-Pedersen
(1993: 99) uses the term “thematicity”, which she defines as the number of
repetitions of the nominal indicated with a pointing sign. Thus, when the
 The pointing hodgepodge 41

signer refers repeatedly to an individual this is marked with a pointing sign

directed to a spatial location. The problem is that Engberg-Pedersen (1993)
does not make a distinction between first and further mentions. She only
deals with localisation in space and collapses both the establishment of the
location (and the corresponding association) and the reference back to it.
She attributes the feature of topicality to space. That is to say, she ascribes to
space the capacity of referring back to entities (co-reference ability), which
I consider only to be part of the co-reference mechanism. First mention is
usually related with semantic attributes such as (non-)specificity and (in)defi-
niteness, and further mentions with co-reference. As she does not distinguish
between first and further mentions, she cannot attribute semantic attributes to
space, only functional attributes, namely that of co-reference (or thematicity,
following her terminology). Also definiteness is in many cases the product
of subsequent mentions (see Chapter 5). There is thus a strong connection
between definiteness and co-reference that Engberg-Pedersen seems to
neglect. As will become clear in Chapter 6, this issue is better defined in
terms of specificity.

2.6.3. Sign language pronouns

Sign Language pronouns have received a great amount of attention in the

literature (see Cormier 2012 for an overview). They are a contentious issue
and a matter of debate concerning its spatial nature and the person informa-
tion they encode. Pronouns have a spatial nature because they are always
directed to an area of the signing space. First person pronouns are directed to
the chest of the signer; second person pronouns are directed to the addressee,
and third person pronouns are directed to the area in the space associated with
that referent in case of non-present referents. As for second person and third
person present references, they are always dependent on the location that the
person occupies and according to some researchers they do not seem to show
linguistic distinctions. The multiple and even infinite number of locations
where pronouns can be directed to turn the integration of these infinite loca-
tions on a discrete linguistic system into an impossible task. This is the reason
why, according to some analyses, spatial locations are considered to be not
phonologically specifiable and to belong to the realm of gesture for second
and third (person) elements (see Liddell 1990 and a series of subsequent
publications; Lillo-Martin 2002; Meier 1990; Meier and Lillo-Martin 2010,
among others). In contrast, the location for first person pronouns is fixed,
since the pointing is always directed to the chest of the signer (Meier 1990).
42 Space in sign languages: background

The so-called “infinity problem” has led some researchers to propose an

analysis against the traditional account based on a distinction between first/
non-first pronominal distinction (Engberg-Pedersen 1993; Meier 1990; Lillo-
Martin 1995, inter alia). The different perspectives on the person features
that SL personal pronouns encode can be divided among five groups, which
are defined and described below.

(i) Three-person distinction

The first analyses of SL pronouns consider that the same features described
for spoken language pronouns could be applied to visual-gestural
languages (Berenz and Ferreira Brito 1990; Friedman 1975; Padden 1988;
Sandler 1989). However, this traditional view was discussed in the late
eighties/beginning of the nineties, when signing space became the main
focus of attention and it was questioned in analyses that argued for a two-way
distinction (see (ii) below). A second wave of research appeared in reac-
tion of the two-way distinction analysis (Alibašić and Wilbur 2006; Berenz
1998; Meurant 2008; Neidle and Lee 2006). The three-person distinction
encoded in SL pronouns is shown by including the nonmanual component
in the analysis. Berenz (1998, 2002) argues for the existence of first, second
and third person pronouns distinction. She presents the Body Coordinated
Model (BCM), which is used to analyse second and third person pronouns.
In the BCM four coordinates, namely eye gaze, head, handshape and chest
are aligned. In the case of pronominal reference to second person the
angle of the four coordinates will line up along the midline of the signer’s
body and they will all be directed to the addressee. In case of third person
pronouns, disjunction of some of the coordinates will occur, and at least
one of the coordinates will not be aligned. Also reference to third person is
made with a briefer eye gaze than when directed to second person. Alibašić
and Wilbur (2006) base their analysis on Berenz’s BCM and also argue for
a three person distinction in Croatian SL. They conclude, though, that the
chest is not a reliable feature.
In a different spirit, Neidle et al. (2000) consider ASL locations to be the
overt manifestation of phi features related to pronouns and verbal agree-
ment. Neidle and Lee (2006) argue for a formal distinction between second
and third person by analysing the head tilt. Although in both cases, the
head moves in the direction of the phi-location, the salient part of the head
involved in the movement is different. For third person it is the temple, and
for second person it is the centre of the forehead. Finally, eye gaze has also
 The pointing hodgepodge 43

been analysed as a nonmanual marking to distinguish among the three gram-

matical persons (Meurant 2008, for Southern Belgium SL).

(ii) Two-person distinction

Meier (1990) is the first author to argue for a first/non-first person distinc-
tion in ASL. The main argument comes from the impossibility to distin-
guish the features of second and third person pronouns because they are
both directed to space and they depend on the actual location of the referent
referred to. That is, the actual direction of the index sign can be the same in a
context referring to a second person and a context referring to a third person,
depending on the position the reference being pointed at is. Other authors
have followed Meier’s claim for different SLs, such as Engberg-Pedersen
(1993) for Danish SL, Lillo-Martin (1995) for ASL, Meir (1998) for Israeli
SL, Rathmann (2000) for German SL, and Smith (1990) for Taiwanese SL.

(iii) No person distinctions

Ahlgren (1990) for Swedish SL and McBurney (2004) for ASL consider that
there is in fact no grammatical category of person encoded in the index signs
directed to space because they primarily function as demonstratives rather
than personal pronouns. Demonstrative index signs are used to deictically
identify referents in a discourse by their location and they do not encode
semantic notions of first and second person reference. Rather, they only
localise entities in signing space.

(iv) One single pronoun

On a very different analysis, Lillo-Martin and Klima (1990) argue for a unique
pronominal form in ASL. This work argues that ASL pronouns are expressed
by a unique pronominal form that goes with a referential index that is overtly
manifested. The authors’ hypothesis regarding ASL pronouns is as follows:
“Pronouns marked with an r-locus represent the physical sign pronoun directed
toward the r-locus “a”. This sign is interpreted with respect to the discourse
referent assigned to it (for example xj). In this theory referential signs (such as
pronoun signs and other indexed nominals) are interpreted as pairings of the
sign with a discourse referent.” (Lillo-Martin and Klima 1990: 199).
44 Space in sign languages: background

(v) Spatial pronouns

A completely different analysis considers that points in space are indeed SL

pronouns (Lacy [1974] 2003; Kegl [1976] 2003). Lacy considers the manual
handshape to be an indicator of the spatial location, while Kegl considers
the handshape to be an agreement marker.

2.7. Acquisition and emergence of new sign languages

This last section is a very short note on two claims related to the acquisition
of signing space and to the emergence of new SLs. The main goal is to briefly
provide some evidence for the linguistic status of signing space and to show
that signing space is part of the grammatical structure of the language.
Focusing on acquisition, deaf children acquire the ability to direct agree-
ment verbs and index signs towards objects that are present in the physical
context by age 3;0 to 3;6 (Emmorey 2002a). But the ability to direct verbs
towards locations in signing space takes a longer acquisition route. The
process of referring to a non-present entity faces some difficulties and in
order to be successful the deaf children must learn to a) associate a referent
with a spatial location; b) use different locations for different referents; c)
use verb agreement or pronouns with non-present referents; d) remember the
association of referents with locations over a fragment of discourse (Lillo-
Martin 1999). The most common errors are, for instance, using one spatial
location for several different referents or using inconsistent locations for a
single referent (Bellugi et al. 1990). They also go through different stages
of acquisition as argued in Lillo-Martin (1999). They respect the syntactic
restriction that null arguments must be identified. They initially direct agree-
ment verbs towards objects that are present without lexically specifying the
arguments. Then, they go through a stage in which they use overt arguments
with unmarked verbs and so they do not use unidentified null arguments with
unmarked verbs. Finally, they correctly direct agreement verbs towards loca-
tions and use null arguments (however, cf. also Quadros and Lillo-Martin
(2007) where it is argued that children may also use directionality in gestures
together with agreement).
By the age of 6 the cross-sentential use of pronouns and agreeing verbs
appears to be firmly acquired. This late acquisition is due, as explained
by Lillo-Martin (1999), to non-linguistic cognitive factors, such as spatial
memory. The relevant morphosyntactic principles are mastered by about
the age of 3 but the children have difficulties establishing and remembering
 Acquisition and emergence of new sign languages 45

unique associations between discourse referents and locations. This shows

that deaf children also undergo specific stages while acquiring spatial syntax.
Some more evidence of the linguistic use of signing space comes
from the emergence of two new SLs. One piece of evidence comes from
Nicaraguan SL (ISN), a language birth that could be monitored. The other
comes from Al-Sayyid Bedouin SL (ABSL), a language that developed in a
stable community without influences from other languages, either signed or
spoken. Let us start with the monitored one. The recent emergence of a new
SL in Nicaragua, which could be documented by linguists, allows observing
the very early stages of a language and the steps that pidgins follow until they
are transformed into a full-fledged language. The Nicaraguan Deaf commu-
nity and its language came into existence only since 1979, when a Deaf
school was created in Managua and deaf people from all over the country
would gather together there (Kegl et al. 1999). This had as a result that the
first form of communication (i.e. a pidgin) and its evolution could be moni-
tored and studied.
People entered the school at different moments, and this population was
divided into three cohorts that correspond to the moment of accessing it.
Senghas and Coppola (2001) studied whether spatial modulations (that
is, altering the direction of the signs movement to a non-neutral location)
are indeed in the process of emerging as a grammatical device in ISN. The
results of this study indicate that child learners create ISN, rather than being
a reflex of the first language that was created at the beginning. Their analysis
shows that spatial modulations are more frequent in the signing of early-
exposed signers of the second cohort than among early-exposed signers of
the first cohort. This indicates that the second cohort did not reproduce the
language as it was produced by their first-cohort elders; rather, they changed
the language they learned. Another study by Senghas and Coppola (2001),
which focused on the indication of shared reference and on the fluency
of spatial modulations, also concluded that the youngest members of the
second cohort surpassed their input and established a partially developed
language and systematized it in a specific way. The second cohort reanalysed
the location of spatially modulated signs as indicating something akin to
co-reference. In fact, the increase in the use of deictic signs as they progress
from homesigners through the first three cohorts of ISN signers is relevant.
This increase is due entirely to an increase in non-spatial, referential uses of
deixis, i.e. non-locative (Coppola and Senghas 2010). Indeed, according to
these authors there is a decrease in the spatial meaning attributed to deictic
forms in parallel to their increasingly grammatical uses by ISN signers. The
crucial aspect in the transformation of index signs from pointing gestures
46 Space in sign languages: background

to forms that function grammatically is their loss of their locative function.

They also observe an increase in the production of index signs that refer to
entities rather than to locations, such as places, which also indicate co-refer-
ence relations. Hence, Coppola and Senghas (2010) argue that among the
three cohorts studied and over the span of 30 years a clear distinction is made
between a simple concrete deictic gesture intended to draw attention to a real
world object, and an abstract index sign directed to the empty space which
serves a particular linguistic role in the sentence.
Let us now move to the Negev region of present-day Israel where the
Al-Sayyid Bedouin group was founded about 200 years ago. The group,
considered by outsiders as Bedouins, is now in its seventh generation and
contains about 3,500 members. Within the past three generations, approxi-
mately 150 individuals with congenital deafness have been born into the
community, since consanguineous marriage has been the norm so far
(Aronoff et al. 2008). This unusual situation has led to a situation where
the signing community of Al-Sayyid is actually much larger than the actual
number of deaf members. This community presents a unique situation of
a language that developed in a stable community, without influences from
spoken or signed languages from outside.
According to Aronoff et al. (2004) and Meir et al. (2007), the clause
structure of ABSL is not based on the use of space, but rather on the order of
signs. When looking at predicate forms, no evidence of morphology marking
person was found. Predicates where some movement is included are extended
from the signer’s own body outward and inward (on the horizontal plane, see
§2.3.2) to a small degree. Interestingly, a difference among groups of signers
was found: younger signers show a greater use of space. While older signers
prefer to use to a lesser extent this proximal/distal distinction in front of the
signer’s body, younger signers make more use of it (almost three times more)
than older signers, thus reflecting a greater use of space in younger genera-
tions (Padden et al. 2010). This shows that even if this emerging language
might utilize little space in older generations, the incorporation of space into
the grammar takes time, as the use by young generations shows.

2.8. Proposal

This book claims that signing space should be analysed as a categorical

element and very rarely as a gradual continuum. A distinction between the
linguistic space and real space needs to be argued for. The former is a limited
and discrete construct built among the conversation participants, whose
 Summary 47

different parts in spatial planes are categorically interpreted. The latter is the
space where objects in reality are localised which is perceived by our senses.
Signing space is part of the discourse grammar of the language and once
thoroughly analysed, space is more similar to a linguistic system than to the
realm of gesture. It is undoubtedly when we enter the discourse level that
spatial locations become distinctively important. As will be shown along the
coming chapters, referring terms and other linguistic expressions directed
to signing space display a complex and dynamic use of linguistic space.
However, what is unusual in SLs is the greater potential for expressing refer-
ential distinctions, and the fact that in SLs discourse referents are overtly
manifested, which is an unavoidable imprint of the visual-spatial modality.
This book argues for a three-person pronominal distinction in LSC.
However, it also proposes that points in space are clitics attached to the
manual form (similar to the spatial pronouns analysis). I do not consider it
to be a full pronoun but rather a clitic. Although the consideration of seeing
spatial locations as pronouns is a very interesting one as pointed by Kegl
([1976] 2003), it has an important drawback: if we consider that points in
space are pronouns, and the actual pointing sign is only an agreement marker
or a dummy element, we would need to accept that this dummy element is
a very complex one that allows numeral and plural incorporation. Seen the
complexity of manual pointing signs which can incorporate number and the
direction of the movement varies according to the number incorporated, they
can be only analysed as pronominal forms which are cliticised to a clitic
spatial morpheme (see §3.1).
Finally, I follow the r-locus view, which considers that r-indices are
overtly expressed. However, this claim has not been thoroughly formalised
under a theoretical framework26, and this is precisely the main goal of this
book. As will be proven here, the establishment of spatial locations correlate
with a semantic phenomenon, namely that of scope. The analysis offered
here proves that features like specificity and topicality can be attributed to
spatial locations.

2.9. Summary

This chapter has presented a state of the art about the theme of this book,
namely the use of signing space in SLs. It has described what this use of space
consists in, and it has presented the relationship between the use of space and
the modality of languages. The two main opposing views concerning the
analysis of space have been contrasted, clearly favouring the r-locus view.
48 Space in sign languages: background

However, the lack of formalisation of this view has been noted and it is
indeed one of the aspects this book aims to cover. The different syntactic
and semantic analyses attributed to pointing signs have also been presented.
Some provision of evidence of the linguistic status of space coming from
acquisition and from cases of emergence of new SLs has also been included.
The chapter has concluded with the proposal defended herein.
Chapter 3
A morpheme on spatial planes

The structured use of space in ASL is nowhere more evident than in the
means by which verbs reflect their arguments [...]. What is reflected is not
objectively who or what is referred to by the subject of the verb but rather
whether in the discourse situation the subject is the speaker, the person or
persons addressed, or some subject of the discourse, not restricted to the par-
ticipants in it.
Klima and Bellugi (1979: 276)

3.1. Introduction

This chapter is devoted to the description and analysis of the use of signing
space by focusing on non-descriptive locations. Unlike descriptive locations,
which make a freer use of space, non-descriptive locations are categorically
articulated in the different areas within the three spatial planes that are stand-
ardly projected with respect to the body of the signer. Signs directed to the
different parts of spatial planes contribute to the establishment of a gram-
matical morpheme that consists in an abstract point in space. The direction
where the physical point in space is established is completely irrelevant.
What is relevant is the fact that a point is chosen at all. This spatial point is
categorically defined and interpreted within the linguistic system (Wilbur
2008). I argue that this abstract point in space functions as a clitic pronoun
(Fischer 1975). In Catalan Sign Language (LSC), this clitic pronoun can be
abstractly established in different parts of the three spatial planes. Yet, only
the two directions of the frontal plane, namely upper and lower, are gram-
matically relevant and distinctively interpreted. [lower] is the default feature
which the majority of signs are attached to. In contrast, the upper direction
of signs is instantiated by the feature [upper], which is the marked location
established on the upper frontal plane denoting particular meanings, such
as locatives, hierarchical relations, non-specificity and non-presence in the
immediate physical context. Moreover, the two directions on the horizontal
plane, namely [ipsilateral] and [contralateral], are discursively relevant in
that they codify contrastive topics. Overall, the chapter provides a unified
treatment of different linguistic phenomena that have been hitherto described
50 A morpheme on spatial planes

separately in the sign language literature, namely hierarchical relations, loca-

tives, contrast, as well as other aspects that have not yet been the subject of
research neither in LSC nor in other signed languages, such as the encoding
of specificity. As shown, the features found in two of the three spatial planes
as described in the phonological literature are also relevant beyond the
sentence level and they serve distinctive discourse functions.
This chapter is structured as follows. §3.1 presents the analysis that this
book follows in considering the spatial morpheme localised on the planes to
be a clitic. §3.2 presents the set of mechanisms employed in LSC to establish
this spatial morpheme and hence to localise entities in space. §3.3 is a detailed
description of the three spatial planes used in non-descriptive locations in
LSC. After this overall description, §3.4 outlines the matrix of features that
spatial locations incorporate. §3.5 briefly expands on body-anchored locali-
sations. §3.6 summarises the main findings of the chapter.

3.2. The spatial morpheme

A question that has concerned linguists from the beginning of sign linguis-
tics is what exactly is there in space that allows to direct index and localised
signs to it (see §2.4 and §2.5). I defend that index signs and other local-
isation mechanisms establish a spatial location, which is in fact a spatial
morpheme attached to manual and nonmanual signs. This idea has been
already outlined by previous works, which have influenced and inspired the
shaping of ideas of the present book. As for SL agreement, Fischer (1975)
is, to the best of my knowledge, the first work to argue that points in space
are cliticised pronominal forms attached to verbal roots. This view differs
from the traditional and general idea that spatial locations are agreement
markers of verbal inflection (see §3.2.3). Nevins (2009, 2011), following
Fischer (1975), claims that agreement verbs are formed by morphemes that
are cliticised to the verbal root. In syntactic terms, this view imposes a clitic-
doubling analysis when the arguments of the clause are overt, as defended in
Quer (2009) and Koulidobrova (2010). More specifically, Quer implements
the big-Determiner Phrase (DP) hypothesis (Uriagereka 1995), whereby the
DP and the clitic are generated as a single argument.
As for pronominal forms, Kegl (1976) argues that points in space precisely
function as pronouns in ASL. She considers the manual index handshape
pointing towards the spatial location to be an agreement marker indicator.
Some years later, in her book and a paper based on it (Shephard-Kegl 1985;
Kegl 1986), she argues that spatial locations are more specifically clitic
 The spatial morpheme 51

pronouns, which can be distinguished into two classes. Proclitic pronouns

are established by role prominence and are always connected to the subject.
Enclitic pronouns are established by coindexation and they are in comple-
mentary distribution with full pronouns. Unfortunately, Kegl did not carry on
this view in further research. In fact, the idea of analysing spatial locations as
clitic pronouns is not so far the mainstream view.
In a mixed type of analysis, Padden (1990) considers the spatial modi-
fication expressed on plain verbs to indicate either subject or object of the
predicate to be a pronoun clitic. According to her, this differs from verb
agreement articulated through verb inflection that agreement affixes repre-
sent. An important argument in her analysis is that the pronoun clitic occur-
ring with plain verbs is not very restrictive, while person and number are
highly restricted as they appear only with agreement verbs. This book follows
Fischer (1975), Kegl (1976, 1986), Shephard-Kegl (1985), Padden (1990),
Quer (2009), and Nevins (2011) in considering that spatial locations function
as clitic pronouns. The spatial morpheme can appear across different catego-
ries, such as index signs, spatially modified signs, as well as verb inflection.
According to Zwicky and Pullum (1983), this is a strong argument for a
morpheme to be considered a clitic.
The establishment of a spatial morpheme form in signing space has been
questioned and often bypassed by many works. The problem with spatial
locations is that researchers have focused primarily on the physical point
in space where index signs are directed to (see the spatial mapping view, as
described in §2.4.1), thus neglecting the fact that what matters is not the exact
direction in space but rather its categorical interpretation in the linguistic
system. What is important is the fact that a point is chosen at all. Following
Wilbur (2008), I consider that the spatial direction where index signs can be
directed to consists in an abstract and unique point in space. What matters
is not the exact direction where index signs are directed to, but rather the
abstract end point that is expressed with the localisation of signs and that is
interpreted in the grammar of the language as a categorical element, rather
than a gradient one. Wilbur (2008) clearly defines it as follows:
“The morpheme is not ‘this particular point in space where the sign move-
ment or indicator pointing just stopped’; rather it is the geometric point in
space (p), which indicates an individual (x), no matter where it is made in
space.”
(Wilbur 2008: 239)
Wilbur considers this morpheme to be a geometrical point that has morpho-
logical expression via agreement morphemes and end state. The relevance of
52 A morpheme on spatial planes

her account is that she makes precise that the direction where the physical
point in space is established is completely irrelevant. What it is important
is that the spatial point is categorically defined and interpreted within the
linguistic system. The infinity issue (see §2.4) is thus no longer a problem,
since there is one and only spatial morpheme. A unique spatial morpheme
exists in the grammar of the language, which consists in an abstract point
where indexical signs and other localisation mechanisms are oriented to.
I consider this final end point, represented as (p) for point in space, to be
a grammatical clitic morpheme which stands for the overt manifestation of
a discourse referent (see Chapter 4). It is an abstract spatial morpheme that
is cliticised to the manual handshape as well as to nonmanualarticulations.
Moreover, it is an invariable spatial morpheme, regardless of the direction
of the index sign (see Chapter 7). The so long considered underspecifica-
tion slot for the location feature (Brentari 1998; Kooij 2003; Sandler 1989,
among others) is here taken to be filled in by the abstract clitic (p). The
matrix of features that index signs include is illustrated below, where every
Greek letter corresponds to a certain feature. Importantly, the location slot
has a concrete feature, namely (p), which also determines the orientation
parameter of the sign.

(6) Index sign

– Handshape: α
– Movement: γ
– Orientation: β(p)
– Location: point in space (p)

As shown in the subsequent sections, this abstract point in space is established

by different localisation mechanisms on the three spatial planes in front of
the signer’s body. However, only some areas within these spatial planes are
grammatically relevant in LSC. Before delving into the description of spatial
planes, the localisation mechanisms used in LSC are presented.

3.3. Localisation mechanisms

In Chapter 2, we saw that sign language (SL) discourse referents are asso-
ciated with an area in signing space. An index sign, followed or preceded
by a nominal, indicates that from that moment on the area the pointing is
directed to will be associated with the referent the nominal denotes, as long
as the referential framework is not shifted. An index directed to a location
establishes thus a discourse referent (DR) on a determined spatial area.
 Localisation mechanisms 53

Agreement verbs, index signs or eye gaze directed to that location in subse-
quent discourse are understood as coreferential with the corresponding DR.
Example (7) illustrates this. It is an LSC discourse fragment where the
signer is talking about his son. The first time he talks about him, he utters the
nominal sign for son and directs an index sign to the contralateral part (i.e. the
left area in a right-handed signer, see §3.2.1).27 This is the first mention of the
DR “son” and the index associates the nominal with the contralateral part of
signing space. In the second utterance in (7), two index signs are directed to
the same area and are therefore understood as coreferential.

(7) ix3c laptop 1-offer-3 son ix3cl for new 3-select-3 work ix3cl
need laptop ix3cl.
‘I will offer this laptop to my son, because he has been selected for a
new job and he needs a laptop.’

Figure 12 (previously shown as Figure 8, but repeated here for conveni-

ence) contains a sequence of the two index signs appearing in (7). The stills
in Figure 12a correspond to the nominal sign for son and the index sign to
localise it. As we can observe in Figure 12b, corresponding to an index sign
in the second sentence in (7), further mentions of the DR are expressed with
a pronominal index sign directed to the same area first established.

a. son ix3 b. ix3

Figure 12. First and further mention of a localised discourse referent

In the initial literature on sign language research, the process of establishing

a point in space or pointing to a previously established place to refer to a
person, object or location was known as indexing, or nominal establishment
(Friedman 1975; Mandel 1977; Klima and Bellugi 1979; Poizner, Klima,
and Bellugi 1987). The broad term “indexing” included the first associa-
tion of the nominal with an area in space (i.e. first mention), as well as all
54 A morpheme on spatial planes

those instances used to refer back to the same DR denoted by the nominal
(i.e. further mentions). Indexing is then used as an umbrella term within
which two functions are included: the predicational and the anaphoric one.
As seen in §2.4, location and locus are very widely used terms. Its
definition may vary according to the theoretical framework in which it is
considered. From a spatial mapping perspective, it is defined as points in
space standing for a projection of the referent (Engberg-Pedersen 1993: 97).
However, the r-locus view does not consider this point in space to be such
a projection (see §2.4.1.2 for arguments against the spatial mapping view).
Rather, the area where the fingertip of an index sign points to is a gram-
matical morpheme, which is semantically linked to a DR (see §3.4 below).
In order to keep the terminology clear, the terms and the definitions used in
this book follow below.

(8) Definitions
Localisation: to direct an index sign or spatially modify a sign towards
a spatial area. A point in space is chosen, though it doesn’t matter which
point. By means of this indexing, a DR is established in the spatial area
and gets thus associated with it. It functions both as introduction of the
DR and as anaphoric reference.
Location: spatial morpheme semantically associated with a DR. It is
represented as (p).

The association between a DR and a location may be done by different

means. The set of linguistic mechanisms used in LSC to establish (p) are the
following:28

(9) Localisation mechanisms in LSC

Manual:
– Index signs
– Spatial modification of signs
– Verb agreement

Nonmanual:
– Eye gaze
– Body lean
– Head tilt

Most frequently these mechanisms do not occur alone, but rather combined.
In Figure 13 the signer establishes a DR in signing space by means of a
combination of mechanisms: spatial modification of the plain verb search
 Localisation mechanisms 55

and head tilt are oriented towards the lateral part where the corresponding
DR gets established. In what follows, a subsection is devoted to each mecha-
nism. A state of the art literature is first offered and the LSC particular char-
acteristics are then described on the basis of a qualitative analysis of the
small-scale LSC corpus.

3.3.1. Index signs

Index signs consist in an index finger handshape (fist closed, index extended)
directed to an area in space (Figure 14) (however, see Fenlon et al. 2013
for assimilation processes in the index handshape from neighbouring signs).
They have been described for many SLs and constitute the mechanism most
often referred to from the set listed in (3) (Berenz 1998, 2002, for Brasilian
SL (LIBRAS); Bergman 1982, for Swedish Sign Language (SSL); Bos
1990, for Sign Language of the Netherlands (NGT); Engberg-Pedersen
1993, for Danish Sign Language (DSL); Pfau 2011 for German SL (DGS);
Friedman 1975, McBurney 2002, Neidle et al. 2002, Padden 1988, Zimmer
and Patschke 1991, for American Sign Language (ASL); Zeshan 2000, for
IndoPakistani Sign Language (IPSL); Quer 2004, for LSC, among others).

Figure 13. Localisation mechanisms Figure 14. Index sign

In LSC index signs may function as determiners (10a), demonstratives (10b),

pronominals (10c), and possessives (10d) (Quer 2004; Quer and GRIN 2008).
The same manual handshape can function differently. Importantly, it is
always spatially modified and thus localises the nominals that co-occurs
with, in the case of determiners, demonstratives and possessives, or the DR
that picks up, in the case of pronouns. The spatial modification is indicated
with the subscripts in the examples below.
56 A morpheme on spatial planes

(10) a. ix3a book interesting.

‘The book is interesting.’
b. While referring to a present book
book ix3a name pyjama cl.stripes.
‘The name of this book is “The boy in the striped pyjamas”.’
___br
c. ix3a, ix1 like.
‘I like it.’
_______________br
d. delfina ix3a book interesting.
‘Delfina’s book is interesting.’

Besides pronominal index signs, localisation in LSC can also be expressed

with the sign that consists in a derived form of the lexical noun person.
This sign is articulated with a baby-C handshape and a vertical downward
movement (Figure 15a). It functions as a pronominal index which can be
coreferentially used for the three person distinctions. This sign has been char-
acterised by Meir (2003) for Israeli Sign Language (ISL) as a pronominal
form of case marking. Pronominal index signs and the person-3 sign can be
indistinctly used without affecting the propositional meaning.29 An impor-
tant difference, though, is that while person-3 can only denote [+human]
entities, index signs can denote any kind of entity and it is not restricted to
[+human]. This sign has an emphatic variant that consists in a bimanual B
handshape articulation with the palm of the hands facing each other and a
downward movement (Figure 15b). Importantly for the present account, both
index and person signs establish (p).

a. C-handshape variant b. B-handshape variant

Figure 15. person sign
 Localisation mechanisms 57

3.3.2. Spatial modification

Signs are not always signed neutrally, but are very often spatially modified
as well, when they are not body-anchored (Baker and Cokely 1980; Shepard-
Kegl 1985). Spatial modification of a sign consists in signing the corre-
sponding sign in a non-neutral location in space. That is, not in neutral space
in front of the chest of the signer, but rather towards the ipsilateral (Figure
16a) or contralateral part (Figure 16b). This spatial modification establishes
(p) in a lateral part of signing space.

a. Plain verb: disappear b. Classifier

Figure 16. Spatial modification of signs

In LSC common and proper nouns, determiners, plain verbs and classi-
fiers can be spatially modified. In Quebec Sign Language (LSQ), Rinfret
(2009: 220) finds that the strategy of spatially modifying the signs is used
differently according to the type of data. The author argues that in elicited
data spatial modification of signs is the mechanism more frequently used
to localise entities in space, followed by index signs, body lean and finally
eye gaze. In contrast, in spontaneous data spatial modification of signs is the
least used mechanism, and signers prefer to localise with eye gaze, followed
by index signs and body lean before using the spatial modification. No such
difference has been found in the LSC data and spatially modification of signs
is a strategy equally used in elicited and spontaneous data. However, further
research with different kinds of data is needed.

3.3.3. Verb agreement

Verb agreement in SLs is also dependent on the use of space. A verb agreeing
with subject and object is directed towards the spatial location established
58 A morpheme on spatial planes

with these referential functions (Janis 1995; Mathur 2000; Padden 1988;
Zwitserlood and van Gijn 2006, among others). Agreement is marked with
the direction of the movement, palm or fingers orientation and, according to
some studies, with nonmanual markers. The direction of these mechanisms
indicates coreferential binding with the arguments of the predicate. The three
verbal classes identified by Padden (1988) are plain, agreement and spatial
verbs, which are distinguished among them by affixes, as defined below:

(i) Plain verbs are not inflected for person or number. Agreement is ex-
pressed by means of personal pronouns or the auxiliary agreement sign.
(ii) Agreement verbs are inflected for person and number and the articula-
tion of the verb moves from and towards two areas in space to indicate
the arguments of the predicate. They are divided into two main groups:
regular, in which the path is from subject to object (Figure 17a), and
backwards, where the path is from object to subject (Figure 17b).
(iii) Spatial verbs agree with spatial locations, although in this case inflec-
tion indicates a locative argument or adjunct, that is where the refer-
ent identified with the predicate is located or moves from/to in space.
Verbs of location and movement represented by classifiers are included
within this broad group.

a. Regular agreement verb: support b. Backwards agreement

verb: attract
Figure 17. Agreement verbs

While this classification has a syntactic motivation according to Padden

(1988), a semantic account has been defended claiming that for both agreeing
verbs and backward verbs movement proceeds from the Source argument
towards the Goal of the action (Meir 2002). However, other authors convinc-
ingly argue that this semantic analysis cannot be valid for SLs. A strong argu-
ment comes from the syntactic behaviour of auxiliary signs, which always
 Localisation mechanisms 59

consists in a movement from subject to object, irrespective of the theta-role

they bear (Pfau, Salzmann, and Steinbach 2010; Quer and Quadros 2008).
Also, this tripartite classification has been put into question arguing for a
plain vs. non-plain verb distinction (Janis 1995; Quadros 1999; Quer and
Quadros 2008). It is difficult to maintain whether spatial and agreeing verbs
form separate categories since they agree with locative arguments and
personal arguments and this is often indistinguishable on the surface. The
reader is referred to Quer and Quadros (2008) for fresh evidence coming
from LSC and Brazilian Sign Language (LSB).
Manual movement in signing space is not the only indicator of verb
agreement. It has been argued for ASL that besides manual agreement, the
nonmanual component also expresses inflection, and this is explained in the
next subsection.

3.3.4. Nonmanual mechanisms

The nonmanual component provides very important mechanisms to localise

entities in space. It is responsible for localising nouns and plain verbs, as
well as marking inflection of agreement verbs in ASL under certain analyses.
Concerning verb inflection, it has been claimed that agreement in ASL can
also be expressed nonmanually. According to Aarons (1994), Bahan (1996),
MacLaughlin (1997) and Neidle et al. (2000), transitive constructions are
marked with head tilt and eye gaze: the signer’s head may tilt in the direction
of the location for the subject, whereas his eyes may gaze in the direction of
the location for the object. According to these authors, headtilt is a nonmanual
manifestation of subject agreement, and eye gaze is a nonmanual manifesta-
tion of object agreement. This nonmanual agreement is not obligatory. In a
later study, Neidle and Lee (2006) opt for a reconsideration of the function
of head tilt. Since one unresolved puzzle was the seeming optionality of this
marking, Neidle and Lee argue that head tilt is in fact a focus marker that
incorporates expression of subject agreement. Hence, its occurrence depends
on the information structure of the sentence, and when it occurs, it serves to
mark both focus and subject agreement.
Independently of the complications in the description of nonmanual
agreement (see Sandler and Lillo-Martin 2006: 43), head tilt, eye gaze and
body lean30 serve as mechanisms to indicate that a nominal is established and
associated with a certain location in signing space. Baker and Cokely (1980)
refer to “eye-indexing” when eye gaze can suffice as a strategy for localising
a DR in space. Eye gaze directed to space consists in a break in the eye gaze
60 A morpheme on spatial planes

directed to the addressee, to redirect it to a specific location in signing space.

But nonmanual strategies are not only confined to eye gaze, but also to body
lean and head tilt (Rinfret 2009 for LSQ).

a. Head tilt and eye gaze b. Head tilt

Figure 18. Nonmanual mechanisms of localisation

Most frequently, nonmanual mechanisms occur in combination with manual

ones. Figure 18a shows localisation with a combination of two nonmanual
mechanisms, namely head tilt and eye gaze, co-occurring with the sign
person-3. Figure 18b shows localisation with the spatial modification of
the sign person-3 and head tilt co-occurring with it. The next section is a
description of the spatial planes and their use in LSC discourse.

3.4. Non-descriptive use of spatial planes

As seen in §2.3.1, descriptive locations are used to express spatial rela-

tions among objects and they are not restricted to specific areas within the
three-dimensional space. They are motivated by a mapping with the situa-
tion described and they are represented by meaningful locations. This is the
reason why their use of space is freer. In contrast, non-descriptive locations
are localised arbitrarily with respect to the situation described. They identify
the arguments of the verb and they are categorically defined as occurring in
the three spatial planes projected with respect to the body of the signer. In this
section the spatial planes used in non-descriptive locations, briefly presented
in §2.3.2, and the features contained within them are treated in detail. The
three spatial planes, already shown in §2.3, are graphically reminded below.
 Non-descriptive use of spatial planes 61

a. Horizontal (x) b. Frontal (y) c. Midsaggital (z)

Figure 19. Spatial planes

The current proposal of the planes is based in Brentari (1998), and the major
features distinctions are presented according to Liddell and Johnson (1989)
and Sandler (1989). These features are applied and extended to the LSC
discourse data, and a whole section is devoted to the specialised use in LSC
discourse of each plane, namely horizontal, frontal and midsaggital.31
The geometrical units in which space may be divided into are points, axes
and planes. Points are zero-dimensional elements, which intersect with the
three spatial planes. Axes are one-dimensional lines, which consist in a set
of points whose coordinates satisfy a given linear equation. Finally, planes
are a set of points, which extend in a two-dimensional area. Although points
and axes have been previously used to analyse pronominal and agreement
verbs in signing space (Padden et al. 2010; Wilbur 2008), the present work
uses the notion of spatial plane because it allows focusing on the different
features contained within each two-dimensional area. In mathematics, planes
are defined according to two perspectives: (i) in terms of the planes where a
point intersects (i.e. a point intersects at a position on the three planes, namely
x, y, z); and (ii) according to all the points contained on the specific plane.
Although perspective (i) is very relevant when dealing with spatial points,
to refer to spatial locations established in signed discourse I use perspective
(ii) because it allows to focus on the different features contained within each
plane (Sandler 1989). The features established on each plane are the result of
a particular direction of index signs or other localisation mechanisms. What
matters is not the particular point in space, but rather the area in a plane
that gets activated through the direction articulated with the index sign, as
already explained in §3.1.
62 A morpheme on spatial planes

3.4.1. Horizontal

The horizontal plane stands perpendicularly to the body of the signer and
since the beginning of SL linguistics research it is commonly considered
as the default plane where the majority of signs are localised (Klima and
Bellugi 1979). According to Sandler (1989), the horizontal plane can be
divided into [ipsilateral] and [contralateral]. In Liddell and Johnson (1989)’s
model the horizontal plane is further divided into another [centre] feature.
This tripartite distinction is the one found in LSC, and (p) may be established
in three areas as shown below.

Figure 20. Horizontal plane

Following Liddell and Johnson, the features are distinguished in accord-

ance with the signer’s body: [centre] is in line with the breast; [ipsilateral]
is in line with the outside edge of the dominant shoulder, and [contralateral]
with the non-dominant shoulder. Figure 20 is an example of the divisions
within the horizontal plane for a right-handed signer, in which the ipsi-
lateral part coincides with the right hand part. With respect to the other
planes, the horizontal one has more divisions, as three distinct directions
are established.
The features within this plane and the axis they form are grammatically
relevant in the expression of plurality and temporality. As for plurality, signs
directed towards the axis unifying the contralateral and the ipsilateral part
and repeated up to three times denote reduplication. This reduplication
is only expressed in the line from the contralateral to the ipsilateral part.
This same axis also expresses sequences of temporal units. The units are
logically ordered from the contralateral to the ipsilateral part, as shown in
(11), where the days of the week are directly localised to the contralateral
and ipsilateral parts.
 Non-descriptive use of spatial planes 63

(11) ix1 tuesdaycl fridayip work office.

‘I work at the office from Tuesday to Friday.’

A temporal axis is also established from the contralateral to the centre part.
This is the anaphoric axis in the time lines described by Engberg-Pedersen
(1993: 81). The anaphoric axis is used to establish events with respect to a
point of reference. It does not have a default time and thus it is always estab-
lished in the context. In (12) the temporal sign before is articulated in the
anaphoric axis from the centre to the contralateral part. But before the articu-
lation of the sign, the point of reference needs to be established.

(12) takecl+++ abandonip not before there-is selectioncl.

‘Before taking [them to the place where they were exterminated] there
was a selection.’

As far as the discourse level is concerned, the ipsilateral and contralateral

areas on the horizontal plane are used as unmarked locations where DRs
are established. There does not exist any grammatical norm, but the general
tendency is that signers use their corresponding ipsilateral part to establish
the first spatial location (i.e. right-handed signers use the right side of signing
space, whereas left-handed signers use the left side). This is most surely
motivated by economy reasons since the ipsilateral spatial location is always
closer to the dominant hand. Importantly, discourse and sidedness (i.e. the
side in space where DRs are localised) can override handedness, since the
active hand can be shifted at a certain point for discourse reasons. Due to
the setting of spatial locations, the hand closer to the lateral part where the
DR has previously been localised may become the preferred hand along a
discourse segment and the signer may use the non-active hand since it is
closer to the contralateral part. That is to say, a right-handed signer may use
his left hand (instead of the right one) as the active one within a discourse
section when the entity is localised in the contralateral part, and thus reverse
the hands dominance (Frishberg 1983). In this book I focus on the localisa-
tion of entities in space without regard to the articulator used and neglecting
the articulation of dominance reversals. I leave this interesting issue for
future research.
Interestingly, the three main areas distinguished on the horizontal plane
are not equally used. On the one hand, the kind and attributes of the entities
that can be localised in space are different. On the other, there is a contrastive
import when using the lateral parts of signing space. These two aspects are
the main concern of the following two subsections.
64 A morpheme on spatial planes

3.4.1.1. Kinds of spatial entities

Localisation can be used for DRs denoting present and non-present enti-
ties in the immediate physical context, as well as abstract objects. When
denoting present entities, DRs are introduced into the universe of discourse
via a deictic demonstrative pronoun, i.e. an index sign pointing to the direc-
tion that the present object occupies, as shown below. In such cases, a fixed
eye gaze co-occurs with the introduction of the NP denoting the entity.

Figure 21. Reference to present objects

Since first and second person roles are required for the conversation to
take place, the physical location of signer and addressee are used as default
discourse locations. The index sign that localises the DR for first person is
directed to and contacts the chest of the signer. The index sign that localises
second person is directed towards the position the addressee occupies. The
location for second person is commonly established on the horizontal plane
[centre], frontal plane [lower], and midsaggital plane [distal]. However,
as noted by Bhat (2004), pronominal reference to first and second person
functions differently from third person reference. First and second person
pronouns function as shifters rather than referring terms. They indicate the
two principal conversation roles, namely that of being the sender and that
of being the addressee, respectively. First and second person pronouns are
local pronouns that directly point to their meaning. They act as shifters that
indicate the involvement of conversation roles (Bhat 2004). This contrasts
with reference to third (person) entities that identify the thing the conversa-
tion is about by locating it with reference to the spatio-temporal location of
the event. Obviously, the thing the conversation is about can be centred on a
first or second person, but in this case it functions as the thing the discourse
is about, rather than as the conversation participant.
The distinction between conversation roles and the entity the conversa-
tion is about is also found in LSC in relation to establishment of DRs in
 Non-descriptive use of spatial planes 65

signing space. First and second person pronouns are directed towards the
central part of the horizontal plane, while third person pronouns are directed
towards the lateral parts with respect to the coordinates of the body of the
signer.32 The pronoun used to mean second person is directed to the central
part, and in contexts of role shift, the second person is by default established.
However, first and second person pronouns are not used to denote what the
discourse is about and they do not denote discourse entities. As they are
considered to be conversation participants, they are left aside in the present
account, which only deals with discourse entities.
Concerning non-present entities, a random location is used and only
a very brief eye gaze is directed to the location in space (Figure 22). For
third (person) entities, the presence and absence of the entity in the physical
surrounding affects the direction of the index sign and the spatial setting
of locations, but this is only an epiphenomenon. As shown in Barberà and
Zwets (2013), in present references the location that establishes the anchor
for further coreferential chains and the actual position of the object coincide.

Figure 22. Index signs directed to non-present entities

Not all the entities established in discourse are equally localised in space,
and in LSC there is a clear restriction of the kinds of entities that occupy
a spatial location. The motivation for the difference in the kind of entities
localised in each spatial part is found in the semantic ontology. Natural
language semantics categorises entities into different ontological classes.
The distinction between events or eventualities, states, propositions and facts
turn the semantics of sentences somewhat more complex (Parsons 1990).
Events are considered to be spatiotemporal entities that denote an action.
Propositions are objects of belief, and they can receive a true or false value.
A third category, very much discussed in the philosophical logic, is facts,
which can be considered to be expressions of propositions. That is, proposi-
tions are kinds, and facts, which are truth-evaluated propositions, consist in
66 A morpheme on spatial planes

their instantiation. These ontological categories are considered to be abstract

objects used in natural language to refer and they can be ordered from less to
more abstract: events and states are placed at one end of the continuum, facts
are situated in the middle, and propositions are placed at the most abstract
extreme of the continuum (Asher 1993).
In this section I am not going to go deeply into the features that each class
has in LSC. However, it is worth noting that LSC shows a distinction between
the expression of entities (i.e. DRs) on the one hand, and the expression of
other classes which include facts, propositions, and events, on the other. For
the sake of simplicity, I call this second cluster of classes “non-entities”. Hence
in this informal description, I define non-entities as being negatively identi-
fied with respect to DRs. Since DRs are described as the entity the conversa-
tion is about (see §4.1.3 for a detailed definition of DRs as entities), there is
no need here for a fine-grained definition of facts/propositions/events. All that
is not a DR, falls into the category of non-entities. As described before, DRs
occupy a location in signing space, which is always established in the lateral
parts. This contrasts with non-entities, which are established instead in the
central position on the horizontal plane. The areas on the horizontal plane are
thus specialised: DRs always occupy a spatial location on the lateral parts and
non-entities occupy a location in the central part.
A further distinction between DRs and non-entities is found in its anaphoric
behaviour. Unlike entities, which are localised during first mention and can
be picked up by distant and non-distant resumptive pronouns, non-entities
are never localised during first mention, but they are rather introduced into
the discourse without being spatially established. However they can have
non-distant anaphoric pronouns referring back to them. Although being char-
acterised as non-entities, they can also serve as antecedents towards which
resumptive pronouns refer back to (for instance, in the case of propositions).
An example of a localised entity is (7), repeated here for convenience as
(13) and with subsequent context, which contains a non-distant and a distant
pronoun, marked in boldface below.

(13) ix3c laptop 1-offer-3 son ix3cl for new 3-select-3 work ix3cl
need laptop ix3cl.
[...]
ix1 sure ix3cl happy.
‘I will offer this laptop to my son, because he has been selected for a
new job and he needs a laptop.
[...]
I’m sure he will be very happy.’
 Non-descriptive use of spatial planes 67

Example (13) contrasts with (14) where a non-entity is introduced although

not localised. However, in the subsequent sentence a resumptive pronoun
that consists in a lax pointing directed to the centre is used (ix3c). In (14)
the event of Hitler becoming the German chancellor is not established in
space. But a non-distant resumptive pronoun refers back to the just intro-
duced non-entity (i.e. the proposition). This is articulated with a lax pointing
sign directed to the centre, as graphically shown in Figure 23.

(14) year 1933 hitler person-3ip start 1-appoint-3ip equal/same re-

sponsible maximum germany zone.ix3c novelty law.
‘In 1933 Hitler was appointed chancellor of Germany. This (issue) en-
tailed the creation of a new law.’

Figure 23. Coreferential index referring to a non-entity in LSC

There is thus a difference between the localisation of DRs and non-enti-

ties on the horizontal plane, and also in the referring back process: while
DRs are localised on the lateral parts and later on referred back to by both
distant and non-distant resumptive pronouns, non-entities are not localised
but directly referred back by non-distant anaphoric index signs directed to
the centre. When introduced into the discourse, non-entities do not occupy
any spatial location. This is logical according to our account in which only
entity-like DRs have its corresponding spatial location on a side of the hori-
zontal plane (see §4.2). However, when referring back to non-entities in
short distance contexts a lax pointing to the centre can be used. Hence there
is a clear distinction between the entity-like properties of the lateral parts
of LSC signing space, and the non-entity-like properties of the central part.
Since this book focuses on the entity-like properties signing space has, I only
deal with the lateral parts and will leave the non-entity-like properties of the
central part for future research.
68 A morpheme on spatial planes

3.4.1.2. Contrastive topics

In LSC the features on the horizontal plane used to localise entities, namely
[ipsilateral] and [contralateral], are not grammatically relevant themselves.
Whether (p) is precisely established on the ipsilateral or on the contralateral
part does not mark any difference on the grammar of LSC. (15a) and (15b)
are equivalent and the interpretation of the sentence is the same regardless
of which nominal is localised in which lateral part. The denotation of the
nominal is not affected by the localisation side, as the translation in (15)
shows.

(15) a. yesterday joanip 3ip-tell-1 pilar ix3cl sick.

b. yesterday joancl 3cl-tell-1 pilar ix3ip sick.
‘Yesterday Joan told me that Pilar was sick.’

This phenomenon contrasts with the facts discussed in the following section
(§3.3.2), where we will see that the two features on the frontal plane, namely
upper and lower, have different grammatical denotations and correspond
to different specificity interpretations. In the small-scale LSC corpus, the
only motivations that force the localisation of (p) on the ipsilateral or on
the contralateral side is due to assimilation processes and economy reasons,
which escape the grammatical restrictions of the language. However, it is
important to note that when two lateral locations are established within a
concrete fragment of discourse (i.e. two locations are chosen) a contrastive
relation arises, and this is so regardless of the exact lateral location assigned
to each DR. Thus, even if no grammatical relation is established between the
entity localised and the exact opposite part of the horizontal plane, a contras-
tive relation arises when the two of them are established. This is also shown
in (15) where two entities that are contrasted are established in the two lateral
parts.
In LSC, when both the ipsilateral and the contralateral parts are used in
the same fragment of discourse to localise two entities, a contrastive relation
is overtly expressed. This is an overt marking of the expression of contras-
tive topics (see Barberà 2007 for LSC contrastive topics; but also Büring
2003; Vallduví and Vilkuna 1998; for contrast in the spoken language litera-
ture, and Wilbur 2012, for a general overview of ASL contrastive topics).
Engberg-Pedersen (1993: 74) descriptively defines this use as a convention
of comparison, used when two entities need to be compared or contrasted. In
LSC, this contrastive use of the lateral parts coincides with double contrast
as defined in Mayol (2009, 2010). That is, two clause discourses in which
 Non-descriptive use of spatial planes 69

two DRs are introduced in each clause and their respective verbs predicate
two different, contrasting actions (see §7.4.1 for examples of contrastive
topics in LSC).
Furthermore, two or more DRs holding an affinity relation may be local-
ised on the same area (Figure 24). This kind of organisation of the frame
of reference in signing space has already been described for Danish Sign
Language by Engberg-Pedersen (1993). She calls it the “semantic affinity
convention” and it is the convention that covers different relations contrib-
uting to the organisation of the spatial frame of reference. In LSC contexts of
parent-child, person-place, and also different possession relations, the DRs
are localised on the same area, as long as they do not need to be distinguished
for discourse reasons (i.e. contrastively marked).

Figure 24. Discourse referents with a semantic affinity

3.4.2. Frontal

The frontal plane, according to Brentari (1998)’s terminology, extends verti-

cally to the body of the signer.33 The features [lower] and [upper] described
by Sandler (1989) are clearly distinguished in the LSC data. The phonolog-
ical distinction between [lower] and [upper] cannot be made in accordance
with the angle of the arm since the forearm cannot be taken as indicative of
the direction shown by the index sign. When the forearm is parallel to the
ground both the lower and the upper area can be indicated, because the wrist
can be oriented to the two parts. Likewise, when the forearm is not parallel
to the ground and the angle formed between the forearm and the ground is
bigger than 90º, also both the upper and the lower area can be signalled by
the different directions of the wrist. Thus the angle formed by the ground and
the forearm cannot be taken as a reliable clue to distinguish the two parts of
70 A morpheme on spatial planes

the frontal plane, since the wrist also plays an important role. Even if the arm
makes a specific angle, the wrist can point differently. The two parts on the
LSC frontal plane are clearly distinguished when we consider the shoulder
and the head. The space from the height of the shoulder and upwards is
considered to be the upper part. The lower part extends below the height of
the shoulder (Figure 25).

Figure 25. Frontal plane

In LSC the lower part of the frontal plane is the default area where (p) is
established. In contrast, when (p) is established on the upper part, which
is a marked area, it is associated with some particular and very concrete
meanings, namely hierarchical relations, locatives, expression of grammat-
ical specificity, and also absence from the immediate physical context. These
marked meanings differentiated from the default marking are presented
below.

3.4.2.1. Hierarchical relations

The upper part of the frontal plane is used to denote social hierarchical rela-
tions, and more specifically superiority. The contrast between upper and lower
frontal plane is associated with asymmetrical relations such as parents-child,
boss-worker, professor-student, etc. In such contexts, (p) established on the
upper part of the frontal plane denotes the individual who is higher in the
social hierarchy. This use has been previously described for LSC (Morales-
López et al. 2005), for Indo-Pakistani Sign Language (Zeshan 2000), and for
ASL (Liddell 1990; Schlenker and Lamberton 2012).
Within this use only definite NPs referred by pronouns and namesigns
(i.e. signs used as proper names within the deaf community) are localised
 Non-descriptive use of spatial planes 71

on the frontal plane. In fact, this is a crucial difference with another use
that I will discuss later on that denotes non-specificity and which is only
operative when localising indefinite NPs (see §3.2.2.3). Definite NPs
formed by common nouns such as ministry, government, boss, dean,
father+mother and university are always associated with the upper part
of the frontal plane. Also name signs referring to someone higher in the social
hierarchy are also localised towards an upper spatial location. Depending on
whether they have contact with the signer’s body, they are localised with an
index sign co-occurring with them (Figure 26a). Non-body anchored nouns
may be spatially modified and thus articulated at a higher spatial location
(Figure 26b).

a. ix3 government

b. university
Figure 26. Expression of hierarchical relations

The upper locations form a system of honorific speech, which are a morpho-
logical way of encoding the relative social status of the DRs appearing in the
discourse. They express social characteristic distinctions among the entities
the discourse is about. What is important to note is that, in contrast with other
Indo-European languages where honorific pronouns are encoded through
second person pronouns, in LSC honorificity is only marked on third person
pronouns, with a marked location towards the upper frontal plane.
72 A morpheme on spatial planes

According to the spatial mapping view (see §2.4.1), the localisation of

entities towards an upper spatial location is an instance of a projection of
the iconic properties, as well as the mental representation of the referent in
signing space. However, I argue that the use of the upper part of the frontal
plane to denote hierarchical relations does not reflect real heights that occur
in real world situations and it is thus not always iconic. In fact, there is not a
transfer from real world space to linguistic space since LSC does not convey
the exact heights of entities. I will report an example in order to show this
clearly. It is quite normal that when a teenager grows he becomes taller than
his parents. In an LSC conversation between a father who measures 1,60
metres and his son who measures 1,90 metres, the father will be always
localised in an upper part of the frontal plane and the son on the lower part.
Regardless of real height, parents are localised on the upper part while chil-
dren are localised on the lower one. Hence real height difference is not trans-
ferred to space. Signing space represents thus a linguistic convention that
follows the premise that referents that occupy a prominent position in the
hierarchical social scale are localised up on the frontal plane, rather than a
real-world situation convention. A great number of examples that show this
preference towards the linguistic convention could be reported. To further
illustrate this point, I add another example related to topographical loca-
tion of the referent in real world. At UPF, our university, the dean’s office is
located on the third floor of the building. The office of the SL crew happens
to be on the seventh floor. In every reference to the dean the localisation to
refer to her is established with an upper spatial location in the frontal plane.
The signer does not present the information differently according to what the
addressee knows about the layout of the building, but rather there is a strict
compliance with the linguistic convention. Hence no projection of the real
world situation into signing space takes place.

3.4.2.2. Locatives

Locative NPs denote spatial locations, such as places, cities, regions and
physical locations in the world. In LSC the locative noun is usually accompa-
nied with an index sign (Quer et al. 2005). This index sign used in a locative
NP tends to be localised on the upper frontal plane when denoting countries
and bigger regions (Figure 27a). Also, locatives mark plurals with points in
space (Figure 27b), rather than arc-shaped movements, which are character-
istic of plural pronominal forms. In contexts denoting areas within a small
region or a city, the imaginary map can be extended on the horizontal plane.
 Non-descriptive use of spatial planes 73

a. Singular locative b. Plural locative

Figure 27. Locative signs directed towards the upper frontal plane

When more than one locative is used in a fragment of discourse, they are
localised on the frontal plane, which is used as if it were a map. The distance
between the places and the location is considered to be at a certain scale on
the plane. This use is reminiscent of the absolute localisation where real-
world locations are transferred to signing space. According to this use, if
a signer is talking about an event of moving from Germany to France, the
localisation of the two countries will be established in the frontal plane and
within a certain scale. In this example, Germany will be localised in the
upper and ipsilateral part and France will be localised in a lower part of the
frontal plane and towards the contralateral part. Countries are established
according to their positions in a frontal map and within a certain scale. Such
contexts are a conflation of descriptive and non-descriptive locations. Once
the spatial location for the country is established the agreement relations are
established. Another example of conflation between the two uses of space
occurs when we have sentences such as “The neighbour from upstairs sent
me a fax”, where the expression of the subject location of the verb send is
localised at an upper part denoting the upper floor and the path moves to first
person location. Hence descriptive and non-descriptive locations in LSC are
often conflated.
The use of space for locatives has been considered to be an iconic conven-
tion according to Engberg-Pedersen (1993: 74), since signers organise the
frame of reference setting according to an imaginary map when denoting
geographical places. It is true that signers localise countries and cities
according to the location they occupy in the map represented in signing
space with a certain scale. However, LSC tends to lose this iconicity when
more than one use of space is at play. In a context where two locative NPs
denoting two cities in a country having different positions on the hierarchy
scale (i.e. one has more administrative/political power than the other) are
74 A morpheme on spatial planes

presented, the representation of the imaginary map is blurred. As shown in

Barberà (2007), some instances in the small-scale LSC corpus show that the
linguistic convention overrides the iconic one. When, for instance, a signer
is talking about Spain and Catalonia, although Catalonia is the north-eastern
autonomy of the country and Barcelona is geographically to the North of
Madrid, the capital of the country is always localised at an upper location
rather than at a lower one, as it would be expected according to the iconic
convention. Since Madrid is the place where the central government is, it is a
powerful entity. Hence it occupies an upper location regardless of its southern
position with respect to Catalonia (Figure 28). The linguistic convention is
preferred over the locative one, which is considered to be more iconic. Loss
of iconicity is shown in these contexts.

a. Sign include referred to b. Sign include referred to Spain

Catalonia (Madrid)
Figure 28. Non-iconic convention

3.4.2.3. Specificity

The two parts of the frontal plane are also used when the signer wants to
convey the specificity of the entity being talked about. The denotation of the
same nominal localised on the upper and the lower frontal plane results in
different interpretations. While the nominal localised on the lower part of the
frontal plane is interpreted as specific (16), the nominal on the upper part is
understood as non-specific (17).

(16) ix1 interview ix3l woman.

‘I have an interview with a womanspec.’

(17) ix1 interview ix3u woman.

‘I have an interview with a womannonspec.’
 Non-descriptive use of spatial planes 75

Hence, when (p) is established on the lower part of the frontal plane it overtly
expresses specific entities (Figure 29a), while (p) established on the upper
part is circumscribed to non-specific entities (Figure 29b).

a. Lower location denoting a b. Upper location denoting

specific DR a non-specific DR
Figure 29. Specificity marking on the frontal plane

Importantly, this non-specific use is distinguished from the hierarchical one,

as only indefinite NPs are marked for non-specificity. Only indefinite NPs
can be used to denote specificity or non-specificity by location marking (see
Chapter 5 for tests of indefiniteness). In contrast, when denoting hierarchical
relations, definite NPs such as namesigns, pronouns and definite descrip-
tions are used to localise the corresponding entity. In this chapter, I am only
offering a brief description of the specificity use on the frontal plane, and
Chapter 6 presents a detailed analysis with respect to the expression of speci-
ficity in LSC.
Interestingly, the localisation mechanisms listed in (9) which are
directed to the frontal plane are used differently and some restrictions
apply differently to the upper and the lower part. Within the lower part of
the frontal plane any kind of manual and nonmanual mechanism can be
directed to it (Figure 29a). Concerning linguistic categories, both lexical
and functional elements may be directed to the lower part and establish (p).
As shown in (18) and (19), common nouns and plain verbs can be localised
on the lower part.

(18) housel red

‘The red house’

(19) four person++ there-isl.

‘There are four people.’
76 A morpheme on spatial planes

As shown in (20) the arguments of a verb may be also associated with a

lower location. Verb inflection is thus grammatical in lower locations.

(20) ix3l 3l-advise-1 better not.

‘He advised me not to do it.’

Last but not least, determiners may also be directed to a lower location when
establishing a NP. Note that without regard of the kind of linguistic mecha-
nism used to localise the NP, whenever the localisation is established on the
lower frontal plane a specific interpretation arises.

(21) house alll white.

‘All the houses are white.’

(22) house somel red.

‘Some of the houses are red.’

However, the upper frontal plane presents some restrictions. As for the
nonmanual component, only eye gaze can be directed to an upper location.
Head tilt and body lean cannot be oriented towards the upper part of the
frontal plane, arguably because of phonological restrictions. Such a restric-
tion not only operates on the kind of localisation mechanisms, but also on
the kind of linguistic elements that are localisable on the upper frontal plane.
As for manual signs, only a specific set of signs can be oriented towards the
upper part, such as weak determiners (23) and verb inflection (24).

(23) house someu cle.b-house-destroyed.

‘Some housesnon-spec were destroyed.’

(24) 3u-advise-3 better not.

‘Someonenon-spec advised him not to do it.’

In contrast, there are some clear restrictions on the kind of linguistic cate-
gories. Bare common nouns (25), plain verbs (26) and strong determiners
(27)34 cannot be spatially modified towards an upper location. In fact, upper
localisations with the above mentioned lexical categories are considered to
be ungrammatical in LSC. (p) isthus only established on the upper part with
a restricted set of linguistic elements.

(25) *houseu cle.b-house-destroyed

(26) *there-isu four person++

 Non-descriptive use of spatial planes 77

(27) *house allu white

A further restriction is found on the sign person-3 discussed in §3.1.1. The

sign person-3 in LSC may have two functions: a nominal and a determiner/
pronominal one. When used as a nominal, it cannot be localised in the upper
frontal plane (28a). When used as a pronoun or as a determiner, it may be
directed towards the upper part and establishes the corresponding NP (28b).
In the latter case, the interpretation that we get is a non-specific one as shown
in the translation of the glosses.

(28) a. *person-3u
‘A mannon-spec’
b. ix3u man
‘One mannon-spec’

However, when the sign person-3 is used as a noun it is considered to be

grammatical to localise it on the upper part.

(29) one personu

‘One personnon-spec’

In this second case, note that without regard on the kind of linguistic mecha-
nism used to localise the NP, whenever the localisation is established on the
upper frontal part a non-specific interpretation arises. The distinction of the
two parts of the frontal plane denoting specificity is the main concern of
Chapter 6 and, as will be shown, only functional categories can be localised
on the upper frontal part.
In some contexts, two different uses of the frontal plane denoting different
meaning may co-occur. This is the case when, for instance, a lower loca-
tion marked in one element is conflated in the same NP with an upper loca-
tion marked in another element. In such cases the two opposed locations are
articulated, although minimised for phonological reasons. That is, the lower
location tends to be marked loosely and with a tendency towards an upper
direction. In (30) the determiner denotes a specific entity and hence it has a
direction towards the lower part. The nominal denotes an entity higher in the
social hierarchy, which is commonly localised on the upper part. The two
opposed directions are marked, although the upper direction of the nominal
starts before the onset of the articulation of the nominal. The determiner
some is articulated towards the lower part of the frontal plane, but before
78 A morpheme on spatial planes

the end of the articulation, it is directed towards the upper part of the frontal
plane where the nominal university is also directed to.

(30) somespec university

‘Some universities’

In principle, no iconic rule operates on the specificity use of the frontal plane.
If iconicity were a major criterion, LSC could in principle also convey the
expression of specificity using for instance the proximal and distal features
on the horizontal plane. Since the proximal area is closer to the body of the
signer it could be used to represent specific entities (since they are closer,
they are better known by the signer). In addition, the distal area within the
midsaggital plane could be used to denote non-specific entities, that is enti-
ties not known or not identifiable by the signer. However, this is not how
specificity in LSC is manifested. Another iconic possibility could be found
by representing specificity on the horizontal plane (see §3.2.1). Everything
that is known and identifiable by the signer is localised on the ipsilateral
part, which is the lateral part close to the active hand of the signer. All those
entities neither known nor identifiable by the signer could be localised on the
contralateral part, which is the side in signing space used by the non-active
hand. Again, this is not how specificity is marked in LSC and the iconicity
hypothesis is thus blurred.
The frontal plane to denote specificity is a major spatial distinction that
corresponds to a grammatical function, and this is precisely the main concern
of Chapter 6. Now, let us move to the last meaning assigned to the upper part
of the frontal plane.

3.4.2.4. Absence in the physical context

A final use of the frontal plane to be noted is the one that denotes absence of
the entity, which is always [+human], within the immediate physical context.
This is especially notorious in LSC when the entity talked about is a person
who is not present in the conversation environment. Hence namesigns used
to refer to someone who is not around co-occur with an index sign pointing
towards the upper part of the frontal plane.
As shown so far, the uses of the upper part of the frontal plane in LSC
split into four main functions. First, it is the area where hierarchical relations
are distinguished. Second, it is the place where locative signs are mainly
directed. Third, non-specificity marking is overtly expressed when DRs are
 Non-descriptive use of spatial planes 79

established in this area. And fourth, non-presence in the immediate physical

context, especially when denoting human individuals, is also marked with an
index sign towards the upper part. Importantly, it has been shown that when
a conflict of locations arises, the linguistic convention is preferred over the
iconic one. This leads to a preference for relative localisation constrained by
linguistic conventions over iconic and absolute ones.

3.4.3. Midsaggital

The midsaggital plane extends vertically and perpendicularly to the body of

the signer. Two features are found, namely [proximal] and [distal]. [proximal]
“is defined as a distance a few inches from the specified place, and [distal]
is a comfortable arm’s length away from the place” (Sandler 1989: 136).35
Hence the distinction proximal vs. distal is established in accordance with
the angle the elbow forms: the [proximal] feature occurs when the angle of
the elbow is smaller than 90º, and the [distal] feature occurs when the angle
is bigger than 90º, as shown in Figure 30.

Figure 30. Midsaggital plane

The features proximal and distal unify the axis used to express temporal
information, where present tense is signed in the proximal area, and future
tense is signed in the distal area. They are also relevant at the lexical level.
The sign for tomorrow is signed in the proximal area, and the sign for
the-day-after-tomorrow is signed in the distal area. This axis also forms
the mixed temporal axis (Engberg-Pedersen 1993: 81), which conflates the
anaphoric and the deictic axis. That is to say, the temporal information in this
axis is marked in the discourse and anchored in the context, and lexical signs
like from-now-onwards and until-now are articulated starting in the
proximal area and moving towards the distal area.
80 A morpheme on spatial planes

However, as for the discourse level, the dual distinction on the midsag-
gital plane is not found when establishing entities in space. Entities are not
abstractly established in LSC in the proximal as opposed to the distal part.
Rather, the midsaggital plane is used as a single extension and no distin-
guishable areas can be established when localising entities. Thus a singleton
feature [front] is distinguished and (p) is established in it without conveying
further distinctions. Nonetheless, when a demonstrative sign is used deicti-
cally, and it is thus pointing to an object present in the physical environment
there is a conflation of a descriptive use of space (i.e. because of the deictic
component) and a non-descriptive use. The direction of the demonstrative
pointing towards a present object is always precisely oriented towards the
direction where the present object is found in the physical environment. This
descriptive use is conflated with a non-descriptive one, since once the entity
is established in the discourse, it is possible to refer back to it. As already
mentioned, descriptive uses of space are freer and categorical distinctions
are established with difficulty. Hence due to the descriptive component of
these conflated structures, no distinction between [proximal] and [distal] can
be straightforwardly made. This is why the midsaggital plane is treated as a
single extension where no further distinctions are found. In the next section,
I focus on the clusters of features spatial planes have.

3.5. Features on spatial planes

As said at the beginning of §3.2, a location is an intersection among the

three spatial planes. Hence, when (p) is established, a cluster of features,
which characterise the intersection, coincide. So far we have seen that six
features are established in LSC discourse. The possibility of combinations
among these features adds up to six, since we have three possibilities: three
features on the horizontal plane ([ipsi], [contra], [centre]); two features on
the frontal plane ([low] and [up]), and one feature on the midsaggital plane
([front]) (i.e. 3x2x1). Although these six combinations could potentially be
the spatial areas used in the grammar of the language, in LSC only five direc-
tions formed by clusters of features are indeed found. The [front] area on
the midsaggital plane is not further divided into [low] and [up] parts, but it
is rather considered to be one and only direction without further division.
As for the frontal plane, the two features [low] and [up] can be combined
with the lateral horizontal features, namely [ipsi] and [contra] features. That
is, the [ipsi] feature can be combined with [low] and [up], and the [contra]
feature can be combined with [low] and [up]. According to this, only the
following combinations in (31) are possible.
 Features on spatial planes 81

(31) a. (p): {[front], [ipsi], [low]}

b. (p): {[front], [ipsi], [up]}
c. (p): {[front], [centre]}
d. (p): {[front], [contra], [low]}
e. (p): {[front], [contra], [up]}

These five clusters of features represent the directions where (p) may be
established. However, no grammatical difference has been found in LSC in
using [ipsi] and [contra] features apart from motivations due to assimila-
tion processes and economy reasons (see §3.3.1). Concerning these lateral
parts, what matters is the establishment of the two opposing sides, rather
than the concrete side of localisation. Hence, although five directions are
possible, only three clusters of features are relevant in LSC grammar. The
reason is that the cluster formed by [[front], [ipsi], [low]] is the mirror image
of [[front], [contra], [low]], and [[front], [ipsi], [up]] is the mirror image of
[[front], [contra], [up]].

(32) a. [[front], [ipsi], [low]] ≡ [[front], [contra], [low]]

b. [[front], [ipsi], [up]] ≡ [[front], [contra], [up]]

While the features [ipsi] and [contra] do not imply any contrastive difference
in the grammar of LSC, the [up] and [low] contrast in the frontal plane does
imply a grammatical distinction in LSC. As detailed in 3.3.2, NPs localised
on the upper part are associated with some particular and marked meanings,
while the lower part is the default marking. Hence the two features on the
frontal plane are relevant and play a very specific role in LSC grammar,
whereas this is not the case for the lateral features. The three clusters of
features which are relevant for LSC and which characterise (p) are the
following:

(33) a. (p): [[front], [ipsi]/[contra], [low]]

b. (p): [[front], [ipsi]/[contra], [up]]
c. (p): [[front], [centre]]

These three clusters of features are each specialised in the contribution of gram-
matical and referential aspects. As shown in (33), the referential properties and
82 A morpheme on spatial planes

the anaphoric behaviour of the [centre] area diverges from the [ipsi] and [contra]
sides. Not only the kind of entities, but also the referring back process is very
different. Entity-like properties of space are only found with (34a) and (34b),
and the semantics of non-entities in (25c) is left aside in the present book.

(34) a. (p): [[front], [ipsi]/[contra], [low]]: {discourse referents}

b. (p): [[front], [ipsi]/[contra], [up]]: {discourse referents (hierarchical
position, locative information, non-specificity, absence in the physi-
cal context)}
c. (p): [[front], [centre]]: {non-entities (events, propositions, facts)}

Only two features concerned with entity-like properties of signing space are
relevantly distinct in LSC grammar, namely lower and upper. Thus, the clus-
ters of features in (35a) and (35b) can be abbreviated as [low] and [up] to
keep the denotation simple.

(35) a. [[front], [ipsi]/[contra], [low]] (p)[low]

b. [[front], [ipsi]/[contra], [up]]  (p)[up]

Hence [low] and [up] are the two grammatically relevant features that can
be attached to the morpheme (p) in LSC. But since (p)[low] is the default
morpheme, I keep the denotation even simpler and distinguish between (p),
which is the unmarked spatial morpheme, as opposed to (p)[up], which is the
marked spatial location having a concrete meaning.
The spatial morpheme (p) consists in an abstract and unique point in
space, which is interpreted in the grammar of the language as a categorical
element. This morpheme is semantically associated with an individual from
the model (as will be shown in Chapter 4), and has the feature [low] as the
default marking. The marked feature [up] denotes a particular meaning.
Morphophonologically, I consider the feature [up] added to the spatial
morpheme to be an LSC homomorph, which denotes four concrete meanings.
Homomorphs are morphemes with the same form but different meaning.
An English example is the morpheme –er which can denote comparative
meaning, as in bigger; human agentivity, as in teacher, and inanimate instru-
ment, as in screwdriver. In the case of LSC, the homomorph [up] is special-
ised with four meanings, indicated below:

(36) [up]: {hierarchical position, locative information, non-specificity,

absence in the physical context}
 Body-anchored locations 83

The grammatical difference between (p) and (p)[up] is also notable when
looking at the set of mechanisms that can localise in the different spatial
directions. As seen in §3.3, while no difference is manifested on the midsag-
gital or on the horizontal plane, the frontal plane imposes some restrictions
on the kind of localisation mechanisms. Concerning nonmanuals, only eye
gaze can be attached to the [up] affix. This restriction could be motivated by
a phonological restriction, since it is physically hard or impossible to direct a
body lean or a head tilt towards the upper part as opposed to the lower part.
Moreover, only a specific set of elements can be attached to the affix [up].
The localised signs in (37), as previously seen, prove that only functional
categories, such as weak determiners (37a) and verb inflection (37b), can be
felicitously localised on the upper frontal plane. However, nouns (37c), plain
verbs (37d) and strong determiners (37e) are considered to be ungrammatical
when localised on the upper part.

(37) a. house someu cle.b-house-destroyed.

‘Some housesnon-spec were destroyed.’
b. 3u-advise-3 better not.
‘Someonenon-spec advised him not to do it.’
c. *houseu cle.b-house-destroyed
d. *there-isu four person++
e. *house allu white

The restriction of directing functional elements only towards the upper part
is a very interesting one, which shows that the feature [up] is grammatically
relevant. As shown in the translation of the glosses in (28a) and (28b), the
interpretation that arises is a non-specific one. How the spatial morpheme
(p) is associated with meaning is the main focus of Chapter 4, and the non-
specificity marking of the feature [up] is further treated in Chapter 6.

3.6. Body-anchored locations

In the preceding sections, I have mainly focused on locations established on

the three spatial planes, which extend in front of the signer’s body. However,
locations can also occur as body-anchored forms, since the body of the signer
and the spatial area immediately in front of it is also considered to be part of
signing space. Hence, DRs may also be introduced into the discourse without
84 A morpheme on spatial planes

being localised in the three-dimensional extent. In such contexts the signer’s

body is used as the location parameter. Body-anchored locations are carried
out with verbs agreeing with the body of the signer (generally assigned to
first person). As a further fact, this can be combined with role shift. When
this happens the DR is picked up again through a role shift construction (see
note 25). This referential shift is indicated by modifications on the signer’s
facial expression and body position. When the shift takes place, the body of
the signer is used as a default location for the displaced first person.
In (38) the signer introduces a new entity into the discourse, which is a
man who is walking. Although he is talking about a third person he does not
localise it on a lateral part, but rather uses an entity classifier and role shift,
and the location is established in the body. When the body-anchored location
is established, the verb WALK and the entity classifier are articulated from
the body and moving forward. To denote the shift, there is a change on the
signer’s facial expression (Figure 31). Hence body-anchored locations occur
when the spatial morpheme (p) is articulated at the signer’s location.

___________________________________rs
(38) one man walk CLe.long-thin-entity-moving-forward.
‘There is a man walking.’

Figure 31. Body-anchored location

Importantly, body-anchored locations need to be distinguished from signs

with an internal phonological specification where the location parameter is
phonologically specified at a determined location closer or in contact with
the body of the signer. The location parameter is a fixed element, which
differs from body-anchored locations, which are a more abstract location
 Body-anchored locations 85

near the body of the signer where entities are established. Even more, body-
anchored locations do not have to be confused with lack of establishment of
a spatial location. Signers may sign NPs according to its citation form. In this
case, there is no establishment of the entity denoted with a spatial location,
neither in signing space nor in the signer’s body. Whenever the citation form
is expressed there is no overt connection with the DR denoted. In Figure 32a
the citation form of the sign person is illustrated. As it can be observed, no
manual or nonmanual mechanisms localise the nominal. This contrasts with
Figure 32b where the signer directs eye gaze and a slight spatial modification
to the ipsilateral part.

a. Citation form b. Localisation in signing space

Figure 32. Citation form vs. localisation

Also weak referential elements are not localised in LSC. That is, a bare noun
not spatially localised can yield a generic reading (see §4.3 and Quer 2005a,
2012). The minimal pair in (39) shows this distinction. While the non-local-
ised bare noun in (39a) has a generic interpretation, the localised noun in
(39b) has a referential interpretation.

(39) a. woman read like.

‘Women like to read.’
b. woman ix3l read like.
‘This/the/aspec woman likes to read.’

In fact a generic interpretation is not obtained when a body-anchored loca-

tion agrees with the signer’s location. As (40) shows, when the verb is local-
ised at the signer’s body, only the referential readingis felicitous (40a).
86 A morpheme on spatial planes

(40) woman 1-walk like.

a. ‘A woman likes to walk.’
b. *‘Women like to walk.’

Hence, generic and weak bare nouns are not assigned a spatial location in
LSC discourse, as will be shown in the following chapter. This contrasts with
localised entities. The spatial morpheme (p) may be localised on signing
space or the signer’s body, and it stands for the overt manifestation of DRs.
How the spatial morpheme is connected to the DR which denotes is precisely
the main concern of the next chapter.

3.7. Summary

This chapter has described the three spatial planes used for localisation, the
features within each plane, as well as the grammatical correlations found in
non-descriptive locations. It has been shown that localisation mechanisms can
be modelled through a grammatical morpheme (p) that consists in an abstract
point in space regardless of the direction of the localised sign. The direction
in space towards the horizontal plane where (p) is established is irrelevant
for the grammar of LSC. This spatial morpheme is invariably established
in the ipsilateral or contralateral direction without implying a contrastive
meaning in the grammar of LSC. It rather implies a contrastive import at the
discourse level. Concerning the frontal plane, the features [low] and [up] are
grammatically relevant when attached to (p). The clitic morpheme used by
default has the feature [low], and the notation used here is (p). The marked
feature [up] is used to denote concrete meanings, namely locatives, nouns
denoting entities in a higher position in the hierarchy, absence in the physical
context, as well as non-specificity. As for the notation, (p)[low] is used for
this marked use.
Chapter 4
Spatial locations and discourse referents

Entities may be thought of as hooks on which to hang attributes.

(Webber 1979)

4.1. Introduction

It is clear from Chapters 2 and 3 that sign languages (SLs) in general, and
Catalan Sign Language (LSC) in particular, offer the possibility of estab-
lishing in signing space entities the discourse is about. However, how spatial
locations are associated with meaning and which their precise function is
has not been thoroughly formalised under any theoretical framework. In
this book it is considered that locations undertake a semantic function: that
of being the overt manifestation of discourse referents (DRs). Under the
specific formalisation I use, the DR established in space corresponds to a
variable established in the main universe of discourse. Hence, locations are
directly related with DRs that are attached to quantifiers with wide scope.
This chapter offers new evidence in favour of the r-locus view, according
to which spatial locations stand for the representation of DRs. §4.1 is an
overview of the theoretical framework used in this book, namely dynamic
semantics. §4.2 offers new and detailed arguments for the claim that loca-
tions are the overt manifestation of DRs. §4.3 shows that the new arguments
which are related to scope induce a revision of this claim according to which
locations stand only for DRs attached to quantifiers that have wide scope.
§4.4 summarises the main findings of this chapter.

4.2. Dynamic semantics

In this section the theoretical framework used in this book is presented. I do

not intend to present a complete review of the proposals related to dynamic
semantics, but rather I give an overview of the most influential aspects that
are relevant for the data to be covered and my analysis.
88 Spatial locations and discourse referents

4.2.1. Discourse and discourse model

A discourse is more than a sequence of isolated sentences. Sentences within

a discourse are interpreted with respect to their truth conditions, but they also
need to be interpreted in connection with the context. Every new sentence
introduced into a discourse is connected to the preceding ones, but at the
same time every new sentence adds information and increments the context.
The context is thus changing constantly and formal theories of discourse
treat sentences as denoting functions from contexts to contexts.
According to Heim (1982), context is a file of information held in
common by the participants of the discourse. Heim bases her theory on
Stalnaker (1978)’s notion of common ground, which includes the domain of
the discourse. The common ground is technically a set of indices considered
to be DRs that the interlocutors in the conversation know, which includes
linguistically given information, common educational and cultural expe-
rience, as well as sensory input (Heim 1982; Stalnaker 1978). Stalnaker
construes the common ground as a set of possible worlds, i.e. the set of all
and only those possible worlds compatible with everything that the sender
presupposes. Common grounds change from one context to the next because
the words that the sender utters and any obviously observable change in
the conversation’s physical surroundings potentially change the presumed
common background of the participants in the conversation. Hence, during
a conversation the participants keep adding the content of what is asserted to
what is presupposed. This is summarised as follows:
“To make an assertion is to reduce the context set in a particular way, pro-
vided that there are no objections from the other participant in the conversa-
tion. The particular way in which the context set is reduced is that all of the
possible situations incompatible with what is said are eliminated.”
(Stalnaker 1978)
Hence when a proposition γ is added to context c, the context obtained is
the derived context c’, as shown in (41). By adding new propositions, the
context set is reduced and the amount of possibilities is narrowed down.

(41) γ ˄ c = c’

A discourse includes discourse internal relations that allow to coher-

ently connect sentences by means of rhetorical relations (Mann and
Thompson 1988). However, a discourse also includes all those events and
actions that take place in the surrounding physical context. Suppose that in
the middle of a conversation a goat walks into the room. From that moment
 Dynamic semantics 89

on, it is presupposed that a goat is in the room and it is part of the common
ground. This presupposition can then be exploited by uttering “How did
it get in here? It is stinky!”. Since the goat enters the common ground it
may be referred to by a definite Noun Phrase (NP). Hence Stalnaker (1998)
proposes to identify context with the body of information that is presumed,
at a particular point in a discourse, to be common to the participants in the
discourse.
A discourse is thus defined as a sequence of sentences connected coher-
ently among them, linked to a context that is constantly changing. Every
discourse includes a discourse model, which is a mental representation
of the entities involved in it and the attributes and relations among them
(Webber 1979). In the previous example, for instance, the discourse model
includes an entity that corresponds to the goat that the conversation is about,
and we refer to this entity by means of a pronominal expression. How the
attributes among entities of a discourse model are related is the topic of the
following section.

4.2.2. Discourse representation theories

The formal representation of natural languages via the use of predicate logic
after Montague Grammar faces several problems when the representation
of larger chunks of discourse is needed. While Montague Grammar aims at
analysing the conditions under which a sentence is true, relying on reference
and truth, dynamic semantics theories regard the meaning and interpretation
of an expression as its potential to change the context of interpretation in the
discourse domain. Non-complex sentences can be easily represented through
predicate logic as shown in (42a) and (42b). However, complex sentences
face some difficulties, as shown in (42c).

(42) a. The girl likes the owl.

∃x∃y (girl (x) ˄ owl (y) ˄ like (x, y))
b. The girl doesn’t like the owl.
∃x (girl (x)) ˄ ¬ ∃y (owl (y) ˄ like (x, y))
c. If the girl has owls around, she caresses them.
∃x∃y (girl (x) ˄ owl (y) ˄ be-around (x, y)) caress (x, y)

While in (42a) and (42b) the existential quantifier has scope over the vari-
ables, in (42c) the variables (x, y) applied to the condition caress (x, y) are
90 Spatial locations and discourse referents

outside the scope of the existential quantifier. The problem posed by this
kind of sentences is called the problem of “donkey anaphora”. This phenom-
enon was named after the famous sentences by Geach (1962), where he used
donkeys and farmers in his examples. The so-called “donkey sentences”
show the impossibility of predicate logic to represent sentences where an
indefinite NP and an anaphoric pronoun are outside the regular scope domain
of the NP, as shown in (43).36

(43) a. If a farmer owns a donkey, he beats it.

∃x∃y (farmer (x) ˄ (donkey (y) ˄ own (x, y))  beat (x, y)
b. Every farmer who owns a donkey beats it.
∀x∃y (farmer (x) ˄ (donkey (y) ˄ own (x, y))  beat (x, y)

The earliest formal dynamic semantic theories were independently devel-

oped by Kamp (1981) and Heim (1982), who present two similar theories of
discourse models. Kamp presents Discourse Representation Theory (DRT)
and Heim introduces File Change Semantics (FCS). Heim (1982) uses the
notion of file-card as a metaphor for how the information about entities
is being tracked and she views discourse as a management of files. DRT
(Kamp 1981; Kamp and Reyle 1993) uses the notion of “discourse referent”.
Each DR in a model (or file-card in a file) corresponds to a discourse entity.
That is, the thing the discourse is about. Every time that a participant in a
conversation talks about a new entity, a discourse referent is added to the
model. Entities that are picked up from prior discourse correspond to DRs
that are updated. As discourse progresses new DRs are added and already
existing ones are updated. Each DR’s content is a contribution to the creation
of discourse context that interlocutors share. Thus the context is constantly
being built by the additional information that interlocutors keep adding to
the conversation. According to both FCS and DRT, a model is formed by
a set of entities which form the common ground (Heim 1982: 286). Both
FCS and DRT present similar analyses and the term “discourse representa-
tion theories” is commonly used to refer to the two approaches. This book
uses both as theoretical framework. However, DRT (Kamp and Reyle 1993;
Kamp, Reyle, and Genabith 2007) is the approach which has received more
attention and has been updated more. Therefore it is the technical tool used
to represent the semantics of LSC discourse.
According to DRT, interpretation involves a two-stage process: first, the
construction of semantic representations called Discourse Representation
Structures (DRSs), which represent larger linguistic units and discourses
 Dynamic semantics 91

rather than single sentences; second, a model-theoretic interpretation of those

DRSs. The construction of a DRS for such larger units proceeds sentence by
sentence, and the semantic cohesiveness among the sentences dynamically
contributes to the incremental nature of interpretation. To process a sequence
of sentences S1, S2...Sn, the construction algorithm starts with a syntactic
analysis of the first sentence S1 and transforms it into a DRS K1, which
serves as the context to process the second sentence S2 from which DRS
K2 arises. Simple monoargumental sentences have a combination of a noun
phrase (NP) and a verb phrase (VP), which semantically signifies that the
individual indicated by the NP (the so-called “discourse referent”, see 3.1.3),
satisfies the predicate expressed by the VP. Individuals satisfy the predicate
and in a DRT formal representation DRs are substituted for the NP and repre-
sented by a variable.
A DRS is a pair of sets <U, C>, where U is a set of DRs, called the
universe of discourse, and displayed at the top of the diagram; and C is a
set of DRS conditions (i.e. predicates followed by variables). The semantic
representation of the first clause in (44) is represented in the DRS K1 in
(45), where the DRs are represented by variables at the top of the diagram,
and the conditions, such as lali (u), book (v), and read (u, v), are represented
below.

(44) Lali is reading a book. She likes it.

(45) u v
lali (u)
book (v)
read (u, v)

(46) u v x y
lali (u)
book (v)
read (u, v)
she (x)
it (y)
like (x, y)
x=u
y=v
92 Spatial locations and discourse referents

In K1 (45), the DR that stands for the variable u represents the individual indi-
cated by the NP “Lali”. And in K2 (46) the variable x represents the individual
indicated by “she”. They both point to the same object in reality, namely a
SL linguist from Barcelona called Lali who happens to be my friend. The
two variables are equated by the identity relation x=u. However, this identity
equation is not obviously resolved. In fact, the central problem that theories
of discourse anaphora have faced is that of defining and explaining the rela-
tion which holds between the anaphoric pronoun and its antecedent, i.e. the
referentially independent expression from which the anaphoric expressions
gets its reference. The perspective adopted by classical semantics is that NPs
and pronouns correspond less directly to quantifiers and variables than the
dynamic view posits. The perspective adopted by dynamic semantics (Heim
1982; Kamp and Reyle 1993; Roberts 2005, among others) is that natural
language allows for dependencies that are far more flexible in standard
logical styles. Dynamic semantics assumes that anaphora is not a relation
between pronouns and other NPs, but rather one between pronominal refer-
ring expressions and DRs that are present in the semantic representation
under construction in the discourse progression.
Nevertheless, natural languages provide some features that constrain
identity relations, i.e. the mapping of two referring terms standing for DRs,
which point to the same object in reality. In K2, u and x are considered to
denote the same object, but x is expressed by an underspecified element.
This identity relation is obtained following semantic and syntactic informa-
tion. The first identity equation in K2 is x=u, where x is a new DR and u is a
suitable DR chosen from the universe of discourse. The suitable qualification
depends on all sorts of considerations, both semantic-syntactic, with features
such as gender, number, case, and grammatical role, as well as pragmatic.37
The matrixes in (47) show the features included in each NP considered for
the identity equation.

(47) Lali +subj She +subj

+fem +fem
+sg +sg
+nom +nom

The features of Lali and she coincide. Hence when the features coincide, the
first NP established becomes a suitable antecedent for the pronominal form.
Lali is a suitable antecedent for she, as well as book is a suitable antecedent
 Dynamic semantics 93

for it as the equation y=v indicates. The corresponding pronominal construc-

tion rule used in DRT that contributes to the DRS composition is as follows
(see Appendix for the complete list of construction rules used in this book):

CR.PRON: Upon encountering a pronominal form,

1. trigger the syntactic configuration [s NPα [VP]] or [s VP [NPα]], and
2. introduce a novel discourse referent α into the main DRS, and
3. check which variable in the main DRS shares the features α has, and
4. if no suitable variable is found, go to CR.PRON2; if the suitable variable
is found introduce an identity equation α = γ
5. go to CR.PROM

The matrix of features allows the establishment of the identity equation.

4.2.2.1. Donkey anaphora in DRT

Discourse representation theories (DRT and FCS included, henceforth

DRTs) resolve the problem of donkey anaphora. Both definite and indefinite
NPs are not treated as existential operators, but rather as variables, which
correspond to a DR as defined by Karttunen (1976) (see §3.1.3 below). DRTs
consider that there is existential quantification, which takes scope over the
entire discourse, and unselectively binds all the free variables in it. For this
treatment, DRTs adopt unselective restricted quantification, as proposed in
Lewis (1975). Given a quantifier, the construction rule instructs to create a
complex condition, which contains a pair of DRSs as the following represen-
tations of a donkey sentence below show.

(48) If a man is rich, he owns a donkey.

(49)

x y
man (x) → donkey (y)
rich (x) own (x, y)

(50) Every man who owns a donkey beats it.

94 Spatial locations and discourse referents

(51)

xy beat (x, y)
man (x)
donkey (y)
own (x, y) →
beat (x, y)

These complex conditions represent a tripartite quantificational structure.

The antecedent DRS (i.e. the left-hand box, called K1) represents the restric-
tion; the arrow represents an unselective universal quantifier, and the conse-
quent DRS (i.e. the right-hand box, called K2) represents de nuclear scope.
Since there are no variables in the main universe of the DRSs (49) and (51),
the sentences are true iff it is verified by the empty embedding function.
Every function which verifies the antecedent can be extended to a function
which verifies the consequent. The result is that (9) comes out as true iff
for every rich man in the model it is possible to find in the model a donkey
which he owns. Similarly (51) comes out as true iff for every man who owns
a donkey in the model, there is a donkey which he beats.
Hence, donkey sentences do not pose a problem for DRTs. The indefinite
NP in the restriction has no quantificational force and it is instead treated
as a variable. This variable occurs in the scope of the universal quantifica-
tion (and thus it is bound by it) that is associated by the semantics of the
pair of embedded DRSs K1 and K2. Donkey sentences illustrate the utility
of DRs (see §3.1.3). Both (9) and (11) include subordinated variables in the
discourse model, which allow keeping track of the entities mentioned in the
if-clause or relative clause in an embedded context. However, no reference
needs to be implied by the two sentences.

4.2.2.2. Accessibility

The structure of a DRS plays a crucial role in pronoun resolution. In DRT,

anaphoric pronouns are only allowed to refer to DRs that are accessible.
Accessibility is defined in terms of how the DRSs are nested into each other.
The positioning of a variable within a DRS has consequences for co-refer-
ence possibilities: a variable inside a subordinate DRS may not corefer with
a subsequent variable. In (53) z is embedded in the DRS K2, which is bound
by the negation operator. Since K2 is embedded in K1, z is accessible to y.
 Dynamic semantics 95

(52) Marta has a cat. It isn’t black.

(53) x y
marta (x)
cat (y)
have (x, y)

¬ z
it (z)
black (z)
z=y

However, in (55), z is included in DRS K1. K2 is embedded in K1, which is

bound by an operator. In this case, z in K1 is not accessible to y in K2, because
K2 is nested into K1.

(54) Marta doesn’t have a cat. #It is black.

(55) x

¬ x, y
marta (x)
cat (y)
have (x, y)
it (z)
black (z)
z=y

While (53) correctly predicts that the coreferential relation is grammatical,

the semantic representation in (55) also predicts that the second sentence in
(54) is a non-felicitous continuation. Let us turn now to the features of DRS’
variables.

4.2.3. Discourse referents

DRT considers that linguistic reference is not a mapping from linguistic

expressions to real world objects, but rather from linguistic expressions to
96 Spatial locations and discourse referents

constructs that are built along the discourse progression which represent
an object that exists in the real world. Discourse referents (DRs) are enti-
ties that denote the object of thought or the thing the conversation is about.
Once established in the discourse they can be referred back to by a pronoun
or retrieved by a definite description (Karttunen 1968, 1976). The concept
of DR does not have to be identified with the thing in the real world and
they can denote concrete, abstract and non-referential entities. For instance,
a sentence like “I saw a unicorn in the forest” can be felicitously uttered
in a world where unicorns do not exist. Also there are cases where the DR
does not have a unique counterpart in reality and it can have a non-specific
reading (i.e. not refer to a specific individual), as in cases like “Harvey courts
a girl at every convention”. DRs are thus the concepts that we have in mind,
that is objects of thought about which we say something.
The notion of DR resolves many discourse based concepts that can be
problematic for a syntactic theory of referential indices, such as the distinc-
tion between definite vs. indefinite NPs, generic vs. non-generic NPs, specific
vs. non-specific NPs, anaphoric vs. deictic NPs (Karttunen 1968). Below
some representative examples of these problematic issues are shown. In (12)
the indefinite NP “a book” has an ambiguous reading between denoting a
specific or a non-specific DR. In English the resumptive pronoun disambigu-
ates the sentence (Partee 1970). When it is followed by option (56a) it refers
to a specific DR, while in (56b) it refers to a non-specific one.

(56) I am looking for a book.

a. Here it is.
b. Here is one.

Other instances of indefinite NPs also imply the existence of an individual

in the discourse when they are picked up again except for two cases, which
happen to be also problematic for reference-based discourse theories, namely
indefinite NPs in predicate nominal position, and singular indefinite NPs in
generic sentences. The sentence in (57) is a statement about an individual
and one of his properties, but the indefinite predicate nominal (“an engi-
neer”) does not refer to a DR. In Catalan, in contrast, there is no article in
front of the indefinite NP as shown in (58), which minimises the possibilities
that the predicate nominal could have some referential properties. Generics
also cannot be interpreted as referring expressions, and thus they cannot
introduce new individuals since there does not exist a specific engineer to
which (59) is referring to. I come back to indefinite NPs in predicate position
and generic statements in §3.3.
 Dynamic semantics 97

(57) Francesc is an engineer.

(58) El Francesc és enginyer.

the Francesc is engineer

(59) An engineer is a problem solver.

Some entities introduced into the discourse do not necessarily have a real
counterpart in reality. Yet they are still introduced into the discourse implying
existence in the model and have the potential of being referred back to. This is
the reason why it is then more appropriate to talk about “discourse referent”,
rather than “referent”. In (60) and (61) the two indefinite NPs establish a DR,
but they do not imply real reference: first, in (60) the indefinite NP is a vari-
able bound to a quantifier, and second, in (61) the indefinite appears under
the scope of negation.

(60) Every engineer found a book and kept it.

(61) Francesc didn’t see a book.

An excerpt of discourse can contain more than one DR but only one real
corresponding object in reality. In (62) there are three DRs, namely “lali”,
“she” and “her”, but only one referent, namely the real person in reality
named Lali.

(62) Lali is reading a book. She likes it very much but her boss does not
allow her to read during breaks.

Hence the notion “discourse referent” as coined by Karttunen (1968, 1976)

avoids many claims about reference. As definite or indefinite NPs do not
necessarily have reference, the term “discourse referent” was used to denote
the entities forming the discourse model. Thus “discourse referent” and
“referent” do not coincide. The first one refers to the entity present in the
discourse model, the thing that participants are talking about.38 They are NPs
that may have or not a real-world reference, but they still introduce an indi-
vidual into the discourse. The second refers to the property of denoting an
entity that exists in the real world and hence it has some reference. Existence
within the common ground must be differentiated from existence in the real
world. As we have seen, a NP may introduce a DR into the model even if
it has no referent in the real world. In (56) above, the indefinite “a book” is
98 Spatial locations and discourse referents

under the scope of negation and does not correspond to an entity in the real
world. However, it still introduces a DR.
Nevertheless different terminology has been used (see Prince 1981). The
terms “discourse referent” and “discourse entity” are used as synonyms to
denote the same linguistic construct. Another widespread term is “s-topic”,
which also denotes the object of thought the sentence is about. Even if this
book uses the term “discourse referent”, the following section is devoted to
the s-topic concept.

4.2.3.1. S-Topic

Under some analyses the term “topic” has been used as a broad term to denote
different but related notions, such as what the sentence is about, the informa-
tive part of the sentence, and opposition to focus, among others.39 Some
authors include in this notion the abstract object or what the sentence is about
(Reinhart 1981), while others only include the linguistic marking which signals
the abstract topic (Büring 1999, 2003). According to Reinhart, although in
most cases the topics tend indeed to represent old information, this is neither
a sufficient nor a necessary condition for topichood (Reinhart 1981: 73).
Topics are better analysed in terms of their effect on the ongoing discourse and
considering the effects of previous discourse on the given sentence, rather than
as old information. Stalnaker (1978) and Reinhart (1981) base their analysis of
s-topics in pragmatic assertions. As explained in §3.2.1, Stalnaker defines the
context set of a given discourse at a given point as the set of the propositions
that both sender and addressee accept to be true at that point. These proposi-
tions are classified in the discourse as detailed below.
“Sentence topics are one of the means available in the language to organise,
or classify the information exchanged in linguistic communication – they are
signals for how to construct the context set, or under which entries to classify
the new proposition.”
(Reinhart 1981: 80)
For Reinhart, a topic represents thus an entry under which the oncoming
information is stored. And as will be shown below, this is reminiscent of
Webber (1979)’s hooks, Heim (1982)’s file-cards  and Vallduví (1992)’s
addresses. Büring (1999, 2003), in contrast, uses the term topic to refer to
a linguistic category expressed by linguistic means, which in English for
instance is prosodically manifested by a fall-rise pitch accent. Due to this
 Dynamic semantics 99

different conception, some authors have argued that the marking expres-
sion must be kept distinct from the abstract object. Hence according to
Vallduví (1992) and McNally (1998) there is a clear-cut distinction between
s-topics and the linguistic marking used to signal a specific s-topic. Such a
distinction is also assumed here and DRs and linguistic markings expressed
through referring terms are teased apart. S-topics (and also DRs) are linguistic
constructs that denote the abstract unit where the information is entered. They
correspond to the conceptual entity that we refer to and they are the entity
within the discourse model that we are talking about (see §3.1.3). S-topics
are made explicit in the discourse by means of referring terms. Referring
term is the linguistic material that point to an abstract s-topic and it can
be instantiated by different linguistic markings, such as intonation, morpho-
logical marking, or a marked syntactic configuration. As shown in §7.2.2, the
distinction between DRs and referring terms must necessarily be applied to
the analysis of index signs and locations in LSC.
As just stated, the notions of s-topic and topic have received different
labels in the literature and also different treatments. The following table
shows the differences in terms and treatments according to each author. While
Gundel, Vallduví and McNally make a distinction between the linguistic
material and the abstract element, Karttunen, Webber, Reinhart and Heim
do not make a distinction but their definition of DR and entity, s-topic and
file-card respectively is closely related to the abstract element. Büring does
not make a distinction either, but his definition of topic corresponds to the
linguistic material only.

Table 2. Terminology and treatments of s-topic

Linguistic marking Semantic entity
Karttunen (1968, 1976) discourse referent
Webber (1979) discourse entity
Reinhart (1981) s-topic
Heim (1982) file-card
Gundel (1988) topic (relational sense) topic (referential sense)
Vallduví (1992) link address
McNally (1998) link s-topic / topic
Büring (1999, 2003) contrastive topic
100 Spatial locations and discourse referents

4.2.3.2. Referential status

According to Vallduví (1992: 59), links are pointers in the sense that they
direct the addressee to the given DR where the propositional content of the
sentence is entered. However the term link is avoided in this book since links
are conceived as relational elements opposed to focus, which are very much
connected with information packaging. As Prince (1981) and Vallduví (1992)
show, information-packaging and referential status naturally reflect the
sender’s hypothesis about the receiver’s assumptions, beliefs and strategies.
While information packaging contributes to the update of DRs, referential
status is responsible for creating new referents or activating existing ones in
the discourse model. Referential status is an absolute property that reflects
the status of an entity with respect to the discourse model, which is expressed
through referring terms. Even if both modules are closely related, this book
focuses on referential status only.
In order to clarify the distinction between information packaging and
referential status of a DR, let us have a look at the following example. In
(63) the two instances of third person pronouns “him” reflect a prominent
referential status. Since both DRs are prominent entities the referring terms
used to denote them are pronominal forms. However, in terms of informa-
tion packaging the two pronouns are distinguished. The first one is the focus
(i.e. new information) while the second is the topic (i.e. old information).
Thus their information structure is different, while their referential status is
the same.

(63) I saw hím but not him.

Referring terms (i.e. the formal marking of NPs) reflect the referential status
of DRs. Although not all referring terms of a given sentence can be consid-
ered DRs simultaneously, a fragment of discourse can certainly have more
than one DR. Which of the referring expressions of a given DR counts as
topic is determined, in most cases, by differences in prominence. The reader
is referred to Chapter 7 for a detailed treatment of referring terms and promi-
nence issues. The general terminology used in this book is as follows.

(64) - Discourse referent is used to mean the construct, the semantic entity
or object of thought the discourse is about (known as s-topic under
some analyses).
- Variable is the construct used in dynamic logic that corresponds to a
discourse referent.
 Dynamic semantics 101

- Referring term is the morphophonological marking from natural lan-

guages that refers to a discourse referent.
- Object is used to refer to the real thing existing in the world.

4.2.4. Desiderata for a DRT application to sign language

Sign languages make a great use of deictic pronominal forms because of

its characteristic face-to-face interaction. Deixis is used in sign language
discourse, as well as in spoken language oral conversations, as shown below
in the English (65a) and LSC (65b) counterpart.

(65) While entering the office and seeing a man standing on the top of a lad-
der who is fixing something in the ceiling.
a. What is he doing here?
br bf
b. ix3ip do-what.

The two pronominal forms directly refer to the man who is present in the
physical environment without having been previously introduced into the
linguistic context (see Chapter 5 where cases of pragmatic anaphora, in
which deixis and anaphora converge, are treated). However, how underspec-
ified forms without a corresponding linguistic antecedent are incorporated
into the semantic structure of the discourse has not been treated in classical
DRT. Kamp (1981: 197, footnote 5), Kamp and Reyle (1993: 66) and Kamp,
Reyle, and Genabith (2007) explicitly ignore the use of deictic pronouns.
Their proposal is only concerned with written language and hence it only
affects anaphoric pronouns rather than deictic uses.
Also, whenever a semantic representation of a discourse is given, only in
very few cases has the incorporation of prominence been deeply treated in
depth (see Pinkal 1986; Roberts 1998). For instance, in the discourse previ-
ously presented in (22), and repeated here as (66) for convenience, it is perti-
nent to ask why one DR (in this case ‘Lali’) is more prominent than the other
(‘book’) and how this affects the ongoing discourse.

(66) Lali is reading a book. She likes it very much but her boss does not al-
low her to read during breaks.

This book offers an innovative approach that classical DRT lacks. It incor-
porates the properties that a visual-spatial language has which also affect
102 Spatial locations and discourse referents

the semantic representation. The desiderata for a DRT application to sign

language, and more concretely to LSC, are two-fold:

(i) To properly analyse the role that signing space plays in the semantic
representation of discourse.
(ii) To address deictic pronominal uses, and hence to build the correspond-
ing construction rules needed in contexts with deictic elements.

Moreover, this book also aims at implementing a semantic representation

of discourse that incorporates a prominence level of the entities within the
model by analysing how prominence is integrated. Thus, a general goal (not
specifically applied to sign language) is the following:

(iii) To integrate a theory of discourse structure with special focus on prom-

inence to the representational semantic level.

The following section presents how spatial locations are incorporated in a

DRT approach to LSC (goal (i)). As will be shown, locations allow resolving
the identity equation between two variables. Deixis incorporation is treated
in Chapter 5, where a specific construction rule for deictic elements is offered
(goal (ii)). Chapter 7 revises this treatment by incorporating prominence also
into the picture (goal (iii)).
Once the basis of the most influential aspects of dynamic semantics rele-
vant for this book has been presented, I turn now to the relation between
DRs and locations in LSC. In the following it is shown that locations stand
for the overt manifestation of DRs, but they are restricted to some semantic
constraints.

4.3. Locations and discourse referents

As reported in Chapter 2, within the SL literature there is a controversy

about the grammatical status of locations, since they may depend on the
actual position of present objects (see §2.4.1). While the spatial mapping
view claims that index signs directed to space are formed by a linguistic and
a gestural component, the latter motivated by the impossibility to integrate
spatial locations into a finite system, the r-locus view argues for a grammat-
ical analysis of spatial locations. The main goal of this section is to provide
new arguments in favour of the r-locus view. The main claim of the r-locus
view is defined in (67).
 Locations and discourse referents 103

(67) The discourse referent hypothesis (first version)

(p) is the overt manifestation of the DR the referring term denotes.

In what follows new and original arguments are provided in favour of this hypoth-
esis. However, as we will see at the end of this chapter this hypothesis needs
to be revised in order to fully account for the behaviour of locations in LSC.

4.3.1. Locations as variables

The ideas in this chapter have received a great amount of inspiration from
the works that formulated the r-locus view, which have been crucial for the
sharpening of the main claim (see §2.4.2). Lillo-Martin and Klima (1990)
analyse pronominal forms as being interpreted as a pair formed by a pointing
sign and a DR. Hence, both the interpretation of pronouns and the inter-
pretation of indexed nominals (that is, those nominals which are spatially
modified) are obtained by means of assimilation between locations and DRs.
Likewise, Wilbur (2008) assimilates the established spatial location (p) with
a semantic individual.
As seen at the beginning of this chapter, a discourse model is a set formed
by a subset of DRs and conditions applied to these DRs. In the dynamic logic
that DRT uses, DRs are formally represented with variables. The proposal
that underlies this book is to establish a correspondence between the spatial
morpheme (p) and a DRS variable. (p) has a semantic function, since it is
the overt manifestation of a semantic construct, namely a DR.40 The spatial
location expressed with (p) is in fact the overt manifestation of a DR: (p) is
associated with an individual from the discourse model. While in spoken
language the link between a referring term and an individual from the model
is done implicitly and ambiguously, in SL the connection is overt. SL refer-
ring terms are formal markings that are commonly directed to signing space.
This direction towards signing space establishes a location (p) that overtly
refers to a DR from the discourse model. This is shown in example (15)
in the preceding chapter, repeated here as (68) for convenience. The two
pronominal index signs in the second sentence are associated with the DR
“son”, established in the contralateral part in the first sentence.

(68) ix3c laptop 1-offer-3 son ix3cl for new 3-select-3 work ix3cl
need laptop ix3cl.
‘I will offer this laptop to my son, because he has been selected for a
new job and he needs a laptop.’
104 Spatial locations and discourse referents

The covert referential indices of spoken languages are manifested overtly in

SLs (Lillo-Martin and Klima 1990). R-indices are the semantic constructs,
which allow making the coreferential relations, and they are overtly expressed
by means of referential locations, which are the specific directions towards
signing space that signs take. In (68), for instance, it is the orientation of
index signs towards the contralateral part. At the level of the morphophono-
logical form, the representation of r-indices is overtly expressed as distinct
locations in signing space. As shown in the next subsection, locations are
thus the overt manifestation of DRs.

4.3.2. Identity features

As argued in §3.1, dynamic semantic theories associate DRs with NPs which
denote a nominal. DRs are represented within DRSs through variables. More
than one variable in the discourse model can point to the same object in
reality. Let us look at an LSC fragment, which includes three DRs that are
linked to the same object in reality.

(69) now ix1pl want 1-explain-2 theme history person-3ip woman

name a-n-n-a f-r-a-n-k.
ix3ip woman person-3ip character is/exact jew.
‘Now we want to explain the story of Anna Frank. This girl was a Jew.’
[…]
ix3ip hide during take-opportunity everyday write++.
‘During the time she was hidden, she took the opportunity to write a
diary.’

In this excerpt we find three DRs, referred as “Anna Frank” (expressed

through the manual alphabet fingerspelled), “ix3 woman” (‘this girl’),
and ix3 (‘she’). The three DRs are linked to the same object in the world,
namely a young little girl called Anna Frank. Each DR is localised in
space using different mechanisms. The first one is localised with the sign
person-3 (Figure 33a). The second one is localised by a determiner index
sign co-occurring with the nominal woman (Figure 33b). Finally, the third
one in the excerpt is localised with a pronominal index sign (Figure 33c). As
shown below, the three DRs are localised towards the same spatial location,
namely the lower ipsilateral part of signing space.
 Locations and discourse referents 105

a. person-3 b. ix3 woman c. ix3

Figure 33. Localisation of discourse referents

The semantic representation of this excerpt yields a DRS with three vari-
ables, namely x, y and z. The three variables (i.e. the logic constructs that
are identified with DRs) point to the same discourse referent in this specific
discourse. This is why in the simplified DRS in (70) the three of them are
equated under the identity equation.

(70) x y z
anna frank (x)
explain (x)
girl (y)
jew (y)
y=x
she (z)
write-diary (z)
z=x

The identity equation between an underspecified element and its antecedent

is obtained by means of suitability motivations based on semantic-syntactic
and pragmatic criteria. Identity features in SLs are obtained through location
information (Zwitserlood and van Gijn 2006: 213). The coincidence in the
spatial direction of index signs contributes to the identification of underspec-
ified expressions with their antecedent. As seen, in Figure 33 for instance,
the ipsilateral location of (p) is the criterion used as suitability in order to
identify the underspecified referring expression with its corresponding ante-
cedent. As for LSC, the antecedent and the underspecified referring term
need to have the same spatial direction on the horizontal plane in order for an
identity relation to arise.41 LSC referring terms have an orientation towards
the horizontal plane and establish a location (p), which corresponds to the
overt marking of the variable in the corresponding semantic representation.
106 Spatial locations and discourse referents

DRS’ variables are thus overtly expressed in LSC with the establishment
of spatial locations. Importantly, the coincidence in the direction towards
signing space where the spatial location is established identifies the two vari-
ables through the identity equation in the DRS. In (24) the two equations y=x
and z=x are encoded by coincidence in the direction towards space. Hence,
in short fragments of discourse in LSC, the identity equation in the construc-
tion rule for pronouns (see §4.1.2 and Appendix) is resolved by coincidence
in the direction in signing space where (p) is established. In fact, this spatial
use is with no doubt a unique possibility that SLs have with respect to spoken
languages. This is stated in (71).

(71) The spatial point hypothesis (first version)

The identity condition in the DRS is encoded through coincidence in
direction of spatial establishment of (p).

At this point of the presentation it is enough to accept that the identity

features in LSC rely on spatial location in order to identify a DR with its
corresponding antecedent. However in Chapter 7, I will revise The spatial
point hypothesis. We will see that when dealing with connected discourse
and when considering long stretches of discourse there is not a one-to-one
mapping between the direction of (p) in space and the DR associated. In
§7.4, it is shown that whenever an index sign refers to the most prominent
DR of the model the exact location in space where it is established can vary
along the discourse. (p) is thus a very abstract point in space, which is identi-
fied with a spatial direction in the horizontal plane and that is categorically
interpreted. Hence, the identity equation in LSC is not always resolved via
coincidence in the direction, but rather via semantic-syntactic and pragmatic
motivations as with spoken languages. But at this point of the presentation,
it is sufficient to accept that the identity condition is encoded through in the
direction towards space. Let us move now to the differences that the scope of
quantifiers attached to DRs have and how this is represented in the establish-
ment of (p) in signing space.

4.4. Scope of discourse referents’ quantifiers

The term “discourse referent” has been defined in §4.1.3 as the object of
thought the conversation is about. However, not all DRs are equally stable
in the discourse. While some DRs are permanent entities in a (fragment of)
discourse, some others may not be permanent at all and only discursively exist
  Scope of discourse referents’ quantifiers 107

within a certain discourse span. This distinction is noted by both Karttunen

(1969, 1976) and Heim (1982). Karttunen assumes that the appearance of
an indefinite NP establishes a DR just in case it justifies the occurrence of a
coreferential pronoun or a definite NP later on in the discourse. However, the
scope of the quantifier attached to a DR (i.e. the discourse segment in which
an introduced DR can be accessed) can vary. When a DR is introduced under
normal circumstances as a permanent DR, it remains available for being
picked up latter on in the discourse. Compare (72) with (73) below:

(72) Celia must write a postcard to Marta from Venice.

a. It must be mailed right away.
b. # It has a picture of Murano on it.

(73) Celia wrote a postcard to Marta from Venice. It has a picture of Murano
on it.
(Inspired by Karttunen 1969)

The utterance in (72) establishes a DR which is under a narrow scope quan-

tifier. A coreferential pronoun can only occur under the scope of the same
modal operator, in this case the modal verb “must” (72a). Without such an
operator, the DR cannot be referred back by a resumptive pronoun (72b). In
contrast, (73) establishes a DR that is attached to a wide scope quantifier.
Since there is no operator restricting the scope of the indefinite, it is acces-
sible in further discourse. The resumptive pronoun in the second sentence in
(73) is felicitous. Hence while (72) introduces a short term DR attached to a
narrow scope quantifier, (73) establishes a long term one, attached to a wide
scope one.42
As previously stated, DRs cannot be exactly identified with referential
indices, because there are some NPs which bear a referential index but fail
to set up a well-established DR (see §4.1.3). Previously, (54) was provided
as an example of indefinite NP bound by a narrow scope quantifier. This
example is repeated below as (74) for convenience. Because of its narrow
scope determined by the negation operator, no coreferential pronoun can
access the indefinite and thus the resumptive coreferential pronoun is not
felicitous.

(74) Marta doesn’t have a cat. #It is black.

Scope amounts thus to the existence of an NP bound by an operator which

binds the extension of its accessibility. Hence if an NPj is bound by an
108 Spatial locations and discourse referents

operator x, then the DRj that NPj introduces ceases to exist outside the scope
of x. If NPk is free, then the DRk lives throughout the entire text. The defini-
tion of scope used in this book is schematised in (75).

(75) Scope of DRs:

Narrow scope: Op <NPj ....NPj>
Wide scope: NPk ... NPk ... Op <NPj .... NPj>

Karttunen and Heim use different notions (DRs and file-cards, respectively),
but importantly, they highlight the distinction of the two scopes, which
predicts some coreferential restrictions. For instance, in (34) the pronoun
cannot be anaphoric to the indefinite NP because the scope of the DR is
bound by the negation operator, as defined in (75).
The narrow/wide scope distinction is formally represented in DRT. Those
variables attached to a quantifier with wide scope appear in the universe
of discourse of the main DRS (Figure 34a). However, variables attached to
a narrow scope quantifier appear in a subordinate DRS (Figure 34b). The
occurrence in a subordinate DRS as in Figure 34b is the result of being under
the scope of an operator, as shown in (35).

y
x

a. DR with wide scope b. DR with narrow scope

Figure 34. Scope representation in DRT

Recall that the positioning of the corresponding variable within the DRS
has consequences for coreferential possibilities, as both Karttunen and Heim
predict. Their positioning stands also in direct relation with the semantic and
referential properties that DRs have. In the following subsections, examples
of dependent variables, that is variables introduced into the model the value
assigned to which co-vary with those assigned to another variable (Farkas
1997), are presented. I take examples of non-argumental NPs, LSC donkey
sentences, non-specific indefinites and generic statements to analyse the
behaviour of these variables by studying their semantic representation. These
contexts establish a variable in the DRS. However, the quantifier attached to
the variable has narrow scope and it is only established in a subordinate DRS.
As a consequence, they are not available for further coreferential relations
  Scope of discourse referents’ quantifiers 109

outside the scope of the relevant operator.43 As we will see below, there is
a correlation between scope of the quantifier attached to the variable and
establishment of a spatial location in LSC signing. While DRs attached to
wide scope quantifiers have a corresponding location in LSC, DRs attached
to narrow scope ones do not establish such location.

4.4.1. (Non-)argumental NPs

As already stated in §4.1.3, nominals in predicate position do not introduce

a DR. They do not refer to an individual, but rather it is a predication about
an individual. Equative sentences are a good example to show the differ-
ence in the introduction of DRs. Specificational and predicational sentences
have received a great amount of attention among generative syntacticians.44
The felicity conditions of these types of sentences are different and from a
semantic point of view they also have different DRS’ representations. As
shown below, in LSC the syntactic configuration of predicational and speci-
ficational sentences is very different. LSC clearly distinguishes between
argumental and non-argumental NPs and this is reflected in the localisa-
tion process. This difference proves the distinction in the introduction of
each nominal. In order to compare the two different structures, first I give
a Catalan sentence and then the LSC counterpart.(76) is a Catalan equa-
tive sentence. It introduces an individual called “Francesc” and it predicates
something about him, namely the property of being an engineer.

(76) El Francesc és enginyer.

the Francesc is engineer

(76) is a predicational sentence where the argumental NP “Francesc” picks

out a DR, which has the following semantic representation. The vari-
able introduced by the individual is applied to the property.

(77) x
francesc (x)
engineer (x)

The equivalent of a predicational sentence as in (76) is signed with the

nominal francesc localised in space (78a, 78b, 78c). Any attempt of local-
ising the non-argumental NP engineer results in an odd construction as
shown in (78d, 78e).
110 Spatial locations and discourse referents

(78) a. francesc ix3a/person-3a engineer.

b. ix3a francesc, engineer.
c. francesca engineer.
d. # francesc engineer ix3a.
e. # francesc, ix3a engineer.
‘Francesc is an engineer.’

As for specificational sentences, they also introduce an individual and a prop-

erty (Rosselló 2008). In (79) the property of being an engineer is applied to
the individual “Francesc”. In this Catalan specificational sentence, the non-
argumental NP “enginyer” is not a DR, but rather a predicate ascribed to the
argumental NP “Francesc”.

(79) L’enginyer és el Francesc.

the engineer is the Francesc

In LSC, specificational sentences equivalent to (79) have a different syntactic

structure. The precopular nominal is instantiated as a rhetorical question with
the corresponding nonmanual marker.45 The rhetorical question is followed
by the argumental NP, which is localised in space (80a, 80b, 80c). But the
non-argumental NP in the rhetorical question cannot be localised in space
(80d, 80e).

br
(80) a. engineer who, francesc ix3a.
br
b. engineer who, ix3a francesc.
br
c. engineer who, francesca.
br
d. # engineer ix3a who, francesc.
br
e. # ix3 engineer who, francesc ix3a.
‘The engineer is Francesc.’

The corresponding DRS of a specificational sentence is represented by a func-

tion denoting asserted identity in classical DRT (Kamp and Reyle 1993: 257).
  Scope of discourse referents’ quantifiers 111

This is shown in (81), where the condition x is y asserts that the individuals
represented by x and y coincide.
(81)
xy
francesc (x)
engineer (y)
x is y

As shown in (78a), (78b), (78c), (80a), (80b), and (80c), argumental NPs in
LSC are grammatically localised in signing space and establish thus (p). In
contrast, non-argumental NPs do not refer to an individual, but rather they
attribute some property to it. In (78a), (78b), (78c), no variable is established
in the DRS but rather the property denoted by the non-argumental NP is
ascribed to the variable introduced by the argumental NP. Predicational and
specificational sentences in LSC show that non-argumental NPs cannot be
localised in signing space, as in (78d), (78e), (80d), and (80e).
However, there is one context where non-argumental NPs can have a
localisation in space and this is when there is a contextually determined
group of people. In such contexts the nominal engineer can be localised in
space, as shown below.

br
(82) engineer ix3pla who francesc.
‘Among those, the engineer is Francesc.’

Indeed, localisation also denotes individuals, which are contextually deter-

mined and constitute a subset of a non-empty set. This is further treated in
§6.2.2. For now let us assume for the sake of the present argumentation that
non-argumental NPs not belonging to a contextually determined set are not
localised.
A counterargument to what has been said so far could be that in (78)
and (80) there appears a proper name (“Francesc”). In DRT proper names
directly establish a variable in the main DRS because by uttering the proper
name the existence of such an individual is asserted. However if you slightly
modify the sentence and substitute the proper name for a definite description
the same semantic representation is obtained. (83) shows again that the non-
argumental NP attributes some properties to the argumental NP and hence
only the argumental NP is established in space.
112 Spatial locations and discourse referents

br
(83) a. ix1 friend ix3a, engineer.
‘My friend is engineer.’
br
b. engineer who ix1 friend ix3a.
‘The engineer is my friend.’

The discussion in this section is indicative that only argumental NPs have a
corresponding spatial location. Argumental NPs project a variable into the
main DRS and this corresponds with the establishment of (p) in LSC signing.
The upcoming arguments also show that lack of a variable in the main DRS
corresponds to a lack of spatial location establishment in actual signing.

4.4.2. Donkey sentences

Donkey sentences were previously introduced in §4.1.2.1 and they illus-

trate the usefulness of DRs. The corresponding semantic representation of a
donkey sentence shows that the position the variables occupy in the universe
of the DRSs is very important. In general, nominals introduced by donkey
sentences are represented by variables which are unselectively bound by a
universal quantifier. They do not appear in the main DRS but rather in a
subordinate one.
In LSC donkey sentences, nominals do not occur with a determiner index
sign directed to space to establish a location, but rather are uttered as bare
nouns and hardly ever localised. Verb agreement is articulated in a neutral
articulation (Quer 2010a). That is, agreeing verbs are not directed to any
particular location but rather to the centre of signing space. They are thus
an example of expressions of neuter forms. In the glosses in (84) below the
lack of localisation is shown by the absence of subindices, which signal the
spatial localisation.

(84) example/if town farmer horse there-is, sure 1-take-care-3c.

‘If a farmer owns a horse, he certainly takes care of it.’

Neither the farmer nor the horse in (44) occur with an index sign or eye
gaze that localises the nominal in space, and they are also articulated in an
unmarked position in space, as Figure 35a and Figure 35b show. Since the
individuals are not introduced in any spatial location, the verb in Figure 35c
  Scope of discourse referents’ quantifiers 113

does not agree with any location either since it is expressed through an unin-
flected form.46

a. farmer b. horse c. 1-take-care-3c

Figure 35. Donkey sentence in LSC

The semantic representation of (84) is as follows in (85). The complex condi-

tion formed by the antecedent and the consequent constrain the variables to
appear in embedded boxes (44). They appear thus in a subordinate DRS and
not in the main DRS. Hence the corresponding quantifiers attached to the
variables have narrow scope.

(85)

xy take-care
(x, y)
farmer (x)
horse (y) →
own (x, y)

Interestingly, a sentence like (84) could be felicitously followed with a

pronominal form referring back to one of the DRs introduced under the scope
of an operator such as a modal verb (86a). However, when the resumptive
pronoun in following discourse is not bound by an operator, the continuation
is not considered to be felicitous by native signers (86b).

(86) a. ix3ip must good-person.

‘He must be such a good person.’
b. # ix3ip good-person.
‘He is such a good person.’
114 Spatial locations and discourse referents

Contexts where variables with narrow scope can be further referred back to
as long as they are under the scope of an operator are known as modal subor-
dination (see Roberts 1989, 1990), and they are further treated in §6.3.3.2.
Also resumptive pronouns in subsequent sentences can be directed to the
centre of space. In such contexts they refer to the whole proposition (87).
This is in fact coherent with the entity-like properties attributed to the lateral
parts of the horizontal plane, namely ipsilateral and contralateral, and the
non-entity-like properties assigned to the central part, as described in the
previous chapter (see §3.3.1).

(87) but ix3c normal not.

‘But this is unlikely to happen.’

The correspondence between the lack of location in signing space in LSC

and the lack of variable in the main DRS shows a direct relation between the
DRS variable setting and the establishment of spatial locations. Nominals
corresponding to variables appearing in embedded boxes are not represented
by locations in LSC signing. The third argument in this line, referred to quan-
tified NPs and distributivity, offers another proof to support this claim.

4.4.3. Distributivity and quantification

Bound variable and quantified readings of NPs share a feature: both uses
have non-referential antecedents. That is, they do not identify concrete indi-
viduals. Non-referential antecedents can be further referred back to with
anaphoric elements, but since they are non-specific they have narrow scope.
As for SLs, it has been largely noted in the literature that distributive and
quantificational NPs show a different behaviour concerning the use of space.
Since they do not denote DRs, they do not establish a fixed location. Rather
space is used to denote plurality or quantificational relations.
Klima and Bellugi (1979) observe that in ASL grammatical categories
such as number and distributivity are very interrelated and there are a set
of verbal inflections that overtly express those relations. The form of some
verbs reflects both a distinction in the number of actions as well as quantifi-
cational distinctions, as the ASL examples in (88) show.

(88) a. meeting time-ten, supervisor inform[dual].

‘The supervisor informed each of the two about the ten o’clock
meeting.’
  Scope of discourse referents’ quantifiers 115

b. homework, teacher give[multiple].

‘The teacher gave out homework to them.’
c. diploma, principal give[exhaustive]; (me) none.
‘The principal gave out a diploma to each one, except for me.’
(ASL, Klima and Bellugi 1979: 281)

The verb in (88a) has a double movement, which overtly expresses the quan-
tification relation of duality. The verbs in (88b) and (88c) express plurality.
While the former has a circular movement expressing multiplicity, the move-
ment of the latter expresses exhaustivity by articulating the action back and
forth several times.
Petronio (1995) looks at the interaction between bare NPs in ASL and the
three classes of verbs (Padden 1988). She argues that bare NPs expressed
together with plain verbs are interpreted as either singular or plural. In
fact the interpretation is influenced by pragmatic, discursive, and contex-
tual factors. Also, affixes on agreement verbs as well as the morphological
information included in classifier constructions determine the quantifica-
tional value of the corresponding bare NP argument. What is interesting for
the present reasoning is that the arguments of these verbs, which co-occur
with bare nouns, do not occupy a particular location in space, as the glosses
indicate. Rather, the inflection of agreement verbs towards space is used to
denote singularity or plurality. The following sentences from ASL cited in
Petronio (1995) contain typical examples of an inflected agreement verb.
inform is inflected for singular in (89a), dual in (89b) and multiple in (89c).
The verbal inflection of inform determines the quantificational value of the
bare noun nurse. However, the nominal is not localised in space and no
spatial location is set up.

br
(89) a. nurse, ix1 finish inform[singular].
‘I informed the nurse.’
br
b. nurse, ix1 finish inform[dual].
‘I informed two nurses.’
br
c. nurse, ix1 finish inform[multiple].
‘I informed the nurses.’
(ASL, Petronio 1995: 609)
116 Spatial locations and discourse referents

Similar constructions are found also in LSC (Quer 2005a, 2012). In (90) the
verbal morphology influences the quantificational interpretation of the bare
noun student (see also Barberà and Mosella 2014 for further LSC exam-
ples of number agreement).

(90) a. student three ix1 1-ask-3[mult].

‘I asked three students.’
b. student three ix1 1-ask-3[exh].
‘I asked each one of the three students.’
c. student two ix1 1-ask-3[dual].
‘I asked two students.’
(LSC, Quer 2005a)

The agreement verb ask is inflected for multiple, exhaustive and dual. This
inflection is marked with a direction of the verb towards signing space to
denote plurality. But again it does not refer to a concrete individual and the
nominal does not occupy a spatial location. If the nominals occupied a loca-
tion in space, a singular resumptive pronoun could be used in following
discourse to refer back to one of the students. But this is not the case as
shown by the non-felicitous continuation in (91).

(91) #ix3ip clever.

‘He is very clever.’

However, plurals have a different behaviour. Plural referents in LSC can

be viewed as a group and be treated as a collective location. In such cases
although index signs are not articulated without an arc-shaped movement
but rather as a singular use, they are interpreted as plurals. In this case, the
uptake is felicitous as shown in (92a).47 Whenever a plural pronoun is used
to refer back to the set of students (i.e. a collective plural, with an arc-shaped
movement) the uptake is also felicitous (92b). Hence only when the resump-
tive pronoun is treated as referring back to a plural entity, but without a
specific singular location, the continuation is felicitous.

(92) a. ix3ip clever.

‘They are very clever.’
b. ix3plip clever.
‘They are very clever.’
  Scope of discourse referents’ quantifiers 117

According also to Quer (2005a, 2012) there is a strong relation between

quantifier-bound readings and distributivity. Quantifier-bound readings are
also expressed by means of lexical quantification. In this case, the redupli-
cation of quantifiers (each-one in (93a)), verbs (ask (93a), help (93c)),
the agreement marker glossed as agr (93b) and a specific LSC marker to
denote kin relations (kinship in (93c)) express distributivity and also bind
the quantifier.

br
(93) a. student each-one+++ teacher ask+++.
‘Each pupil asked his/her teacher.’
br
b. student each-one+++ teacher agr+++ respect.
‘Each pupil respects his/her teacher.’
br
c. mother son/sibling kinship+++ help+++.
‘A mother helps her siblings.’
(LSC, Quer 2005a)

Verbal morphology is used to mark plurality and space is used to denote

distributivity. However no specific spatial location is established. The repre-
sentation of (93a) is shown below, where the quantifier ‘every’ binds the
variable x and no variable is introduced in the main universe of discourse.

(94)

y
x every teacher (y)
student (x) x ask (x, y)

On the basis of the above argumentation concerning non-argumental NPs,

donkey sentences and quantified sentences, it has been shown that there is a
direct relation between lack of location establishment in LSC signing space
and subordinate DRS variables which have narrow scope. When no location
is established, there is a lack of variable appearing in the main universe of
discourse in the corresponding semantic representation. The fourth argument
is related to genericity.
118 Spatial locations and discourse referents

4.4.4. Genericity

Generic statements express general claims about kinds, rather than claims
about particular individuals, as well as propositions, which denote general
properties (Krifka et al. 1995). In English, generics can be expressed using
a variety of forms. Definite and indefinite singulars are two possible forms
(95a, 95b), as well as bare plurals (95c).

(95) a. The dog is a mammal.

b. A dog is a mammal.
c. Dogs are mammals.

These constructions are impossible to interpret in a spatiotemporal context.

And queries such as “How many dogs?”, “When are they mammals?” cannot
be answered as they just do not make sense.
De Vriendt and Rasquinet (1989) observe that SLs generally do not
make use of determiners in generic NPs. Since the expression of index signs
attribute some referential properties to the NP, generic statements do not
co-occur with an index sign, and hence the entity is not localised in space. In
LSC, bare nouns can assume a generic interpretation if they are not localised
in space (Quer 2005).

br
(96) woman play like not.
‘Women do not like to play.’
(LSC, Quer 2005a)

Any attempt to localise the DR “woman” in space is understood as referen-

tial (i.e. as denoting a specific woman).

br
(97) woman ix3ip play like not.
‘A/the/this/that woman does not like to play.’

Interestingly, in LSC there is a lexical sign that denotes genericity

(Quer 2005a). It is used when a general claim is stated about a kind, and it
occurs either pronominally or postnominally. This sign (Figure 36) does not
act as a mechanism of localisation and cannot co-occur with an index sign.
Hence the co-occurring nominal is never established in space.
  Scope of discourse referents’ quantifiers 119

Figure 36. Sign to denote genericity

Generic statements are represented according to the idea that the generic
operator binds particular variables in its scope. Variables appear in the
complex construction represented by a subordinate DRS bound by the
generic operator, as shown below.

(98) Men like to play.

(99)

x like-play (x)
GEN
man (x) x

Variables in generic statements are not main variables but rather subordinate
ones. The subordination setting in the DRS implies lack of establishment of
(p) in LSC signing space. However when the generic statement refers to an
object present in the immediate physical environment by a kind-example,
an index sign can be directed to it. Hence generic statements can co-occur
with an index sign whenever a token of that kind is present. This is not a
counter-example, but rather an example of a proper deictic reference towards
an object from the physical context. Last but not least, the fifth argument is
related to reference to kinds.

4.4.5. Kinds

Kind reference expresses reference involving entities related to specimens

(Carlson 1977; Krifka et al. 1995), as in (100) where “potato” refers to the
kind tuber tuberosum.
120 Spatial locations and discourse referents

(100) The potato was first cultivated in South America.

In such contexts, LSC nominals are never localised in space. The following
examples show that when denoting kind reference in LSC, the nominal is not
localised in space. That is, “doctor” in (101a) and “professor” in (101b) are
not spatially established.

(101) a. doctor don’t-do.

‘A doctor would never do this.’
b. teacher frisk never.
‘Teachers can never frisk (his students).’

When we try to localise the nominal in space, the reading we get turns out to
be a referential one since it refers to a concrete and identifiable DR, as shown
in (102).

(102) a. ix3 doctor don’t-do.

‘This doctor would never do this.’
b. teacher ix3 frisk never.
‘This teacher can never frisk (his students).’

Concerning the semantic representation, kinds do not have a corresponding

variable in the main DRS. One possible way to semantically represent kind
reference follows the idea that a generic operator binds particular variables
in its scope. Variables appear in the complex construction represented by a
subordinate DRS bound by the generic operator.48 The simplified semantic
representation of (101a) is represented in (103), where the variable is bound
by a generic operator.
(103)

x GEN this (y)

doctor (x) x ¬ do (x, y)

The examples in this section have shown that nominals referring to kinds are
not localised in LSC. In the corresponding semantic representation, the vari-
able appears subordinated in an embedded context.
 Summary 121

So far, the five previous arguments show that semantically DRs with
narrow scope do not have a corresponding variable in the main universe
of the DRS but rather in a subordinated one. They have been grouped into
four different types, namely donkey sentences, quantified NPs, generic state-
ments and kinds. Also non-argumental NPs have provided some evidence
towards the hypothesis defended in this chapter. On the one hand, non-argu-
mental NPs are nominals in predicate position. They do not establish a DR,
but rather they function as a predicate, which is applied to the DR introduced
by the argumental NP. Hence the nominal does not project a variable in the
DRS. On the other hand, variables in contexts such as donkey sentences,
quantified NPs, generic statements and kind reference do not establish a DR
with wide scope and the corresponding variable appears in the subordinate
DRS. The variable only exists within the scope of the operator which binds
it. Outside this scope, the variable does not have existence anymore and it
cannot be accessed by further pronominal forms in discourse. Variables with
wide scope do not have any restriction of existence and their accessibility in
discourse is not restricted by any operator.
Concerning the form level, DRs with narrow scope do not establish a
spatial location in the actual LSC signing. This contrasts with DRs with wide
scope, which formally establishes a location (p), since signs are directed to
a concrete area in space. This location is available for further co-reference.
Hence locations are semantically represented by main variables in the corre-
sponding representation that are not restricted in a discourse segment by any
operator. As shown by these new arguments, The discourse referent hypothesis
presented in (27) needs to be revised. It is not that spatial locations establish
a DR, but rather that only DRs with wide scope have a corresponding spatial
location. DRs with narrow scope, which appear in subordinate DRS, do not
establish a location.

(104) The discourse referent hypothesis (second version)

(p) is the overt manifestation of DRs attached to a quantifier that has
wide scope.

4.5. Summary

This chapter has introduced the theoretical framework used in this book,
namely dynamic semantics. Spatial locations in LSC have been associated
with DRs, as understood in dynamic semantics. The implementation in DRT
122 Spatial locations and discourse referents

has allowed us to see that the positioning of the variable in the DRS is the
representation of the scope of the variable. Narrow scope DRs (i.e. bound
by an operator and represented by an embedded variable) do not occupy a
spatial location in LSC. Only DRs with wide scope (i.e. those DRs not bound
by any operator which can be accessed in further discourse that are repre-
sented by a variable in the main DRS) are formally represented by location
(p) in LSC actual signing. Hence the phenomenon of establishing entities
in LSC space is directly associated with the establishment of DRs into the
model. Interestingly, only a specific set of DRs can be localised in space;
specifically, only those which appear in the main universe of discourse and
have thus wide scope.
However, as will be shown in Chapter 6 the positioning of the variable in
a DRS is also relevant for specificity marking. The main/subordinate DRS
distinction is overtly encoded on the LSC spatial frontal plane and on the
amount of morphophonological information directed to it. As we will see,
some DRs can be localised on the lower part of the frontal plane, while others
are localised on the upper part. This distinction is directly connected to the
expression of specificity, which is the main topic of Chapter 6. But before
delving into the specificity domain, let us focus on the relationship between
space and definiteness marking in LSC in the next chapter.
Chapter 5
Deixis and familiarity

The deictic use and the anaphoric use have at least this in common: both
are among the possible uses of definite NPs, but neither is possible with an
indefinite. One possible explanation for this might be that the pragmatics of
deixis and anaphora are intrinsically similar, and definiteness correlates with
the property they share.
Heim (1982: 309)

5.1. Introduction

Whether definiteness is encoded in signing space has been a matter of debate

in the sign linguistics literature. The main goal of this chapter is to analyse
the connection between locations and the expression of definiteness. Its main
aim is not to describe how definiteness is encoded in Catalan Sign Language
(LSC), but rather whether spatial location establishment denotes definite-
ness. The goals of this chapter are two-fold. On the one hand, it shows that
definiteness is not formally encoded in LSC signing space and that both
asserted and presupposed discourse referents (DRs) are spatially established.
The establishment of (p) does not distinguish between definite and indefinite
noun phrases (NPs). On the other hand, it shows that deictic uses may have
a corresponding (p) established. Deictic, as well as anaphoric uses, have in
common that both are possible uses of definite NPs, but neither is possible
with indefinite NPs. Both presuppose that the DR is familiar and that it
belongs to the common ground.
The theoretical background in §5.1 presents the ingredients needed for
the purposes of the chapter from theories about definiteness, deixis, as well
as the description of definiteness marking in sign language. §5.2 shows that
deictic uses are another means to introduce DRs, which are newly intro-
duced to the common ground. Thus, when (p) is established it does not need
to have an overt antecedent but it refers to an entity from the model. The
identification between the variable being overtly expressed in discourse with
the default variable associated with entities introduced without an explicit
linguistic antecedent is a very common use in LSC. These contexts of weak
familiarity show that reference in LSC is always referred with respect to the
124 Deixis and familiarity

discourse model and it is thus anaphoric. In §5.3 it is shown that the distinc-
tion between definiteness and indefiniteness is not marked in LSC locations
and this is analysed with respect to the status of the DR in the model. That
is, whether the DR is presupposed or asserted. Also, when (p) corresponds
to a presupposed DR, this is better analysed in terms of familiarity. Finally,
§5.4 briefly shows that indefiniteness marking may be overt expressed with
a restricted set of determiners and a nonmanual marking. §5.5 concludes the
chapter.

5.2. Definiteness: background

The current state of research on definiteness usually associates definiteness

with uniqueness and familiarity. On the one hand, uniqueness approaches are
built on the insight that a definite description is used to refer to entities which
have a role or a property which is unique. These theories are more focused
on logical semantics (see Abbott 1999 and subsequent work; Kadmon 1990;
Löbner 1985; Russell 1905; Strawson 1950). On the other hand, pragmatic
theories tend to treat familiarity as the central notion for definiteness. They
are based on the idea that definite descriptions serve to pick out DRs that are
in some sense familiar to the discourse participants (among the most recent
work, see Heim 1982; Kamp 1981; Kamp and Reyle 1993; Prince 1981,
1992; Roberts 2003). An important feature that marks a distinction between
these two theoretical options is that whereas uniqueness-based theories treat
(indefinite) NPs as quantificational, familiarity-based theories treat them as
variables. In the following, a subsection is devoted to each approach.

5.2.1. Uniqueness

Uniqueness indicates that there is one and only one entity of some property.
The representation of this notion in mathematical logic uses the existential
quantifier followed by the exclamation mark which turns the formula into an
indication of uniqueness, as shown in (105) below. It can be read as “there is
exactly one x, such that x has the property P”.

(105) ∃!x P(x)

An example of such a sentence in English is (106a), which denotes that there

is exactly one entity of the kind book in the universe of discourse and that it
  Definiteness: background 125

is on the table. This differs from (106b), which implies that there are some
more books elsewhere but that there is exactly one on the table.

(106) a. The book is on the table.

b. A book is on the table.

The domain of quantification needs to be restricted to the relevant context

in order for the utterance to be felicitous. With a sentence such as (106a)
the claim must be about a salient and relevant book in the specific context.
This salient entity does not need to be the same throughout the discourse
and the focus of salience can be shifted to different entities at different
points in discourse (Lewis 1979). However, which domain of reference the
uniqueness condition applies to has been a matter of debate. Kadmon (1990)
and Roberts (2003) propose to limit the uniqueness of the definite to the
specific universe of discourse, and hence to restrict the utterance to the rele-
vant context in order for it to be felicitous. To use Roberts’s term, prag-
matic enrichment of the descriptive content of the NP is an instance of the
phenomenon of domain restriction in the interpretation of logical operators
(Roberts 2003: 292). A domain of reference is required for the uniqueness
condition. When a definite NP is uttered, it does not generally apply to the
set of DRs denoted by the NP existing in the whole universe, but rather it
applies to a restricted domain. If a sentence like (107) is uttered, the inter-
locutor does not think of all the books existing in the universe, but rather of
the set of books which are relevant in the specific situation. This reading is
obtained by restricting the domain of interpretation of the set. The formula in
(108) shows that this can be achieved by the intersection of the set of books
and the domain variable (C), which refers to the relevant discourse familiar
set of books.

(107) All the books are on the table.

(108) ∀x(book (x) ˄ C (x))  on table (x)

5.2.2. Familiarity

The notion of definiteness has been established within a familiarity theory

according to which the essential function of definiteness is to signal that the
intended DR of an NP is already familiar to the audience at the current stage
of the conversation (Heim 1982). Definiteness is understood as identifiability
126 Deixis and familiarity

of the DR that can be found in the universe of discourse. While a definite NP

is used to signal existence in the model, an indefinite NP is used to signal
that the DR being introduced is yet unfamiliar, i.e. novel. Modern accounts
of familiarity develop a new theory based on a more formal semantic and
pragmatic account (Heim 1982; Kamp 1981). Both Heim and Kamp argue
against the view that definite and indefinites are quantificational expressions
and treat both of them as free variables that are bound to an existential oper-
ator (as seen in §4.1.2). Their representation of an indefinite is shown in
(109), and for a definite, in (110) below.

(109) A man came in. (110) The man came in.

x x
man (x) man (x)
come (x) come (x)

Definite and indefinite NPs have the same semantic representation and both
have the same open formula. The difference does not lie in the quantifier
attached to them, but rather in the different conditions attached to each one:
indefinites are associated with a novelty condition, whereas definites have a
familiarity condition (Heim 1982). The novelty condition indicates that the
DR and its descriptive content are not presupposed to be satisfied by any
individual in the domain of the common ground of the context. Definites and
indefinites also have different descriptive content: in indefinites, the descrip-
tive content is asserted, while in definites it is presupposed. In the Heimian
account, familiarity is determined by whether there is information about a
corresponding DR already in the local context of interpretation. For every
indefinite, a new DR is created, the descriptive content of which is novel with
respect to the model. For every definite, a suitable old DR is updated, the
descriptive content of which is already familiar with respect to the model (see
§4.1.3.1 for the equivalence of terms among the different accounts, e.g. under
Heim’s account, a DR is labelled “file-card”). This is formalised as in (111).

(111) Extended Novelty-Familiarity Condition (Heim 1982): For a logical

form Ф to be felicitous w.r.t. a context C it is required for every NPi in
Ф that:
(i) if NPi is [-definite], then i Dom(C)
(ii) if NPi is [+definite], then
a. i Dom(C), and
b. if NPi is a formula, C entails NPi.
  Definiteness: background 127

The formula in (111i) refers to the novelty condition, (111iia) refers to

the familiarity condition and (111iib) to the descriptive content condition.
However, Heim does not explain how suitability is specified in condition
(111iib). In §4.2.1 suitability features were presented for LSC, and this issue
is further treated in Chapter 7.
In order to account for definites appearing for the first time and also for
associative uses of definites, a process of accommodation is used. Heim’s
operation of accommodation (based on Lewis 1979) involves a linking or
bridging operation (Clark 1975). Accommodation is defined as a non-mono-
tonic process which forces to review the previous record of the discourse and
to adapt it to the new demands. Novel definites are rendered felicitous by
accommodation, which is defined as follows:
“If at time t something is said that requires presupposition P to be acceptable,
and if P is not presupposed just before t, then –ceteris paribus and within
certain limits – presupposition P comes into existence at t.”
(Lewis 1979)
By this process, the interlocutor accepts the information as given and revises
his interpretation of the context accordingly. Hence, if at some point of a
conversation the sender utters (112) without previously having introduced
the information that she has a brother, the interlocutor assumes, all other
things being equal, that the sender has a brother.

(112) I gave the book to my brother.

The need of this additional operation weakens Heim account by making

only possible to explain first mention definites through accommodation. For
this reason, Roberts (2003) extends Heim’s familiarity and offers a revised
version of the theory by making a distinction in terms of the introduction of
the DR into the model. Roberts (2003) argues for a distinction of definites
depending on where the antecedent may be found.

5.2.2.1. Weak/strong familiarity

Roberts (2003) argues for a re-elaboration of the Heimian notion of famili-

arity. She proposes a distinction between strong and weak familiarity, the
latter being a broader notion very well suited to account for definite NPs
presupposing existence in the discourse. Roberts defines the two distinct but
related notions of strong and weak familiarity as follows. Strong familiarity
128 Deixis and familiarity

refers to those DRs that are explicitly introduced in the preceding linguistic
context by an antecedent, whereas weak familiarity refers to those DRs
whose existence is entailed in the context. Roberts’ proposal is summarised
in (113).

(113) Taxonomy of familiarity

a. Strong familiarity: A linguistic antecedent exists in the preceding
discourse.
b. Weak familiarity: Existence is entailed in the context.

Strong familiarity denotes instances of definite NPs that have a coreferential

NP in the preceding linguistic context.49 This contrasts with weakly familiar
DRs that do not have a coreferential linguistic antecedent and hence their
existence is entailed in the context. Some motivations allow DRs falling into
this group to be first mention definite NPs. They can be accessed by the
interlocutors since they are present in the immediate context of utterance and
are thus perceptually accessible. They can also belong to the group of DRs
which are known from the general culture or at least known from the relevant
common knowledge surrounding the conversation, as shown in (65a) in the
previous chapter repeated here as (114) for convenience. In addition, they
can also be accommodated, as described in the previous section.

(114) While entering the office and seeing a man standing on the top of a 
ladder who is fixing something in the ceiling:
What is he doing here?

It is important to note that it follows from Roberts’s analysis that weak famil-
iarity subsumes strong familiarity and it is more inclusive, encompassing
explicitly introduced DRs and also DRs introduced non-linguistically on the
basis of contextual entailment alone (via perceptually accessed information).
Weak familiarity is thus a broader concept that includes strong familiarity.
It is also worth mentioning that the distinction between weak/strong famili-
arity is not equivalent to prominent vs. non-prominent DRs. Both weakly
and strongly familiar DRs can be prominent or not. For instance, there can
be strongly familiar DRs that are not prominent because they are picked up
again far away from its antecedent. Also not all prominent DRs are only
strongly familiar, since they can also be weakly familiar. In Chapter 7, a
deeper treatment of prominence is offered. Instead of being equivalent to
prominence, Roberts’s distinction is equivalent to Prince’s fine-grained
distinction of discourse/addressee familiarity described below.
  Definiteness: background 129

5.2.2.2. Discourse/addressee familiarity

Prince (1981, 1992) considers definiteness to be a formal phenomenon that

can be analysed as a discourse phenomenon, related to the information-status
of the entities in the discourse. She argues for a distinction between discourse
new/old entities and addressee50 new/old entities, which is based on the
sender beliefs about the addressee’s beliefs about an entity. The distinction
between discourse-familiarity and addressee-familiarity can be combined in
a four-celled matrix of possible information statuses. Of these four cells,
only three occur in natural discourse.51 They are the following:

(115) a. Addressee-new/discourse-new: Information which has not been

evoked in the current discourse, and which the sender assumes to
be also unknown by the addressee.
b. Addressee-old/discourse-new: Information that has not been
evoked in the current discourse, but that the sender assumes it is
known by the addressee.
c. Addressee-old/discourse-old: Information that has previously been
evoked in the current discourse, and that the sender therefore be-
lieves is known to the addressee.
d. Addressee-new/discourse-old: Theoretically, information which
has been previously evoked in the current discourse, but which the
sender assumes to be unknown by the addressee. As Prince notes,
this information status does not occur in natural discourse.

The following example will help us work out to which kind of NP refers each
information-status.

(116) Last night a friend called to tell me that on March 19th 2011, the moon
will be the closest it has been to the Earth in 18 years. It will also be
at its fullest. He proposed to go out to the mountain to try the new
camera he just bought.

In (116) a friend is an instance of addressee new/discourse new DR. The

moon, the Earth, and the mountain are three instances of addressee old/
discourse new DRs. And it, its, and he are examples of addressee old/
discourse old DRs. Prince argues that definites do not presuppose that the
DR they denote is discourse-old but rather addressee-old. This resembles
Heim’s claim that DRs that satisfy the familiarity presupposition of definites
130 Deixis and familiarity

need not be introduced by prior mention. Thus Prince’s addressee-oldness is

equivalent to Heim’s familiarity even if the second author does not make this
further distinction. In fact, this distinction is precisely exploited by Roberts
(2003), who argues for a fundamental difference between the “localisation”
of the linguistic antecedent, i.e. whether the antecedent can be found in the
linguistic context or not, which is in fact a consequence of whether the DR
is asserted or presupposed.
Robert’s taxonomy of familiarity is equivalent to Prince’s distinction.
Both pairs of concepts show a distinction between antecedents appearing
or not in the linguistic context, i.e. between existence being entailed from
the context and existence being asserted by means of a full referential NP.
The table below shows this equivalence. Although the terms are equivalent,
in this book I use “weak familiarity” following Roberts to denote cases in
which the DR is definite by virtue of having the corresponding object in
the present environment and no linguistic antecedent is explicitly introduced
into the model.

Table 3. Equivalence of information-status w.r.t. definiteness

Information-status
Prince 1981, 1992 Discourse-oldness Addressee-oldness
Roberts 2003 Strong familiarity Weak familiarity

Data from natural languages provides evidence that different languages have
different requirements on the type of familiarity required for use in their definite
articles. For example, Hidatsa and Ewe use the definite article only anaphori-
cally, i.e. when strong familiarity is satisfied (Lyons 1999: 158). German
also distinguishes between two types of definite articles. Non-contracted
forms, which consist in bare prepositions followed by regular forms of defi-
nite articles, are used in contexts of strong familiarity, whereas contracted
forms (contraction between a preposition and definite article) are only used in
contexts of weak familiarity (Puig-Waldmüller 2008; Schwarz 2009).
Some approaches argue for a theory of definiteness that combines the
two notions, i.e. uniqueness and familiarity.52 Based on corpora work,
Fraurud (1990), Birner and Ward (1998) and Poesio and Vieira (1998) claim
that in order to account for all definite NPs occurrences found in corpus, both
uniqueness and familiarity together must be taken into account. A similar
claim is made by Farkas (2002) and Farkas and de Swart (2007), who assume
that both uniqueness (“maximality”, in their own terms) and familiarity play
a role in definiteness. To include these two semantic properties they use
  Definiteness: background 131

the umbrella term “determined reference”. In the analysis of the possible

encoding of definiteness in LSC locations, the theory needs to be based on
the notion of familiarity, as will be shown in §5.3. Now, the next section is
devoted to the description of deixis, as one of the main means to introduce
definite NPs into the model.

5.2.3. Deixis

Deictic elements directly refer to objects present in the real world and they
have an interpretation related to the spatiotemporal coordinates of the actual
context of utterance, such as I-here-now. By deixis is meant the location
and identification of persons, objects, events, processes and activities being
talked about, or referred to, in relation to the spatiotemporal context involved
in an act of utterance and with the participation of one sender and at least one
addressee (Anderson and Keenan 1985; Lyons 1977). The two properties of
deixis are referentiality, since it is used to refer to, as well as egocentricity,
since it is dependent on a centre of coordinates. Deixis is thus understood in
relation with the canonical situation of utterance in which the communica-
tion occurs in face-to-face interaction.
Traditional grammar distinguishes between deictic and anaphoric uses
of pronouns. The terms are defined as follows. A pronoun is deictic when it
receives its reference from an extralinguistic element, and it draws the atten-
tion to some new object of discourse. And it is anaphoric when it picks up
a DR from the preceding text. However, the distinction is not so clear-cut.
There are contexts where the difference between a deictic and an anaphoric
element is blurred. For instance, if we think of a possible utterance like (13),
we realise that the use of “she” is deictic, since it refers to someone who was
present in the immediate physical context, but it is also an anaphoric use
since it picks up a referent, although not previously introduced. The current
view is that anaphora and deixis should not be distinguished (Heim and
Kratzer 1998; Recanati 2005). This is shown below, where a weakly familiar
DR is referred to.

(117) After someone left the room:

I am glad she finally left.

As (117) shows, not all the information in the context is always linguisti-
cally given. Information in the common ground can sometimes be there by
virtue of the common experience and background of the interlocutors, but
132 Deixis and familiarity

also it can be accommodated on the basis of deixis or inference. Under some

accounts, anaphora is considered to function once DRs have already been
established in the universe of discourse either by being explicitly mentioned
or implicitly mentioned also and then accommodated. Hence anything in the
immediate environment of the sender and addressee towards which the atten-
tion is directed becomes a DR whether it has been explicitly introduced with
a full NP before or not. In fact anything occurring in the surrounding context
that is linguistically mentioned is information added to the common ground
(Heim 1982: 309; Karttunen 1968: 16; Prince 1981; Vallduví 1992: 68). The
knowledge included in the common ground is not null at the beginning of a
discourse since weakly familiar DRs that the addressee has in his knowledge-
store are part of the common ground, including also all the objects that are in
the immediate surrounding context. These contexts have received different
terminology, such as “pragmatic anaphora”, by Partee (1978); “indexicality”
by Nunberg (1993); “hearer-old”, by Prince (1981, 1992), “weak familiarity-
entities perceptually accessible”, by Roberts (2003); or “bridging” by Clark
(1975). As already stated, in this book I use “weak familiarity” following
Roberts to mean the knowledge included in the common ground that has not
been explicitly introduced into the model.
This chapter shows that deixis and anaphora, traditionally considered
different -although related- phenomena, are in fact the same phenomenon.
They are just different means of introducing entities into the discourse model
(Roberts 1998). Both deictic reference and anaphoric reference presuppose that
the DR is already familiar to the audience. In the case of deictic reference, it has
attained familiarity by being pointed at, being perceptually prominent, or being
otherwise salient. In the case of anaphoric reference, it has been made familiar
by previous mention. The deictic use is shown in (118) where the definite
description “the goat” is licensed not by an antecedent NP in prior discourse,
but by the common experience of A and B at the moment of utterance, which
entails the existence of a single, perceptually prominent goat. Both interlocu-
tors are certain that the other has the same goat in mind, so the utterance can
be assumed to be felicitous. In a context like (118) the use of a pronominal
form such as “It stinks” instead of the definite NP would also be felicitous. In
(119) the demonstrative “this” is accompanied by a deictic gesture that brings
the indicated object to the attention of other interlocutors and so introduces the
corresponding DR, which will satisfy the familiarity presupposition of the defi-
nite. It is worth noting that in the previous contexts there is a deictic use (118)
and a deictic term (119). Deictic terms change their reference from context to
context, while this is not the case in deictic uses like (118).
  Definiteness: background 133

(118) A goat walks into the room noisily. A says to B:

The goat stinks!

(119) This [accompanying deixis] is the tool you need to use.

(Roberts 1998: 367)

In line with Roberts (1998) there is no need to establish different types

of definites on the basis of how they find their antecedent (e.g. deictic vs.
nondeictic pronouns). Instead, we consider that they are all free variables with
familiarity presuppositions. DRs satisfying familiarity presuppositions may
get introduced into the discourse context in different ways, one possible way
being deixis. Thus deixis and weak familiarity uses should not be seen as the
means to refer to DRs which have a corresponding object in the immediate
context, but rather the means to introduce DRs into the universe of discourse.
These DRs that may be put in the universe of discourse through deixis are
entities from the physical context and shared knowledge. Anaphoric and
deictic uses are then special cases of the same phenomenon: the pronoun
refers to an individual that, for whatever reason, is highly prominent at the
moment when the pronominal referential expressions is uttered. Prominence
is obtained from different means, namely by recent previous mention and
by being present in the physical context. Indeed, weak familiarity cases do
not differ that much from indirect anaphora as in (120), where the pronoun
is connected to an expression found in the previous context to which it indi-
rectly refers to.53

(120) a. Susan went to the surgery. He gave her pills for her headache.
b. I’ve just been to a wedding. The bride wore blue.
(Consten 2003)

Deixis and anaphora are related to definiteness. They both presuppose that
the DR is already familiar to the audience. Familiarity is obtained in deixis
by pointing at something that is perceptually prominent and in anaphora by
pointing at some previous linguistic mention. So far the fundamentals about
definiteness relevant for the present account have been exposed and they
are summarised in (121). They are the basis on which the present analysis
applied to LSC relies.

(121) a. Definites and indefinites have different conditions attached: Indef-

inites are novel to the discourse, and definites are familiar.
134 Deixis and familiarity

b. Asserting the existence of the DR in the model creates strong fa-

miliarity (or discourse- oldness), while presupposing the existence
of the DR in the model creates weak familiarity (addressee-old-
ness).
c. Deixis is another means to introduce a DR into the model, which is
presupposed to be already in the common ground.

To describe how definiteness is distinguished in the use of LSC signing space

and how deictic uses also play a role in the introduction of DRs in the model
in LSC is the main goal of the chapter. But before delving into it, this section
concludes with a state of the art of the description of definiteness marking
in SLs.

5.2.4. Definiteness in sign language

Whether definiteness is grammatically encoded in sign languages (SLs) is

still a matter of debate among SL linguists. Studies on sign language definite-
ness are very scarce. So far, only descriptions of how definiteness is expressed
in ASL and Hong Kong Sign Language (HKSL) are available. According to
some works, in ASL an index sign directed to space in a prenominal position
is considered to be the formal marking of definiteness (Bahan et al. 1995;
Bahan 1996; MacLaughlin 1997; Wilbur 2008). ASL marks indefiniteness
with an upward direction of manual and nonmanual mechanisms, which
establish a spatial region rather than an area (MacLaughlin 1997). Indefinite
NPs are established on the upper part of the frontal plane with the determiner
something/one, which is an index finger pointing upwards very similar to
the numeral one. The difference is that something/one involves a slight
circular movement of the forearm and hand. Definiteness is marked with an
index pointing towards the lower part of the frontal plane, whereas indefi-
niteness is marked with an index sign co-occurring with a darting eye gaze
directed towards the upper part of the frontal plane. A further (in)definiteness
distinction is that while definite determiners in ASL access a point in space,
indefinite determiners involve an articulatory movement within a spatial
region rather than a point (MacLaughlin 1997: 129). As for HKSL, Tang and
Sze (2002) describe a similar indefinite determiner as the one described for
ASL. It is articulated with the same handshape used for the definite deter-
miner (e.g. index handshape), but the index finger points upward. Unlike
the indefinite determiner in ASL, in HKSL it does not involve a tremorous
motion. When this sign is articulated, eye gaze is never directed to space
  Definiteness: background 135

but instead towards the path of the hand, suggesting that there is no spatial
location established for the DR. This is important and we will see that LSC
shares this upward darting eye gaze, as well as the weak establishment of
a spatial location. However, in LSC this formal marking does not denote
indefiniteness, but rather non-specific reference, as shown in Chapter 6.
Definiteness distinctions are not only marked in the manual component,
but also in the nonmanual one. According to Tang and Sze, the (in)definite-
ness distinction is marked in the eye gaze behaviour. The indefinite deter-
miner in HKSL is distinguished from the definite one following the eye gaze
which co-occurs with it: while definite determiners co-occur with an eye
gaze directed to the spatial location, for indefinite specific DRs eye gaze is
directed towards the addressee (Tang and Sze 2002: 303). Hence in HKSL
the (in)definiteness distinction is formally marked in the nonmanuals, and
more concretely by eye gaze.
Figure 37 is a representation of the definiteness distinctions projected into
space that the above-cited works on ASL and HKSL describe. The upper
part of the frontal plane is the extended area (represented with a big ellipse)
where indefinites are localised. The lower frontal plane is the more reduced
area (represented with a point) where definites are established. As will be
shown in this book, definiteness marking in LSC differs from the picture in
Figure 37, since both definites and indefinites can occupy a location on the
lower frontal plane. The upper frontal plane in LSC is in fact reserved for a
subtype of indefinites, namely non-specific only (see Chapter 6).

Figure 37. Definiteness marking on the frontal plane in ASL and HKSL

However, other authors have questioned the definiteness marking of index

signs. In fact, according to some works, definiteness is not encoded in SLs
(Engberg-Pedersen 1993, 2003, for Danish SL (DSL); Rinfret 2009 for
Quebec SL (LSQ); Winston 1995 for ASL). Engberg-Pedersen (1993: 101)
136 Deixis and familiarity

argues that in DSL DRs with high discourse value are more likely to be repre-
sented by a location than DRs with a low discourse value.54 According to her,
discourse value in DSL is measured following the number of repetition of
mentions of the DR. Winston (1995: 109) also ascribes to spatial locations
in ASL the potential of marking discourse value. The location in space itself
is a marking of topic continuation as a consequence of its discourse-status
marking. If the entity is not established, it means that it is an unimportant
entity and the discourse will not be centred on it.
As will be shown in this chapter, in LSC the mere localisation does
not stand for the marking of definite NPs, since also indefinite NPs can be
established in space. The argumentation used in this book moves away from
discourse value, and it is based on definiteness and specificity phenomena, as
well as topicality. Although I agree with Engberg-Pedersen and Winston that
locations denote topicality of the entity, I also introduce specificity marking
into the picture. My explanation is more indirect but also more complete,
since definiteness, specificity and topicality are considered when analysing
the semantic attributes DRs may have in order to have a corresponding
spatial location established. Furthermore, the above-cited works do not
provide a formalisation of discourse value. The theoretical background DRT
used in this book provides a detailed framework that provides the tools for an
implementation to concretely define these notions without having to resort to
vague notions such as “discourse value”. One of the goals of this book is to
offer a clear formalisation of how signing space is used in discourse.
After this state of the art of definiteness marking in space in SLs, we now
turn to the language object of this study, namely LSC. As it will be shown in
the next section, both definites and indefinites are established in signing space,
and hence both presupposition of existence (for weakly familiar DRs) and
assertion of existence (for strongly familiar DRs) are equally marked in space.

5.3. Anaphoric deixis in LSC

In this section we will see that newly introduced DRs establishing (p) do
not need to have an overt antecedent. As shown in §5.1.2, weak familiarity
applies to contexts where the antecedent is not overtly expressed, but rather
inferred from the contextual environment. The DR is incorporated into the
model by means of accommodation.
To the best of my knowledge, there is only one work in sign language where
deictic references are considered to be first mention references (Pizzuto et al.
2008, following Lyons 1977). Other works consider that in deictic frames of
 Anaphoric deixis in LSC 137

reference, signers point in the direction of objects in the context of utterance.

In these contexts, the frame of reference is determined by the actual locations
of the objects to which the signer refers (Cormier 2007; Engberg-Pedersen
1993). The more extreme description is that presented by Liddell (2003), who
argues that the directionality of pointing signs is crucial for the understanding
of reference. According to him, ASL pronouns physically point to their DRs
and their significance can only be determined by the directionality in signing
space (see §2.4.1.2, for the constructive criticism of the spatial mapping view).
The present account claims that the interpretation of index pronominal
signs is not done by the directionality or the action of the physical index sign,
but by the selection of a suitable DR among a set of context-available infor-
mation. The selection of the specific DR that will be chosen in every context
depends solely on the linguistic context, where syntactic-semantic-pragmatic
motivations are included. As mentioned in §2.4.1.2, many counterexamples
are found in which the directionality of index signs does not precisely match
the intended DR. Also contexts of indirect reference show that the physical
directionality cannot be a reliable clue when linking the pronominal form
with its corresponding DR. Hence, linguistic and non-linguistic motiva-
tions, such as syntactic-semantic-pragmatic, shared knowledge and physical
environment need to be considered. To show that directionality cannot be a
reliable clue, let us consider the example shown in Figure 11 in Chapter 2,
repeated here as Figure 38 for convenience. The context is the following:
the two main deaf clubs in Barcelona, namely Cerecusor and Casal, are
organising a joint party with all the members. The president of the Catalan
Federation for the Deaf is telling the organising committee the tasks that
each club must undertake. It is a mixed committee formed by four members
of Cerecusor and three members of Casal. The setting is shown in Figure 38,
where y corresponds to Cerecusor members and x to Casal members. The
signer in front of them is the president, who tells the following:

President
Figure 38. Mixed position of the members of two deaf clubs
138 Deixis and familiarity

(122) association cerecusor responsible food drink. association

casal responsible organisation.4-ix2 buy thing++. 3-ix2
stay sign theme organise strategy how.
‘Cerecusor will be responsible for food and drinks. Casal will take
care of the organisation. You-four go and buy everything. You-three
may stay here and talk about organisational issues.’

In such contexts the incorporated pronouns cannot be understood according

to the directionality of signs because, as shown in the setting in Figure 38,
the members are mixed and not seated next to each other. When the president
utters 4-ix2 (‘you four’), he directs a four-handshape with an arc-movement
towards the front. This pronominal form refers to entities marked as y in the
figure above. But they are not seated together, and thus the arc-movement
is directed towards the whole group. The same goes for the pronoun 3-ix2
(‘you three’), which denotes the entities marked as x. Thus the arc-movement
must be a circular one but not directed to any concrete area. The interpre-
tation of these pronominal forms must be done following linguistic clues.
Indeed, this is a case of weak familiarity since no linguistic antecedent for
the pronouns is previously introduced. As already mentioned, this deictic
use is similar to indirect anaphora previously exemplified in (120), where an
element from previous discourse functions as an anchor. In (122), the two
DRs“Association Cerecusor” and “Association Casal” serve as anchors to
which pronouns are linked to, although they have not been previously local-
ised in signing space. The directionality of the pronominal form is in fact
misleading for the right interpretation and the interpretation cannot only rely
on direct deictic cues.
As already claimed in §2.4.1, Liddell (1995) argues that when surro-
gate and token space are used grammatical reference functions in the same
way as with real space because signers imagine surrogates and tokens as
being in an unlimited number of locations and therefore treat them as if they
were physically present. Thus, any kind of reference for Liddell is articu-
lated like reference to real space: deictic and not anaphoric. His account
does not make a distinction between presence and absence of objects in
the immediate context, and the actual world and the discourse model are
fused. However, in order to account for reference to entities which do not
exist (i.e. like unicorns) or also for quantified expressions, a clear distinc-
tion must be made between what exists in the real world and what exists
in the discourse. While Liddell’s account considers all the references to be
deictic, this book considers all the references to be anaphoric to the discourse
model. According to the present account, reference is always anaphoric to
 Anaphoric deixis in LSC 139

the model. That is, referential expressions refer to entities present in the
discourse model, without regard to the mechanism used introduce them into
the model. This is shown with the minimal pair (123) and (124) below. In
the first example the goat entering the room is referred to with a definite
description, while in the second example it is referred to by a demonstra-
tive pronoun. (123) is a context where the pronoun in the second sentence is
linked to the linguistic antecedent introduced.

(123) Uttered while a goat is entering the room.

ix3a goat smell-bad. but ix3a beautiful.
‘The goat stinks. But it is beautiful.’

(124) ix3a smell-bad. but ix3a beautiful.

‘It stinks. But it is beautiful.’

Since a noun is introduced, the corresponding construction rule determines

the establishment of a DR in the DRS and the corresponding predicative
condition. The pronoun in the second sentence in (123) derives from the use
of the construction rule for pronouns (see Appendix for the complete list of
construction rules used in this book):

CR.N: Upon encountering a common noun co-occurring with a determiner,

1. trigger the syntactic configuration [s NPα [VP]] or [s VP [NPα]], and
2. introduce a novel discourse referent α into the main DRS, and
3. introduce the predicate condition β(α)

CR.PRON: Upon encountering a pronominal form,

1. trigger the syntactic configuration [s NPα [VP]] or [s VP [NPα]], and
2. introduce a novel discourse referent α into the main DRS, and
3. check which variable in the main DRS shares the features α has, and
4. if no suitable variable is found, go to CR.PRON2; if the suitable variable
is found introduce an identity equation α = γ
5. go to CR.PROM55

A suitable antecedent is found because of the coincidence in the direction

of (p), according to the hypothesis mentioned in the preceding chapter and
repeated below.

(125) The spatial point hypothesis (first version)

The identity condition in the DRS is encoded through coincidence in
direction of spatial establishment of (p).
140 Deixis and familiarity

Hence the identity condition is created and resolved. The resulting DRS is
shown in (126).

(126) x y
goat (x)
stink (x)
it (y)
beautiful (y)
y=x

But the same goat within the same context can also be referred to by a
pronoun (20). This is a context of a weak familiar DR, where no linguistic
antecedent is introduced. Instead the goat is accommodated into the model
by being perceptually accessible to the two conversation participants. For
the construction of the semantic representation, first the construction rule
for pronouns is used. But since no suitable variable is found, the algorithm
requests to go to the second construction rule for pronouns CR.PRON2.

CR.PRON2: Upon encountering a pronominal form,

1. go to the main DRS and take the default variable δ,
2. introduce an identity equation α = δ
3. go to CR.PROM

Weakly familiar reference corresponds to contexts where the DR is very

prominent in the surrounding context. That is, it is directed to an entity that
attracks the attention of the conversation participants. As claimed in §5.1.3,
many authors accept that anything occurring in the surrounding context is
information added to the common ground (Heim 1982; Karttunen 1968;
Nunberg 1993; Partee 1978; Prince 1981; Roberts 2003; Vallduví 1992).
However, the formal representation of weak familiarity in DRT has not been
implemented yet. As CR.PRON2 in the present account states, there is a
default variable δ that stands for all the DRs introduced into the common
ground without an explicit linguistic antecedent. This default variable is
identified with the variables in deictic contexts and with weakly familiar
DRs. The final semantic representation is shown in (127).
 Familiarity 141

(127) y, w
it (y)
stink (y)
it (w)
beautiful (w)
y=δ
w=δ
w=y

So far, deictic uses have been proven to be another means to introduce DRs
into the model, which are expressed by definite NPs by virtue of having
the corresponding object present in the immediate physical situation. The
identification between the variable being overtly expressed in discourse with
the default variable associated with entities introduced without an explicit
linguistic antecedent is a very common use in LSC. These contexts of weak
familiarity show that reference in LSC is always referred to the discourse
model and it is thus anaphoric.

5.4. Familiarity

Leaving intensional contexts aside, indefinite NPs generally denote an asser-

tion of existence. That is, they establish a DR into the model by asserting that
such an individual exists. On the other hand, definite NPs imply a presuppo-
sition of existence. That is, the interlocutor is able to infer that the DR refers
to a non-empty set. Recall that a broad definition of familiarity is considered
here (see §5.1.2), including not only strongly familiar DRs for which there
exists a linguistic antecedent in the discourse, but also weakly familiar ones
where no linguistic antecedent is present. In LSC the mere localisation of
a DR in space does not formally denote a definite reading. The distinction
to show both definiteness and indefiniteness marking established in signing
space is implemented here with respect to the status of the DR in the model.
That is, whether the DR is presupposed or asserted. As shown below, both
possibilities establish (p) in LSC.
In this section we will see that the establishment of (p) in LSC does not
denote a difference between definite and indefinite NPs as shown for other
sign languages (unlike in ASL and HKSL, as seen in §5.1.5). It is also shown
that when (p) is established for a presupposed DR, this is better explained in
terms of familiarity.
142 Deixis and familiarity

5.4.1. Assertion of existence

An example of assertion of existence marked in space is (128). The utterance

is restricted to a novelty condition, since it is the first introduction of the
DR “man” into the common ground. The first sentence is a body-anchored
localisation co-occurring with role shift (see §3.5). The NP “one man” is
not localised on the lateral parts but it is rather a body-anchored location
(see §3.5). The predicate walk and the entity classifier move forwards from
the body of the signer to the centre of signing space. The introduction of
the second DR “another man” in the second sentence is done by means
of a spatial location established on the ipsilateral part, which also co-occurs
with role shift. They are both indefinite NPs, which are introduced with two
indefinite determiners, namely one and another (128). However they are
both localised: the first is body-anchored and the second is localised on a
lateral part. This shows that indefinite NPs in LSC asserting the existence of
the DR into the discourse may be spatially established.

br rs
(128) one man 1-walk cle.long-upward-entity-advancing 1-see-3 tree
cle.tree bike cle.tree/cle.bike.
‘There is one man and he is walking. He sees a bike leaning against
a tree.’
[...]
br rs
anotherip-l man 3-walkip-l cle.long-upward-entity-advancingip-l
1-see-3 bike ix1 poss.
‘There is another man who also walks there and realises that the bike
is his.’

The corresponding semantic representation in (129) shows the introduction

of the two corresponding variables for the two men, namely x and z. Its
assertion in the discourse is done in the two instances by means of a topical-
ised clause as shown in (128). The first one with the NP “one man” local-
ised as body-anchored is marked with brow raise, and the second sentence
is introduced with the NP “another man” and (p) corresponding to this
DR is localised on the ipsilateral side and co-occurring with brow raise and
a pause after the NP.
 Familiarity 143

(129) x, y, z
man (x)
walk (x)
bike (y)
lean-against-tree (y)
see (x, y)
man (z)
see (z, y)

Both DRs denoted by the NPs in (128) are not presupposed to exist in the
model previously to its introduction. They follow a novelty condition. It is
the beginning of a story in which the signer introduces the characters that
participate in it through two localised NPs.
Localised indefinite NPs may also be uttered as non-topicalised elements.
In (130) an indefinite NP is the internal argument of the verb offer. It is
localised on the ipsilateral side. This is the first time the signer introduces
this DR and she does so by means of an indefinite NP containing the indefi-
nite determiner one.

eg:ip-l
(130) ix1 1-offer-3 one person-3ip-l pen-drive computer
eg:ip-l
pen-drive 1-offer-3ip-l.
‘I will offer a pen-drive to a person/someone.’

(131) x, y, z
person (y)
pen-drive (x)
offer (1, x, y)

Indefinite NPs in LSC can thus be localised in signing space and establish
(p). This shows that assertion of existence in LSC can occur with the nominal
established in signing space.
144 Deixis and familiarity

5.4.2. Presupposition of existence

The establishment of (p) also occurs with DRs which are presupposed to
exist in the model. Both definite NPs with a previous explicit antecedent
(strongly familiar DRs) and definite NPs without a previous antecedent
(weakly familiar DRs), can establish (p). By the mere appearance in the
discourse, the signer is intending to refer to the non-empty set to which
the nominal refers to. Hence their existence is presupposed in the common
ground and it is thus familiar to the conversation participants rather than
referring to a unique DR.
The example in (132) shows a combination of the two possibilities
(i.e. assertion and presupposition). The first sentence asserts that a DR that
has not been previously mentioned is introduced into the model. The DR is
localised on the ipsilateral side. The second sentence contains a resumptive
pronoun referring back to the same DR. Hence it presupposes that the DR
exists in the model and it is a familiar entity for the conversation participants.
Both weakly and strongly familiar DRs are spatially established.

(132) today interview one person-3ip woman. ix3ip knows english.

‘Today (I) have an interview with a woman. She knows English.’

As far as weakly familiar DRs are concerned, their existence is entailed in

the context, either because the corresponding object in reality is perceptually
accessible or because the DR belongs to the general encyclopaedic knowl-
edge. An example of the former is given in (133). The signer is talking about
a laptop that is in front of him. Even if the object is around it must be intro-
duced into the discourse model and so he introduces the nominal for the
DR occurring with an index sign articulated with pronated palm, emphasis
and a slight repetition of movement. The predicative index sign alone in the
indexing clause is sufficient to access a DR that is perceptually accessible
and hence familiar to the discourse participants.

(133) ix3a laptop.

‘There is a/this laptop.’

Concerning DRs that belong to the general encyclopaedic knowledge, a

spatial location is also established. In daily conversations and in Webvisual56
many references to DRs that are introduced for the first time in the discourse
(i.e. discourse-new) but which are known by the participants (i.e. addressee-
old) are established in signing space. In such cases, the nominal denotes a
 Familiarity 145

DR that belongs to the general knowledge or that is contextually entailed.

Thus DRs for “the Pope”, “Laporta” and “Guardiola” (the two previous pres-
idents of Football Club Barcelona), “Montilla” (the previous president of the
Generalitat de Catalunya), and “Antonio Martínez” (the current president of
the Catalan Federation for the Deaf) are localised in space when appearing
for the first time in discourse. In (134), for instance, the DR “Hitler” is local-
ised on the ipsilateral side when first introduced into the discourse.

(134) ix3c found organise mateix person-3ip hitler.

‘This was founded by Hitler himself.’

Other examples containing proper names which by themselves already imply

a presupposition of existence can also be established in space. Proper names
also co-occur with an index sign in LSC and establish (p). As shown below,
the DR denoted by the proper name can be established with a spatial location
on the horizontal plane.

(135) yesterday joan ix3ip 3ip-tell-1 ix3ip thesis soon finish.

‘Yesterday Joan told me he will finish his thesis soon.’

The previous examples from (128) to (135) show that in LSC there is no
distinction between assertion and presupposition of existence of DRs
concerning the spatial localisation, since both asserted and presupposed DRs
can be established in space. Both asserted and presupposed DRs are intro-
duced as variables that occupy a localisation in space as shown in Figure 39,
and hence the novel/familiar condition is thus not formally distinguished in
LSC spatial marking.

Figure 39. (In)definiteness marking in LSC signing space

146 Deixis and familiarity

The examples denoting a presupposition of existence of the DR in the

common ground show that when analysing LSC locations in terms of defi-
niteness, it is better to account for them as considering the phenomenon of
familiarity. By the mere appearance in the discourse of the DR, the signer
intends to refer to the non-empty set to which the nominal refers to. Hence its
existence is presupposed in the common ground and it is thus familiar for the
conversation participants rather than referring to a unique entity. Uniqueness
contexts are not always constrained by the localisation of the DR, since both
unique and non-unique DRs do not need to be localised. As shown, in (136)
the DR corresponding to the sun, which is considered to be a unique entity,
is not associated to a spatial location. A uniqueness account does not apply
when localisation of DRs needs to be accounted for.

(136) ix morning sun cle.sun-raise time six half.

‘Today the sun has risen at half past six.’

5.5. Indefiniteness marking

Although definiteness marking is not expressed in space, LSC has other

strategies to denote indefiniteness, such as the use of a concrete set of deter-
miners, as well as a specific nonmanual marking, as briefly exposed below.

5.5.1. Indefinite determiners

LSC has a specific set of determiners, which denote indefiniteness. They indi-
cate that the DR of an expression is presumed to be not identifiable or familiar.
The entity is thus not part of the common ground of the discourse. Some of
these determiners are for instance, any, some, one, one+++(‘few’), just to
list some of them (this is, of course, not an exhaustive list). They can have a
strong and a weak reading depending on the part of the frontal plane where
they are established. They are treated in detail in §6.3.2. Besides the manual
marking, indefiniteness has a particular nonmanual marking, as described in
the following subsection.

5.5.2. Nonmanual indefiniteness marking

Nonmanual marking is also a crucial part of the grammar of sign languages

(see Pfau and Quer 2010, for an overview of nonmanuals). In LSC nonmanuals
 Summary 147

also play a role in the encoding of information structure, since indefiniteness

is expressed with a specific nonmanual marking. It is articulated on the lower
part of the facial expression and it consists in sucking the cheeks in and
pulling the mouth ends down. This is sometimes combined with a shrug. The
facial expression is shown in Figure 40.

Figure 40. Indefiniteness nonmanual marking

Interestingly, this nonmanual articulated on the lower part of the facial

expression provides semantic-pragmatic information (see Wilbur (2000),
who claims that the nonmanuals from the upper part of the face are used
for syntactic information (i.e. affirmation, negation, topics, conditionals),
whereas the nonmanuals from the lower part of the face are used to provide
adverbial modification (i.e. adjectives, adverbials)).
An interesting question that arises and which is outside the scope of this
book is how this indefinite nonmanual marking is compositionally combined
with other markings, as well as how the pulling down of the cheeks is
combined with shrug, or whether there is a different meaning attributed to
the two of them. Since the main focus of research of the present book is how
(in)definiteness with respect to the use of signing space is expressed, this
interesting issues have only been treated descriptively and will have to be
treated in deep in future research.

5.6. Summary

In this chapter the main features characterising the localisation of DRs in

signing space in LSC and their relation with the expression of definiteness
have been analysed. We have seen that, although (in)definiteness distinctions
can be expressed with a restricted set of determiners and a specific nonmanual
marking, the establishment of the DR in signing space does not distinguish
between definite and indefinite NPs. The features of localisation of (p) are
the following. First, objects referred to by means of weak familiarity do not
148 Deixis and familiarity

have an explicit linguistic antecedent, but they have a corresponding DR in

the common ground. The pronominal form in weak familiarity contexts does
not deictically refer to the present object, but rather to the DR they are linked
to. This shows that all references are anaphoric to the discourse model and
that even deictic uses are anaphoric to the entities present in the discourse
model. Second, the localised DR is not marked as definite or indefinite, and
both asserted and presupposed DRs are localised in space. As for presup-
posed DRs, the establishment of (p) is better explained in terms of familiarity
in the discourse model. LSC differs from other SLs, like ASL and HKSL,
since no definiteness distinction is found with respect to the localisation of
DRs.
However, concerning indefinite NPs there is a further categorisation,
which is that of specificity. Specific indefinites are used to indicate that
the DR is known to the sender, though not to the addressee. Non-specific
indefinites are used when neither the sender nor the addressee know the DR.
Hence while definiteness implies givenness for both interlocutors, specificity
implies accessibility to the addressee alone. Definiteness and specificity are
closely connected to scope. Generally, while definite and specific indefinites
correspond to permanent DRs, non-specific indefinites introduce temporary
DRs, which survive only under the scope of an operator. This is precisely the
main topic of the next chapter, where we will see that in fact LSC locations
mark distinctions denoting specificity.
Chapter 6
Specificity

The notion of specificity in linguistics is notoriously non-specific.

(Farkas 2002)

6.1. Introduction

This chapter is devoted to the expression of specificity in relation to signing

space. As seen in Chapter 5, in Catalan Sign Language (LSC) there is no
distinction between the marking of definite and indefinite noun phrases
(NPs) upon signing space localisation. However, in this chapter it is shown
that in LSC there exists a particular marking to denote (non-)specificity. The
phenomenon analysed here is the possibility of establishing spatial locations
on the upper frontal plane as opposed to being localised on the lower frontal
plane. I will argue, as already advanced in §3.2.2.3, that this distinction is
related to the expression of specificity. In LSC two kinds of localisation on
the frontal plane are found, namely a strong and a weak localisation, which
correlate with a specific and a non-specific interpretation, respectively. Strong
localisation is instantiated by the feature [low], which is formally marked by
the default location (p), while weak localisation is instantiated by the feature
[up], formally marked as (p)[up]. The formalisation offered to explain this
distinction is framed within the distinction between main and subordinate
variables in a Discourse Representation Structure (DRS): whereas main
DRS variables represent a specific interpretation and they are expressed with
spatial locations established on the lower frontal plane, subordinate DRS
variables represent a non-specific interpretation and they are expressed with
spatial locations on the upper frontal plane. Hence the main/subordinate DRS
distinction, which is associated with wide and narrow scope respectively, is
overtly encoded in the use of signing space in LSC. The chapter is organised
as follows. §6.1 presents the background concerning specificity relevant for
the present account. In §6.2 the different properties encompassed by speci-
ficity with respect to LSC data are presented. The LSC localisation pattern,
which is associated with specificity marking, is offered in §6.3. §6.4 summa-
rises the main findings of the chapter.
150  Specificity

6.2. Specificity: background

Noun phrases can be categorised with respect to definiteness and indefinite-

ness, as seen in Chapter 5. Indefiniteness can be further divided with respect
to specificity depending on the knowledge that the sender and the addressee
have about a Discourse Referent (DR). Specific indefinites encode that the DR
is known only by the sender, but not by the addressee. Non-specific indefinites
encode that the DR is not known by the sender or the addressee. Specificity
is encoded differently in each language. Some languages encode it in the
article system, others encode it with affixes and others lack encoding of this
semantic-pragmatic notion. Samoan and Maori are two Polynesian languages
with an article system that distinguishes specificity rather than definiteness
(Lyons 1999). Samoan uses the article le with specific DRs, which indicates
that the NP refers to one particular entity regardless of whether it is definite
or indefinite. A different article (se) is used with non-specific DRs, which
do not refer to a particular, specified item (Mosel and Hovdhaugen 1992,
cited in Lyons 1999: 57). In Maori, the article he (which does not distinguish
number) is used when the kind of entity is crucial, and teetahi/eetahi when
the number is significant (Bauer 1993, cited in Lyons 1999: 59). The mean-
ings and patterns of use of Maori articles are not yet established, but it seems
that its article system relates partly to the distinction between specific and
non-specific, rather than definite and indefinite.
Another way of marking specificity is by means of affixes. Turkish,
for instance, encodes specificity with an accusative affix. NPs with overt
case morphology are specific, and NPs without case morphology are non-
specific (Enç 1991). This is shown in §6.1.2, where Turkish specificity is
treated in detail. This overt distinction differs from languages like Catalan
or English where specificity is not overtly marked in the determiner system.
Indefinite NPs in Catalan and English are thus ambiguous between having
a specific or a non-specific interpretation. It is only in coreferential chains
that the resumptive pronoun disambiguates the two readings. Let us look
at some examples. In English the indefinite determiner a is used both for
specific and non-specific NPs. (137) has two possible readings: a specific
and a non-specific one. Yet specificity in English has observable effects on
co-reference, and the resumptive pronoun disambiguates the two possible
readings (Partee 1970). Under the specific reading, the indefinite NP refers to
an identifiable book (137a). Under the non-specific reading, Celia is looking
for an element of the kind “syntax book”, but there is not any concrete book
that the speaker has in mind when uttering (137b).
  Specificity: background 151

(137) Celia wants to read a book about syntax…

a. but she cannot find it.
b. but she cannot find one.

Specificity encompasses different but related properties, such as scope, parti-

tivity and identifiability. In the following a subsection is devoted to each
property in detail.

6.2.1. Scope

An expression α is in the scope of an expression β iff the interpretation of

α is affected by the semantic contribution of β (Farkas 2000). Scopal speci-
ficity is defined in terms of the interpretation of the indefinite NP outside the
scope of an operator. According to this view specificity is equated with wide
scope (Farkas 1994, 1997, 2002; Ionin 2006). Indefinite NPs that are outside
the scope of an operator are considered to have wide scope, while indefinite
NPs under the scope of an operator are considered to have narrow scope. As
previously seen, indefinite NPs in English are ambiguous between a definite
and an indefinite reading. In (137a) the indefinite NP has wide scope, as
represented in (138a) below, and a specific interpretation arises. In contrast,
in (137b) the indefinite NP has narrow scope, as shown in (138b), and thus it
yields a non-specific reading.57

(138) a. ∃x(book (x) ˄ (celia (y) ˄ want-read (y, x)) ˄ cannot-find (y, x))
b. want-read (y, x)(∃x(book (x) ˄ celia (y) ˄ cannot-find (y, x))

In Discourse Representation Theory (DRT) specificity is treated as a scope

phenomenon (Kamp and Bende-Farkas 2006; Kamp and Reyle 1993). The
narrow/wide scope divide is implemented with the positioning of the vari-
ables in the boxes. Wide scope is represented when the variable is inserted in
the main DRS and all the variables contained in it are under the scope of the
main universe of discourse. Narrow scope is represented when a variable is
inserted in an embedded box. To illustrate this point let us look at an instance
of a sentence with two readings.

(139) Every boy in Joana’s class likes a girl.

152  Specificity

Within the specific reading, which can be paraphrased as “There is a girl

such that every boy in Joana’s class likes her”, the DR corresponding to the
indefinite NP is represented with a variable in the universe of discourse, and
it is in a higher position in the DRS. All the embedded variables are within
the scope of the main variable.
(140) z u y
joana(z)
z’s class (u)
girl (y)
x like (x, y)
boy (x) →
x in u

Under the specific interpretation the variable is represented in the main DRS
and it is a global DR. It can be picked up by further pronominal reference and
(141) can be a felicitous reading of (138).

(141) Every boy in Joana’s class likes a girl. She flirts with a different boy
every day.

Differently, within the non-specific reading, paraphrased as “For every boy

in Joana’s class there’s a girl such that he likes her”, the non-specific DR
does not appear in the main DRS, but only embedded in the DRS for the
consequent (i.e. the embedded right box).

(142) z u
joana(z)
z’s class (u)

x y
boy (x) → girl (y)
x in u like (x, y)

The variable for the non-specific reading of the indefinite is a local DR,
which is represented in an embedded box. Thus it is only available to be
further picked up as long as it is under the scope of an operator.58 Outside this
  Specificity: background 153

scope, a continuation with a resumptive pronoun is not felicitous (143). Let

us continue with the second property related to specificity, namely partitivity.

(143) Every boy in Joana’s class likes a girl. # She flirts with a different boy
every day.

6.2.2. Partitivity

A partitive indefinite is an indefinite NP that has a restricted set as a possible

value. Indefinite NPs receive a partitive interpretation when the denotation
of the NP is included within a given set. In English, for instance, sentences
like (144) are examples of overt partitives. The partitive and non-partitive
pairs in (144) and (145), respectively, are quite similar in interpretation. The
main difference is that in the case of overt partitives (144), the quantification
necessarily ranges over some specific, non-empty, contextually fixed set.

(144) a. Three of the books

b. One of the books
c. Some of the books

(145) a. Three books

b. One book
c. Some books

Enç (1991) views specificity as partitivity. She argues that in some languages
NPs in certain positions are always unambiguous with respect to speci-
ficity. The ambiguity is resolved through case marking: NPs with overt case
morphology are specific, and NPs without case morphology are non-specific.
An example of this phenomenon is Turkish where specific indefinites are
marked with accusative case. Such indefinites denote members of a previ-
ously mentioned set. For instance, the presence of accusative case on an
indefinite yields a partitive interpretation (146), as opposed to the minimal
pair without the accusative case (147).59 The indefinite NP with accusa-
tive case has a covert partitive reading, and it introduces into the domain of
discourse individuals from a previously given set.60

(146) Iki kiz-i taniyordum. (147) Iki kiz taniyordum.

Two girl-Acc I-knew Two girl I-knew
‘I knew two of the girls’ ‘I knew two girls’
(Enç 1991: 6)
154  Specificity

In short, indefinite partitives such as “three of the books” refer to a subgroup

of the referent of the NP contained in the partitive, in this case “the books”.
Partitive specifics induce a presupposition that there is a non-empty and
contextually salient set. Under this view, the specificity of the NP places
a constraint on the structure of the domain of discourse in addition to the
constraint placed by the definiteness of the NP. Partitivity quantifies over
contextually given sets. By contextually given it is meant ‘already in the
domain of discourse’ (i.e. in the common ground, see §4.1.1), since the
contextually relevant individuals are those that have been previously estab-
lished in the discourse, or also incorporated to the model by means of accom-
modation. Hence the set may be accommodated, explicitly mentioned or part
of a contextually determined set. The third and last property is related to
identifiability.

6.2.3. Identifiability

Identifiability, also known as epistemic modality, is another phenomenon

related with specificity. It is defined as the property of those indefinite NPs
that are identifiable by the sender, i.e. those entities that are known and/
or inherently identifiable (Fodor and Sag 1982; von Heusinger 2002, 2008,
2011b; Kamp and Bende-Farkas 2006). The following example shows this
distinction. While (148a) corresponds to an epistemically specific DR and
it is thus identifiable by the sender, (148b) corresponds to an epistemically
non-specific and thus unidentifiable DR.

(148) a. A student cheated on the syntax exam. It is the blond lady that al-
ways seats on the back row.
b. A student cheated on the syntax exam. I wonder who it was.

It is commonly assumed that in English adjectives such as certain, specific,

and particular form specific NPs. The insertion of these adjectives in (148a)
makes the sentence felicitous, as in (149a). However this is not the case in
non-specific readings like (148b). The insertion of the adjective blocks the
non-specific interpretation (149b).

(149) a. A certain student cheated on the syntax exam. It is the blond lady
that always seats on the back row.
b. A particular student cheated on the syntax exam. #I wonder who
it was.
  Specificity: background 155

Partee (1970) proposes to collapse the specific use of indefinites with a

referential use in the sense of Donnellan (1966), and the non-specific use
of indefinites with an attributive use. Donnellan argues for a distinction
between referential and attributive uses of a NP also related with identifi-
ability. Referential NPs are used to refer to particular individuals (i.e. with a
specific reading), whereas attributive NPs refer to non-particular individuals
(i.e. with a non-specific reading). However, the existence of such an indi-
vidual is presupposed in both interpretations. The well-known example by
Donnellan in (150) is analysed as having two readings. In the first reading, the
NP is interpreted as referential and hence as specific. In the second reading,
the definite description is interpreted as predicative, and as a consequence
there is no such specific individual in the mind of the person uttering the
sentence, but rather it is implied that the task this individual has undertaken
is that of having murdered Smith.

(150) Smith’s murderer is insane.

(Donnellan 1966)

The identifiability property is then based on knowledge of the DR and on

referential and attributive uses of NPs. However, as Geurts (1999) claims,
the identifiability view of specificity based on knowledge of the DR is quite
vague, since it is very difficult to determine what a sender has in mind. Of
course to determine what is part of the common ground is also a difficult
task, but it is nothing compared to achieving a definition of having some-
thing in mind. At least, to be part of the common ground can be diagnosed
by means of some tests (such as being available for anaphoric uptake), but
to the best of my knowledge no diagnostic test has been established to deter-
mine whether some referent is in the mind of someone (if it is not just by
directly asking him). Since epistemic identifiability is closely connected to
the scope of the variable, in order to distinguish between identifiable and
non-identifiable DRs I use scope as criterion. Scope can be formally proven
without having to resort to the opacity of the mind of the sender that identifi-
ability by itself encompasses.
It is important to note that von Heusinger (2008, 2011b) considers
noteworthiness to be another property encompassed by specificity. Since
remarkable information about the specific DRs is usually provided along
the discourse, wide scope variables have a noteworthy feature. However
cases of modal subordination show that narrow scope can also be related to
noteworthiness, and thus both wide and narrow scope variables can be note-
worthy. This is the reason why I do not consider noteworthiness to be only
156  Specificity

a property of specificity, but rather a property related to discourse structure

that is orthogonal to specificity, since both specifics and non-specifics may
be noteworthy. This is further argued for in the next chapter, in §7.3. The list
under (151) summarises what has been laid out so far.

(151) - Scope is related to a dependence on an operator. Wide scope arises

when the variable is outside the scope of an operator and appears
in the main universe in the semantic representation. Narrow scope
arises when the interpretation of an indefinite is under the scope of
an operator.
- Partitivity is linked to indefinite NPs the denotation of which is in-
cluded within a given set. The set may be accommodated, explicitly
mentioned or part of a contextually determined set.
- Identifiability is understood as the interpretative property of those
indefinite NPs known by the sender. Identifiable DRs have a cor-
responding wide scope variable, while non-identifiable ones have a
narrow scope one.

6.2.4. Specificity in sign language

Studies on sign language specificity are very limited. So far, only a descrip-
tion of how specificity is expressed in American Sign Language (ASL) and
Hong Kong Sign Language (HKSL) is available, as already seen for defi-
niteness marking (see §5.1.4). While the studies on ASL indefiniteness and
specificity focus on the description of direction of signs on spatial planes
and nonmanual marking, the study on HKSL concentrates on the nonmanual
behaviour only. As already pointed out in the previous chapter, ASL marks
indefiniteness with the determiner something/one, an index sign pointing
upwards which involves a slight circular movement of the forearm and hand.
This articulation correlates with the degree of identifiability of the DR:
when the DR is identifiable, and hence specific, the tremoring motion of the
manual sign is minimised. When the DR is not identifiable, and hence it is
non-specific, the movement is bigger and intensified and the hand moves
through a larger area in space (MacLaughlin 1997: 131). However, the
concepts indefiniteness and specificity are collapsed in her account and she
uses “indefinite” and “specific” interchangeably without establishing a clear
categorisation (MacLaughlin 1997: 129–137).
  Specificity in LSC 157

Nonmanual marking also contributes to the expression of specificity. As

described for ASL in Bahan (1996: 272), eye gaze to mark agreement also
differs according to the (non-)specificity interpretation of the DR. While
the expression of specific referents involves a direct eye gaze to the spatial
location, non-specific referents involve a darting gaze generally towards an
upward direction. This is important and we will see that LSC shares this
upward darting eye gaze for non-specific reference.
Concerning HKSL, specificity is marked with the sign one, articu-
lated with an upwards index finger moving from left to right with a trem-
oring motion involving the wrist. This sign is accompanied with round
protruded lips, lowered eyebrows and an audible bilabial sound (Tang and
Sze 2002: 304). When this sign is articulated, eye gaze is never directed to
a spatial location but instead towards the path of the hand, suggesting that
there is no location established for the DR.
On a different view, Zimmer and Patschke (1990) for ASL and
Bertone (2007, 2009) for Italian Sign Language (LIS) explicitly claim that
an index sign directed to signing space specifies the noun it co-occurs with.
However, no further comment of what is meant by specificity nor which
properties are encompassed by it are mentioned. To further refine the notion
of specificity in relation to signing space is precisely the aim of this chapter.
As will be shown along it, specificity marking in LSC is slightly different
from the forms described for ASL and HKSL. Specific indefinites are local-
ised on the frontal plane, as well as definites. In fact, the upper frontal plane
is reserved for non-specific DRs and, similarly to ASL and HKSL, the loca-
tion is very weakly established.

6.3. Specificity in LSC

As seen in the previous chapters, in LSC the introduction of DRs into the
model is done by means of localisation, which derives into the creation of a
spatial location (p). In §4.2, it has been shown that variables being in the scope
of an operator imply lack of establishment of spatial location in actual signing.
Only variables attached to wide scope quantifiers have a corresponding spatial
location, as stated in The discourse referent hypothesis repeated below.

(152) The discourse referent hypothesis (second version)

(p) is the overt manifestation of DRs attached to a quantifier that has
wide scope.
158  Specificity

In this chapter, I focus on specificity contexts in LSC and this hypothesis is

slightly refined and extended, due to the fact that in specificity contexts the
presence of an operator can also give rise to a weakly established spatial
location on the upper part of the frontal plane. While definiteness distinc-
tions are not overtly encoded in space marking (see §5.3), LSC does encode
the marking of (non-)specificity, which is overtly expressed with different
directions of index signs directed to space. The two relevant directions for
(non-)specificity marking are the lower and the upper part of the frontal
plane, previously described in §3.2.2.3, and now graphically reminded in
Figure 41.

Figure 41. Upper and lower features of frontal plane

This chapter argues that specific indefinites are characterised by a location on

the lower part of the frontal plane and a strong coincidence of morphophono-
logical features directed to it. Here I offer a more fine-grained version of The
discourse referent hypothesis, which focuses on specificity marking, defined
below.

(154) The discourse referent hypothesis: specificity version

(p) is the overt manifestation of wide scope denoting specificity.

This contrasts with non-specific DRs, which are characterised by a loca-

tion on the upper part of the frontal plane and a weak morphopho-
nological marking of signs directed to space. This is stated in the hypothesis
below.

(154) The discourse referent hypothesis: non-specificity version

(p)[up] is the overt manifestation of narrow scope denoting non-spec-
ificity.
  Specificity in LSC 159

Figure 42 is a representation of the duality of specificity marking expressed

in LSC signing space, which this chapter focuses on.

Figure 42. Definiteness and specificity marking on LSC frontal plane

A thorough and intensive analysis of our LSC small-scale corpus data allows
distinguishing between two different kinds of localisation according to the
morphophonological mechanisms directed to spatial locations. These differ-
ences result in two distinct localisation processes that I call strong and weak
localisation, which correlate with the expression of specific and non-specific
DRs, respectively. Specific DRs, instantiated by strong localisation, are
referred to with a clear establishment of a spatial location on the lower part.
Non-specific DRs, instantiated by weak localisation, are characterised by
a diffused marking of a big area (that is, not a clear marking, but rather
marked with a fuzzy direction) on the upper part of the frontal plane. The
upper spatial location is thus weakly established and this is represented as a
bigger and wider dotted circle in the upcoming figures and stills (Figure 42).
In what follows we will see that the three properties encompassed by speci-
ficity can be assigned to the locations established on the two areas within the
frontal plane, namely the upper and the lower areas. These differences yield a
localisation pattern. But before the exposition of the localisation pattern (see
§6.3), I revise the properties encompassed by specificity that were previously
introduced in §6.1, and I exemplify each one with an LSC minimal pair.

6.3.1. Scope

As presented at the beginning of this chapter (see §6.1.1), specific DRs are
formally represented with wide scope variables, while non-specific DRs
are represented with narrow scope ones.61 In intensional contexts, English
160  Specificity

indefinite NPs can have a double reading, namely a specific and a non-
specific one. This differs from indefinite NPs in LSC, which do not have an
ambiguous reading, because specificity is overtly encoded. To see this more
clearly, I will exemplify it with a minimal pair.
The example in (155) is about a particular, specific cat the signer has in
mind. The signer first localises the nominal by means of an eye gaze towards
the contralateral part, which co-occurs with the expression of the verb want
(Figure 43a) and the verb buy (Figure 43b). Next, the pronominal index sign
co-occurs with a body lean towards the contralateral part again (Figure 43c).
Importantly the direction of localisation signs coincides in the three mecha-
nisms of localisation, namely eye gaze, body lean and index signs, and this is
precisely what is required for the proper creation of the contralateral spatial
location (p). The clear marking of both manual and nonmanual signs directed
to space in (155) allows the creation of a spatial location, which constitutes
the overt manifestation of a specific DR corresponding to the cat the signer
is talking about.

eg:cl bl:cl
(155) ix1 cat want buy. ix3cl-l character obedient.
‘I want to buy a catspec. It is very obedient.’62

a. want b. buy c. ix3cl

Figure 43. Localisation of a wide scope DR

The implementation of specificity marking is formally represented with a

variable appearing in the main DRS. The existential quantifier associated
with the variable has wide scope over the other possible embedded variables
in the subordinated DRS (156). This is a global DR that appears in the main
DRS.
  Specificity in LSC 161

(156) xy
cat (x)
□ buy (1. x)
it (y)
obedient (y)
y=x

This example contrasts with the minimal pair in (157) in which the signer
is referring to a non-specific, unidentifiable cat. Non-specificity is directly
marked in the signs directed to space in LSC. When the signer utters the
nominal cat she only directs a single eye gaze to the ipsilateral upper part
(Figure 44a). She then directs the plural indefinite determiner ix3pl to the
ipsilateral upper part (Figure 44b). In following discourse no eye gaze or
body lean is directed to any direction in space. When the modal verb must
is uttered, it does not co-occur with other nonmanuals directed to any spatial
direction (Figure 44c). Due to the few morphophonological marking towards
the upper part, the upper spatial location (p)[up] in (157) is very weakly
established. This constitutes the overt manifestation of a non-specific DR
bearing narrow scope.63

eg:ip-u
(157)64 cat ix3plu-ip ix1 want buy. must character obedient.
‘I want to buy a catnonspec. It must be obedient.’

a. cat b. ix3plu-ip c. must

Figure 44. Localisation of a narrow scope DR
162  Specificity

This is formally implemented with a subordinate variable in (158), which

is embedded under the necessity operator corresponding to a local DR. The
variable can only occur within the embedding of an operator and it is thus
infelicitous outside its scope.
(158)

xy
cat (x)
buy (1, x)
□
it (y)
obedient (y)
y=x

This minimal pair shows that scope differences are marked on the LSC
frontal plane. Signs localised within the lower part mark wide scope which
results into a specific reading, while signs directed towards the upper
part contribute to a weakly established location analysed as narrow scope
denoting non-specificity.

6.3.2. Partitivity

Indefinite NPs receive a partitive interpretation when the denotation of the

NP is included within a given set (see §6.1.2). In LSC there is a difference
between NPs with a specific restriction of a quantified NP and those without
such a restriction. This is marked in LSC with a difference in the two opposed
directions of the frontal plane. Under the restriction of a quantified NP, LSC
localisations use the lower frontal plane. When there is no such restriction,
the upper frontal plane is used (Quer 2010a). In (159) the determiner some
quantifies over the common noun friends. The determiner is articulated
towards the ipsilateral side of the lower frontal plane by means of eye gaze
and body lean directed to it. This complex NP denotes a specific DR, namely
a set formed by a concrete group of people.

(159) groupip-l friend someip-l inside ix3c hide during year-two.

‘Some of the friends were hidden there for two years.’
  Specificity in LSC 163

a. friend b. someip-l
Figure 45. Quantified NP for a specific DR

The quantifier some denotes a subset of the set referred by the NP. This is
shown in the corresponding DRS (160) by the relation x ∈ X, where X corre-
sponds to a non-atomic variable that is projected to the main universe. x is an
atomic variable and hence a subset of X. It does not appear in the main DRS
but it belongs to the set, as the formula x ∈ X indicates. In the actual LSC
signing a set that belongs to another set involves the creation of a (p) on the
lower frontal plane, as shown in Figure 45b.

(160) X
friends (X)

x some x
x∈X x hide (x)

In contrast, in (161) the signer is referring to a non-concrete, non-specific

DR. He refers to a group of people that does not belong to a determined set.
To refer to it, he directs an index plural sign towards the upper ipsilateral
part. The localisation is only articulated manually with the index sign, and
no eye gaze or body lean is directed to the upper location, as illustrated in
Figure 46.

(161) ix3plip-u some 1-denounce-3ip-u ix3c there-is.

‘Someone denounced they were there.’
164  Specificity

ix3plip-u
Figure 46. Quantified NP for a non-specific DR

The sentence in (161) denotes a non-specific DR not belonging to a contextu-

ally determined set. The lack of contextual determinacy is manifested in the
actual LSC signing by establishing (p)[up] on the upper frontal plane. In the
corresponding DRS, this is represented with an embedded variable under the
scope of the quantifier, which does not belong to any set from the main DRS,
as represented in (162).

(162) y
they (y)
y
they (x)
x some denounce
people (x) x (x, y)

The previous minimal pair is a representative example of the distinction between

partitive cases found in our LSC small-scale corpus. The lower frontal plane is
not only used to denote wide scope cases, but also partitivity. In contrast, the
upper frontal plane is used to denote both non-partitivity and narrow scope. As
we will see in the following subsection, the expression of identifiability is also
distinguished by the use of the two opposed parts of the frontal plane.

6.3.3. Identifiability

Identifiability has been defined as the interpretive property of those indefinite

NPs known by the sender (see §6.1.3). Interestingly, LSC overtly encodes
  Specificity in LSC 165

the information that is identified by the signer and distinguishes it from what
is not known by the signer. As seen in Chapter 5, no distinction is made
between the knowledge of the signer and that of the addressee. Rather the
establishment of lower and upper locations mark the distinction between
what the signer knows and what the signer does not know, respectively.
In the fragment in (163) the signer is talking about a pen-drive. After
explaining what this gadget is used for, she explains to which person she will
offer it. The person the signer is talking about is an identifiable person. That is,
a particular individual that the signer has in mind. She knows that the person
she is talking about works with computers and that it will be a very appro-
priate present. The DR is first referred to with an indefinite NP introduced
by ‘one person-3ip-l’. The localisation process to the ipsilateral part is done
both with manual and nonmanual mechanisms. On the one hand, she directs
the agreement verb 1-offer-3ip-l and the sign person-3ip-l towards the ipsi-
lateral part. On the other, eye gaze and body lean are also directed to the same
side (Figure 47). Hence, (p) is established in the lower ispilateral part.

eg:ip-l
(163) ix1 1-offer-3ip-l one person-3ip-l pen-drive computer pen-drive
eg:ip-l eg:ip-l
1-offer-3ip-l, because person-3ip-l always++ work theme
is/same computer.
_____eg:ip-l
pen-drive adequate ix1 1-offer-3ip-l ix3ip-l Pen-drive.
‘I will offer the pen-drive to someone, since he/she/this person always
works with computers. I find it very adequate to offer the pen-drive to
him/her.’

a. 1-offer-3ip-l b. person-3ip-l c. ix3ip-l

Figure 47. Localisation of an identifiable DR
166  Specificity

Figure 47 shows that in this LSC fragment the signer establishes (p) in the
ipsilateral part that corresponds to an identifiable DR. This spatial location
is available for further co-reference, as exemplified in the second utterance
in (163) where a resumptive pronoun is directed back to it (Figure 47). The
corresponding semantic representation places a wide scope variable in the
main DRS, which is available for further co-reference as the identity relation
z=x shows. Again, there is a correlation between placing a variable in the main
DRS and establishing a lower spatial location in actual signing (Figure 47).

(164) x, y, z
person (y)
pen-drive (x)
offer (1, x, y)
work-computer
(y)
offer-adequate
(1, x, z)
z=y

The establishment of this global DR contrasts with a minimal pair example

found in the small-scale corpus. In (165), the signer is talking about a book
and explains to which person he would offer it. The indefinite NP the signer
uses to refer to the person has a clear non-specific interpretation, as marked
in the glosses and in the corresponding translation. This non-specific inter-
pretation is derived from the localisation process. The first instance of
person-3 (Figure 48a) is localised with a very slight eye gaze co-occurring
with it and directed towards the ipsilateral part. In the second instance of the
sign person-3 (Figure 48b) the nominal is localised with eye gaze towards
the centre of signing space, instead of towards the ipsilateral side as in the
first instance. The subsequent signs are directed towards the contralateral
part (Figure 48c) and then towards the centre again (Figure 48d). Hence eye
gaze is not directed to a particular spatial direction, but instead it moves in
different directions towards the upper frontal plane (Figure 48). As a conse-
quence, (p)[up] is very weakly established.

eg:book eg:ip-u
(165) ix1 think ix3 book 1-offer-3 adequate person-3ip-u
eg:cl-u eg:c
must person-3cent like hobby is/same traditional past
  Specificity in LSC 167

eg:ip-l
same/always. ix3ip-l ix1 1-offer-3 person-3ip-l ix3ip-l.
‘I think that I would offer this book to someonenon-spec...
It must be someone who likes traditional things. Definitely, I would
offer it to him/her.’

a. person-3 b. person-3 c. like d. hobby

Figure 48. Introduction of a non-identifiable DR

The corresponding DRS places a variable for the unidentifiable DR in an

embedded DRS. Hence it does not appear in the main universe of discourse,
but rather in a subordinated context corresponding to the embedded box for
the antecedent. This local variable is dependent on the operator.

(166) x
book (x)
yz
offer (1, x, y)
person (y) □
trad-things (z)
like (y, z)

As the DRSs in (164) and (166) show, the scope of y is different. While in
(166) y has narrow scope and it is thus interpreted as an unidentifiable DR
(i.e. the sender cannot identify it from other DRs present in the model), in
(164) y has wide scope and it is interpreted as an identifiable one (i.e. the
sender does have a concrete entity in mind). Since non-identifiable DRs
correspond to narrow scope variables, a coreferential pronoun in further
discourse has to be under the scope of the modal operator too, as shown in
(167a). Outside the scope, the uptake is not felicitous (167b) (in §6.3.3.2
modal subordination contexts are treated in more detail).
168  Specificity

(167) I would offer this book to a person who likes traditional things.
a. He must be smart.
b. #He is smart.

In short, the direction of eye gaze towards the upper frontal plane and (p)[up]
being weakly established stand in direct relation with the introduction of the
variable in an embedded context. Hence, upper space and weakly established
locations are overt markings for narrow scope contexts to denote non-iden-
tifiable DRs in LSC. For now, it is sufficient to assume that non-identifiable
DRs weakly establish a location on the upper frontal plane, but in §7.3 it will
be proven that embedded variables can also establish a location on the lower
frontal plane as long as they refer to the discourse topic.
On the basis of the minimal pairs presented in this section, it is fair
to say that the position of the spatial location in the frontal plane in LSC
stands in direct relation with specificity. The lower frontal plane is associ-
ated with scopally specific (i.e. wide scope) and identifiable DRs, as well as
with a restriction of the quantified NP. In contrast, within the upper frontal
plane there is no such restriction and this is thus associated with scopally
non-specific (i.e. narrow scope) and non-identifiable DRs. This distinction
derives into a localisation pattern based on the direction of the localisation
of signs (whether they are directed towards the lower or the upper part) and
also according to the amount of morphophonological features directed to
both directions. The presentation of the LSC localisation pattern is the main
concern of the following section but before delving into it, a final note is
devoted to argue in favour of a denotation of non-specificity marking of the
upper frontal plane, rather than narrow scope marking.

6.3.4. Narrow scope marking (or what this chapter is not about)

As stated in the non-specificity version of The discourse referent hypoth-

esis, this chapter argues that in LSC narrow scope denoting non-specificity
is expressed with signs directed to the upper part of the frontal plane. The
immediate question that arises here is whether these upper locations are true
referential (i.e. they stand for non-specificity marking), or they stand for
structural marking in general (i.e. they mark narrow scope). To analyse the
final function of upper locations, other phenomena related to narrow scope
marking are treated here in order to reject (or not) the possibility that we are
dealing with narrow scope marking when using the upper part of the frontal
  Specificity in LSC 169

plane. Narrow scope is also one of the main features of weak definites, such
as “(take) the train” or “(read) the newspaper” in English. In the following
we will see that since the features related with weak referentiality, and more
concretely with weak definites, are not found on the upper locations, we
are definitely dealing with a referential phenomenon, and thus a specificity
distinction.
Although they are very similar in form, weak definites and regular
definites have some distinguishing properties. Weak definites in Dutch and
English (i) take narrow scope (168a); (ii) occur with a restricted class of
nouns which lexically determine the construction (168b); (iii) only allow
modification that yields a subclass (168c) (see Carlson and Sussman 2005),
and (iv) have limited capacities to establish DRs (168d) (Scholten and
Aguilar-Guevara 2010).

(168) a. Every student took the trainwdef.

b. Mary went to the storewdef. vs. Mary went to the desk.
c. Lola is in the psychiatric hospitalwdef. vs. Lola is in the big hospital.
d. ??Lola is still at the schooli because today her class had to help to
clean iti.

(168a) can be paraphrased as “for every student there is a train they took”,
showing that weak definites take narrow scope. (168b) shows that there is a
distinction between the kind of noun which is used in weak definite readings,
such as the pen, the radio, the store or the hospital, which is contrasted with
regular definites, such as the cage, the tape-recorder, the building or the desk.
(168c) shows that the kind of modification co-occurring with weak definites
can only yield a subclass and thus regular modifiers such as new, big, and
green do not derive into a weak definite reading. Finally, (168d) shows that
the sequence where a weak definite establishes a DR that is picked up by a
resumptive pronoun does not sound very natural.
Although in Chapter 5 it has been shown that LSC spatial locations do
not encode definiteness, here I am testing whether the upper locations are
a phenomenon related to weak referentiality. Hence, I am using the distin-
guishing properties attached to weak definites to determine whether or not
we are dealing with such phenomenon. The tests I will use are the following:
(i) the class of nouns localised on the upper frontal plane, (ii) the kind of
modification allowed and (iii) the possibility of establishing a DR.
170  Specificity

(i) Class of nouns

In LSC, spatial locations established on the upper frontal plane are not
restricted to a class of nouns, which have been used in the literature to distin-
guish the nouns that are used as weak definites, such as the pen, the radio or
the hospital (Carlson and Sussman 2005; Schulpen 2011). Nouns considered
forming a restricted class as pen and hospital (169a) and (170a), as well as
the ones considered being regular definite nouns, such as cage and building
(169b) and (170b) can be both localised on the upper frontal plane. LSC
does not show a distinction on the class of nouns that may be localised on the
upper frontal plane. Also, note that all the examples with an upper location
need to include intensional verbs. This fact shows further evidence towards
the referential account of the use of upper locations.

(169) a. ix1 look-for ix3u pen.

‘I am looking for a pen.’
b. ix1 look-for ix3u cage.
‘I am looking for a cage.’
(170) a. ix1 find ix3u hospital need.
‘I need to find a hospital.’
b. ix1 find ix3u building need.
‘I need to find a building.’

(ii) Kind of modification

It has been suggested that weak definites can be modified as long as the
modifier establishes a subkind of the noun it modifies (Aguilar-Guevara and
Zwarts 2010; Schulpen 2011). Unlike weak definites, upper locations allow
any kind of modification, and not only the one that yields a subclass. The
LSC examples below show that subclass modification (171a) and (172a), as
well as regular modification (171b) and (172b) are grammatically localised
on the upper frontal plane. Hence the upper frontal plane does not respect the
criteria established for weak referentiality marking.

_____________br
(171) a. cat persian ix3u ix1 want buy.
‘I want to buy a Persian cat.’
  Specificity in LSC 171

_________br
b. cat fat ix3u ix1 want buy.
‘I want to buy a fat cat.’
____________________br
(172) a. restaurant turkey ix3u ix1 want eat.
‘I want to eat at a Turkish restaurant.’
__________________br
b. restaurant calm ix3u ix1 want eat.
‘I want to eat at a calm restaurant.’

(iii) Possibility of establishing a DR

More importantly, weak definites in LSC are not established in signing space,
but rather articulated in neutral space without having a corresponding spatial
location.65 Examples (169) to (172) prove that the upper frontal plane does
not stand for narrow scope as a structural marking, but it rather stands for
narrow scope marking denoting non-specificity. If the upper frontal plane
would stand for structural narrow scope marking, only half of the exam-
ples would be felicitous. As shown, this is not the case. Moreover the upper
spatial locations are not limited with respect to the establishment of DRs, and
as previously shown in (165) indefinite NPs established in the upper signing
space in LSC can be further referred back to by a resumptive pronoun as long
as the corresponding variable is under the scope of an operator. This I take to
be an important distinction with respect to weak referentiality.
On a final note it is worth mentioning that in §4.3 I have shown some
arguments where narrow scope marking does not have an upper location
establishment. In fact, we have seen that when the DR inserted in contexts of
donkey sentences, generic NPs and quantified NPs there is a lack of spatial
locations establishment. This also points towards the hypothesis that upper
locations stand for non-specificity, rather than narrow scope marking in
general, since narrow scope also entails lack of location establishment. Also
following this reasoning, in §6.3.3.2 I will also show that modal subordina-
tion contexts, which also share the property of narrow scope, are marked
with a lower spatial location, rather than an upper one. Once the arguments
for non-specificity marking are clear, it is now time to present the localisa-
tion pattern this chapter argues for.
172  Specificity

6.4. Localisation pattern

In the previous section it has been shown that the expression of specificity in
LSC can be analysed taking into consideration three properties that specificity
encompasses, namely scope, partitivity and identifiability. In this section we
will see that the expression of specificity in LSC can be accounted through a
localisation pattern. What I call strong localisation consists in a kind of local-
isation established with localised signs directed to the lower frontal plane
and quite a lot of morphophonological mechanisms also directed to it simul-
taneously. As a consequence, (p) is strongly established. This contrasts with
weak localisation in which the spatial location is weakly established with
localised signs directed to the upper frontal plane and very few mechanisms
are directed to it, which do not simultaneously co-occur. As a consequence,
p[up] is very weakly established. This localisation pattern is associated with
specificity marking. Strong localisation expresses specific DRs and can be
accounted through wide scope, partitivity and identifiability. The semantic
representation of strong localisation is shown in (173), where the variable
appears in the main DRS and has wide scope over the embedded variables.

(173) x

Weak localisation expresses non-specificity and it is explained through

narrow scope, non-partitivity and non-identifiability. This is implemented in
DRT with a variable embedded in a subordinated DRS, as shown in (174).

(174)

The localisation pattern is motivated by the analysis of corpus data and

the association with specificity interpretation. In the next subsections the
 Localisation pattern 173

systematic and compositional combination of the morphophonological

features contributing to specificity readings is first presented. Second, the
categories that may be spatially modified within the localisation pattern are
analysed (see §6.3.2). Finally, the dual nature of localisation in LSC is further
motivated by arguing through examples coming from modal subordination
and the combination between non-specificity and partitivity (see §6.3.3).

6.4.1. Compositional analysis of the data

According to our LSC annotated data, the morphophonological features of

weak and strong localisation are divided into five main components, which
may be grouped as follows:

i) Direction of signs towards the frontal plane

ii) Amount of mechanisms directed to space
iii) Eye gaze duration
iv) Co-occurrence of mechanisms
v) Coincidence of direction

These features are systematically and componentially combined to denote

meaning. That is, specificity interpretation is overtly encoded by the combi-
nation of the above-cited features. Each feature contributes to the construc-
tion of specificity interpretation as a whole. In what follows they are treated
in detail one by one.

6.4.1.1. Direction towards the frontal plane

Both manual and nonmanual linguistic elements can be directed to signing

space. In order to indicate different locations established on the frontal plane
many elements can be used, such as index signs, spatially modified signs,
agreement verbs, as well as eye gaze, body lean and head tilt. In LSC the
frontal plane can be marked by some of these mechanisms as shown in the
figures below. Figure 49 shows both manual and nonmanual mechanisms
directed towards the lower part of the frontal plane, which correspond to a
strong localisation. As already stated in §3.3, the distinction between signs
directed towards the ipsilateral and contralateral parts of the horizontal plane
are not contrastively relevant for the grammar of LSC, but rather the lower
and upper areas of the frontal plane.
174  Specificity

Figure 49. Strong localisation and use of lower frontal plane

The upper frontal plane may also be indicated by both manual and nonmanual
features. Figure 50 also shows mechanisms directed towards the upper part,
which correspond to a weak localisation.

Figure 50. Weak localisation and use of upper frontal plane

As seen in the previous section, the distinction between the two directions
within the frontal plane is associated with specificity interpretations. While
localisation on the lower frontal plane is used for specific readings, localisa-
tion on the upper frontal plane is used for non-specific readings. The former
are instances of strong localisation, and the latter are localised by means of
weak localisation. A detailed observation and analysis of our LSC small-scale
corpus shows that both manual and nonmanual mechanisms are directed to
the two parts of the frontal plane. Table 4 below shows the instances of indef-
inites found in the LSC corpus (32), identified according to a familiarity
presupposition implied (Chapter 5). Among the indefinites which corre-
spond to a strong localisation and thus have a specific interpretation, 22 out
of 22 are localised on the lower frontal plane. Concerning the indefinites that
 Localisation pattern 175

correspond to a weak localisation and have a non-specific interpretation, 8

out of 10 instances of indefinites are localised on the upper frontal plane.

Table 4. Indefinites localised on the lower/upper frontal plane

Indefinite NPs / 32 Specific interpret. / 22 Non-specific interpret. / 10
Lower frontal plane 22 2
Upper frontal plane 0 8

Interestingly, the fact that two instances of indefinite NPs are articulated on
the lower frontal plane is not due to a random behaviour of index signs. Their
direction towards space in non-descriptive localisation is very systematic,
as already argued in §3.3. These two instances occur in contexts where the
signer is talking about a non-specific entity and, although it is not an iden-
tifiable and concrete DR, the discourse is centred on it. The DR appears
under an intensional verb and it has a non-specific interpretation but this
is not incompatible with being the entity the discourse is about and hence
with being localised on the lower frontal plane. They are instances of modal
subordination contexts (Roberts 1986, 1989) and section §6.3.3.2 is devoted
to this issue.

6.4.1.2. Amount of mechanisms

The difference between weak and strong localisation is also distinguished

by the amount of morphophonological mechanisms directed to signing
space in each case. Concerning strong localisation, the spatial location is
more properly established due to a coincidence of more than two mecha-
nisms which simultaneously co-occur directed towards the same spatial
area. They are directed towards the lower frontal plane. On the other hand,
concerning weak localisation the spatial location is weakly established
since only very few mechanisms are used, which very rarely co-occur.
Also the movement of the manual sign and eye gaze do not have the same
direction. They are both directed to the upper frontal plane, but there is no
coincidence in the axis direction in signing space. While the manual sign
points towards one direction, the nonmanual consists in a darting eye gaze
that moves around in the upper space. The amount of simultaneous mecha-
nisms used during both localisations is presented in the following table.
The instances of indefinites receiving a specific interpretation incorporate
a higher number of mechanisms during the localisation process. Whereas
176  Specificity

specific interpretations are characterised by a localisation with at least two

mechanisms, which results in a strong localisation, non-specific interpreta-
tions generally have no more than two mechanisms and it thus results in a
weak localisation.

Table 5. Number of mechanisms used simultaneously in indefinites and the corre-

sponding interpretation
Number of mechanisms Specific interpret. / 22 Non-specific interpret. / 10
1 1 3
2 7 6
3 14 1

Strong localisation that corresponds to a specific interpretation is character-

ised by two or more mechanisms, whereas weak localisation corresponding
to non-specificity is featured by two or less mechanisms.

6.4.1.3. Eye gaze duration

As said in §3.2, nonmanual mechanisms are crucial for the localisation

process. When an indefinite NP has a specific interpretation, eye gaze
towards the strongly established location is longer than when it has a non-
specific interpretation. In fact, indefinites interpreted as non-specific only
have a weakly established location and eye gaze directed to it is very brief.
But once this brief eye gaze is directed to the weakly established location, a
darting eye gaze is directed afterwards towards the upper frontal plane from
one lateral part to the other. This darting eye gaze lasts longer but it is not
fixed to any particular spatial area.
The following table presents the difference in eye gaze duration for indef-
inite NPs. Eye gaze towards the spatial location in specific readings doubles
the time of eye gaze directed to a spatial location within non-specific read-
ings. However, eye gaze in non-specific interpretations, after being briefly
directed to the upper location, moves around in signing space from one
lateral part to the other, without being fixed in a single direction. This darting
eye gaze is quite long and in our small-scale corpus it is never articulated
within specific interpretations. The longer eye gaze directed to a spatial loca-
tion contributes to the strong establishment of a location, which stands for
the manifestation of a DR interpreted as specific. In contrast, the short eye
gaze fixed to a location and the subsequent darting eye gaze, which moves
 Localisation pattern 177

around on the upper frontal plane, contributes to the weak establishment of a

location that stands for a non-specific DR.

Table 6. Eye gaze duration

Indefinite NPs Specific interpret. Non-specific interpret.
Eye gaze duration to 431 268
spatial location (ms)
Darting eye gaze (ms) - 2015

6.4.1.4. Simultaneity and coincidence in direction

Localisation is often expressed with manual and nonmanual marking. In

our small-scale LSC corpus, spatial locations yielding to a specific interpre-
tation are established with simultaneous articulation between manual and
nonmanual marking. This simultaneity is less frequent when the interpreta-
tion is non-specific. Also whenever there is more coincidence in the direc-
tion towards space of manual and nonmanual marking, it corresponds to a
specific interpretation. When the coincidence in direction is almost absent,
the DR is interpreted as non-specific.

Table 7. Instances of simultaneous co-occurrence with manual and nonmanual com-

ponent
Indefinite NPs Specific Non-specific
interpretation interpretation
Simultaneous occurrence of manual and 20 12
nonmanual marking
Coincidence of direction of manual and 30 2
nonmanual marking

Due to the great coincidence in direction and simultaneity, a spatial loca-

tion is established which corresponds to a specific interpretation. This spatial
location is strongly established on the lower frontal plane, and this is the
direction towards which index signs, eye gaze and other localisation mech-
anisms are directed to. On the contrary, the lack of coincidence in direc-
tion and of simultaneity between the manual and the nonmanual component
in non-specific interpretations results in a lack of a discrete, identifiable
spatial location. Eye gaze is very short and, since there is not a concrete
178  Specificity

and well-established spatial location, it is only directed briefly to the weakly

established spatial location. Moreover, eye gaze darts around towards the
upper frontal plane denoting that a concrete spatial location is only weakly
established. These phonological differences result in the distinction this
chapter argues for, namely a strong vs. a weak localisation, which is charac-
terised by the pairing between the features described so far and the semantic
interpretation (i.e. expression of specificity).
In the first chapter of this book (see §1.4) I already indicated that the
data used is taken from a small-scale LSC corpus. The main aim has been to
develop a qualitative study by observing and analyzing the tendencies that
naturalistic, semi-spontaneous and elicited data provide. The quantification
shown in the previous tables should be taken as a strong tendency of real
data, which is associated with a theoretical model to account for specificity
marking. A follow-up study based on a yet non-existent large LSC corpus
should help confirm the generalisations and the analysis provided here on
the basis of naturalistic data. Yet this strong tendency can be summarised as
follows:

(175) Localisation pattern: Morphophonological data analysis

a. Strong localisation
– Direction towards the lower frontal plane.
– Movement of manual signs and nonmanual (especially eye gaze)
have the same direction.
– Eye gaze is longer fixed and targets a spatial location.
– High number of localisation mechanisms (x ≥ 2) simultaneously
combined.
– Spatial location well established.
b. Weak localisation
– Direction towards the upper frontal plane.
– Movement of manual signs and nonmanual signs (especially eye
gaze) do not have the same direction.
– Eye gaze is very short and does not target a spatial location; it
rather darts around.
– Few localisation mechanisms are used (x ≤ 2), and lack of coin-
cidence among them.
– Spatial location weakly established.

As is clear at this point of the argumentation, each localisation encompasses

the specificity properties previously defined (see §6.1). On the one hand,
strong localisation is used to denote specific and identifiable DRs. It is used
 Localisation pattern 179

in partitive constructions, to denote contextual determinacy. On the other

hand, this contrasts with weak localisation, which is used to denote non-
specificity and non-identifiability. Within weak localisation, DRs do not
belong to a contextually determined set. This is summarised in below.

(176) Localisation pattern: Specificity expression and properties

a. Strong localisation
– Wide scope
– Partitivity
– Identifiability
b. Weak localisation
– Narrow scope
– Non-partitivity
– Non-identifiability

Moreover, weak localisation is instantiated by the feature (p)[up] and strong

localisation is instantiated by the feature (p), already introduced in §3.4.
The establishment of these two features is instantiated by the two kinds
of localisation defined so far. In fact, weak and strong localisation are not
only distinguished by the direction on the frontal plane and the amount of
morphophonological features directed to it, but also by the kind of gram-
matical elements that may be localised on each part. Below the restrictions
that each localisation imposes on signs are presented.

6.4.2. Spatially modified categories

The morphemes (p) and (p)[up] established with strong and weak locali-
sation respectively add different constraints on the categories they can be
attached to. The morpheme (p) established with strong localisation does not
add any restriction on the kind of grammatical elements that can be related
to it, since both lexical and functional categories can be directed to it. In
contrast, the morpheme (p)[up] established in weak localisation imposes
some restrictions and, as will be shown below, only functional elements can
be associated with it.
As for the (p) morpheme established with strong localisation, there is no
restriction concerning the linguistic categories that can be spatially modified.
Non-anchored common nouns, plain verbs and entity classifiers, as well as
determiners and verb agreement can be strongly localised. More specifically,
the sign itself may be spatially modified, or it may also be established on
180  Specificity

space with nonmanuals or with the co-occurring determiner. In (177a) the

noun is spatially modified, and in (177b) it is the determiner that is spatially
modified, as the subindices in the glosses indicate.66

(177) a. housel
b. house ix3l

Plain verbs (i.e. the type of verbs which do not agree for subject and object
(see §3.2.3) can also be spatially modified and hence localised on the lower
frontal plane.

(178) a. constructl
b. there-isl

However, lexical categories cannot be localised on the upper part, and any
attempt to spatially modify non-anchored nouns, plain verbs or entity clas-
sifiers towards the upper frontal plane feature results in an ungrammatical
construction.67 Non-anchored common nouns or plain verbs cannot be
incorporated to the clitic (p)[up] to denote non-specificity, as shown in the
ungrammatical examples in (179). Rather, only the determiner co-occurring
with the noun can be weakly localised (180).

(179) a. *houseu
b. *constructu
c. *there-isu

(180) ix3u house

A distinction is thus established: while the feature [low] can be attached to

lexical categories, [up] cannot. However, the situation is slightly different
when we focus on functional categories. In LSC functional elements such
as determiners and verb inflection can establish (p), as well as (p)[up]. As
for inflection, agreement verbs (i.e. the type of verbs which are inflected for
subject and object; see §3.2.3) can be both localised on the lower and on the
upper frontal plane, but a different interpretation is conveyed. The interpre-
tation that we get with (p) corresponds to a specific reading (181a), while
the interpretation that we get with (p)[up] corresponds to a non-specific one
(181b). Unlike with plain verbs, the feature [up] can be attached to verb
 Localisation pattern 181

inflection denoting a non-specific interpretation. Hence while lexical catego-

ries cannot have the [up] feature, functional categories can.

(181) a. 3l-advise-1.
‘Some specific person advised me’
b. 3u-advise-1.
‘Some non-specific person advised me’
(adapted from Quer 2010a)

Another inflectional category that can be linked to [up] is a restricted set of

determiners. Weak determiners (Milsark 1974) are grammatically localised
both on the lower and the upper frontal plane. Weak determiners in LSC,
such as one, any, some, one+++ (‘few’) can be attached to both [low] and
[up].

(182) house somel, house fewl, house anyl

‘Some of the houses’, ‘few of the houses’, ‘any of the houses’

(183) house someu, house fewu, house anyu

‘Some houses’, ‘few houses’, ‘any house’

Weak determiners are ambiguous between denoting presupposition and

assertion (Diesing 1992). In LSC this ambiguity is resolved by the spatial
modification. When the weak determiner is established on the lower frontal
plane, a presupposition reading arises, and hence the DR denotes that there
exists an entity under discussion. Examples in (182) can thus be paraphrased
as “some of the houses”, “few of the houses” and “any of the houses” since a
restriction on the quantified NP is overtly expressed with a localisation on the
lower part. When no such restriction is present, this is also overtly expressed
in the determiner system with a weak determiner directed towards the upper
part. The non-presupposition of existence is marked with the spatial modifi-
cation of determiners towards the upper part. Hence the upper part denotes
non-contextual determinacy and examples in (183) are paraphrased as “some
houses”, “few houses”, “any house”.68 This shows that the morphophonolog-
ical feature [low] cliticised to the determiner as well as on the verbal inflec-
tion denotes contextual determinacy and specificity, whereas the feature [up]
is associated with non-contextual determinacy and non-specificity (184).

(184) a. friend someu

‘Some friendsnon-spec’
182  Specificity

b. student oneu
‘One studentnon-spec’

As for strong determiners, they are much more restricted in that they can
only occur on the lower frontal plane (185a). As shown in (185b), strong
determiners spatially modified towards the upper part are considered to be
ungrammatical.

(185) a. friend mostl, house halfl

b. *friend mostu, *house halfu

This restriction shows that the presupposition of existence that strong deter-
miners imply cannot be grammatically encoded on the upper part. Moreover,
whenever a strong determiner is spatially modified towards a lower spatial
location, only a restricted set of elements is denoted. In (186) not most
students in the universe are intended but rather only the set under discussion.
This restriction of the set is overtly encoded with the determiner most being
spatially modified and hence with the spatial morpheme [low].

br
(186) student, mostl come.
‘Most students came.’

Since the morpheme [low] marks domain restriction, a universal determiner

localised on the lower frontal plane denotes that the set referred to is not
empty, and thus it refers to a contextually determined set.
In LSC the spatial morpheme (p) established on the lower frontal plane
denotes a DR. It is used to denote definites as well as specific indefinites. It
marks that the DR is in the model, and no distinction between knowledge
of the sender or the addressee is made. Once the entity is introduced into
the model, this is not formally distinguished in the use of signing space in
LSC. Hence since it is included in the model, it is restricted within a domain,
and this domain restriction is in fact marked by the spatial morpheme estab-
lished on the lower frontal plane. On the contrary, (p)[up] is established by
some functional categories only. [up] is a marked feature denoting absence
of domain restriction. Since there is no restriction in the domain, NPs local-
ised with (p)[up] are used to denote non-specific DRs. Hence the feature
[low] combines with lexical and functional elements to denote specificity
and domain restriction, and forces the associated determiner to refer to a set
 Localisation pattern 183

of relevant elements. The feature [up] combines with weak determiners and
verb inflection to denote absence of domain restriction. As shown in Barberà
(2012c), domain restriction is thus a necessary part of the denotation of the
spatial morpheme. The context narrows down the domain where the func-
tion will choose any individual or sets of individuals (see recent work on
quantifier domain restriction expressed on the frontal plane in American Sign
Language (Davidson and Gagne 2014)).
On the basis of the above argumentation, it is fair to say that the two
parts of the frontal plane are not equal. (p)[up] established by weak localisa-
tion has more restrictions: only weak determiners and verb agreement with
a non-presuppositional interpretation can be directed to the upper frontal
plane. The feature [up] functions thus as non-contextual determinacy of the
DR denoted by the weak determiner or by agreement inflection. This differs
from (p) established by strong localisation, where both lexical and functional
categories can be localised. The feature [low] functions as contextual deter-
minacy and domain restriction. Interestingly, in LSC the two localisations
can be combined under the same construction, as shown in the next section.

6.4.3. Dual nature of localisation

The localisation pattern shows a different use of the establishment of DRs.

While strong localisation is associated with the expression of specificity,
weak localisation is associated with non-specificity. Moreover, some partic-
ular constructions incorporate these two kinds of localisation, showing the
dual nature of such a process. This is the case of some partitives that are
combined with determiners denoting non-specificity, as well as contexts of
modal subordination.

6.4.3.1.  Non-specific partitives

Although partitivity is commonly associated with specificity, in some

contexts partitive constructions in some particular languages can co-occur
with a determiner having a non-specific interpretation, as shown in the
following English (187a) and Catalan (187b) equivalent examples.

(187) a. I need one of those books.

b. Necessito un d’aquells llibres.
184  Specificity

The determiner in (187a) and (187b) is ambiguous between a specific and a

non-specific reading. Within the specific interpretation, there is one identifi-
able and concrete book from the set of a determined group of books that the
speaker needs. In the semantic representation, the determiner attached to the
corresponding variable has wide scope over the set of books.

(188) x
book (x)
need (1, x)

books (X) →
x∈X

Within the non-specific interpretation, from the set of those books there is
an unidentifiable, non-concrete book that the speaker needs. In the corre-
sponding semantic representation the quantifier attached to the set of books
has wide scope over the book. Since it is a non-atomic variable it is repre-
sented with a capital X. The variable attached to the non-specific DR has
narrow scope and thus it appears in an embedded context in the DRS.

(189) X
book (X)
need (1, x)

books (x) →
x∈X

LSC partitive constructions generally denote specificity, but they may

co-occur with determiners denoting both specific and non-specific DRs. In
such constructions, the partitive first establishes the domain of quantifica-
tion, and then the determiner that ranges over it. In (190), the domain of
quantification is first strongly established on the lower frontal plane ‘book
ix3l’ (Figure 51a) and the specific determiner that ranges over it is uttered
afterwards ‘onel’ (Figure 51b).

(190) book ix3l, ix1 need onel.

‘I need onespec of those books.’
 Localisation pattern 185

a. ix3l (‘those’) b. onel (‘onespec’)

Figure 51. Partitive construction with specific determiner

The combination of a non-specific determiner with a partitive construction

is grammatical in LSC. In (190) the domain is also first established (‘book
ix3l’) and afterwards the non-specific determiner is uttered (‘oneu’). Again,
strong localisation is characterised by the signs directed to the lower frontal
plane (Figure 52a) and weak localisation is marked with the non-specific
determiner directed to the upper frontal plane (Figure 52b).

(191) book ix3l, ix1 need oneu.

‘I need onenon-spec of those books.’

a. ix3l (‘those’) c. oneu (‘onenon-spec’)

Figure 52. Partitive construction with non-specific determiner
186  Specificity

(191) shows that although a partitive construction is used and hence a strong
localisation is established, a weak localisation can also be established in the
same utterance when a non-specific determiner co-occurs with it. A single
utterance can combine the denotation of specificity and non-specificity and
this is formally marked on the direction of signs on the frontal plane and
the corresponding localisation. This shows thus that the equivalent signed
sentences to (187a) and (187b) are not ambiguous. The two different forms
obtained for each reading show that partitives can be combined with non-
specific determiners and that these combined constructions show the dual
nature that localisation has in LSC.

6.4.3.2. Modal subordination

Another context where the weak and strong nature of localisation is combined
is that of modal subordination. It is generally considered that anaphoric
contexts generally take place when variables are attached to quantifiers with
wide scope. However, there exists a context in which anaphora can occur
within intensional contexts. These are called “modal subordination” contexts
and consist in noteworthy entities introduced into the model the existence of
which is not presupposed. Modal subordination is instantiated in anaphoric
contexts which occur under the scope of a modal operator or a propositional
attitude predicate, which display anaphoric relations that at first glance appear
to violate generalisations about scope operators and anaphoric potential
(Roberts 1986, 1989). As shown in (192a) the resumptive pronoun not bound
by an operator is an infelicitous continuation. Once the pronoun is under a
narrow scope context it is considered a felicitous continuation (192b).

(192) If John bought a book, he will be home reading it by now.

a. # It is a murder mystery.
b. It will be a murder mystery.
(Roberts 1986: 19)

Modal subordination consists in contexts that combine narrow scope with

anaphoricity, a property generally attributed to wide scope. Interestingly, this
dual nature of modal subordination is also overtly expressed in LSC instances
of modal subordination contexts. They consist in narrow scope entities that are
embedded under an operator and are thus introduced into discourse by means of
a weak localisation. But also, since the entity is noteworthy and further resump-
tive pronouns refer back to it, a spatial location is also strongly established.
 Localisation pattern 187

The previous example (165), repeated for convenience as (193), shows the
dual establishment of the DR in signing space. In (193) the signer is talking
about a non-specific, unidentifiable DR. As shown in the glosses, the DR is first
weakly established towards the upper frontal plane. But in the last sentence a
resumptive pronoun directed towards a lower location refers back to it.

eg:book eg:ip-u
(193) ix1 think ix3 book 1-offer-3 adequate person-3ip-u
eg:cl-u eg:c
must person-3cent like hobby is/same traditional past
eg:ip-l
same/always. ix3ip-l ix1 1-offer-3 person-3ip-l ix3ip-l.
‘I think that I would offer this book to a personnon-spec
It must be someone who likes traditional things. Definitely, I would
offer it to him/her.’

While introducing the antecedent (“someone who likes traditional things”)

the signer establishes a very weak spatial location on the upper frontal plane.
Immediately after that, a darting eye gaze is directed to an upper direction
that goes from the ipsilateral, to the centre and to the contralateral part. This
eye gaze moves around along the upper frontal plane, without being directed
to an area. However, when the antecedent is introduced, although being
under an intensional context, a spatial location is established on the lower
frontal plane in order to have a location set up where to direct the corefer-
ential signs to. In the subsequent sentence the signer directs a pronoun, an
agreement verb and a pronoun again towards an established spatial location
on the ipsilateral side. A strongly established location is then used where
coreferential signs are directed to (Figure 53).

Figure 53. Resumptive pronoun within an intensional context

188  Specificity

Previously, Table 4 showed that two instances of indefinites with a non-

specific interpretation were localised on the lower frontal plane, rather than
on the upper part. In fact, these are two instances of modal subordination
found in the corpus. (193) is such an example. Although the variable occurs
embedded under an operator, the discourse is about that DR, which is the
topic of the discourse fragment. Hence a lower spatial location is established
and further resumptive pronouns can be directed back to it (see §7.3 for a
further treatment of the notion of discourse topic).
Within modal subordination contexts, the discourse may continue
focusing on the DR as long as the operator binds the variable. Following
Roberts (1986, 1989), in the semantic representation of modal subordination
the necessity operator is included in the DRS both for the antecedent and the
consequent (194).
(194)
x
book (x)
yz
offer (1, x, y)
person (y) □
trad-things (z)
like (y, z)

Resumptive pronouns can be used along the discourse as long as they are
embedded within the corresponding modal operator. Hence a felicitous
continuation of (193) may be (195) which is semantically formalised as (196).

(195) He would be very happy.

(196) x
book (x)
yz offer (1, x, y)
person (y) □
trad-things (z)
like (y, z)

yz
happy (w)
person (y) □ w=y
trad-things (z)
like (y, z)
 Localisation pattern 189

The translation for the DRS in (197) can be paraphrased as follows:

(197) In all of the possible worlds where there is someone who likes tradi-
tional things, then I offer the book to this person.

The antecedent functions as a hypothetical common ground that narrows

down the context set over which the necessity force of the modal operator
will range. Intensional contexts, in fact, introduce possible worlds since
they move us to the universe of desires of an individual. In English, these
contexts are introduced by verbs such as “want”, “desire”, “seek”, “look
for”, etc. English indefinite NPs introduced by such verbs are ambiguous
between having a specific reading (i.e. a de re reading which focuses on the
referent), or a non-specific reading (i.e. a de dicto reading that reproduces
the propositional attitude of the individual). In LSC, in contrast, no such
ambiguity is found. Intensional verbs co-occur with a nonmanual operator.
This operator is overtly expressed as a darting eye gaze directed towards the
upper frontal plane, which is not fixed at a concrete location but rather moves
around (Figure 54). This darting eye gaze is systematically found in LSC
intensional contexts and also contexts where a non-specific entity is denoted.
It is a marking for narrow scope, which is associated with modal subordina-
tion and non-specificity.

Figure 54. Darting eye gaze

This eye gaze functions as an overt nonmanual intensional operator. It has

scope over the introduction of the antecedent in an intensional context, which
narrows down the context set over which the modal operator may range.
190  Specificity

6.5. Existence in the model

Definiteness and specificity encode the difference of knowledge the interloc-

utors of the conversation have about entities in the discourse. Definites are
used when both sender and addressee know the DR; and indefinites, when
the DR is not known by the addressee. Indefinites can be subcategorised
between specific and non-specific. On the one hand, specific indefinites are
used when the entity is known by the sender, but not known by the addressee.
On the other, non-specific indefinites encode that the DR is known by neither
the sender nor the addressee. Note, however, that specific indefinites do not
differ that much from definites once introduced into the model. In fact, Enç
(1991) and von Heusinger (2002) claim that definiteness and specificity is
related to a DR linked to a previously entity or accommodated. Both definites
and specific indefinites are dependent on a domain of discourse, and this
dependence is formally marked in LSC by spatial locations established on
the lower frontal plane. Since they introduce a noteworthy DR the discourse
will focus on, the addressee is not aware of its existence before its introduc-
tion into the model. But once introduced into the model and as the discourse
develops, the DR is incorporated as part of the common ground. Hence both
definites and specific indefinites encode that the entity is part of, or will be
immediately incorporated into, the common ground. This is in opposition to
non-specific indefinites, which encode that the entity has a rather minor role
in the common ground, since it will not be present in a long fragment (again,
modal subordination contexts are an exception, as stated in §6.3.3.2).
As shown along the chapter, this is represented in DRT with the posi-
tioning of the variables in the corresponding DRS. Those DRs that are
introduced into the model and that will have a discourse importance are
represented with a variable in the main DRS. In LSC there is no need for
encoding the difference in what the signer knows that the addressee is aware
and unaware of. What is encoded is rather that the introduced DR will have
importance in the model and that the discourse will develop around it. This is
marked with a location established on the lower frontal plane, and semanti-
cally it corresponds to a global variable. In contrast, DRs that are introduced
but for which no further information will be provided are marked differently,
namely with index signs weakly localised towards the upper frontal plane.
Semantically, this is represented with a subordinate variable.
A formal difference between definites and specific indefinites is absent
in the spatial marking in LSC (it is rather only marked with nonmanuals).
Hence, spatial locations stand for an entity present in the model be it definite
or indefinite (but see Chapter 7 for an incorporation of discourse structure
 Existence in the model 191

into the analysis of spatial locations). Locations on the lower plane encode
what the sender knows (and thus they encode only the sender’s assumption
or belief that P, as in Stalnaker (1974)), but also what the addressee will
know immediately. Hence, they encode what is present in the model, rather
than the status the entity has. In LSC, strong localisation marks that the DR is
present in the model, since it is known by the sender, and that it will be soon
known by the addressee. In contrast, weak localisation marks that neither the
sender nor the addressee knows the DR, and the discourse will not be about
that DR (again, leaving aside cases of modal subordination). In a nutshell,
spatial locations are the overt encoding of what is in the model of discourse,
rather than the distinction between sender and addressee’s acquaintance with
DRs.
Whether embedded variables denote existence in the discourse is a
controversial issue. Along with Karttunen and Heim, I consider that all
variables, independently of embedding conditions, denote existence in the
model. The different nature of embedding corresponds to difference in scope
behaviour. Embedded variables can only exist in the model as long as they
are under the scope of an operator. Non-embedded variables do not have
such restriction. This vision differs from Matthewson (1998), according
to which only variables in the main universe of discourse assert existence.
In the Salish language St’át’imcets neither definiteness nor specificity are
encoded in the language, but rather existence. Matthewson revises the
DRT mechanism, which formalises the distinction between specificity/non-
specificity. She redefines this formalisation into an assertion/non-assertion
of existence, which behaves similarly to specificity. Indefinites under the
scope of an operator do not allow a coreferential pronoun outside it, whereas
indefinites outside the scope of an operator do allow coreferentiality. The
former are considered not to assert existence, while the latter do assert it
(Matthewson 1998: 159–165).
Following Karttunen and Heim, I claim that the main difference is not
in the assertion of existence but rather in the scope of the variable. Both
embedded and non-embedded variables assert existence, and they denote
non-specific and specific DRs, respectively. Yet, only the latter allow coref-
erentiality without any restriction. But as we have seen in the previous
section, modal subordination contexts show that also in narrow scope
contexts coreferentiality can arise as long as the variable is under the scope
of an operator. Hence the notion of “existence” is applied to one of the
universes under consideration in the discourse. Modal subordination cases
verify truth conditions by looking for a possible world where the ante-
cedent is true and then apply the consequent to it. Once the antecedent has
192  Specificity

narrowed down the context set, and chosen a possible world where that is
true, the necessary or possibility force will range over it. This means that in
the particular possible world the entity corresponding to the variable must
necessarily exist. Even if the variable is embedded under an operator, it
has existence in a possible world. Hence, along with Karttunen and Heim,
in this book it is considered that variables denote existence in the current
model and they can be further distinguished between being embedded or not
under an operator.
A spatial location set up in LSC discourse implies existence in the model.
This existence in the discourse can be presupposed or asserted, and it can be
attributed to DRs that are free in their domain and to DRs that are bound to
an operator, such as necessity and possibility. This scope (in)dependence for
wide and narrow scope variables is overtly marked in the two directions of
the frontal plane. Narrow scope denoting non-specificity is marked on the
upper part, while wide scope is marked on the lower part. As shown along
this chapter narrow scope is marked with a weak localisation, whereas wide
scope is marked with strong localisation. This is summarised below.

(198) Existence in the discourse:

– Presupposition  Definite NP
– Assertion  Indefinite NP: Strong localisation
– Specific NP
– Non-specific NP Weak localisation

6.6. Summary

As a general discussion, the LSC data used in the present work raises an
interesting aspect, namely the overt marking for specificity distinctions,
which so far has not been treated in as much detail in other works of SLs. The
previous pieces of research studying the spatial definiteness marking in ASL
and HKSL (§5.1.4) show a difference with respect to LSC. Two hypotheses
arise in this respect:

1. We are facing a language variation phenomenon

2. The analysis of ASL and HKSL has been considered under the prisma
of definiteness distinctions widely attested in English and romance lan-
guages, where no specificity distinctions in the determiner system are
found.
 Summary 193

Future cross-linguistic studies may help figuring out whether this is a

language specific variation and different SLs have a particular codification
for definiteness and specificity in the determiner system, or rather western
sign languages have a similar codification. Furthermore, it will be interesting
to look at the differences of marking in the use of signing space as maybe
other SLs codify (non-)specificity distinctions in the horizontal and midsag-
gital plane, as opposed to the frontal plane.
Summarising the findings so far, the analysis of the data and the correla-
tion with the expression of specificity shown in this chapter can be taken as
evidence of good motivation for the localisation pattern. The results from
the analysis of the small-scale LSC corpus data have led us to the distinction
between a strong localisation and a weak localisation. Strong localisation
uses more co-occurring mechanisms directed to the lower part of the frontal
plane and a well-established spatial location is set up with the feature [p].
Weak localisation uses fewer mechanisms that do not necessarily overlap
and are directed to an upper part of the frontal plane. Hence, a spatial loca-
tion is weakly established, which corresponds to the feature (p)[up]. This
distinction motivated by the data has a direct correlation with the semantics
of the language, and more concretely with specificity marking. On the one
hand, strong localisation is associated with specificity, and this is supported
by the argument that the three properties specificity encompasses, namely
wide scope, partitivity and identifiability are found on strong localisation.
On the other hand, weak localisation is associated with non-specificity. This
is also supported by the fact that the three properties, namely narrow scope,
non-partitivity and non-identifiability are applied to weak localisation. Recall
that the arguments provided in §4.2.2 concerning narrow scope and binding
by an operator were instantiated with lack of spatial location establishment.
However, only isolated sentences were treated then. Now, once connected
discourse is considered, the binding by an operator denoting specificity is
instantiated in LSC grammar with a weakly established location.
Some works about specificity (von Heusinger 2008, 2011b) consider
noteworthiness to be a property encompassed by specificity. According to
these works, since with specific DRs remarkable information about the DR is
usually provided along the discourse, the variable has a noteworthy feature.
However, this chapter has proved that in cases of modal subordination
narrow scope can also be related to noteworthiness. As long as the variable
is embedded under a modal operator, further information can be provided.
Thus I do not consider noteworthiness to be only a property of specificity,
but rather a property related to discourse structure and that it is orthogonal to
194  Specificity

specificity. Both specifics and non-specifics can be noteworthy. In LSC, for

instance, there is a close relationship between strong localisation and note-
worthiness (understood as topicality of the DR). As shown in the following
chapter, both wide and narrow scope variables can occupy a spatial location
in signing space as long as they are connected to the most prominent DR. In
Chapter 7, discourse structure is incorporated in the analysis of the properties
that DRs may have in order to establish a spatial location.
Chapter 7
Discourse structure and prominence

As discourse progresses, the discourse entities that are the object of our atten-
tion constantly change and, therefore, the attentional structure has to model
these changes. Grosz and Sidner’s proposal is that the attentional states of the
interlocutors can be modelled by a set of so-called focal spaces.
(Vallduví 1997)

7.1. Introduction

This chapter builds on the principles concerning (p) exposed in the previous
chapters and presents fine-grained hypotheses which enhance the analysis of
Catalan Sign Language (LSC) spatial locations specifically on the horizontal
plane. The goals are two-fold. First, it is shown that lower spatial locations
correspond to discourse prominence, defined as variables with backward
looking properties as well as forward looking properties, independently
of the scope of the determiner attached to the variable. Previously, I have
defended that narrow scope variables correspond to a lack of spatial location
establishment, as presented in Chapter 4, or also to an upper established loca-
tion when denoting non-specificity, as presented in Chapter 6. However, here
it is shown that narrow scope variables that are linked to the most prominent
Discourse Referent (DR) at a specific point in a discourse behave like wide
scope ones and establish a lower spatial location. The second goal is related
to the nature of spatial locations. (p) is an abstract point in space, which does
not correspond to an exact point nor it is related to a specific direction on
spatial planes. In connected discourse, locations associated with the most
prominent DR can be shifted in space, showing that the exact direction on
planes is irrelevant for the nature of (p). What is relevant is that the spatial
location (p) is associated with a discourse referent (DR) from the model
independently of the direction of the referring term in signing space.
The structure of this chapter is as follows. In §7.1 the theoretical ingre-
dients for the presentation of the fine-grained hypotheses are presented,
namely discourse anaphora and underspecification, prominence and note-
worthiness. In §7.2 sign language pronominal issues are laid out focusing on
an important distinction between pronominal forms and DRs. §7.3 presents
196 Discourse structure and prominence

an implementation of a representational semantic level, which integrates a

theory of discourse structure with special focus on prominence. §7.4 anal-
yses spatial locations established in long stretches of discourse and it claims
that they are underspecified forms. §7.5 summarises the main findings of the
chapter.

7.2. Background

This section presents the theoretical background related to discourse structure

relevant for the present account. First, some notes on discourse anaphora and
underspecification are presented. Then, prominence issues are introduced.
Finally, noteworthiness aspects conclude this framework section.

7.2.1. Discourse anaphora and underspecification

Coreferential pronouns function as free variables and are not bound at the
syntactic level. In this respect, pronouns can be used without a linguistic
antecedent (such contexts are cases of weak familiarity seen in Chapter 5),
since the antecedent does not need to be always overt (199). It can also
happen that the antecedent is overt but occurs in a preceding sentence (200).

(199) While walking into a room:

What is he doing here?

(200) I couldn’t reach Santi last night through Skype. He was probably at
the Deaf club.

In dynamic semantics it is considered that “an anaphoric expression

harkens back not to another expression, such as a noun phrase (NP) in prior
discourse, but to a DR, an element of that structured information” (Roberts
2005: 1). This chapter is devoted to coreferential uses. That is, it is centred
on discourse anaphora of the kinds similar to (199) and (200), where there is
an explicit linguistic antecedent in previous discourse, or a direct reference
to an object present in the physical context, as already seen in Chapter 5.69
To determine the semantic value of he in (199) and (200) we need to know
what its antecedent is and what its semantic content is. In (200) Santi is
the antecedent, i.e. the referentially independent expression from which the
anaphoric expressions gets its reference. The pronoun he is the anaphor, that
 Background 197

is to say, the referentially dependent expression. Both Santi and he refer

to the same individual, thus they are co-referential. This is what is called
discourse anaphora, defined as the phenomenon whereby the interpretation
of an occurrence of one expression depends on the interpretation of an occur-
rence of another element, which has been explicitly or implicitly introduced
in sequences of sentences.
When studying discourse anaphora, it is also relevant to look at the
kind of referring terms that are used. DRs are encoded in different ways in
discourse. In fact, senders make different assumptions about the informa-
tion-status of that DR in the addressee’s mind and then choose a referring
expression accordingly (Prince 1992; Vallduví 1992). DRs may be promi-
nent in the discourse, they may be known but not prominent, they may be
new by the addressee, or inferrable from what the addressee knows. All these
factors determine the choice of the different linguistic elements that a sender
may make in order to refer to a DR, such as a definite NP, an indefinite NP,
a pronoun or a null argument, depending on the language. Natural languages
provide a wide range of referring terms and sign languages are not an excep-
tion. To refer to objects of thought, signers can also use a definite description,
a proper noun, but also a pronominal form.
Referring terms pick up or refer to DRs of the universe of discourse,
i.e. the objects of thought the conversation is about. The different kinds of
referring terms present different properties. Full NPs and proper names have
reference independently and they select a specific DR from the universe of
discourse. In contrast, pronouns and demonstratives do not select inher-
ently a DR from the universe of discourse. Rather they restrict the entities
to which they can refer to but they do not specifically pick one DR up. An
example of this is the proper name Francesc, which rigidly picks up the
DR for “Francesc”, which is ontologically connected to the real human man
who is my partner. Hence between the proper name and the DR there is a
direct reference connection. In contrast, pronouns have some features that
allow restricting the amount of potential antecedents. The specific features
of English pronouns, for instance, are gender, number and case. For example,
the pronoun he has the features [+masc], [+sg], and [+nom]. He does not
directly identify the specific DR to which it refers, but rather it selects a
subgroup of possible entities, namely those that are [+masc], [+sg], and
[+nom] from the universe of discourse domain. Once the selection is done,
syntactic and pragmatic constraints allow to pick up the corresponding DR
among the set of possible DRs.
As underspecified and definite expressions, pronouns imply a familiarity
presupposition because its mere existence in a discourse presupposes that
198 Discourse structure and prominence

there is already a DR corresponding to that variable. As seen in cases of

weak familiarity in Chapter 5, this variable corresponds to a default vari-
able δ when the DR is accommodated into the common ground. The use of
a pronominal form also implies that they occupy a prominent position in the
discourse structure. For instance, a sentence like (201) uttered out of the blue
is only felicitous if there is a prominent female DR in the linguistic context,
or in the immediate physical context.

(201) She is a very smart linguist.

Underspecified referring terms, such as pronominal forms, are characterised

by two main features:

(i) They inherently specify some properties of the DR and they act as set
restrictor devices among the entities from the universe of discourse.
(ii) Their referential interpretation is dependent on a prominent DR they
pick up.

As argued in Chapter 5, the DR can have a corresponding overt form in the

previous linguistic context, or it can be absent from the linguistic context.
That is, no antecedent is found in the linguistic context, but rather the
pronominal form gets its reference from the connection with the DR that has
a corresponding object in the physical context of the conversation.

7.2.2. Prominence

The structure of discourse is a composite of three distinct but interacting

components. Following Grosz and Sidner (1986), these components are a
structure of the actual sequence of utterances in the discourse, a structure
of intentions (i.e. purposes) and an attentional state (i.e. focus of attention).
The interaction of these three components contributes to the dynamism of
discourse where DRs appear and disappear within discourse fragments. In
different fragments, the prominence of DRs can vary and if at some point in
a discourse an entity is very prominent, at another point it may not be that
prominent (Lewis 1979). Prominence is defined as the degree of relative sali-
ence of a unit of information, at a specific point in time, in comparison to the
other units of information (Chiarcos, Claus, and Grabski 2010). Nowadays,
Centering Theory (CT, Brennan, Friedman, and Pollard 1987; Grosz 1981;
Grosz et al. 1995; Walker, Joshi, and Prince 1998) represents probably the
 Background 199

most influential account of entity-based prominence in discourse. Centering

is a processing model that relates the local utterance-by-utterance context
and discourse anaphoric reference, which has been applied to study different
languages, such as Italian (Di Eugenio 1998), Japanese (Walker et al. 1994),
Turkish (Turan 1995), and Chinese (Qinan 2008), among others. It consti-
tutes a basis to theorise about local coherence, prominence and choice of
referring expressions, such as the difference of use between pronominal
forms and definite NPs (Grosz, Joshi, and Weinstein 1983). The parame-
ters of the theory have been proved but also questioned in some works (see
Poesio et al. 2004).
CT assumes that attention has to be focused or centred in discourse. It
introduces independent terminology, which is adapted here. For instance,
centres are defined as entities that serve to link an utterance U with other
utterances. They actually have the same properties as “discourse referents”
(see §4.1.3), and this is the reason why the terminology of CT is adapted
to the one used in the present book. The most important notions are the
following:

(i) Forward Looking DRs: set of DRf(Uk) which appear in the DRS K
and that can be referred to in subsequent utterances.
(ii) Backward Looking DR: a unique entity DRb(Uk), defined for each
utterance Uk (except for the initial segment) that refers back to a for-
ward looking DR of the preceding utterance Uk-1, and that, intuitively
represents the DR which is the centre of attention at utterance Uk.
(iii) Preferred DR: DRp(Uk) is the one that is on the top of the hierarchy of
the set of DRs in the main DRS.

The backward-looking DRb(Uk) is selected from the set of DRf (Uk-1). Hence
DRb(Uk) connects utterances with preceding discourse. DRf(Uk) are the set
of variables present in the DRS and they are organised in a prominence
ranking, which serves as likelihood to serve as backward-looking centre
of DRb(Uk+1). In the present account, I argue that the ordered ranking of
DRf(Uk) is not dependent on the position the variable occupies in the DRS,
namely whether the variable is attached to a wide scope quantifier and hence
inserted in the main DRS, or rather attached to a narrow scope quantifier and
then inserted into a subordinated DRS. As we will see below in §7.3, subor-
dinate variables can also be promoted to DRp(Uk) as long as they are under
the scope of the corresponding modal operator.
The degree of prominence of a DR directly affects the referring term that
will be chosen to denote such a DR. The form chosen reflects the prominence
200 Discourse structure and prominence

of the entity within a specific fragment and, according to the literature,

pronominal and weaker forms are the expressions used when the entity
is actively prominent in the consciousness of the addressee (Ariel 1988,
1990; Gundel, Hedberg and Zacharski 1993;70 Prince 1981).71 Accessibility
theory (AT, Ariel 1988, 1990) is the account that has influenced the present
book most, since it offers a procedural analysis of referring expressions, as
marking varying degrees of mental accessibility. The basic idea is that refer-
ring expressions instruct the addressee to retrieve a certain piece of given
information from his memory by indicating how accessible this piece of
information is to him at the current stage of the discourse. One important
contribution of Ariel’s approach is that AT takes into account that natural
languages provide senders with the means to code the accessibility of the DR
to the addressee. As Ariel (1996: 22) claims:
“AT predicts that when antecedents are nonprominent/distant/in less cohe-
sive units to the potentially anaphoric expressions, a relatively lower acces-
sibility marker will be chosen. When the antecedent is prominent/recent/in
highly cohesive unit, it will be coded by a relatively high accessible marker.”
AT is articulated in a hierarchy where NPs formed by full noun and a modi-
fier are considered to be low accessibility markers, and verbal person inflec-
tions and null arguments are considered to be high accessibility markers.
These form-function correlations on the accessibility marking scale are not
arbitrary and there are three partially overlapping criteria, namely (i) inform-
ativity (the amount of lexical information), (ii) rigidity (the ability to pick
a unique DR, based on the form), and (iii) attenuation (phonological size/
reduction). The more informative, rigid and unattenuated expressions will
code a lower degree of accessibility; and the less informative, rigid and more
attenuated expressions will code a higher degree of accessibility. Both the
notions of informativity and rigidity are used in §7.4 in the analysis of index
signs in LSC.
Accessibility factors may be influenced by different issues related to
the particular discourse model. For example, if most references go to the
global discourse topic, pronouns may be used even if there is a great distance
between the DR and the anaphor. It seems that rather than viewing the factors
separately, we should think of a combination of them for the prediction of
the occurrences. Any particular instance may be coded by values both high
and low in the hierarchy, and only the combination of the factors will deter-
mine the degree of accessibility used by the sender. In order to determine the
degree of accessibility of a DR, Ariel proposes that different factors interact
with prominence (determined by many other factors: grammatical function,
 Background 201

high vs. low physical prominence in the context, order of mention, definite-
ness and quantification of the NPs), competition among other potential ante-
cedents for an anaphoric expression, distance (recently mentioned entities
are more accessible than remotely mentioned ones) and cohesion unity. One
important point in the AT is the importance given to mental representations.
Ariel argues that it is the discursive rather than the physical prominence of
the entities involved that determines the degree of accessibility assigned to
a particular mental representation. Although the physical context does affect
the discourse model, mental representations are the only direct product of the
discourse model. LSC, and SLs in general, consitute a good example for this
statement, as seen in deictic contexts in Chapter 5. The following section is
devoted to the last phenomenon related to discourse structure.

7.2.3. Noteworthiness

Noteworthiness requires the statement of something remarkable about the

individual denoted by an indefinite NP. It has been claimed that notewor-
thiness depends on scope since only wide scope allows noteworthiness to
arise (von Heusinger 2011b). However, modal subordination contexts where
narrow scope variables can be further referred back to contradict this claim,
as seen in §6.3.3.2. This is the reason why in this book noteworthiness is not
considered to be a property of specificity, but rather an orthogonal phenom-
enon between specificity and discourse structure, because both wide and
narrow scope variables may be noteworthy.
Spoken (informal) English makes use of an existential construction
co-occurring with an indefinite-this for DRs introduced for the first time
(Ionin 2006; Prince 1981). This-NP functions as an indefinite and it intro-
duces a novel individual into the discourse as in (202). When we try to use it
as a definite it results in an odd construction such as (203) below.

(202) A colleague complaining about having problems with quantification:
a. There is this book on mathematics that may be helpful.
b. I realised that you are having problems with the quantification of
your data. Maybe you could check for some hints in a book on math-
ematics and there is this book on mathematics which may be helpful.

Both Prince (1981) and Ionin (2006) claim that this-NPs are found in construc-
tions denoting specific DRs since specificity tests for coreferentiality apply
202 Discourse structure and prominence

to them (see §6.1). Consider the well-known example in (203) and compare
it with its analogue in (204):

(203) A: John wants to marry a Norwegian.

B: Is she tall?
B’: Must she be tall?

(204) A: John wants to marry this Norwegian.

B: Is she tall?
B’: #Must she be tall?

An NP can be anaphoric to a nonspecific DR, as exemplified in (203B’). But

(204B’) shows that the nonspecific continuation is not felicitous when the DR
is introduced by an indefinite-this NP. Hence this-constructions in English
are considered to be specific and indefinite (Gundel, Hedberg, and Zacharski
1993; Ionin 2006; Prince 1981; see von Heusinger 2011a for similar claims
about German dies). Indefinite-this NPs introduce new information into the
model that is mentioned for the first time. But the new topic they introduce
is information that is going to be talked about in the following discourse.72
Noteworthiness is defined by Ionin (2006: 180) as those elements for which
“the speaker has a particular DR in mind, about which further information
may be given” (following Maclaran 1982). The property of noteworthiness
requires thus the statement of something noteworthy about the individual
denoted by the indefinite (Ionin 2006). It is used to index persistent DRs
in the discourse model, but also to keep track of antecedent-pronoun rela-
tions confined to the scope of quantifiers. Noteworthiness, in fact, coincides
with the definition of forward looking DR, presented in the previous section
according to which it is the DR from the set of DRf(Uk) which is referred to
in subsequent utterances.
As we will see, when establishing (p) in LSC this is very much connected
to noteworthiness. That is, all DRs localised in signing space are referred
back by a resumptive pronoun in most of the cases, which implies that the
DR denoted is a prominent and a noteworthy entity. It is the entity on which
discourse is focused and most likely it is the one that will be referred back
to. Hence it constitutes the discourse topic. Although in most cases topics
tend indeed to represent old information, this is neither a sufficient nor a
necessary condition for topicality. Topics are better analysed in terms of
their effect on the ongoing discourse and considering the effects of previous
discourse on the given utterance, rather than as old information (Reinhart
1981; Vallduví 1992). This is why discourse topics that are prominent at a
 Sign language pronominal issues 203

specific discourse fragment have both backward looking properties, as well

as forward looking properties (see §7.3,). It is now time to turn to the appli-
cation of these concepts to LSC. First, the relation between LSC pronouns
and DRs is presented.

7.3. Sign language pronominal issues

Since this book focuses on DRs and what the discourse is about, I will only
concentrate on third person pronouns. In fact, this distinction is already
made in the literature about pronouns. The term “third person” is negatively
defined with respect to “first person” and “second person”: it does not corre-
late with any positive participant role (Lyons 1977: 638). First and second
person pronouns are the most basic deictic elements because they “point to”
and directly refer to the speech participants, which need to be present in order
the speech act can occur.73 Third person pronouns have a different nature and
should not be included under the same category as pronouns (Bhat 2004;
Lyons 1977). On the one hand, the function of first and second person
pronouns is primarily to indicate the two principal speech roles, namely that
of being the sender and the addressee, respectively. On the other hand, third
person pronouns are used to refer to the thing the discourse is about, the non-
person. Of course, it may be that in some contexts the thing the discourse is
about is a first or a second person. However, in such a context they do not act
as conversation participants, but rather as the topic of the discourse. As my
interest draws on the entities in the discourse (that is, on what the discourse is
about) I will narrow the focus of attention to third person pronouns, leaving
aside first and second personal pronouns, which are always used locally.

7.3.1. Infinity and unambiguity

It is widely considered in the sign linguistics literature that since pronouns

unambiguously point at their DRs, reference in sign language is unambig-
uous, at least within the grammatical structure of the individual sentence or
of the discourse frame (see §2.4). In contexts of sloppy readings of pronouns,
Lillo-Martin and Klima (1990) show that overt pronouns in ASL make the
sentence non-ambiguous. However, the fallacy of the unambiguous use of
space can be easily challenged by three contexts. The first one is the use of
the same spatial location for both the DR and the physical location that the
DR occupies (Emmorey 2002a:56). The second one is the context where a
204 Discourse structure and prominence

semantic affinity convention is used (Engberg-Pedersen 1993, and see also

§3.3.1.2). That is, when two DRs sharing a semantic relation are localised
in the same area of signing space. The third context stands for attested cases
in which two locations may be associated with a single referent (Johnston
1991; van Hoek 1992). In fact, some works show that there is not necessarily
a one-to-one relationship between (p) and DRs (Janis 1992; Nilsson 2010;
Russell and Manitoba 2008), since in ASL and Swedish Sign Language (SSL)
a single location may be associated with multiple absent DRs. This is strong
evidence against the supposed referential non-ambiguity in SL pronouns, as
shown in Emmorey (1997).
It is also widely accepted that since signs can be directed to an infinite
number of directions in signing space, there is an infinite number of distinct
pronominal forms (Janis 1992; Liddell 2003; Lillo-Martin and Klima 1990;
Meier 1990, and see also §2.4). Indeed, as will be shown below, there is
a potential infinite number of DRs, but not of pronominal forms. Third
person pronouns consist in an index directed to signing space with specific
nonmanuals, which differs from second person form. Importantly, the argu-
ment concerning the infinite number of possible locations is related to the
unambiguity argument. Since there are an infinite number of possible loca-
tions, each location is assigned a single DR. Therefore, the use of a pronoun
will uniquely identify a DR and there will be no ambiguity (see also Janis
1992). But if one thinks of a discourse model with an infinite use of locations,
unambiguity and infinity are related with difficulty since it is impossible to
keep track of referents unambiguously when such a big number of DRs are
present in the model. In such a situation, an infinite number of locations
are associated with many different DRs and hence a one-to-one relationship
will be very difficult to be established and impossible to interpret. Thus, no
unambiguity can be claimed here. This shows that in fact, infinity and unam-
biguity are opposed phenomena and they cannot co-exist. It is indeed diffi-
cult to understand why it has been argued that these two opposed concepts
are present when analysing sign language pronouns.

7.3.2. Pronouns versus DRs

In the SL literature there is an overlapping between the notions of “discourse

referent” and “pronominal form”. For instance, Janis (1992: 169) argues “as
more forms are established and thus entered into the memory, the degree of
 Sign language pronominal issues 205

preciseness of location needed by the forms is greater. When there is only

one established pronoun form, the use of any pronoun is always closer in
form to it than to other pronouns, since others are nonexistent”. Frequently in
the literature the expression “pronoun” is used to mean “discourse referent”.
A pronoun is not established in discourse, but it is used in discourse to
establish a DR. Pronouns and DRs are two related but distinct notions and
they need to be kept apart. As defined in §4.1.3.2, pronouns are referring
terms from natural languages expressed with a concrete morphophonolog-
ical marking that denote DRs. DRs are semantic entities or objects of thought
the discourse is about. It is not that there are an infinite number of possible
pronominal forms, but rather an infinite number of possible associations with
DRs. While pronouns are discrete and restricted units provided by natural
languages, DRs are provided in the discourse model and they can be infinite
depending on each discourse model, as we will see below.
In line with the three-person distinction analysis (Alibašić and Wilbur 2006;
Berenz 1998; Meurant 2008; Neidle and Lee 2006, and see also §2.5.3 for the
controversial analysis of person features encoded in sign language pronouns),
I claim that there is a single third person pronominal form once the nonmanual
component is included in the morphophonological form itself. Pronominal
forms are a combination of a manual index sign with specific nonmanuals,
which establish (p). This establishment is articulated with an alignment of
hand, eye gaze, head and chest (for second person pronouns) towards (p), or
with a non-alignment (for third person pronouns). The nonmanual compo-
nents are morphological elements included in the pronominal forms. In
Table 8 the morphological features included in the pronouns considered in
this book and inspired in Berenz et al.’s analysis are shown. Index handshape
is the default handconfiguration used for the three pronominal forms.74 But
still there is a further distinction. On the one hand, “B” handshape can also
be used for second person in polite and formal conversations. On the other,
thumb handshape can only be used for third person when the person referred
to is absent in the physical environment. Concerning signing space, second
and third person pronouns establish a spatial morpheme that happens to be
body-anchored in first person pronouns. Nonmanual alignment is what distin-
guishes second from third person pronouns. While the former follow an align-
ment of manual and nonmanual components, the latter show a non-alignment
in the features. Obviously the cell of nonmanual alignment with first person
pronouns is empty since there is not an establishment of a location in space
towards which nonmanuals may be directed, but rather on the signer’s body.
206 Discourse structure and prominence

Table 8. Features of pronominal person distinction

Person Handshape Spatial Nonmanual
distinctions morpheme alignment
First person -Index Body-anchored75
pronoun
Second person -Index Point-space +Alignment
pronoun -B
Third (person) -Index Point-space –Alignment
pronoun -Thumb

Once the nonmanual component is included in the pronominal form, the

controversy on the three-person distinction is minimised. A pronoun sign
may be decomposed in different parts: a handshape, and a direction and a
hold, which manifest the established spatial location (p). But (p) does not
indicate person distinction since person features are indicated by the (non-)
alignment of manual and nonmanual components. The actual location
in space indicates a DR, which can be one among an infinite number of
possible DRs. A discourse model may contain an infinite number of possible
DRs (i.e. variables) but the referring expressions chosen to refer to them are
among a limited set that natural sign languages provide. Thus the infinity
issue is transferred from pronouns to DRs and this is a permissible move
since DRs are constrained by the discourse model, which can only be limited
by perceptual and memory limitations, but not by purely linguistic reasons.
This reasoning shows that the number of DRs in a discourse model may be
infinite and they may be referred to by a single third person pronominal form.
To see more closely how pronouns are distinguished from DRs we should
also look at some examples. In §7.1.1 we have seen that pronouns are under-
specified elements that can refer to an infinite number of DRs. They have a
set of features that restrict the entities to which they can refer. The English
third person masculine pronoun he in nominative case can refer to an infinite
number of male individuals, depending on the number of male individuals
present in the discourse model. But we have a unique nominative form to
denote third person male, namely the referring expression he. Likewise,
in SL there is a unique third (person) pronominal form, which is a manual
index that establishes a location in space and a disjoint alignment with the
nonmanual component. This unique pronominal form may also be associated
with a potential infinite number of DRs depending on the variables included
in the discourse model. However, this has passed unnoticed among some
researchers. Lillo-Martin (2002) and Meier and Lillo-Martin (2010) argue
 Sign language pronominal issues 207

for a combination of linguistic and gestural components in the pronouns in

line with the perspective held by the infinity view. Considering the English
sentence in (11), Lillo-Martin argues that the speaker may point via a gesture
to any location in space in order to disambiguate the pronouns. These loca-
tions depend on the actual position of the present DRs and thus are not
listable in the language. Lillo-Martin’s proposal is that non-first singular
SL pronouns (following Meier’s account) are lexically and syntactically
ambiguous, just as the English him used deictically is ambiguous in (205).
However, the combination with a gesture makes the reference unambiguous.

(205) I saw him and him, but not him.

(Lillo-Martin 2002)

As said before, this kind of analysis shows confusion between pronouns and
DRs in SL and this book argues for a dissociation of the two categories.
Pronouns are the linguistic material used to refer to a DR, that is the semantic
individual to whom it is associated (see the distinction between linguistic
marking and semantic entity presented in §4.1.3.1). In (8) the unique
pronominal form him can denote at least three different male individuals.
But we only have one pronoun, that is one grammatical accusative form that
has the potential to be associated with the three individuals in the present
model. The same goes for LSC. A single element formed by a manual index
sign with an establishment of (p) and a disjoint alignment with nonmanuals
is used to refer to third person. This single form can be used to refer to an
infinite number of possible DRs depending on the variables present in the
model.
Below there is a sentence similar to (205) uttered in English and accom-
panied with co-speech gesture. The representation in Table 9 shows that
the same pronominal form him can be associated with two DRs from the
discourse model, namely Frank and Paul. Since it is an example of direct
reference, a pointing gesture co-occurring with the two instances of him is
needed in order to disambiguate the sentence.

Table 9. English sentence with co-speech gesture

“I saw him but not him.”
Referring expression (third person “him”
pronoun singular)
Discourse referents him-1 (= ‘Frank’)
him-2 (= ‘Paul’)
208 Discourse structure and prominence

Table 10 is the same context but without co-speech gesture. Since it is again an
instance of direct reference and there is no co-speech gesture accompanying
the sentence, it turns out to be ambiguous. Let us accept for the interest of
the argumentation that this unusual context can be found. The sentence is not
felicitous in a context with two male DRs, and in an unmarked context where
no DR is more prominent than the other one, the sentence is ambiguous.

Table 10. English sentence

“I saw him but not him.”
Referring expression “him”
(third person pronoun singular)
Discourse referents him-1 (= ‘Frank’)
him-2 (= ‘Paul’)

Table 11 corresponds to the LSC sentence. The unique third person pronom-
inal form in LSC is also used to denote the two DRs from the model. The only
difference between the English and LSC counterparts is the overt connection
with the DR manifested with the establishment of the spatial location (p).
The location is associated with a DR and disambiguates the reference of the
third person pronoun.

Table 11. LSC sentence

ix1 see ix3a, ix3b not. (“I saw him but not him”)

Referring expression ix + nonalignment + (p)

(third person pronoun singular)
Discourse referents ix3a (= ‘Frank’)
ix3b (= ‘Paul’)

Importantly, both the English and the LSC pronouns are proforms and as
such they always need the linguistic, as well as the extra-linguistic context,
to recover their meaning. But if the LSC sentence is not inserted under a
discourse model, the sentence turns out to be as ambiguous as it is in the
spoken language version (Table 10). Without a discourse model to interpret
it, the sentence remains ambiguous. The similar behaviour between contexts
in Table 10 and Table 11 shows that there is no gestural component bound to
the manual form in LSC. The two sentences are similarly ambiguous without
a discourse model. This shows that the two languages behave the same way.
 Prominence 209

So far we have seen that (p) stands for the overt manifestation of a DR, as
presented in (206) below.

(206) The discourse referent hypothesis (second version)

(p) is the overt manifestation of DRs attached to a quantifier that has
wide scope.

In Chapter 4, we have also seen that the coincidence in the direction of the
establishment of (p) resolves the equation identity in a DRS through the
construction rule for pronouns (see §4.2.2), as stated in the already intro-
duced hypothesis:

(207) The spatial point hypothesis (first version)

The identity condition in the DRS is encoded through coincidence in
direction of spatial establishment of (p).

According to (206) and (207), the main striking difference between spoken
and signed languages is that to refer to a DR the latter have a feature expressed
overtly, which in the former is expressed covertly. In the following section,
we will see that when studying spatial locations from a discourse perspective,
they are also involved in factors that derive from the dynamics of discourse
and they incorporate attributes related to prominence. As we will see, the
exact direction in the horizontal plane where (p) is established is irrelevant
for the association with the DR. In contexts of prominent DRs, (p) referring
to the same DR may be established in different areas on the horizontal plane,
namely on the ipsilateral and the contralateral areas, without affecting the
propositional meaning. This shows that (p) consists indeed in an abstract
point in space that does not depend on the direction of spatial planes where
it has been established and it is still categorically interpreted. Hence, in the
following sections it is claimed that two above mentioned hypotheses need to
be revised. I first start with the revision of The discourse referent hypothesis.

7.4. Prominence

7.4.1. Global discourse structure

The arrangement of spatial locations associated with a DR for a signing utter-

ance or discourse fragment is called “frame of reference” (Lillo-Martin and
Klima 1990: 193). The number of DRs that may be localised in space within
a frame of reference has been a frequently asked question that researchers
210 Discourse structure and prominence

have tried to answer (Janis 1992; Loew 1984; McBurney 2002, among
others). McBurney attributes the limitations to general cognitive abilities
and suggests that the number of DRs in space depends on the capacity of
the working memory and claims that the limit is somewhere between five
and seven (based on Miller 1956). However, according to Janis (1992: 103)
in ASL the number of DRs that may be expressed through role shift, which
is up to 2, is smaller than the number of DRs that may be expressed by
personal pronouns, which is up to 3. Ahlgren and Bergman (1990: 258) claim
that in Swedish Sign Language (SSL) narratives the marking for reference
on spatial locations is made for the number of DRs minus one. One DR is
always referred to through role shift, and hence not localised in space. This
distinction on the kind of referring term indicates that the constraint is a
linguistic one, rather than a constraint on memory or perception.
When a very large number of DRs is established within a fragment of
discourse, the list sign is used. This has been labelled “list buoy” (Liddell
2003; see also Liddell, Vogt-Svendsen, and Bergman 2007; Vogt-Svendsen
and Bergman 2007 for a cross-linguistic comparison among ASL, Norwegian
and Swedish SL). It consists in an open non-active hand, with the fingers
extended horizontally. According to Liddell, the fingertips stand for associa-
tions with entities and they are a substitute for spatial location establishment.
However, although there is an association between a fingertip and an entity,
it mainly serves to enumerate a list of entities. The main function of the list
sign is to establish an order among entities introduced.
Our data from the small-scale LSC corpus shows a clear contrast between
the use of the list sign and the use of spatial locations. When two DRs are
present in the model they are established in space. Whenever more than two
DRs are present in the same fragment of discourse the list sign is used to
enumerate them and establish an order. However, it is usually the case that
one DR is more prominent than the others. In order to analyse prominence,
the DRT semantic representation of the discourse used so far in the book is
not enough. A new version of DRT that combines a representational semantic
level with a theory of discourse structure (with special emphasis on promi-
nence) such as Centering theory is presented here.

7.4.2. Topicality

To the best of my knowledge, the first and only attempt to combine a

semantic and a discourse structure level of analysis are Pinkal (1986) and
 Prominence 211

Roberts (1998). These proposals combine DRT with the first works on
Centering Theory (CT, Grosz and Sidner 1986). But since CT has been the
focus of later research, I take as a basis subsequent works (Grosz, Joshi, and
Weinstein 1995; Walker, Joshi, and Prince, 1998). As previously said in §7.1,
the discourse topic, which is the most prominent entity at a specific point in
discourse, is better analysed in terms of its effects on the ongoing discourse,
as well as considering the effects of previous discourse on the given utter-
ance (Reinhart 1981; Vallduví 1992). The present proposal claims that a DR
is linked to the discourse topic of the fragment of discourse (i.e. it is the
most prominent entity of that specific fragment) if it verifies the following
formula:

(208) DRb(Un)=DRb(Un-1) ˄ DRb(Un)=DRp(Un)

In (208) the first argument verifies the previous utterance

(DRb(Un)=DRb(Un-1)), and the second argument verifies the subsequent
utterance (DRb(Un)=DRp(Un)). The intersection between the backward
looking DRb(Uk) and the preferred DRp(Uk) of the current utterance yields
the most prominent DRk at a specific point in a discourse. The preferred
DRp(Un) is the most likely to become the discourse topic at that point in the
discourse. Importantly, the DR that verifies the formula in (209) combines
the backward properties as well as the forward properties to determine the
prominent DR. Since the prominence of DRs may vary in different contexts,
the formula in (209) needs to be verified in every fragment of discourse.

7.4.3. Topical variables

The set of forward looking variables DRf(Uk) are not only restricted to the
ones appearing on the main DRS, but also to subordinated variables as long
as they are embedded under the corresponding operator. In LSC, the DRp(Uk)
chosen among the DRf(Uk) will be correlated with a spatial location as long
as it verifies (208) and independently of the scope of the quantifier attached
to the variable. An example of a wide scope variable that is prominent in
this fragment of discourse is shown in (209), which is the continuation of
example (163) in Chapter 6.

eg:ip-l
(209) ix1 1-offer-3ip-l one person-3ip-l pen-drive computer pen-
drive
212 Discourse structure and prominence

eg:ip-l eg:ip-l
1-offer-3ip-l, because person-3ip-l always++ work theme is/
same computer.
eg:ip-l
pen-drive adequate ix1 1-offer-3ip-l ix3ip-l pen-drive.
ix3ip-l happy, enjoy.
‘I will offer the pen-drive to someone, since he/she/this person always
works with computers. I find it very adequate to offer the pen-drive to
him/her. And he will be very happy and enjoy it a lot.’

The semantic representation for the first sentence is built and the corre-
sponding DRS is shown in (210).

(210) x y z
pen-drive (x)
person (y)
offer (1, x, y)
work-computer (y)

In the second and third sentence pronominal forms appear. Thus the construc-
tion rule for pronouns is used. The variables are established and the identity
equation is introduced.

CR.PRON: Upon encountering a pronominal form,

1. trigger the syntactic configuration [s NPα [VP]] or [s VP [NPα]], and
2. introduce a novel discourse referent α into the main DRS, and
3. check which variable in the main DRS shares the features α has, and
4. if no suitable variable is found, go to CR.PRON2; if the suitable variable
is found introduce an identity equation α = γ
5. go to CR.PROM

The algorithm tells then to go to the following construction rule for promi-
nence. According to it, the variables in the equation need to verify the formula
presented above as (208).

CR.PROM: Upon encountering an identity equation α = γ,

1. check the variable that verifies the prominence rule:
DRb(Un)=DRb(Un-1) ˄ DRb(Un)=DRp(Un)
2. assign the superindex p to the variable,
 Prominence 213

3. check the conditions in the DRS equated to the variable β = α,

4. assign the superindex p to the suitable variable

The variables that verify the formula are assigned a superindex p in the
semantic representation. For the sentence in (209), it is variable z and w, as
shown in (211).

(211) a. [DRb(Un)=DRb(Un-1) ˄ DRb(Un)=DRp(Un)] ≡ z

b. [DRb(Un)=DRb(Un-1) ˄ DRb(Un)=DRp(Un)] ≡ w

The resulting DRS in (15) shows the semantic representation for the frag-
ment of discourse, which is incorporated to the discourse structure and a
superindex p is assigned to the most prominent variables.

(212) x y z
pen-drive (x)
person (y)
offer (1, x, y)
work-computer (y)
he (z)p
happy (z)
z=y
he (w)p
enjoy (w)
w=y

Interestingly, it is also possible that a variable attached to a narrow scope

quantifier may be the prominent DR at a specific point in a discourse, as
shown in the following minimal pair in (213), which is the continuation of
(165) in the preceding chapter.

eg:book eg:ip-u
(213) ix1 think ix3 book 1-offer-3 adequate person-3ip-u...
eg:cl-u eg:c
must person-3cent like hobby is/same traditional past
eg:ip-l
same/always. ix3ip-l ix1 1-offer-3 person-3ip-l ix3ip-l
ix3ip-l happy, enjoy.
214 Discourse structure and prominence

‘I think that I would offer this book to someonenon-spec...

It must be someone who likes traditional things. Definitely, I would
offer it to him. He would be very happy and he would enjoy it a lot.’

Again, first the semantic representation is built in the corresponding DRS.

The variables contained in both the antecedent and the consequent need to be
bound by the corresponding operator.

(214) x
book (x)
yz
offer (1, x, y)
person (y)
□
trad-things (z)
like (y, z)

The continuation of the fragment of discourse contains two pronominal

forms. Since the modal operator binds the variable it is a felicitous continu-
ation. The construction rule for pronouns is used. According to the construc-
tion rule for prominence, it is checked which variables verify the formula for
prominence. Then superindex p is assigned in the DRS for the narrow scope
variables that are prominent, as in (216).

(215) a. [DRb(Un)=DRb(Un-1) ˄ DRb(Un)=DRp(Un)] ≡ w

b. [DRb(Un)=DRb(Un-1) ˄ DRb(Un)=DRp(Un)] ≡ z

(216) x
book (x)
yz
offer (1, x, y)
person (y)
□
trad-things (z)
like (y, z)

he (w)p
yz happy (w)
w=y
person (y)
□ he (z)
p
trad-things (z) enjoy (z)
like (y, z) z=y
 Prominence 215

In LSC, thus, the establishment of lower spatial locations is represented with

variables attached to both narrow and wide scope quantifiers, as long as they
denote the most prominent DR. Both narrow and wide scope quantifiers
attached to variables can be linked to the discourse topic and hence represent
the most prominent DR. In such a context, a lower spatial location is estab-
lished in signing space in LSC, independently of the scope of the quantifier
attached to the variable. As for the semantic representation, the most promi-
nent variable is assigned a superindex p in the DRS. Correspondingly in LSC
a spatial location established on the lower frontal plane corresponds to the
establishment of superindex p in the DRS.
Hence, The discourse referent hypothesis presented in (206) above (and
also in Chapter 4) needs to be revised. Since variables attached to narrow
scope quantifiers may also establish a spatial location (p) when they denote
the most prominent DR, the following hypothesis is formulated here:

(217) The discourse referent hypothesis (final version)

(p) is the overt manifestation of DRs established on the lower frontal
plane when the corresponding variable, without regard to its scope,
denotes the most prominent DR.

The establishment of (p) marks that the DR denoted is the topic of that frag-
ment of discourse and also a noteworthy entity, since the discourse will be
centred on it, even if it is a variable attached to a narrow scope quantifier.
Importantly, when connected discourse is considered, it is observed that
there is not a one-to-one relation between the area in signing space where
spatial location (p) is established and the DR referred to. In some contexts,
spatial locations can be reversed by factors related to discourse prominence.
When the DR is the discourse topic, spatial locations associated with the
same DR can be established in different directions of spatial planes, or even
the same location is associated with more than one DR. In the first case the
location is established in different lateral directions and shifts between ipsi-
lateral and contralateral parts. In the second case, the most prominent DR is
established on a spatial location previously established for another DR. This
shows that (p) is not associated with a concrete direction but it is actually an
abstract point established in signing space. As seen in Chapters 3 and 6, in
LSC only the two features on the frontal plane, namely upper and lower, are
relevant for the grammar of the language, but this is not the case concerning
the lateral parts, namely ipsilateral and contralateral. In the following section
it is shown that the direction of index signs towards the horizontal plane is
not relevant, since it may be directed to the two opposed directions to refer
216 Discourse structure and prominence

to the same DR in different moments of the discourse without affecting the

propositional meaning. Hence, what is relevant is that the index sign estab-
lishes an abstract point in space (p) that is associated with an individual from
the discourse model, independently of the direction on the horizontal plane
as already explained in §3.4. The following section delves into this issue and
provides examples from the LSC small-scale corpus.

7.5. Underspecification

In the sign linguistics literature, spatial locations are traditionally viewed as

points in space. Only some authors talk about areas or regions established
in space through signs directed to it (Engberg-Pedersen 1993; Russell and
Manitoba 2008). Russell and Manitoba (2008) argue that locations of ASL
pronouns are regions established within the articulatory space, instead of
points. They compare these regions with another linguistic system of cate-
gorical regions such as spoken-language vowels. They conclude that the ASL
pronominal system is not analogue, infinite, non-categorical, non-linguistic,
since it shares many of the features that the spoken language vowel system
shows. Also, Baker and Cokely (1980) use clouds instead of points in space
to represent spatial locations.
In this section it is shown that weak referring terms are used when the DR
is very prominent. Weak pronouns are anaphoric, they are not adjacent to the
verb, they cannot be reduplicated and their duration is in between a strong
pronoun and a clitic (Bertone and Cardinaletti 2011; Cardinaletti and Starke
1999). From the careful analysis of our LSC corpus, it should be added to
this list of properties that weak pronouns also have a very weak articulation,
and the direction towards space can be shifted to the lateral parts on the hori-
zontal plane. This shows that the direction on the horizontal plane where (p)
is established is in fact irrelevant for the grammar of LSC.
It is important that the shifting in space I am referring to here is not
confused with what I call “temporal locus-shifting” described in the litera-
ture as “locus-shifting”. Emmorey (2002: 56), Janis (1992: 83) and Padden
(1988: 185) present some ASL examples of spatial location shifting expressed
through classifiers. In these examples, the subject of a classifier predicate
appears in position x and in the following sentence the location of the coref-
erential pronoun shifts to the position of the end point of the verb, namely y.
In such cases, the motion of the verb has a linear movement, from one side
  Underspecification 217

of signing space to the other. The location position is thus shifted from x to y.
Van Hoek (1992) also presents examples in which a person is moving from
one city to another and this is assigned two spatial locations, each associated
with the aspects of the life of the DR in each corresponding city. Because
classifier constructions imply spatiotemporal information in the construction
itself, I consider these examples of location shifting to be motivated by spati-
otemporal information reasons. If we sign a classifier predicate at a different
location in space than previously assigned, the connection with the DR will
be the same but it will be implied that a spatiotemporal setting change has
taken place. A spatial location shift involves time elapse (that is, the duration
time of the event), as well as association with different temporal moments,
as shown in van Hoek (1992). Although this shifting is very relevant for the
study of discourse structure, it is not the kind of shifting I am dealing with
here. Rather here I am focusing on cases where the shift does not contribute
any temporal change in meaning, as we will see below.
The main focus in this section is the relation between prominence of
the DR and underspecification of the referring term used. When studying
connected discourse, DRs can be associated with different directions on
the horizontal plane without further implying any temporal information.
The referring term used is a much-underspecified one. This shows that the
exact direction on the horizontal plane where the location is established is
not relevant for the grammar of LSC. Hence in connected discourse there is
not a strict one-to-one mapping in the spatial direction and the DR associ-
ated since in some contexts the establishment of (p) can be reversed on the
horizontal plane by factors related to discourse prominence and signers may
not exactly use the same area for the same referent.76 My claim is that the
prominence of DRs can override spatial location setting given the dynamic
nature of discourse. Here I revise The spatial point hypothesis, previously
introduced.

(218) The spatial point hypothesis (first version)

The identity condition in the DRS is encoded through coincidence in
direction of spatial establishment of (p).

The study of connected discourse shows that location is not one of the
SLs features used for the disambiguation of the identity condition (contra
Zwitserlood and van Gijn 2003, and in line with Quer 2009), as it is
shown below.
218 Discourse structure and prominence

7.5.1. Informativity

In §7.1.2 we have seen that as discourse progresses, the linguistic elements

used to refer to entities can vary. The more prominent an entity is at a specific
point in the conversation, the more attenuated the referring form will be.
Hence an important characteristic of anaphoric elements is that as discourse
unfolds, the referential expressions that function anaphorically become more
attenuated, less informative and less rigid as long as the DR remains acti-
vated (Ariel 1988, 1990). Informativity concerns the amount of information
that referring expressions have in terms of lexical content. The information
load of pronominal forms is indeed very small. In fact, there is a correlation
between degree of accessibility and information load: the more accessible an
entity is, the less informative the referring expression used will be.
If we consider SL pronouns, an index handshape directed to space provides
very little information about the entity that we are talking about. According
to the general assumption, pronouns do not include gender information, but
only location and number (Sandler and Lillo-Martin 2006; Zwitserlood and
van Gijn 2006). However, the information that location provides is not very
informative. As shown in this section, in some contexts there is not a one-
to-one mapping between the DR and the location establishment, and this
relation is sometimes modified by discourse prominence reasons. Hence in
LSC connected discourse, location is not a reliable cue for the identification
of the DR associated.
This chapter revises the hypothesis presented in §4.2.2 where it has
been claimed that in SLs the association between the pronominal form and
the DR is overt due to the establishment in signing space. While this has
been commonly accepted when considering isolated sentences, connected
discourse forces us to revise this hypothesis. When the discourse contains
two variables (i.e. two DRs), the horizontal plane is usually divided into
two parts, namely ipsilateral vs. contralateral parts. These opposed locations
in space distinguish the two DRs and are interpreted as contrastive topics
(see §3.3.1.2). They are in fact two clauses in which two DRs are introduced
in each clause and their respective verbs predicate two different, opposing
eventualities. Signing space is then restricted to the two variables and refer-
ences to one or the other will be represented by a direction in the two opposed
lateral parts on the horizontal plane, as shown in Figure 55.
  Underspecification 219

a. Contralateral location for DR (x) b. Ipsilateral location for DR (y)

Figure 55. Contrastive locations

Unless the DR is reintroduced by the nominal, the distinction ipsilateral-

contralateral is kept throughout the discourse as long as there is no shift in
the frame of reference. However, in some contexts this general tendency
is reversed and the prominence of the DR affects the one-to-one relation
between the direction of establishment of the spatial location and the DR.
This change in the usual tendency is motivated by the prominent status that
a DR has. This is shown in the example below taken from the Aesop’s fables
“The lion and the mouse”. Along the discourse the two main DRs (i.e. the
lion and the mouse) are localised in opposing sides of signing space. The
referential shift is indicated by modifications in the signer’s facial expres-
sion and body position. In the fragment glossed in (219) the two DRs are
not distinguished by two contrasted spatial areas, but rather they are set in
the same area, namely the contralateral part. As shown in (219) and in stills
Figure 56a it is the lion that is first localised on the contralateral part. In the
third utterance in (219) the signer directs a pronoun to the same direction to
denote “mouse” (Figure 56b). As a consequence of this lack of a one-to-one
relation between a spatial direction and the DR, the coreferential subindices
and area of signing space do not coincide in the glosses in (219).

(219) then/so mousek already 1-favor-3cl-j ix3cl-j lionj.

before (null)k 1-tell promise then/clear.
imagine lionj die almost ix3cl-k 3cl-k-save-3ip-j able.
‘And so the mousek had favoured the lionj.
Before (null)k had promised so.
Imagine! The lionj almost died, but hek was capable to save himj.’
220 Discourse structure and prominence

a. ix3cl lion b. ix3cl (mouse)

Figure 56. Same location for two DRs

The relation between the two DRs does not follow the convention of semantic
affinity as described by Engberg-Pedersen by which DRs have a close rela-
tionship and they are grouped in the same area (as already explained in
§3.3.1.2). Rather, it is a clear case of contrastive use in which the signer is
comparing and contrasting the performance of the two animals in the story.
However since one of the DRs (i.e. the mouse) is more prominent than the
other, the corresponding direction of the spatial location associated is shifted.
Thus the two DRs are localised on the same side without being ambiguous.
The pronoun direction towards the contralateral part does not affect the prop-
ositional meaning of the sentence. On the one hand, the high prominence of
the DR “mouse” helps us understand that an overt pronoun directed to an
area already established for a referent can be re-established by an accessible
one. On the other hand, the semantics of the verb save also triggers a disam-
biguation of the pronoun.77
The disposition of spatial locations shown in (219) is not the most canon-
ical one, but it happens to appear especially in spontaneous signing. The
most canonical situation would be to direct the pronoun ix3 for the second
DR to the area opposed (i.e. ipsilateral) in order to contrast it with the first
entity previously localised on the contralateral part, and also to agree the
verb keeping the contrastive disposition, as exemplified in (220).

(220) imagine, lionj die almost, ix3ip-k 3ip-k-save-3cl-j able.

‘Imagine! The lionj almost died, but hek was capable to save himj.’

Another canonical option would be to have a null argument with the verb still
agreed according to the contrastive location. In (221) the subject of the second
sentence is a null argument. This null argument is licensed by verbal agreement
  Underspecification 221

(Lillo-Martin 1986), because the plain verb of the second sentence (save) is
articulated as an agreeing verb articulated from one location (i.e. ipsilateral) to
the other (i.e. contralateral). Since the verb is expressed with the body of the
signer78 on the ipsilateral side which was previously associated with the DR
“mouse”, the verb will be interpreted as having this DR as the subject.

(221) imagine, lionj die almost, 3ip-k-save-3cl-j able.

‘Imagine! The lionj almost died, but (null)k was capable to save himj.’

Another possibility would be to have a null argument but to keep the agreeing
location disposition of the verb as it is in the original example in (219), in
which case the subject of the sentence would be the same as in the first
one but the meaning would be the opposite, as shown below. (222) is the
canonical utterance that we would get if we did not consider the prominence
of the DR. Note that the prominence of the DR and the accessibility of refer-
ring expressions used need to be considered to get the right interpretation of
a pronominal form directed to space.

(222) imagine, lionj die almost, 3cl-j-save-3ip-k able.

‘Imagine! The lionj almost died, but (null)j was capable to save himk.’

With the previous examples it has been shown that the presence or absence of
the pronoun and the agreeing features of the verb contribute to the meaning.
However what matters most in the original version (219) is that there is an
overt pronoun that indicates a contrastive topic and thus a different subject
from the first sentence. Although the spatial direction of the index sign is the
same, the overt pronoun triggers a different subject interpretation due to the
contrastive use. (p) established towards the same direction on the horizontal
plane may be associated with two DRs and more than one DR can be thus
localised on the same horizontal direction. This shows that space appears to
have few informative properties.

7.5.2. Rigidity

The criterion of rigidity consists in the ability to pick a unique DR, based on the
form. Anaphoric forms referring to prominent entities are rigid. Since the DR
is very prominent in the discourse, the anaphoric form does not have to be very
rigid. Again the accessibility scale is in correlation with the degree of rigidity:
the more prominent the entity is, the less rigid the anaphoric element will be.
222 Discourse structure and prominence

As said before, the general tendency is that signers pick one particular
lateral direction to localise a DR. However, in contexts where the entity is
very prominent, the direction in signing space used to localise can be shifted
without affecting the propositional meaning. In contexts of prominent enti-
ties, different locations in space can be used to denote the same DR. The
entity can be localised on one side and later on be picked up on the other
side.79 In such cases either there is only one variable in the model or in case
there is more than one variable present, one is more prominent than the
others. For the prominent entity, (p) may be established in different direc-
tions on the horizontal plane.
Discourse fragments with one DR have only one variable (x) and thus
only one location is used. Because the DR is accessible enough and there is
no competition among other variables, the location (p) is usually localised
in one area of the signing space. But in informal settings there is not always
a one-to-one mapping and (p) can be localised in different areas, sometimes
even switching between ipsilateral/contralateral, which affects neither the
propositional meaning nor the interpretation of the utterance, as also seen in
the previous section. The fragment cannot be qualified as incoherent because
space is used consistently between an index directed to a spatial location (p)
which is associated with the most prominent DR. The prominence of the DR
makes the localisations towards opposed parts on the horizontal plane possible.
This is shown in Figure 57, where the signer established (p) denoting DR
x towards the two lateral parts of signing space. In the first two mentions,
he directs the index sign towards the contralateral part (Figure 57a and
Figure 57b). But later on in the discourse, he directs the index sign referring
to x towards the ipsilateral part (Figure 57c). Importantly, in this fragment of
discourse DR x is the most prominent entity in the fragment, since it is the
thing the fragment is about. He is not contrasting x to any other entity.

a. DR(x) - Contralateral b. DR(x) - Contralateral c. DR(x) - Ipsilateral

Figure 57. Two lateral directions for the same DR
  Underspecification 223

Another example is shown in Figure 58 where the signer establishes the

same DR in two lateral directions within the discourse fragment. In the first
two mentions, agreement verbs are directed towards the contralateral part
(Figure 58a and Figure 58b). But in the following sequence of utterances,
the same DR is established on the ipsilateral part and referred back to with
an agreement verb (Figure 58c). However, it is important to note that in this
fragment two DRs are present. One is localised in space, and the other is
referred to by a role shift construction.

a. 1-give-3k b. 3k-explain-1 c. 3k-explain-1

Figure 58. Two lateral directions for the same DR II

The examples in the two preceding figures show that (p) is a very abstract
point which can be established in different directions on the horizontal plane,
but importantly it is categorically interpreted as being associated with the
most prominent DR.80 Hence, connected discourse is the domain where it
is more evident that (p) is indeed an abstract point in space, no matter the
direction on the horizontal plane where it is established. The definition of the
nature of (p) is described as follows:

(223) The spatial point hypothesis (revised)

(p) is an abstract point in space no matter the direction on the horizon-
tal plane where it is established.

It is important to note that the LSC examples shown in this section are not
very frequent in the small-scale corpus. This is mainly due to two reasons.
First, most of the signers participating in the corpus are LSC teachers and
they are thus aware of what has been said of the grammar of SL. Hence,
whenever they are in front of a camera and ready to sign the more common
situation is that they use a quite standard register of LSC. Second, the data
comprising the small-scale corpus may not be as naturalistic as the contexts
224 Discourse structure and prominence

where these examples may usually appear. The more relaxed and the more
natural the context is, the more chances that spatial locations may be shifted
on the horizontal plane, even with LSC teachers. The principles stated so far
have paved the area of the use of spatial locations in connected discourse. A
follow-up study based on a large LSC corpus should help confirm the gener-
alisations and the analysis provided here on the basis of naturalistic data.

7.6. Summary

This chapter has presented two fine-grained versions of hypotheses

concerning (p), which have enhanced the analysis of LSC spatial loca-
tions by offering a discourse structure perspective. First, we have seen that
lower spatial locations correspond to discourse prominence, defined as vari-
ables with backward looking properties as well as forward looking proper-
ties, independently of the scope of the quantifier attached to the variable.
Also, the nature of spatial locations has been presented by considering data
from connected discourse. (p) is an abstract point in space which does not
correspond to an exact point nor it is related to a specific direction in spatial
planes. In connected discourse, locations associated with the most prominent
DR may be shifted in space, showing that the exact direction on the hori-
zontal plane is irrelevant for the nature of (p). What is relevant is that the
spatial location (p) is associated with a DR from the model independently of
the direction of the referring term in signing space.
In the review of literature in Chapter 5 concerning the definiteness infor-
mation spatial locations encode (see §5.1.4), I have revised the hypotheses of
some works concerning the lack of definiteness. Thus I agree with Engberg-
Pedersen (1993), Winston (1995) and Rinfret (2009) that spatial locations
denote discourse prominence, rather than definiteness. However, the route
I have taken in the overall book is more indirect, but also more interesting,
because specificity marking has also been included in the analysis of LSC
locations.
Chapter 8
Final remarks

This bookhas offered a description of the uses and functions of spatial planes to
express discourse categories, which contributes to the characterisation of the
abstract import of signing space. It has shown that non-descriptive locations
are categorically defined as being expressed in the different areas within the
three spatial planes projected with respect to the body of the signer. Spatially
associated signs contribute to the establishment of a grammatical morpheme
that consists in an abstract point in space, which is categorically interpreted
within the linguistic system. In LSC, the spatial location (p) can be abstractly
established in different parts of the three spatial planes. The different features
within each plane are specialised in their meaning and, more importantly,
they belong to the grammar of LSC. As stated in Chapter 1, three goals led
the direction of this research, which claimed the following:

G1. To show that spatial locations are integrated into the grammar of
LSC and, even more, they denote specificity.
This book has shown that spatial locations undertake a semantic func-
tion: that of being the overt manifestation of discourse referents (DRs). This
assumption has been first presented as (224).

(224) The discourse referent hypothesis (first version)

(p) is the overt manifestation of the DR the referring term denotes.

Under the particular formalisation of Discourse Representation Theory, the

discourse referent established in space corresponds to a variable established
in the main universe of discourse. The establishment of (p) is the marking
for DRs that have wide scope. That is, those variables which are not bound.
In contrast, variables attached to narrow scope quantifiers lack a spatial loca-
tion establishment. Therefore (224) has been slightly revised and extended
to (225).

(225) The discourse referent hypothesis (second version)

(p) is the overt manifestation of DRs attached to a quantifier that has
wide scope.
226 Final remarks

Interestingly, the notion of specificity plays an important role in the estab-

lishment of (p) and it has been proposed that (p) encodes specificity. The
frontal plane is grammatically relevant for specificity marking: lower spatial
locations correlate with specific discourse referents, whereas upper spatial
locations correlate with non-specific ones. In LSC two kinds of localisa-
tion on the frontal plane are found, namely a strong and a weak localisation.
Strong localisation is instantiated by the feature (p), while weak localisa-
tion is instantiated by the marked feature (p)[up]. This is framed under the
following hypothesis in (226), which is a fine-grained version of (225),
related to the marking of specificity.

(226) The discourse referent hypothesis: specificity version

(p) is the overt manifestation of wide scope denoting specificity.

But also, weak localisation establishes an upper location on the frontal

plane, for those variables being under the scope of a quantifier denoting
non-specificity.

(227) The discourse referent hypothesis: non-specificity version

(p)[up] is the overt manifestation of narrow scope denoting non-spec-
ificity.

Moreover, the incorporation of discourse structure into the analysis has

offered a new and interesting perspective to the analysis of spatial locations.
I have addressed the fact that lower spatial locations correspond to discourse
prominence, defined as the intersection between backward looking proper-
ties and forward looking properties. It has been shown that, independently
of the scope of the quantifier attached to the variable, narrow scope variables
that are linked to the most prominent DR at a specific point in a discourse
behave like wide scope ones and establish a lower spatial location. The wide-
scope hypothesis is revised and transformed into (228).

(228) The discourse referent hypothesis (final version)

(p) is the overt manifestation of DRs established on the lower frontal
plane when the corresponding variable, without regard to its scope,
denotes the most prominent DR.

Although these hypotheses seem at first sight to be contradictory, along the

book it is shown that they are in fact complementary once the proper ingre-
dients are incorporated to the analysis.
 Final remarks 227

G2. To analyse how spatial locations are set, given the dynamic nature
of discourse.
We have also seen that (p) is an abstract point in space which does not
correspond to an exact point nor it is related to a specific direction in spatial
planes. Coreferential DRs are commonly associated with the same direction
on the horizontal plane. The coreferential link is done through the coinci-
dence in the direction of establishment of (p). At a first stage of the book, this
has been formulated as hypothesis (229) below.

(229) The spatial point hypothesis (first version)

The identity condition in the DRS is encoded through coincidence in
direction of spatial establishment of (p).

However, in connected discourse, locations associated with the most promi-

nent DR can be shifted in space, showing that the exact direction on the
horizontal plane is irrelevant for the nature of (p). What it is relevant is that
the spatial location (p) is associated with a DR from the model independently
of the direction in signing space and this is categorically interpreted within
the linguistic system. The spatial point hypothesis as defined in (229) is thus
developed to (230).

(230) The spatial point hypothesis (revised)

(p) is an abstract point in space no matter the direction on the horizon-
tal plane where it is established.

G3. To apply a dynamic semantic theory, such as classical Discourse

Representation Theory, to a visual-spatial language like LSC.
This book has offered an innovative approach that classical DRT lacks.
It has incorporated the properties of a visual-spatial language that also
contribute to the semantic representation. First, signing space has been
adequately incorporated into the semantic representation of LSC discourse.
On the one hand, the sentence level has been considered by claiming that the
identity condition in the DRS is resolved through coincidence in direction
of spatial establishment of (p). But on the other, the analysis of connected
discourse has shown that (p) is in fact an abstract point in space no matter
the direction on the horizontal plane where it is established, as stated in
G2. Second, deictic pronominal uses, typical of face-to-face communica-
tion, have been added to the DRT construction rules needed in contexts of
deictic elements. This has shown that in LSC references are anaphoric to
the discourse model, although in some contexts DRs are introduced to the
228 Final remarks

common ground without an explicit linguistic antecedent. Finally, the present

approach has incorporated discourse structure to the semantic representation
by analysing how prominence is integrated. A theory of discourse structure
with special focus on prominence has been integrated into the representa-
tional semantic level. The prominence of the variables at different fragments
of discourse is determined and marked in the semantic representation, by
considering the backward and the forward properties of the DR.
This book has contributed to enhance our understanding of signing space
in LSC and, more specifically, of the grammatical role that non-descriptive
locations play. The concrete major contributions of this book are basically
two-fold. On the one hand, a semantic formalisation such as DRT has been
used within which the hypotheses have been framed and shown. The theo-
retical background of DRT has provided a detailed framework that supplies
the tools for an implementation to concretely define the grammatical role
of spatial locations. Importantly, the features that characterise visual-spatial
languages, namely the use of signing space and deictic contexts, have also
been taken into account and the corresponding implementations have been
incorporated, as described in G3 above. On the other hand, the analysis of
spatial locations and the referential properties in LSC have been investigated
on the basis of a small-scale LSC corpus, containing semi-spontaneous,
videos recorded for other purposes, and elicited data. In the end, it has been
connected discourse in the language under investigation that has provided
the most important clues for analysis of these grammatical domains.
At the end of such a large project many questions regarding some aspects
touched upon here remain unanswered, mainly due to the fact that a broad
topic has been addressed. As for localisation, it would be interesting to
analyse whether a nominal followed or preceded by an index sign is best
analysed as a whole clause or rather as a phrase. Kegl (1986) makes a clear
distinction between index signs functioning referentially, and those having a
predicative function. The first occur in indexing phrases and the second, in
indexing clauses. However, it should be seen whether topicalised constitu-
ents occur in these structures, which would point towards the phrase anal-
ysis of indexing. Concerning existential clauses, it should also be explored
whether the introduction of DRs into the model can be considered to be a
case of existential sentences. A sentence like (231) indicates that a man is
present in the discourse model.

(231) ix3a man

‘There is a/this man.’
 Final remarks 229

Interestingly, it could be argued that this syntactic construction conflates

two semantic structures. On the one hand, an existence statement where
it is predicated that a DR exists in the current model. On the other hand,
localisation predicts that an entity is at a deictically determined location a,
which is found in signing space. This conflation which has been analysed for
Italian existential sentences (Zamparelli 1995) might be also present in LSC
indexing sentences.
Another issue that could be the topic of future research is the speciali-
sation of morphophonological features denoting specificity presented in
Chapter 6. In LSC there seems to be a specialisation in the property of speci-
ficity the features mark: while eye gaze seems to denote identifiability, signs
localised in space seem to be responsible for wide scope and partitivity. As
shown in Chapter 6, identifiability as one of the properties that specificity
encompasses is mostly marked with eye gaze being directed to a spatial loca-
tion for specific reference, or with an upper darting eye gaze for non-specific
reference. In contrast, scope and partitivity mainly distinguish specific and
non-specific reference with the direction of sign towards the lower and upper
parts of the frontal plane, respectively. In this respect, contexts of interme-
diate scope where a variable receives wide scope with respect to one quan-
tifier and narrow scope with respect to another may also contribute to the
analysis of scope marking specificity and localisation on the frontal plane.
An additional research project would be to refine the kind of motivations
which lead to the prominence hierarchy presented in Chapter 7. Three possi-
bilities could be analysed: whether prominence is a syntactic issue where
the hierarchy goes from subject > object > other, a semantic issue (animate
> inanimate), or a pragmatic issue (topic > comment). But more interest-
ingly, the interaction among the three levels of analysis should be further
considered.
Finally, an additional note concerning the methodology needs to be
made. The main aim in this book has been to develop a qualitative study by
observing and analysing the tendencies that naturalistic, semi-spontaneous
and elicited data from our small-scale LSC corpus can provide. The results
of the small-scale corpus data should be taken as a strong tendency of real
data associated with a theoretical model. A follow-up study based on a yet
non-existent large LSC corpus should help confirm the generalisations and
the analysis provided here on the basis of naturalistic data.
In sum, the results presented in this book have focused on a broad topic,
such as the semantic and pragmatic properties of signing space in LSC. Some
particular sub-topics will have to wait for further determination, but at least
the way has been paved for more fine-grained future proposals.
Appendix

Noun phrase construction rules

CR.PN: Upon encountering a proper noun,

1. trigger the syntactic configuration [s NPα [VP]] or [s VP [NPα]], and
2. introduce a novel discourse referent α into the main DRS, and
3. introduce the predicate condition β(α)

CR.N: Upon encountering a common noun co-occurring with a determiner,

1. trigger the syntactic configuration [s NPα [VP]] or [s VP [NPα]], and
2. introduce a novel discourse referent α into the main DRS, and
3. introduce the predicate condition β(α)

CR.PRON: Upon encountering a pronominal form,

1. trigger the syntactic configuration [s NPα [VP]] or [s VP [NPα]], and
2. introduce a novel discourse referent α into the main DRS, and
3. check which variable in the main DRS shares the features α has, and
4. if no suitable variable is found, go to CR.PRON2; if the suitable variable
is found introduce an identity equation α = γ
5. go to CR.PROM

CR.PRON2: Upon encountering a pronominal form,

1. go to the main DRS and take the default variable δ,
2. introduce an identity equation α = δ
3. go to CR.PROM

CR.PROM: Upon encountering an identity equation α = γ,

1. check the variable that verifies the prominence rule:
DRb(Un)=DRb(Un-1) ˄ DRb(Un)=DRp(Un)
2. assign the superindex p to the variable,
3. check the conditions in the DRS equated to the variable β = α,
4. assign the superindex p to the suitable variable
Notes
1. This name is used since the 80’s. Previous to that, a series of different names,
such as mimics, hands, and signs were used (Frigola 2010).
2. Webvisual is the LSC TV channel on internet in which two daily signed
news and weekly documentaries, news, shows, interviews, etc. are presented
(www.webvisual.tv).
3. Nowadays, numerous deaf clubs, three universities (Universitat Pompeu Fabra,
Universitat de Barcelona and Universitat de Girona) and three vocational
trainings offer LSC lessons.
4. The ECHO webpage, with recordings and many useful annotation guidelines,
can be found at http://www.let.ru.nl/sign-lang/echo/
5. http://www.ru.nl/corpusngtuk/
6. http://www.bslcorpusproject.org/
7. http://www.auslan.org.au/about/corpus/
8. http://www.sign-lang.uni-hamburg.de/dgs-korpus/index.php/welcome.html
9. One year after the manuscript for this book was finished, an LSC pilot corpus
project was set up in November 2012. For more information, see Barberà,
Quer and Frigola, in press, and the website for the project http://blogs.iec.cat/
lsc/corpus/
10. http://www.lat-mpi.eu/tools/elan/
11. http://www.bu.edu/asllrp/signstream/index.html
12. http://www.anvil-software.de/
13. http://www.irit.fr/LS-COLIN
14. http://cosmion.net/jeroen/software/kappa/
15. http://www.sign-lang.uni-hamburg.de/ilex/
16. For the interest of privacy, each signer is identified with a number.
17. The description of space given in the overall book focuses only on LSC, which
shares many features with other Western urban sign languages. However, non-
urban sign languages are being more and more studied in different corners of
the world. As described so far, these non-urban SLs show differences in the
grammar and use of space. For instance, Kata Kolok, a village sign language
in North Bali, and Adamorobe Sign Language, in Ghana, use a much larger
signing space that goes beyond the limits described here (Marsaja 2008: 159;
Nyst 2007; de Vos 2012).
18. In this LSC minimal pair example the nonmanual component is left aside.
19. This does not imply, of course, that syntax does not use space for the construction
of sentences. As argued at the beginning, morphosyntax needs space for the
modulation of signs to express number, person and arguments. Still the area
where space has an important contribution is that of connected discourse.
232 Notes

20. This book uses the terms “index sign” and “pointing sign” to mean the
handshape generally formed by an index finger directed to a spatial location.
Although the handshape configuration may vary due to cliticization (see
Fenlon et al. 2013), we include all index signs under this label irrespective
of the handshape configuration, which possibly assimilates the handshape of
neighbouring signs.
21. Classifiers constructions are complex predicates that express movement,
location and description of a referent, as well as the manipulation of it
(Suppalla, 1986). Handshape classifiers stand for the referent they denote
and have been analysed as agreement markers (see Glück and Pfau, 1998;
Zwitserlood, 2003).
22. Entity classifiers are elements in which the hand configuration indicates
a particular semantic class, and the movement or the location of the hand
indicates the motion or location of the entity denoted (see Zwitserlood 2003).
23. Researchers interested in spatial descriptions in sign languages have attributed
different terminology to the same phenomenon. The two types of descriptive
spatial representation have been labelled as follows: fixed/shifted referential
framework (Bellugi and Klima 1990; Morgan 1999); diagrammatic/viewer
format (Emmorey and Falgier 1999; Emmorey 2002b; Emmorey and Tversky
2002); depictive/surrogate space (Liddell 2003); character/observer viewpoint
(Perniss 2007a). The overlap between the criteria used to distinguish each
representation is strong enough to consider them different labels of the same
phenomenon.
24. Agreeing verbs are called “indicating verbs” in Liddell’s terminology. Because
this terminological use is very theoretically loaded and in order to keep the
coherence in the overall book, I have chosen to use my terminology even if
sometimes it does not coincide with that of the authors the section is about.
25. Role shift is the process whereby a shifting reference is used to reproduce
someone else or his own’s utterances or thoughts, which have occurred at a
moment different from the real utterance context (Lillo-Martin 1995).
26. The only attempt to formalise spatial locations has been Lillo-Martin and
Klima (1990), which is in fact the line of research this book follows. The
book in its globality, and especially Chapters 4 and 6, are crucial for this
formalisation.
27. Nominals can be preceded, followed or both by index signs. §2.5 showed that
different analyses have been provided. The syntactic difference among these
possible configurations in LSC is outside the scope of this book.
28. Mechanisms adapted and extended from Quer et al. (2005) and Barberà (2007).
29. The sign person-3 used pronominally does not have to be confused with the
LSC auxiliary predicate used as an agreement marker analysed by Frigola and
Quer (2006). The auxiliary sign has the same manual handshape and moves
from subject to object.
 Notes 233

30. For a prosodic analysis of body leans, see Wilbur and Patschke (1998) and
Kooij, Crasborn, and Ros (2006) who claim that body leans mark contrast in
ASL and NGT, respectively.
31. The organisation of the frame of reference has been described by Engberg-
Pedersen (1993: 71) for Danish Sign Language (DSL). According to this
author, when more than one referent is present in a fragment of discourse,
the frame of reference is organised according to some conventions guided by
semantic factors. They are divided among four conventions, which are the
semantic affinity, the canonical convention, the convention of comparison
and the iconic convention. Engberg-Pedersen does not analyze them as strict
norms, but rather as a description of the signer’s performance. Instead, I have
chosen to describe the specialised use that each part has within a given plane,
by sorting out the grammatical features contained on each one of those planes.
32. See §7.2 for further treatment of personal pronouns in LSC, where it is also
argued that second person reference is identified not only considering the
spatial location but also with nonmanual alignment.
33. Although Liddell and Johnson also record this plane, called “vertical” plane
in their terminology (Liddell and Johnson 1989: 221), no further features are
distinguished in their model.
34. As attested in Davidson and Gagne (2014), ASL differs from LSC in this
regard, since all quantifiers may be localised in the upper part of signing space.
This fact leads the authors to propose a domain widening account for this kind
of uses. A comparative approach analysing LSC and ASL will have to wait for
future research.
35. Liddell and Johnson also distinguish two more features, which Sandler does
not mention, such as [medial], a position roughly an elbow’s length from
the body location, and [extended], a full arm length from the body location
(Liddell and Johnson 1989: 230). Since they are not attested in LSC, they are
left aside from the present description.
36. Three possible solutions have been offered to this problematic aspect. This
book only deals with dynamic semantics (Heim 1982; Kamp 1981; Kamp
and Reyle 1993; Roberts 2005). The reader is referred to Evans (1980),
Elbourne (2005) and Heim (1990) for an e-type approach of the problem,
where the anaphoric pronoun is analyzed as a concealed description. For a
dynamic predicate logic approach, see Groenendijk and Stokhof (1991).
37. Pragmatic considerations are especially important when an underspecified
referring term can be identified with more than one suitable discourse referent,
as in “John hit Paul. He was mad at him.” In cases of ambiguity, the integration
between syntactic and pragmatic motivations contributes together in order to
get the correct disambiguation of pronouns (Mayol 2009).
38. See Heim (1982) for some problems with using the term “discourse referent”
precisely because it does not coincide with reference.
234 Notes

39. See Vallduví (1992) for a revision of literature about the informational
articulation of the sentence.
40. See also Neidle et al. (2000) and recent work on Kuhn (2014) for a
morphosyntactic feature-based analysis of spatial locations.
41. Person features are not included here, but treated in Chapter 7. The reader is
referred to §7.2 for a three person distinction analysis of LSC pronouns.
42. In fact, Heim redefines the notion of discourse referent’s lifespan (i.e. scope)
introduced by Karttunen. She claims that it is better explained once we think in
terms of file-cards: the lifespan lasts during the period that the card is introduced
and maintained active in the file. As long as the file-card is eliminated from the
file, the corresponding DR’s lifespan is over (Heim 1982: 283).
43. However, see §6.3.3.2 for a description of modal subordination cases and a
further refinement of this claim.
44. The reader is referred to Rosselló (2008) for an interesting syntactic account
distinguishing two different structures for specificational and predicationals
sentences from the start of the derivation. According to her account, while
predicational sentences have an external argument, the precopular DP of
specificational sentences is directly merged by means of an operator.
45. Rhetorical questions are structures formed by a question-answer sequence,
which have a focusing function similar to that of pseudoclefts or Wh-clefts
in other languages. The reader is referred to Wilbur (1994, 1995, 1996, 1997)
and Caponigro and Davidson (2011) for a detailed account of such structure.
46. Interestingly, donkey sentences have been tested in ASL and LSF
(Schlenker 2011a, 2011b). According to these works, nominals in donkey
sentences and in quantified expressions in both ASL and LSF can be localised
in signing space, which leads Schlenker to the conclusion that, in line with
Nouwen (2003), all quantifiers (not only indefinites) can introduce DRs and
can bind variables they do not c-command.
47. However, there seems to be an interpretation difference when the numeral is
non-overt, as in the example “student ix1 1-ask-3[mult].” In this case, a
resumptive pronoun in the following sentence is interpreted as plural only.
The different discourse behaviour plural entities have in discourse in terms of
discourse transparency has been largely noted in the literature (Farkas and de
Swart 2003). Nevertheless, this issue is outside the scope of this book.
48. Another possible analysis, which is not followed here, is to consider that plural
NPs denote properties of individuals instead of kinds, as in McNally (2004).
49. Backwards anaphora, i.e. when the underspecified element precedes the full
NP, which appears afterwards, is also applicable here.
50. Originally, Prince’s terminology uses the term ‘hearer’. However, I have opted
to adapt it into a more generic term such as ‘addressee’ in order to apply it also
to a signed discourse.
 Notes 235

51. The reader is referred to Prince (1981) for a specific taxonomy related to these
properties of referents that distinguish new (equivalent to discourse new/
addressee new), evoked (discourse old/addressee old) and inferrable entities.
52. For a detailed overview with strong and weak points of each line of thought,
see Lazaridou-Chatzigoga (2009).
53. This is in fact what has been said in discourse studies that consider that the
inventories of the referring expressions do not have to be paired with a co-
textually occurring antecedent expression in order to receive an interpretation,
unlike in the traditional account of anaphora. An antecedent, according to Ariel
(1988, 1990), Cornish (1999) and Gundel, Hedberg and Zacharski (1993), is
a mental representation bearing a given prominence, or accessibility level,
and is not a segment of co-text. In Ariel’s theory there is no such antecedent
in the linguistic context, as antecedents are mental representations denoting
prominent entities. The accessibility issue is included in Chapter 7.
54. Engberg-Pedersen (1993: 128) explicitly mentions that in this statement she
leaves aside role shift construction denoting animate referents. As well as in
DSL, role shift in LSC is also a way of assigning discourse prominence to the
entity without regard to whether it is spatially established or not (Barberà 2009;
Barberà and Quer, in press). However, role shift constructions are outside the
scope of the analysis in this book.
55. The construction rule about prominence is treated in Chapter 7.
56. As mentioned in §1.2, Webvisual is the LSC TV channel on the Internet (www.
webvisual.tv).
57. For a covariation analysis of scope, the reader is referred to Farkas (2001,
2002), where it is shown that reduplication of the indefinite article in Hungarian
marks dependency. Dependent indefinites are considered by Farkas to arise
when the DP co-varies with a variable.
58. These are cases of modal subordination, which is treated in §6.3.3.2.
59. Here I am dealing only with semantic partitivity to refer to the interpretation
of partitive NPs. This contrasts with the use of the term “partitivity” in the
syntactic literature where it refers to non-specificity (Belletti 1988).
60. In a similar view, Diesing (1992) argues that specific indefinites
crosslinguistically are always presuppositional since they presuppose the
existence of the set denoted by the restrictor.
61. Cases of intermediate scope, such as “Each teacher overheard the rumour that
a student of mine had been called before the dean.”, where the indefinite NP
“a student of mine” has narrow scope w.r.t. to the DR teacher, and wide scope
w.r.t. the that-clause (Kratzer 1998), are outside the scope of this book and
merit further investigation.
62. I have opted to analyse this segment and followings as two single sentences
because of the prosodic marking. But whether this should be treated as a
236 Notes

relative clause is still a matter of debate which is outside the scope of this
book.
63. Interestingly, in this minimal pair a resumptive pronoun is present in the
specific version (19), whereas there is a null argument in the non-specific
sentence (21). Relating the use of the referring terms with specificity marking
is outside the scope of this book.
64. As noted in the Annotation conventions (page xiii), signs directed towards the
upper frontal plane of signing space are represented with u.
65. As presented in §4.3.5, reference to kinds in LSC does not have a corresponding
spatial location established. This is evidence towards the hypothesis analysing
weak definites as reference to kinds (Aguilar-Guevara and Zwarts 2010).
66. Since in this section only the features [low] and [up] are relevant, the glosses
are only marked with the corresponding indices (i.e. l, u) and ipsilateral and
contralateral are left aside for the interest of simplicity.
67. Recall that here I am only dealing with non-descriptive localisation denoting
non-specificity. In contexts of descriptive localisations, entity classifiers can
be localised in an upper part to denote a descriptively higher location, as
argued in §2.3.1. Descriptive locations, though, are totally unrelated to non-
specificity marking.
68. In LSC, the sign glossed as mateix may be analysed as a determiner (Mosella
2012). It is relevant to note for the present account that this sign can never co-
occur with a nominal established on the upper frontal plane.
69. Bound variable, like “Every student put a screen in front of him.” uses are
another kind of anaphoric relation (Partee 1970). In such contexts, pronouns
do not refer to individuals, as previously seen in quantified expressions in
§4.3.3. In the example below there is not a referential expression, because the
antecedent is not a noun that denotes an individual and hence him is treated
as a bound variable. Bound variables uses are outside the scope of this book.
70. The Givenness Hierarchy by Gundel and her colleagues is the unique proposal,
to the best of my knowledge, supported by an empirical study of the distribution
of referring expressions in naturally occurring discourse in six languages,
namely English, Japanese, Mandarin Chinese, Russian, Spanish and ASL. See
Swabey (2002) for a form-status correlation for referring expressions in ASL
narratives.
71. See Kibrik and Prozorova (2007) for a general analysis of referential choice
in Russian Sign Language (RSL). See also Barberà (2009), for some work
on referential choice in LSC, focused on pronouns and entity classifiers, and
Barberà and Quer (in press), for work on referential choice in LSC, focused on
semantic classifiers and role shift.
72. In fact, Prince states that in her corpus study 86% of the DRs introduced by
indefinite this are referred back again within a few clauses. Ionin (2006: 184)
reports a corresponding statistic for a-indefinites and this-indefinites in order to
 Notes 237

compare both determiners more closely. In a brief, informal search she reports
that 89% of this/these-indefinites instances denoted a DR that was referred to
again a few clauses either explicitly or implicitly. For the a/an-indefinite, 71%
were subject to follow-up mention. Even if this difference is not a great one, it
still suggests that a distinction holds between the two determiners.
73. Role shift contexts in which first and second person pronouns are transferred
are not included in the present analysis.
74. However, assimilation processes can affect this default handshape depending
on preceding and following signs, as shown in corpus work by Fenlon et al.
(2013).
75. As seen in §3.5, body-anchored spatial locations are also available in the
morphophonological array.
76. Concerning the kind of localisation mechanisms used to refer to prominent
DRs, Rinfret (2009) claims that in LSQ there is a difference in use. When a DR
is localised with manual mechanisms, it is considered to be less prominent,
than when it is localised with nonmanual. Nonmanual localisation with body
lean, eye gaze indicates that the DR is very active in the consciousness of
the addressee. Low prominence in LSQ is marked with index signs directed
to space, and the spatial modification of signs (Rinfret 2009: 252). No such
distinction has been found so far in our LSC small-scale corpus.
77. In Figure 56b the signer directs a body lean to the contralateral part. The
simultaneous use of manual forms (the index pronominal sign directed
towards the contralateral part) and nonmanual forms (body lean towards the
contralateral side and marked facial expression) is an interesting avenue for
future research that is not included here.
78. Kegl (1986) does not consider the verb movement to spatial locations as the
agreement that licenses null arguments. Rather, the position of the signer’s
body with respect to the locations, which she considers a clitic, licenses empty
arguments. Both Lillo-Martin (1986) and Kegl (1986) analyses are compatible
with the present argumentation.
79. Engberg-Pedersen (1993: 100) also presents an example in Danish Sign
Language where the same DR is localised in two different spatial locations.
The signer introduces the chairman of the National Association of the Deaf.
She refers to him twice. The first time it is introduced on the ipsilateral part.
The second time she refers to him it is much later but she does not use the same
spatial location and the determiner made with the index finger points towards
the contralateral part.
80. This is coherent with studies focused on role shift in which it has been claimed
that in role shift the most important feature for the referential shift is facial
expression and eye gaze in particular rather than body lean or manual signs
directed to signing space (see Costello, Fernández, and Landa 2008; Herrmann
and Steinbach 2012).
References
Aarons, Debra. 1994. Aspects of the syntax of American Sign Language. Boston,
MA: Boston University dissertation.
Abbott, Barbara. 1999. Support for a unique theory of definite descriptions. In
Tanya Matthews & Devon Strolovitch (eds.), Proceedings from Semantics and
Linguistic Theory IX, 1–15. Ithaca: Cornell University.
Aguilar-Guevara, Ana & Joost Zwarts. 2010. Weak definites and reference to kinds.
In David Lutz & Nan Li (eds.), Proceedings of Semantics and Linguistic
Theory XX, 179–196. Vancouver: British Colombia.
Ahlgren, Inger. 1991. Deictic Pronouns in Swedish and Swedish Sign Language.
In Susan Fischer & Patricia Siple (eds.), Theoretical Issues in Sign Language
Research, Vol. 1: Linguistics, 167–174. Chicago, IL: University Chicago Press.
Ahlgren, Inger & Brita Bergman. 1990. Preliminaries on Narrative Discourse in
Swedish Sign Language. In Siegmund Prillwitz & Tomas Vollhaber (eds.),
Current Trends in European Sign Language Research: Proceedings of the 3rd
European Congress on Sign Language Research, 257–263. Hamburg: Signum
Verlag.
Alba, Celia. 2010. Les interrogatives-Qu en llengua de signes catalana (LSC). Bases
per a una anàlisi. Barcelona: University of Barcelona MA thesis.
Alibašić Ciciliani, Tamara & Ronnie B. Wilbur. 2006. Pronominal system in Croatian
Sign Language. Sign Language & Linguistics 9. 95–132.
AnCora-Co. 2008. Co-reference Guidelines for Catalan and Spanish. Version
2.1. Centre de llenguatge i computació de la Universitat de Barcelona. Ms.,
Universitat de Barcelona.
Anderson, Stephen & Ed Keenan. 1985. Deixis. In Timothy Shopen (ed.), Language
Typology and Syntactic Description, Vol 3: Grammatical Categories and the
Lexicon, 259–308. Cambridge: Cambridge University Press.
Ariel, Mira. 1988. Referring and accessibility. Journal of Linguistics 24(1). 65–87.
Ariel, Mira. 1990. Accessing Noun-Phrase Antecedents. London: Routledge.
Ariel, Mira. 1996. Referring expressions and the +/- co-reference distinction.
In Thorstein Fretheim & Jeannette Gundel (eds.), Reference and Referent
accessibility, 13–35. Amsterdam: John Benjamins.
Arık, Engin. 2009. Spatial language: Insights from signed and spoken languages.
West Lafayette: Purdue University dissertation.
Aronoff, Mark, Irit Meir & Wendy Sandler. 2000. Universal and particular aspects of
sign language morphology. In Kleanthes K. Grohmann & Caro Struijke (eds.),
Proceedings of the 1999 Maryland Mayfest on Morphology, 1–33. University
of Maryland, College Park: Department of Linguistics.
 References 239

Aronoff, Mark, Irit Meir, Carol Padden & Wendy Sandler. 2004. Morphological
universals and the sign language type. In Geert Booij & Jaap van Marle (eds.),
Yearbook of Morphology 2004, 19–40. Dordrecht/Boston: Kluwer Academic
Publishers.
Aronoff, Mark, Irit Meir & Wendy Sandler. 2005. The paradox of sign language
morphology. Language 81(2). 301–344.
Aronoff, Mark, Irit Meir, Carol Padden & Wendy Sandler. 2008. The Roots of
Linguistic Organization in a New Language. Interaction Studies 9. 133–153.
Asher, Nicholas. 1993. Reference to abstract objects in discourse. Dordrecht:
Kluwer Academic Publishers.
Asher, Nicholas & Alex Lascarides. 2003. Logics of Conversation. Cambridge:
Cambridge University Press.
Bahan, Benjamin. 1996. Nonmanual realization of agreement in American Sign
Language. Boston: Boston University dissertation.
Bahan, Benjamin, Judy Kegl, Dawn MacLaughlin & Carol Neidle. 1995. Convergent
Evidence for the Structure of Determiner Phrases in American Sign Language.
In Leslie Gabriele, Debra Hardison, & Robert Westmoreland (eds.), FLSM VI.
Proceedings of the Sixth Annual Meeting of the Formal Linguistics Society of
Mid-Americavol. 2, 1–12. Bloomington: Indiana University Linguistics Club.
Baker, Charlotte. 1977. Regulators and turn-taking in American Sign Language
discourse. In Lynn Friedman (ed.), On the Other Hand: New Perspectives on
American Sign Language, 215–236. New York, NY: Academic Press.
Baker, Charlotte & Dennis Cokely. 1980. American sign language a teacher’s
resource text on grammar and culture. Silver Spring, MD: T. J. Publishers.
Barberà, Gemma. 2007. La cohesió discursiva i l’espai en la llengua de signes
catalana. Barcelona: Universitat Pompeu Fabra MA thesis.
Barberà, Gemma. 2009. Choice of referring expressions in Catalan Sign Language
(LSC): pronouns and classifiers. Paper presented at the Sign Language
Colloquium, Radboud University, Nijmegen.
Barberà, Gemma. 2012a. When wide scope is not enough: scope and topicality of
discourse referents. In Maria Aloni, Floris Roelofsen, Galit Weidman Sassoon,
Katrin Schulz, Vadim Kimmelman & Matthijs Westera (eds.), Proceedings of
18th Amsterdam Colloquium 2011, LNCS 7218, 62–71. Berlin Heidelberg:
Springer-Verlag.
Barberà, Gemma. 2012b. A unified account of specificity in Catalan Sign Language
(LSC). In Rick Nouwen, Anna Chernilovskaya & Ana Aguilar-Guevara (eds.),
Proceedings of Sinn und Bedeutung 16 Volume 1, 43–55. MIT Working Papers
in Linguistics.
Barberà, Gemma. 2012c. Domain restriction in Catalan Sign Language. Poster
presented at the FEAST-Warsaw conference, University of Warsaw, 1 June.
240 References

Barberà, Gemma & Javier Fernández. 2009. Where do pointing signs point to Deixis
versus indexicality. Paper presented at the workshop From gesture to sign:
pointing in oral and signed languages, Université de Lille.
Barberà, Gemma & Guillem Massó. 2009. Keeping track in sign language: Co-
reference annotation using Centering Theory. Paper presented at the workshop
Sign Language corpora: linguistic issues workshop, University College
London.
Barberà, Gemma & Marta Mosella. 2014. Les llengües de signes: Gramàtica i
modalitat gestovisual. Un estudi de cas de la concordança verbal en la llengua
de signes catalana (LSC). Llengua i literatura 24. 111–152.
Barberà, Gemma & Josep Quer. In press. Nominal referential values of semantic
classifiers and role shift in signed narratives. In Annika Hübl & Markus
Steinbach (eds.), Linguistic Foundations of Narration in Spoken and Sign
Languages. Amsterdam: Benjamins.
Barberà, Gemma, Josep Quer & Santiago Frigola. In press. Primers passos cap a la
documentació de discurs signat. El projecte pilot de constitució del corpus de
la llengua de signes catalana. Treballs de Sociolingüística Catalana 25.
Barberà, Gemma & Eulàlia Ribera. 2010. Lexicografia de les llengües de signes i
l’estat de la lexicografia en la LSC. In Joan Martí & Josep Maria Mestres(eds.),
Les llengües de signes com a llengües minoritàries: perspectives lingüístiques,
socials i polítiques, 207–237. Barcelona: Institut d’Estudis Catalans.
Barberà, Gemma & Martine Zwets. 2013. Pointing and reference in sign language
and spoken language: Anchoring vs. Identifying. Sign Language Studies, 13
(4). 491–515.
Barwise, Jon & Robin Cooper. 1981. Generalized quantifiers and natural language.
Linguistics and philosophy 4. 159–219.
Baus, Cristina, Eva Gutiérrez-Sigut, Josep Quer & Manuel Carreiras. 2008. Lexical
Access in Catalan Sign Language (LSC) Production. Cognition 108. 856–865.
Belletti, Adriana. 1988. The case of unaccusatives. Linguistic Inquiry 19(1). 1–34.
Bellugi, Ursula & Edward Klima. 1990. Properties of visual spacial languages. In
Siegmund Prillwitz & Tomas Vollhaber (eds.), Sign Language Research and
Application. Washington D.C.: Gallaudet University Press.
Bellugi, Ursula; Lillo-Martin, Diane; O’Grady, Lucinda & van Hoek, Karen. 1990.
The development of spatialized syntactic mechanisms in American Sign
Language. In William Edmondson & Fred Karlsson (eds.), SLR ‘87: Papers
from the Fourth International Symposium on Sign Language Research, 183–
189. Hamburg: Signum-Verlag.
Benedicto, Elena, Sandra Cvejanov & Josep Quer. 2007. Valency in Classifier
Predicates: A Syntactic Analysis. Lingua 117. 1202–1215.
 References 241

Benedicto, Elena, Sandra Cvejanov & Josep Quer. 2008. The Morphosyntax of
Verbs of Motion in Serial Constructions: A Crosslinguistic Study in Three
Signed Languages. In Josep Quer (ed.), Signs of the time. Selectedpapers from
TISLR 8, 33–53. Hamburg: Signum Verlag.
Berenz, Norine. 1996. Person and deixis in Brazilian Sign Language. Berkeley, CA:
University of California, dissertation.
Berenz, Norine. 2002. Insights into person deixis. Sign Language & Linguistics 5(2).
203–227.
Berenz, Norine & Lucinda Ferreira-Brito. 1990. Pronouns in BCSL and ASL. In
William Edmondson & Fred Karlsson (eds.), SLR ‘87: Papers from the Fourth
International Symposium on Sign Language Research, 26–36. Hamburg:
Signum-Verlag.
Bertone, Carmela. 2007. La struttura del sintagma determinante nella Lingua dei
Segni Italiana (LIS). Venezia: Università Ca’ Foscari dissertation.
Bertone, Carmela. 2009. The syntax of noun modification in Italian Sign language
(LIS). University of Venice Working Papers in Linguistics, Vol. 19.
Bertone, Carmela & Anna Cardinaletti. 2011. Il sistema pronominale della lingua dei
segni italiana. In Anna Cardinaletti, Carlo Cecchetto & Caterina Donati, (eds.),
La lingua dei segni italiana: grammatica, lessico, e dimensioni di variazione,
145–160. Milano: FrancoAngeli.
Bhat, D. N. S. 2004. Pronouns. Oxford: Oxford University Press.
Birner, Betty & Gregory Ward. 1998. Information Status and Noncanonical Word
Order in English. Amsterdam/Philadelphia: John Benjamins.
Bos, Helen. 1990. Person and Location Marking in Sign Language of the
Netherlands: Some Implications of a Spatially Expressed Syntactic System.
In Siegmund Prillwitz & Tomas Vollhaber (eds.), Current Trends in European
Sign Language Research: Proceedings of the 3rd European Congress on Sign
Language Research, 231–246. Hamburg: Signum-Verlag.
Bos, Helen. 1995. Pronoun copy in Sign Language of the Netherlands. In Helen Bos
& Trude Schermer (eds.), Sign Language Research 1994. Proceedings of the
4th European Congress on Sign language Research, München 1994, 121–147.
Hamburg: Signum-Verlag.
Bosch i Baliarda, Marta. 2005. The Phonology of Compounds in Catalan Sign
Language (LSC). Barcelona: Universitat Autònoma de Barcelona MA thesis.
Branchini, Chiara & Caterina Donati. 2009. Relatively different. Italian Sign
Language relative clauses in a typological perspective. In Anikó Lipták (ed.),
Correlatives Cross-Linguistically, 157–191. Amsterdam: Benjamins.
Brennan, Mary. 1983. Marking time in British Sign Language. In Jim Kyle & Bencie
Woll (eds.), Language in sign: An international perspective on sign language,
10–31. London: Croom Helm.
242 References

Brennan, Susan, Marilyn Friedman & Carl J. Pollard. 1987. A centering approach to
pronouns. In Association for Computational Linguistics (eds.), Proceedings of
the 25th Meeting of the Association for Computational Linguistics, 155–162.
Stanford: California.
Brentari, Diane. 1998. A prosodic model of sign language phonology. Cambridge,
MA: MIT Press.
Brentari, Diane. 2010. Sign Languages: A Cambridge Language Survey. Cambridge,
UK: Cambridge University Press.
Büring, Daniel. 1999. Topic. In Peter Bosch & Rob van der Sandt (eds.), Focus.
Linguistic, Cognitive, and Computational Perspectives, 142–165. Cambridge,
UK: Cambridge University Press.
Büring, Daniel. 2003. On D-Trees, Beans, and B-Accents. Linguistics and Philosophy
26(5). 511–545.
Cabeza, Carmen & Ana Fernández-Soneira. 2004. The expression of time in Spanish
Sign Language (LSE). Sign Language & Linguistics 7(1). 63–82.
Caponigro, Ivano & Kathryn Davidson. 2011. Ask, and Tell as Well: Clausal
Question-Answer Pairs in ASL. Natural Language Semantics 19(4). 323–371.
Cardinaletti Anna & Michael Starke. 1999. The typology of structural deficiency:
A case study of the three classes of pronouns. In Henk van Riemsdijk (ed.),
Clitics in the Languages of Europe, 145–233. Berlin: de Gruyter Mouton.
Carlson, Gregory. 1977. Reference to Kinds in English. Amherst: University of
Massachusetts, Amherst: University of Massachusetts dissertation.
Carlson, Gregory & Rachel Sussman. 2005. Seemingly indefinite definites. In
Stephan Kepsar & Marga Reis (eds.), Linguistic Evidence, 71–86. Berlin: de
Gruyter Mouton.
Cedillo, Pepita. 2004. Háblame a los ojos. Barcelona: Octaedro.
Chang, Jung-hsing, Shiou-fen Su & James H-Y. Tai. 2005. Classifier predicates
reanalyzed, with special reference to Taiwan Sign Language. Language and
Linguistics 6(2). 247–278.
Chapa, Carmen. 2001. La variación del registro en LSE. Valencia: Fundación
Fesord CV.
Cecchetto, Carlo, Carlo Geraci & Sandro Zucchi. 2006. Strategies of Relativization
in Italian Sign Language. Natural Language and Linguistic Theory 25. 945–
975.
Cecchetto, Carlo & Sandro Zucchi. 2006. Condizioni di verità, sottospecificazione
e discorso nelle lingue dei segni. In Simona Venezia (ed.), Tradurre e
comprendere. Pluralità dei linguaggi e delle culture. Roma: Rocco Pititto and
Aracne editrice.
 References 243

So, Wing Chee, Marie Coppola, Vincent Liccidarello & Susan Goldin-Meadow.
2005. The seeds of spatial grammar in the manual modality. Cognitive Science,
29. 23–37.
Chiarcos, Christian, Berry Claus & Michael Grabski. 2010. Salience in linguistics
and beyond. In Christian Chiarcos, Berry Claus, & Michael Grabski (eds.),
Salience. Multidisciplinary perspectives on its function in discourse, 1–30.
Berlin: de Gruyter Mouton.
Clark, Herbert. 1975. Bridging. In R. C. Schank & B. L. Nash-Webber (eds.),
Theoretical issues in natural language processing, 169–174. New York, NY:
Association for Computing Machinery.
Coates, Jennifer & Rachel Sutton-Spence. 2001. Turn taking patterns in Deaf
conversation. Journal of Sociolinguistics 5(4). 507–529.
Codorniu, Imma, Josep M. Segimon & M. del Pilar Fernández-Viader. 2005.
Medecines i salut LSC. Barcelona: Fundació privada Illescat.
Consten, Manfred. 2003. Towards a unified model of domain-bound reference. In
Fredrich Lenz (ed.), Deictic Conceptualisation of Space, Time and Person,
223–248. Amsterdam/Philadelphia: John Benjamins.
Coppola, Marie & Ann Senghas. 2010. The emergence of deixis in Nicaraguan
signing. In Diane Brentari (ed.), Sign Languages: A Cambridge Language
Survey, 543–569. Cambridge, UK: Cambridge University Press.
Cormier, Kearsy. 2007. Do all pronouns point? Indexicality of first person plural
pronouns in BSL and ASL. In Pamela Perniss, Roland Pfau & Markus
Steinbach (eds), Visible variation: Comparative studieson sign language
structure, 63–102. Berlin: de Gruyter Mouton.
Cormier, Kearsy. 2012. Pronouns. In Pfau, Roland, Markus Steinbach & Bencie
Woll (eds.), Sign Language. Handbook of Linguistics and Communication
Science, 227–244. Berlin: de Gruyter Mouton.
Cormier, Kearsy, Stephen Wechsler, & Richard P. Meier. 1999. Locus Agreement in
American Sign Language. In Gert Webulhuth, Jean Pierre Koenig & Andreas
Kathol (eds.), Lexical and Constructional Aspects of Linguistic Explanation,
215–229. Stanford: CSLI Publications.
Cornish, Francis. 1999. Anaphora, Discourse and Understanding. Evidence from
English and French. Oxford: Clarendon Press.
Costello, Brendan, Javier Fernández & Alazne Landa. 2008. The non(-existent)
native signer: Sign language research in a small deaf population. In Ronice
Müller de Quadros (ed.), Sign Languages: spinning and unraveling the past,
present and future (TISLR9), Petrópolis/RJ. Brazil: Editora Arara Azul.
Crasborn, Onno. 2007. How to recognise a sentence when you see one. Sign
Language & Linguistics 10(2). 103–111.
244 References

Crasborn, Onno, Harry van der Hulst & Els van der Kooij. 2002. SignPhon: a
phonological database for sign languages. Sign Language & Linguistics
4(1/2). 215–228.
Crasborn, Onno, Johanna Mesch, Dafydd Waters, Annika Nonhebel, Els van der
Kooij, Bencie Woll & Brita Bergman. 2007. Sharing sign langauge data
online. Experiences from the ECHO project. International Journal of Corpus
Linguistics 12(4). 537–564.
Crasborn, Onno, Els van der Kooij, Johan Ros & Helen de Hoop. 2009. Topic-
agreement in NGT (Sign Language of the Netherlands). The Linguistic Review
26(2/3). 355–370.
Davidson, Kathryn. 2014. Scalar implicatures in a signed language. Sign Language
& Linguistics, 17 (1). 1–19.
Davidson, Kathryn. 2013. ‘And’ or ‘Or’: General Use Coordination in ASL.
Semantics & Pragmatics, 6 (4). 1–44.
Davidson, Kathryn. 2012. When Disjunction Looks like Conjunction: Pragmatic
Consequence in ASL. In Maria Aloni, Floris Roelofsen, Galit Weidman
Sassoon, Katrin Schulz, Vadim Kimmelman & Matthijs Westera (eds.),
Proceedings of 18th Amsterdam Colloquium 2011 (LNCS 7218), 72–81. Berlin
Heidelberg: Springer-Verlag.
Davidson, Kathryn & Deanna Gagne. 2014. Vertical representation of quantifier
domains. In Urtzi Etxeberria, Anamaria Falaus, Aritz Irurtzun & Bryan
Leferman (eds.), Proceedings of Sinn und Bedeutung 18, 110–127. Vitoria-
Gasteiz.
DOMAD. 2002a. Els Genís i l’educació. Barcelona: FESOCA.
DOMAD. 2002b. Els Genís i la sanitat. Barcelona: FESOCA.
Diesing, Molly. 1992. Indefinites. Cambridge, MA: The MIT Press.
Dimmendaal, Gerrit. 2001. Places and people: field sites and informants. In Paul
Newman & Martha Ratliff (eds.), Linguistic Field work, 55–75. Cambridge:
Cambridge University Press.
Donnellan, Keith S. 1966. Reference and Definite Descriptions. The Philosophical
Review 75(3). 281–304.
Dryer, Matthew. 2006. Descriptive theories, explanatory theories, and basic linguistic
theory. In Felix Ameka, Alan Dench and Nicholas Evans (eds.), Catching
Language: Issues in Grammar Writing, 207–234. Berlin: de Gruyter Mouton.
Elbourne, Paul. 2005. Situations and individuals. Cambridge, MA: The MIT Press.
Emmorey, Karen. 1997. Non-antecedent suppression in American Sign Language.
Language and Cognitive Processes 12(1). 103–112.
Emmorey, Karen. 2001. Space on Hand: The Exploitation of Signing Space to
Illustrate Abstract Thought. In Merideth Gattis (ed.), Spatial Schemas and
Abstract Thought, 147–174. Cambridge, MA: The MIT Press.
 References 245

Emmorey, Karen. 2002a. Language, Cognition and the Brain. Insights from Sign
Language Research. Mahwah, NJ: Lawrence Erlbaum Associates.
Emmorey, Karen. 2002b. The effects of modality on spatial language: How signers
and speakers talkabout space. In Richard Meier, Kearsy Cormier & David
Quinto-Pozos (eds.), Modality and structure in signed and spoken languages,
405–421. Cambridge: Cambridge University Press.
Emmorey, Karen. 2007. The psycholinguistics of signed and spoken languages: How
biology affects processing. In M. Gareth Gaskell (ed.), The Oxford handbook
of psycholinguistics, 703–721. Oxford, UK: Oxford University Press.
Emmorey, Karen, David Corina & Ursula Bellugi. 1995. Differential Processing of
Topographic and Referential Functions of Space. In Karen Emmorey & Judy
Reilly (eds.), Language, Gesture, and Space, 43–62. Hillsdale, NJ: Lawrence
Erlbaum Associates.
Emmorey, Karen, Hanna Damasio, Stephen McCullough, Thomas Grabowski, Laura
Ponto, Richard Hichwa & Ursula Bellugi. 2002. Neural systems underlying
spatial language in American Sign Language. Neuro Image 17. 812–824.
Emmorey, Karen & Brenda Falgier. 1999. Describing Environments in ASL. In
Elisabeth Winston (ed.), Storytelling and Conversation. Discourse in Deaf
Communities, 3–26. Washington, D.C.: Gallaudet University Press.
Emmorey, Karen & Brenda Falgier. 2004. Conceptual Locations and Pronominal
Reference in American Sign Language. Journal of Psycholinguistic Research
33(4). 321–331.
Emmorey, Karen & Melissa Herzig. 2003. Categorical Versus Gradient Properties
of Classifier Constructions in ASL. In Karen Emmorey (ed.), Perspectives on
Classifier Constructions in Sign Languages, 221–246. Mahwah, NJ: Lawrence
Erlbaum Associates.
Emmorey, Karen, Edward Klima & Gregory Hickok. 1998. Mental rotation within
linguistic and non-linguistic domains in users of American Sign Language.
Cognition 68. 221–246.
Emmorey, Karen & Harlan Lane. 2000. The signs of language revisited. An anthology
to honor Ursula Bellugi and Edward Klima. Mahwah, NJ: Lawrence Erlbaum
Associates.
Emmorey, Karen & Diane Lillo-Martin. 1995. Processing spatial anaphora: Referent
reactivation with overt and null pronouns in American Sign Language.
Language and Cognitive Processes 10. 631–664.
Emmorey, Karen & Barbara Tversky. 2002. Spatial Perspective Choice in ASL. Sign
Language & Linguistics 5(1). 3–26.
Enç, Mürvet. 1991. The semantics of specificity. Linguistic Inquiry 22(1). 1–25.
246 References

Engberg-Pedersen, Elisabeth. 1990. Pragmatics of Nonmanual Behaviour in Danish

Sign Language. In William Edmonson & Fred Karlsson (eds.), SLR’87. Papers
from the Fourth International Symposium on Sign Language Research, 121–
128. Hamburg: Signum-Press.
Engberg-Pedersen, Elisabeth. 1993. Space in Danish Sign Language. The Semantics
and Morphosyntax of the Use of Space in a Visual Language. Hamburg:
Signum-Verlag.
Engberg-Pedersen, Elisabeth. 1995. Point of View Expressed Through Shifters.
In Karen Emmorey & Judy Reilly (eds.), Language, Gesture and Space,
113–154. Mahwah, NJ: Lawrence Erlbaum Associates.
Engberg-Pedersen, Elisabeth. 2003. The functions of a Pointing Gesture. In Sotaro
Kita (ed.), Pointing: Where Language, Culture, and Cognition Meet, 269–292.
Mahwah, NJ: Lawrence Erlbaum Associates.
Di Eugenio, Barbara. 1998. Centering in Italian. In M. A. Walker, A. K. Joshi, & E.
F. Prince (eds.), Centering in Discourse, 115–137. Oxford: Clarendon Press.
Evans, Gareth. 1980. Pronouns. Linguistic Inquiry 11(2). 337–362.
Farkas, Donka. 1994. Specificity and Scope. In L. Nash & G. Tsoulas (eds.), Langues
et Grammaires 1, 119–137.
Farkas, Donka. 1997. Dependent indefinites. In F. Corblin & D. G. Marandin (eds.),
Empirical Issues in Formal Syntax and Semantics, 243–267. Peter Lang
Publishers.
Farkas, Donka. 2002. Specificity distinctions. Journal of Semantics 19. 1–31.
Farkas, Donka & Henriëtte de Swart. 2007. Article choice in plural generics. Lingua
11(9). 1657–1676.
Farkas, Donka & Henriëtte de Swart. 2003. The Semantics of Incorporation.
Stanford: CSLI.
Fauconnier, Gilles. 1985. Mental Spaces. Cambridge, MA: MIT Press.
Fenlon, Jordan, Adam Schembri, Ramas Rentelis & Kearsy Cormier. 2013. Variation
in handshape and orientation in British Sign Language: The case of the ‘1’
hand configuration. Language and Communication 33. 69–91.
Ferrerons, Ramon. 2011. Primer diccionari general i etimològic de la llengua de
signes catalana. Barcelona: Documenta Universitaria.
Fischer, Susan. 1975. Influences on Word Order Change in American Sign Language.
In Charles Li (ed.), Word Order and Word Order Change, 1–25. University of
Texas Press.
Fischer, Susan & Bonnie Gough. 1978. Verbs in American Sign Language. Sign
Language Studies 7. 17–48.
Fraurud, Kari. 1990. Definiteness and the processing of NPs in natural discourse.
Journal of Semantics 7. 395–433.
Friedman, Lynn. 1975. Space, Time, and Person Reference in American Sign
Language. Language 51(4). 940–961.
 References 247

Friedman, Lynn. 1976. The Manifestation of Subject, Object and Topic in American
Sign Language. In Charles Li (ed.), Subject and Topic, 127–148. New York,
NY: Academia Press.
Frigola, Santiago. 2010. La comunidad sorda de Catalunya. In Joan Martí & Josep
Maria Mestres (eds.), Les llengües de signes com a llengües minoritàries:
perspectives lingüístiques, socials i polítiques, 29–54. Barcelona: IEC.
Frigola, Santiago & Josep Quer. 2006. Discurso (In)directo y deícticos en estructuras
de cambio de rol en lengua de signos catalana. 7è congrés de Lingüística
General, Barcelona: Universitat de Barcelona.
Frigola, Santiago, Gemma Barberà, Delfina Aliaga & Joan Gil. 2011. Gramàtica
bàsica LSC: Material d’activitats. Barcelona: FESOCA.
Frishberg, Nancy. 1983. Dominance relations and discourse structure. In William
Stokoe & Viriginia Volterra (eds.), SLR ‘83, Proceedings of the 11th
international Symposium on Sign Language Research, 79–90. Silver Spring:
Linstok.
Fodor, Janet & Ivan Sag. 1982. Referential and quantificational indefinites.
Linguistics and Philosophy 5. 355–398.
García, Carlos & Imma Codorniu. 2007. Aliments, Alimentos, Food. Aprenem LSC:
manual per a l’ensenyament-aprenentatge de la llengua de signes catalana.
Barcelona: Fundació privada Illescat.
Geach, Peter. 1962. Reference and Generality. Ithaca, NY: Cornell University Press.
Geurts, Bart. 1999. Specifics. In Bart Geurts, Manfred Krifka & Rob van der
Sandt (eds.), Focus and Presupposition in Multi-Speaker Discourse, 99–129.
Utrecht: ESSLLI 99.
Glück, Susanne & Roland Pfau. 1998. On classifying classification as a class of
inflection in German Sign Language. In Tina Cambier-Langeveld, Anikó
Lipták & Michael Redford (eds.), ConSole VI Proceedings, 59–74. Leiden:
SOLE.
Gras, Victòria. 2006. La comunidad sorda como comunidad lingüística: Panorama
sociolingüístico de la/s lengua/s de signos en España. Barcelona: Universitat
de Barcelona dissertation.
Groenendijk, J. & M. Stokhof. 1991. Dynamic predicate logic. Linguistics and
Philosophy 14(1). 39–100.
Grosz, Barbara. 1981. Focusing and description in natural language dialogues. In
Aravind Joshi, Bonnie Webber & Ivan Sag (eds.), Elements of Discourse
Understanding, 85–105. Cambridge: Cambridge University Press.
Grosz, Barbara, Araving Joshi & Scott Weinstein. 1983. Providing a unified account
of definite noun phrases in discourse. In Association for Computational
Linguistics (eds.) Proceedingsof the 21st Annual Meeting of the Association
for Computational Linguistics, 44–50.
248 References

Grosz, Barbara & Candace Sidner. 1986. Attention, intentions, and the structure of
discourse. Computational Linguistics 12(3). 175–204.
Grosz, Barbara, Aravind Joshi & Scott Weinstein. 1995. Centering: A Framework
for Modeling the Local Coherence of Discourse. Computational Linguistics
2(21). 203–225.
Gundel, Jeanette. 1988. Universals of topic-comment structure. In Michael
Hammond, Edith Moravcsik & Jessica Wirth (eds.), Studies in Syntactic
Typology, 209–239. Amsterdam: John Benjamins.
Gundel, Jeanette, Nancy Hedberg & Ron Zacharski. 1993. Cognitive Status and the
Form of Referring Expressions in Discourse. Language 69. 274–307.
Gundel, Jeanette & Thorstein Fretheim. 2004. Topic and Focus. In Laurence Horn
& Gregory Ward (eds.), The Handbook of Pragmatics, 175–194. Oxford:
Blackwell.
Heim, Irene. 1982. The Semantics of Definite and Indefinite Noun Phrases. Amherst,
MA: University of Massachusetts dissertation.
Heim, Irene. 1990. E-Type Pronouns and Donkey Anaphora. Linguistics and
Philosophy 13(2). 137–77.
Heim, Irene & Angelica Kratzer. 1998. Semantics in Generative Grammar. Oxford:
Blackwell.
Herrero, Angel. 2009. Gramática didáctica de la lengua de signos española (LSE).
Ediciones SM.
Herrmann, Annika & Markus Steinbach. 2012. Quotation in Sign Languages – A
Visible Context Shift. In Alphen, Ingrid van & Isabelle Buchstaller (eds.),
Quotatives. Cross-linguistic and Cross-disciplinary Perspectives, 203–228.
Amsterdam: Benjamins.
von Heusinger, Klaus. 2002. Specificity and definiteness in sentence and discourse
structure. Journal of Semantics 19. 245–274.
von Heusinger, Klaus. 2011a Specificity, referentiality and discourse prominence:
German indefinite demonstratives. In Ingo Reich, Eva Horch & Dennis Pauly
(eds.), Proceedings of Sinn and Bedeutung 15, 9–30. Saarbrücken: Universaar
– Saarland Unversity Press.
von Heusinger, Klaus. 2011b. Specificity. In Klaus von Heusinger, Claudia
Maienborn & Paul Portner (eds.), Semantics: An International Handbook of
Natural Language Meaning, 1024–1057. Berlin: de Gruyter Mouton.
Hirschman, Lynette. 1997. MUC-7 Co-reference Task Definition. Version 3.0. Ms.
University of Pennsylvania.
van Hoek, Karen. 1992. Conceptual spaces and pronominal reference in American
Sign Language. Nordic Journal of Linguistics 15. 183–199.
van Hoek, Karen. 1996. Conceptual Locations for Reference in American Sign
Language. In Gilles Fauconnier & Eve Sweetser (eds.), Spaces, Worlds, and
Grammar, 334–350. Chicago, IL: University of Chicago Press.
 References 249

Horn, Laurence & Gregory Ward. 2004. The Handbook of Pragmatics. Oxford:
Blackwell.
Huang, Yan. 1994. The Syntax and Pragmatics of Anaphora. Cambridge: Cambridge
University Press.
Huang, Yan. 2004. Anaphora and the Pragmatics-Syntax Interface. In Laurence
Horn & Gregory Ward (eds.), The Handbook of Pragmatics, 288–313. Oxford:
Blackwell.
Illescat. 2002a. Aprenem LSC1! Manual per a l’ensenyament-aprenentatge de la
Llengua de Signes Catalana 1. Illescat: Barcelona.
Illescat. 2002b. Aprenem LSC2! Manual per a l’ensenyament-aprenentatge de la
Llengua de Signes Catalana 2. Illescat: Barcelona.
Illescat. 2004. DILSCAT, Diccionari bàsic de la llengua de signes catalana. Editorial
Fundació Illescat: Barcelona.
Ingram, Robert. 1978. Theme, rheme, topic, and comment in the syntax of American
Sign Language. Sign Language Studies 20. 193–218.
Ionin, Tania. 2006. This is definitely specific: specificity and definiteness in article
systems. Natural Language Semantics 14. 175–234.
Janis, Wynne Dana. 1992. Morphosyntax of the ASL verb phrase. New York, NY:
State University of New York at Buffalo dissertation.
Janis, Wynne Dana. 1995. A crosslinguistic perspective on ASL verb agreement. In
Karen Emmorey & Judy Reilly (eds.), Language, Gesture, and Space, 255–
286. Hilsdale, NJ: Lawrence Erlbaum Associates.
Janzen, Terry. 1997. Pragmatic and Syntactic Features of Topics in American Sign
Language. Meta XLII 3. 502–513.
Janzen, Terry. 1999. The Grammaticalization of Topics in American Sign Language.
Studies in Language 23. 271–306.
Jarque, Maria Josep. 2005. Double mapping in metaphorical expressions of thought
and communication in Catalan Sign Language. Sign Language Studies 5.
3292–316.
Jarque, Maria Josep. 2011. Lengua y gesto en la modalidad lingüística signada.
Anuari de filologia. Estudis de lingüística 1. 71–99.
Jarque, Maria Josep. 2012. Las lenguas de signos: su estudio científico y
reconocimiento legal. Anuari de filologia. Estudis de lingüística 2. 33–48.
Jarque, Maria Josep, Imma Codorniu, Marta Bosch i Baliarda, Maria del Pilar
Fernández Viader, Carlos Garcia, Encarna Serrano & Josep Maria Segimon.
2012. Procesos de lexicalización en la LSC: Procedimientos de combinación.
Anuari de filologia. Estudis de lingüística 2. 141–176.
Jarque, Maria Josep, María Ignacia Massone, Maria del Pilar Fernández Viader, Marta
Bosch i Baliarda & Josep Segimon. 2007. Orden básico de constituyentes en
la lengua de signos de Cataluña. In Pablo Cano López (ed.), Actas del VI
Congreso de Lingüística General, Santiago de Compostela.
250 References

Johnston, Trevor. 1991. Spatial Syntax and Spatial Semantics in the Inflection of
Signs for the Marking of Person and Location in Auslan. International Journal
of Sign Linguistics 2. 29–62.
Johnston, Trevor. 2008. Corpus linguistics and signed languages: No lemmata,
no corpus. In Onno Crasborn, Thomas Hanke, Ernst Thoutenhoofd, Inge
Zwitserlood & Eleni Efthimiou (eds.), Construction and exploitation of sign
language corpora. Proceedings of the 3rd Workshop on the representation and
processing of sign languages. Paris: ELRA.
Johnston, Trevor. 2010. From archive to corpus. Transcription and annotation in
the creation of signed language corpora. International Journal of Corpus
Linguistics 15(1). 106–131.
Kadmon, Nirit. 1990. Uniqueness. Linguistics and Philosophy 13. 273–324.
Kameyama, M. 1998. Intrasentential centering: A case study. In Centering Theory
in Discourse, M. Walker, A. Joshi & E. Prince (eds.), 89–112. Oxford: Oxford
University Press.
Kamp, Hans. 1981. A theory of truth and semantic representation. In J. A. G.
Groenendijk, T. M. V. Janssen, & M. B. J. Stokhhof (eds.), Formal Methods in
the Study of Language, 227–322. Amsterdam: Mathematical Centre.
Kamp, Hans & Ágnes Bende-Farkas. 2006. Epistemic Specificity from a
Communication-theoretical Perspective. Ms. Stuttgart: Universität Stuttgart.
Kamp, Hans & Uwe Reyle. 1993. From discourse to logic. Introduction to
modeltheoretic semantics of natural language, formal logic and discourse
representation theory. Dordrecht: Kluwer Academic Press.
Kamp, Hans, Uwe Reyle & Josef van Genabith. 2007. Discourse Representation
Theory. In Dov Gabbay & Franz Guenthner (eds.), Handbook of Philosophical
Logic. Kluwer Academic Press, Dordrecht: Kluwer Academic Press.
Karttunen, Lauri. 1968. What do referential indices refer to? Paper presented at the
Linguistics Colloquium, University of California, Los Angeles.
Karttunen, Lauri. 1976. Discourse referents. In J. McCawley (ed.), Syntax and
Semantics: Notes from the Linguistic Underground, 363–386. New York, NY:
Academic Press.
Kegl, Judy. 1986. Clitics in American Sign Language. In Hagit Borer (ed.), Syntax
and semantics, Vol. 19: The syntax of pronominal clitics, 285–309. New York,
NY: Academic Press.
Kegl, Judy. [1976]2003. Pronominalization in American Sign Language. Sign
Language & Linguistics 6. 245–265.
Kegl, Judy, Ann Senghas & Marie Coppola. 1999. Creation through Contact: Sign
Language Emergence and Sign Language Change in Nicaragua. In Michel
DeGraff (ed.), Language Creation and Language Change: Creolization,
Diachrony, and Development, 179–237. Cambridge, MA: The MIT Press.
 References 251

Kibrik, Andrej A. & Evgenija V. Prozorova. 2007. Referential Choice in Signed

and Spoken Languages. In Antonio Branco, Tony McEnery, Ruslan Mitkov
& Fatima Silva (eds.), Proceedings of 6th Discourse Anaphora and Anaphor
Resolution Colloquium, 41–46. Porto: Centro de Linguistica da Universidade
do Porto.
Klima, Edward & Ursula Bellugi. 1979. The signs of language. Cambridge, MA:
Harvard University Press.
Kooij, Els van der. 2002. Phonological categories in Sign Language of the
Netherlands: The role of phonetic implementation and iconicity. Utrecht:
LOT, Universiteit Leiden dissertation.
Kooij, Els van der, Onno Crasborn & Johan Ros. 2006. Explaining prosodic body
leans in Sign language of the Netherlands: Pragmatics required. Journal of
Pragmatics 38. 1598–1614.
Kratzer, Angelika. 1998. Scope or pseudo-scope? Are there wide-scope indefinites?
In S. Rothstein (ed.), Events and Grammar, 163–196. Great Britain: Kluwer.
Krifka, Manfred, Francis Jeffry Pelletier, Gregory N. Carlson, Alice ter Meulen,
Gennaro Chierchia & Godehard Link. 1995. Genericity: An Introduction.
In Gregory N. Carlson & Francis Jeffry Pelletier (eds.), The Generic Book,
1–124. Chicago and London: University of Chicago Press.
Kuhn, Jeremy. 2014. Towards a feature-based semantics of ASL loci. In Urtzi
Etxeberria, Anamaria Falaus, Aritz Irurtzun & Bryan Leferman (eds.),
Proceedings of Sinn und Bedeutung 18, 235–252. Vitoria-Gasteiz.
Lacy, Richard.[1974]2003. Putting some of the syntax back into the semantics. Sign
Language & Linguistics 6(2). 211–244.
Lambrecht, Knud. 1994. Information Structure and Sentence Form. Topic, focus
and mental representations of discourse referents. Cambridge: Cambridge
University Press.
Lascarides, Alex & Matthew Stone. 2009. A Formal Semantic Analysis of Gesture.
Journal of Semantics 26(4). 393–449.
Lazaridou-Chatzigoga, Dimitra. 2009. Topics in the syntax and semantics of the
definite  article  in  Modern  Greek. Barcelona: Universitat Autònoma de
Barcelona dissertation.
Lenseigne, Boris & Patrice Dalle. 2006. Using signing space as a representation for
sign language processing. In Sylvie Gibet, Nicolas Courty & Jean-François
Kamp (eds.), Gesture in Human-Computer Interaction and Simulation,
Lecture Notes in Artificial Intelligence, 3881. Berlin: Springer.
Levinson, Stephen. 1996. Frames of reference and Molyneux’s question. In Paul
Bloom, Mary A. Peterson, Lynn Nadel & Merrill F. Garrett (eds.), Language
and Space, 109–170. Cambridge, MA: The MIT Press.
Levinson, Stephen. 2004. Deixis. In Laurence Horn & Gregory Ward (eds.), The
Handbook of Pragmatics, 97–121. Oxford: Blackwell.
252 References

Lewis, David. 1975. Adverbs of quantification. In Edward L. Keenan (ed.), Formal

semantics of natural language, 3–15. Cambridge: Cambridge University
Press.
Lewis, David. 1979. Scorekeeping in a Language Game. Journal of Philosophical
Logic 8. 339–359.
Liddell, Scott. 1990. Four Functions of a Locus: Reexamining the Structure of Space
in ASL. In Ceil Lucas (ed.), Sign Language Research: Theoretical Issues,
176–198. Washington D.C.: Gallaudet University Press.
Liddell, Scott. 1995. Real, Surrogate and Token Space: Grammatical Consequences
in ASL. In Karen Emmorey & Judy Reilly (eds.), Language, Gesture, and
Space, 19–41. Hillsdale, NJ: Lawrence Erlbaum Associates.
Liddell, Scott. 1996. Spatial representations in discourse: Comparing spoken and
signed language. Lingua 98. 145–167.
Liddell, Scott. 1998. Grounded blends, gestures, and conceptual shifts. Cognitive
Linguistics. 9(3). 283–314.
Liddell, Scott. 2000a. Blended spaces and deixis in sign language discourse. In David
McNeill (ed.), Language and Gesture, 331–357. Cambridge: Cambridge
University Press.
Liddell, Scott. 2000b. Indicating Verbs and Pronouns: Pointing Away From
Agreement. In Karen Emmorey & Harlan Lane (eds.), The Signs of Language
Revisited, 303–320. Mahwah, NJ: Lawrence Erlbaum Associates.
Liddell, Scott. 2003. Grammar, Gesture and Meaning in American Sign Language.
Cambridge: Cambridge University Press.
Liddell, Scott & Robert Johnson. 1989. American Sign Language: The Phonological
Base. Sign Language Studies 64. 195–277.
Liddell, Scott & Melanie Metzger. 1998. Gesture in sign language discourse. Journal
of Pragmatics 30. 657–697.
Liddell, Scott, Marit Vogt-Svendsen & Brita Bergman. 2007. A Crosslinguistic
Comparison of Buoys. In Myriam Vermeerbergen, Lorraine Leeson & Onno
Crasborn (eds.), Simultaneity in Signed Languages: Form and Function, 187–
215. Amsterdam/Philadelphia: John Benjamins.
Lillo-Martin, Diane. 1986. Two kinds of null arguments in American Sign Language.
Natural Language and Linguistic Theory, 4. 415–444.
Lillo-Martin, Diane. 1991. Universal Grammar and American Sign Language:
Setting the null argument parameters. Dordrecht: Kluwer Academic Publishers.
Lillo-Martin, Diane. 1995. The point of view predicate in American Sign Language.
In Karen Emmorey & Judy Reilly (eds.), Language, Gesture, and Space, 155–
170. Hillsdale, NJ: Lawrence Erlbaum Associates.
 References 253

Lillo-Martin, Diane. 1999. Modality effects and modularity in language acquisition:

The acquisition of American Sign Language. In Tej Bhatia & William Ritchie
(eds.), Handbook of Child Language Acquisition, 531–567. San Diego, CA:
Academic Press.
Lillo-Martin, Diane. 2002. Syntax in sign: Few or no effects of modality. In Richard
Meier, Kearsy Cormier & David Quinto-Pozos (eds.), Modality and structure
in signed and spoken languages, 237–262. Cambridge: Cambridge University
Press.
Lillo-Martin, Diane & Edward Klima. 1991. Pointing Out Differences: ASL
Pronouns in Syntactic Theory. In Susan Fischer & Patricia Siple (eds.),
Theoretical Issues in Sign Language Research, Vol. 1: Linguistics, 191–210.
Chicago, IL: University Chicago Press.
Lillo-Martin, Diane & Richard Meier. 2011. On the linguistic status of ‘agreement’
in sign languages. Theoretical Linguistics 37(3/4). 95–141.
Löbner, Sebastian. 1985. Definites. Journal of Semantics 4. 279–326.
Loew, Ruth. 1984. Roles and Reference in American Sign Language: a Developmental
Perspective. Minnesota, MN: University of Minnesota dissertation.
Lucas, Ceil & Clayton Valli. 1992. Language Contact in the American Deaf
Community. San Diego CA: Academic Press.
Lyons, John. 1977. Semantics. Cambridge: Cambridge University Press.
Lyons, Christopher. 1999. Definiteness. Cambridge: Cambridge University Press.
MacLaughlin, Dawn. 1997. The structure of determiner phrases: Evidence from
American Sign Language. Boston, MA: Boston University dissertation.
MacSweeney, M., B. Woll, R. Campbell, P. K. McGuire, A. S. David, S. C. R.
Williams, J. Suckling, G. A. Calvert & M. J. Brammer. 2002. Neural systems
underlying British Sign Language and audio-visual English processing in
native users. Brain: a Journal of Neurology 125(7). 1583–1593.
Mandel, Mark. 1977. Iconic devices in American Sign Language. In Lynn Friedman
(ed.), On the other hand, 57–107. London: Academic Press.
Mann, William & Sandra Thompson. 1988. Rhetorical Structure Theory: Toward a
functional theory of text organization. Text 8. 243–281.
Marsaja, Gede. 2008. Desa kolok - A deaf village and its sign language in Bali,
Indonesia. Nijmegen: Ishara Press.
Martín, Ignasi & Marisa Alvarado. 2004. Diccionari temàtic de llenguatge de signes
català. Lleida.
Massó, Guillem & Toni Badia. 2010. Dealing with Sign Language Morphemes in
Statistical Machine Translation. In 4th Workshop on the Representation and
Processing of Sign Languages: Corpora and Sign Language Technologies,
154–157. París: ELRA.
254 References

Massone, M. Ignacia; Marta Bosch i Baliarda & María del Pilar Fernández Viader.
2003. Una aproximación al Sistema de Transcripción Fonológica Liddell-
Johnson: Reflexiones para el análsis de la LSC. Revista española de lingüística
de laslenguas de signos. Madrid: RELLS.
De Matteo, Asa. 1977. Visual Imagery and Visual Analogues in American Sign
Language. In Lynn Friedman (ed.), On the Other Hand, 109–136. London:
Academic Press.
Mathur, Gaurav. 2000. Verb Agreement as Alignment in Signed Languages.
Cambridge: MIT dissertation.
Mathur, Gaurav & Christian Rathmann. 2010. Two types of nonconcatenative
morphology. In Gaurav Mathur & Donna Jo Napoli (eds.), Deaf around the
World: The Impact of Language, 54–82. Oxford: University Press.
Matthewson, Lisa. 1998. Determiner  Systems  and  Quantificational  Strategies: 
Evidence from Salish. The Hague: Holland Academic Graphics.
Matthewson, Lisa. 2004. On the Methodology of Semantic Fieldwork. International
Journal of American Linguistics 70. 369–415.
Mayol, Laia. 2009. Pronouns in Catalan: information, discourse and strategy.
Pennsylvania, PA: University of Pennsylvania dissertation.
Mayol, Laia. 2010. Contrastive Pronouns in Null Subject Romance Languages.
Lingua 120(10). 2497–2514.
Mc Burney, Susan. 2002. Pronominal reference in signed and spoken language:
Are grammatical categories modality-dependent? In Richard Meier, Kearsy
Cormier & David Quinto-Pozos (eds.), Modality and structure in signed and
spoken languages, 329–369. Cambridge: Cambridge University Press.
McEnery, Tony & Andrew Wilson. 2001. Corpus Linguistics. Edinburgh: Edinburgh
University Press.
McNally, Louise. 1998. On Recent Formal Analyses of Topic. In Jonathan Ginzburg,
Zurab Khasidashvili, Carl Vogel, Jean-Jacques Lévyand & Enric Vallduví
(eds.), The Tbilisi Symposium on Language, Logic, and Computation: Selected
Papers, 147–160. Stanford, CA: CSLI Publications.
McNally, Louise. 2004. Bare plurals in Spanish are interpreted as properties. Catalan
Journal of Linguistics 3. 115–133.
Meier, Richard. 1990. Person Deixis in American Sign Language. In Susan Fischer
& Patricia Siple (eds.), Theoretical Issues in Sign Language Research, Vol. 1:
Linguistics, 175–190. Chicago, IL: University Chicago Press.
Meier, Richard. 2002. Why different, why the same? Explaining effects and non-
effects of modality upon linguistic structure in sign and speech. In Richard
Meier, Kearsy Cormier & David Quinto-Pozos (eds.), Modality and structure
in signed and spoken languages, 1–25. Cambridge: Cambridge University
Press.
 References 255

Meier, Richard P. & Lillo-Martin, Diane. 2010. Does spatial make it special? On
the grammar of pointing signs in American Sign Language. In Donna Gerdts,
John Moore & Maria Polinsky (eds.), Hypothesis A/Hypothesis B: Linguistic
Explorations in Honor of David M. Perlmutter, 345–360. Cambridge, MA:
The MIT Press.
Meir, Irit. 2003. Grammaticalization and modality: the emergence of a case-marked
pronoun in Israeli Sign Language. Journal of Linguistics 39. 109–140.
Meir, Irit, Carol Padden, Mark Aronoff & Wendy Sandler. 2007. Body as subject.
Journal of Linguistics 43. 531–563.
Meir, Irit & Wendy Sandler. 2008. A language in space: The story of Israeli Sign
Language. New York, NY: Lawrence Erlbaum Associates.
Metzger, Melanie. 1995. Constructed Dialogue and Constructed Action in American
Sign Language. In Ceil Lucas (ed.), Sociolinguistics in Deaf Communities,
255–271. Washington D.C.: Gallaudet University Press.
Metzger, Melanie & Benjamin Bahan. 2001. Discourse analysis. In Ceil Lucas (ed.),
The sociolinguistics of Sign Languages, 112–143. Cambridge: Cambridge
University Press.
Meurant, Laurence. 2004. L’anaphore syntaxique redéfinie au regard d’une langue
des signes. Silexicales (Linguistique de la LSF. Recherches actuelles) 4. 231–
244.
Meurant, Laurence. 2006. De la deixis en langues des signes: le regard du locuteur.
Studia Romanica Tartuensia Iva, 49–66.
Meurant, Laurence. 2007. The Speaker’s eye gaze. Creating deictic, anaphoric and
pseudo-deictic spaces of reference. In Ronice Müller de Quadros (ed.), Sign
Languages: spinning and unraveling the past, present and future (TISLR9),
403–414. Petrópolis/RJ, Brazil: Editora Arara Azul.
Meurant, Laurence. 2008. Role Shift, Anaphora and Discourse Polyphony in Sign
Language of Southern Belgium (LSFB). In Josep Quer (ed.), Signs of the time.
Selected papers from TISLR 8, 319–352. Hamburg: Signum Verlag.
Miller, Christopher. 1994. Simultaneous constructions in Quebec Sign Language. In
Mary Brennan & Graham Turner (eds.), Word-order issues in sign language,
89–112. Durham: Isla.
Miller, George A. 1956. The magical number seven, plus or minus two: some limits on
our capacity for processing information. Psychological Review 63(2). 81–97.
Milsark, Gary. 1974. Existential sentences in English. Cambridge, MA: MIT
dissertation.
Morales-López, Esperanza, Delfina Aliaga-Emeterio, Jesús Amador Alonso-
Rodríguez, Rosa Maria Boldú-Menasanch, Júlia Garrusta-Ribes &Victoria
Gras-Ferrer. 2002. Deaf people in bilingual speaking communities: the case of
Deaf in Barcelona. In Ceil Lucas (ed.), Turn-Taking, fingerspelling, and contact 
in Signed Languages, 107–155. Washington D.C.: Gallaudet University Press.
256 References

Morales-López, Esperanza, Rosa Maria Boldu-Menasanch, Jesus Amador Alonso-

Rodriguez, Victoria Gras-Ferrer & Maria Angeles Rodriguez-Gonzalez. 2005.
The Verbal System of Catalan Sign Language (LSC). Sign Language Studies
5(4). 441–496.
Morgan, Gary. 1996. Spatial Anaphoric Mechanisms in British Sign Language.
In Simon Botley (ed.), Approaches to Discourse Anaphora: Proceedings
of DAARC96, 500–506. Lancaster: University Centre for Computer Corpus
Research on Language.
Morgan, Gary. 1999. Event packaging in British Sign Language Discourse. In
Elisabeth Winston (ed.), Storytelling and Conversation. Discourse in Deaf
Communities, 27–58. Washington D.C.: Gallaudet University Press.
Morgan, Gary & Bencie Woll. 2007. Understanding sign language classifiers through
a polycomponential approach. Lingua 117(7). 1159–1168.
Mosella, Marta. 2012. Les construccions relatives en llengua de signes catalana
(LSC). Barcelona: Universitat de Barcelona dissertation.
Neidle, Carol. 2007. SignStream Annotation: Addendum to Conventions used for
the American Sign Language Linguistic Research Project. American Sign
Language Linguistic Research Project, Report 13. Boston, MA: Boston
University.
Neidle, Carol, Judy Kegl, Dawn MacLaughlin, Ben Bahan & Robert Lee. 2000. The
syntax of American Sign Language. Cambridge, MA: The MIT Press.
Neidle, Carol & Robert Lee. 2006. Syntactic Agreement across Language Modalities:
American Sign Language. In João Costa & Maria Cristina Figueiredo Silva
(eds.), Studies on agreement, 203–222. Amsterdam: John Benjamins.
Neidle, Carol & Joan Nash. 2012. Noun Phrase. In Roland Pfau, Markus Steinbach
& Bencie Woll (eds.), Sign Languages. Handbooks of Linguistics and
Communication Science, 265–291. Berlin: de Gruyter Mouton.
Nespor, Marina & Wendy Sandler. 1999. Prosody in Israeli Sign Language. Language
and Speech 42(2/3). 143–176.
Nevins, Andrew. 2009. Contributions of ASL Agreement to the Agreement/
Cliticization Distinction. Paper Presented at LSA Meeting, Harvard University,
San Francisco.
Nevins, Andrew. 2011. Prospects and challenges for a clitic analysis of (A)SL
agreement. Theoretical Linguistics 37(3). 173–187.
Newport, Elissa & Ted Supalla. 2000. Sign Language Research at the millennium.
In Karen Emmorey & Harlan Lane (eds.), The signs of Language Revisited,
103–114. Mahwah, NJ: Lawrence Erlbaum Associates.
Nicodemus, Brenda. 2009. Prosodic Markers and Utterance Boundaries in American
Sign Language Interpretation. Washington D.C.: Gallaudet University Press.
Nilsson, Anna-Lena. 2004. Form and discourse function of the pointing toward the
chest in Swedish Sign Language. Sign Language and Linguistics 7(1). 3–30.
 References 257

Nilsson, Anna-Lena. 2007. The Non-Dominant Hand in a Swedish Sign Language

Discourse. In Myriam Vermeerbergen, Lorraine Leeson & Onno Crasborn
(eds.), Simultaneity in Signed Languages: Form and Function, 163–185.
Amsterdam, Philadelphia: John Benjamins.
Nilsson, Anna-Lena. 2010. Studies in Swedish Sign Language: Reference, Real Space
Blending, and Interpretation. Stockholm: Stockholm University dissertation.
Nonhebel, Annika, Onno Crasborn & Els van der Kooij. 2004. Sign Language
transcription conventions for the ECHO Project. Ms. Radboud University
Nijmegen.
Nouwen, Rick. 2003. Plural pronominal anaphora in context. LOT Dissertation
Series 84. Utrecht Institute for Linguistics OTS dissertation.
Nunberg, Geoffrey. 1993. Indexicality and Deixis. Linguistics and Philosophy 16.
1–43.
Nyst, Victoria. 2007. A descriptive analysis of Adamorobe Sign Language (Ghana).
LOT Dissertation Series 151. Amsterdam: LOT dissertation.
Nyst, Victoria. 2012. Shared sign languages. In Roland Pfau, Markus Steinbach
& Bencie Woll (eds.), Sign Language. Handbook of Linguistics and
Communication Science, 552–573. Berlin: de Gruyter Mouton.
Padden, Carol. 1988. Interaction of morphology and syntax in American Sign
Language. New York, NY: Garland.
Padden, Carol. 1990. The Relation between Space and Grammar in ASL Verb
Morphology. In Ceil Lucas (ed.), Sign Language Research: Theoretical Issues,
118–132. Washington D.C.: Gallaudet University Press.
Padden, Carol, Irit Meir, Mark Aronoff & Wendy Sandler. 2010. The grammar of
space in two new sign languages. In Diane Brentari (ed.), Sign languages:
A Cambridge language survey, 573–595. Cambridge: Cambridge University
Press.
Parsons, Terence. 1990. Events in the semantics of English: A study in subatomic
semantics. Cambridge, MA: The MIT Press.
Partee, Barbara. 1970. Opacity, Co-reference, and Pronouns. Synthèse 21(3–4).
359–385.
Partee, Barbara. 1978. Bound variables and other anaphors. In David L. Waltz (ed.),
Theoretical Issues In Natural Language Processing 2 (TINLAP-2), 79–85.
Urbana, IL: University of Illinois.
Perelló, Jorge & Juan Frigola. 1998. Lenguaje de signos manuales. Madrid: CIE
Inversiones Editoriales Dossat.
Perniss, Pamela. 2007a. Space and Iconicity in German Sign Language (DGS). MPI
Series in Psycholinguistics 45, Nijmegen: Radboud University dissertation.
Perniss, Pamela. 2007b. Achieving spatial coherence in German Sign Language
narratives: The use of classifiers and perspective. Lingua 117. 1315–1338.
258 References

Perniss, Pamela, Roland Pfau & Markus Steinbach. 2007. Can’t you see the
difference? Sources of variation in sign language structure. In Pamela Perniss,
Roland Pfau & Markus Steinbach (eds.), Visible variation: Cross-linguistic
studies in sign language structure, 1–34. Berlin: de Gruyter Mouton.
Perniss, Pamela & Asli Özyürek. 2008. Representation of action, motion, and
location in signing space: A comparison of German (DGS) and Turkish (TID)
Sign Language narratives. In Josep Quer (ed.), Signs of the time. Selected
papers from TISLR 8, 353–378. Hamburg: Signum Verlag.
Petronio, Karen. 1995. Bare Noun Phrases, Verbs and Quantification in ASL. In
Emon Bach, Eloise Jelinek, Angelika Kratzer & Barbara Partee (eds.),
Quantification  in  Natural  Language  Vol.  2, 603 –618. Dordrecht: Kluwer
Academic Publishers.
Pfau, Roland. 2011. A point well taken: On the typology and diachrony of pointing.
In Gaurav Mathur & Donna Napoli (eds.), Deaf around the world. The impact
of language, 144–163. Oxford: Oxford University Press.
Pfau, Roland & Josep Quer. 2007. On the syntax of negation and modals in German
Sign Language (DGS) and Catalan Sign Language (LSC). In Pamela Perniss,
Roland Pfau & Markus Steinbach (eds.), Visible variation: Comparative
studies on sign language structure, 129–161. Berlin: de Gruyter Mouton.
Pfau, Roland & Josep Quer. 2010. Nonmanuals: their grammatical and prosodic
roles. In Diane Brentari (ed.), Sign languages (Cambridge language surveys),
381–402. Cambridge: Cambridge University Press.
Pfau, Roland & Markus Steinbach. 2006. Modality-independent and modality-
specificaspects of grammaticalization in sign languages. Linguistics in
Potsdam 24. 5–98.
Pfau, Roland, Martin Salzmann & Markus Steinbach. 2010. Sign language agreement
is syntactic: the importance of auxiliaries. Paper presented at TIN dag 2010,
Utrecht, 6 February.
Pfau, Roland, Markus Steinbach & Bencie Woll. 2012. Sign Language. Handbook of
Linguistics and Communication Science. Berlin: de Gruyter Mouton.
Pinkal, Manfred. 1986. Definite Noun Phrases and the Semantics of Discourse.
Proceeding COLING ‘86 Proceedings of the 11th coference on Computational
linguistics. 368–373.
Pizzuto, Elena, Paolo Rossini, Marie-Anne Sallandre & Erin Wilkinson. 2008.
Deixis, anaphora and Highly Iconic Structures: Cross-linguistic evidence
on American (ASL), French (LSF), and Italian (LIS). In Ronice Müller de
Quadros (ed.), Sign Languages: spinning and unraveling the past, present and
future (TISLR9), 475–495. Petrópolis/RJ, Brazil: Editora Arara Azul.
Poesio, Massimo. 2004. The MATE/GNOME Scheme for Anaphoric Annotation,
Revisited. In Proceedings of SIGDIAL, Boston.
Poesio, Massimo & Renata Vieira. 1998. A corpus-based investigation of definite
description use. Computational Linguistics 24(2): 183–216.
 References 259

Poesio, Massimo, Rosemary Stevenson, Barbara Di Eugenio & Janet Hitzeman.

2004. Centering: A parametric theory and its instantiations. Computational
Linguistics 30. 309–364.
Poizner, Howard, Edward Klima & Ursula Bellugi. 1987. What the Hands Reveal
about the Brain. Cambridge, MA: MIT Press/Bradford Books.
Prince, Ellen. 1981. Toward a taxonomy of given-new information. In Peter Cole
(ed.), Radical Pragmatics, 223–256. New York, NY: Academic Press.
Prince, Ellen. 1992. The ZPG letter: Subjects, definiteness, and information-status.
In William C. Mann & Sandra A. Thompson, Discourse description, 295–325.
Amsterdam/Philadelphia: John Benjamins.
Puig-Waldmüller, Estela. 2008. Contracted Preposition-Determiner Forms in
German: Semantics and Pragmatics. Barcelona: Universitat Pompeu Fabra
dissertation.
Quadros, Ronice M. 1999. Phrase structure of Brazilian Sign Language. Porto
Alegre: Pontifícia Universidade Católica do Rio Grande do Sul dissertation.
Quadros, Ronice M. & Diane Lillo-Martin. 2007. Gesture and the acquisition of verb
agreement in sign languages. In Heather Caunt-Nulton, Samantha Kulatilake
& I-Hao Woo (eds.), Proceedings of the 31st Annual Boston University
Conference on Language Development, 520–531. Somerville, MA: Cascadilla
Press.
Quadros, Ronice M. & Josep Quer. 2008. Back to backwards and moving on: on
agreement, auxiliaries and verb classes in sign languages. In Ronice Müller de
Quadros (ed.), Sign Languages: spinning and unraveling the past, present and
future (TISLR9), 530–551. Petrópolis/RJ. Brazil: Editora Arara Azul.
Quer, Josep. 2004. Efectes de la modalitat de les llengües en la teoria sintàctics:
problemes i perspectives. Presentation at Xarxa Temàtica de Gramàtica
Teòrica, Universitat Autònoma de Barcelona.
Quer, Josep. 2005a. Quantifying across language modalities: generalized tripartite
structures in signed languages. Presentation at the I Workshop on Sign
Languages, EHU Vitoria-Gasteiz.
Quer, Josep. 2005b. Context Shift and Indexical Variables in Sign Languages. In
Proceedings of Semantic and Linguistic Theory, SALT XV.
Quer, Josep. 2009. Further explorations into the morphosyntax of agreement in sign
languages. Presentation at CILS, Namur.
Quer, Josep. 2010a. Signed agreement: Putting some more arguments together.
Presentation at TISLR10. Purdue University, Indiana.
Quer, Josep. 2010b. La normalització de les llengües de signes. In Joan Martí &
Josep-Maria Mestres (eds.), Les llengües de signes com a llengües minoritàries:
perspectives lingüístiques, socials i polítiques, 239–255. Barcelona: Institut
d’Estudis Catalans.
260 References

Quer, Josep. 2011a. Reporting and quoting in signed discourse. In Elke Brendel, Jörg
Meibauer & Markus Steinbach (eds.), Understanding Quotation, 277–302.
Berlin/NY: de Gruyter Mouton.
Quer, Josep. 2011b. When agreeing to disagree is not enough: Further arguments
for the linguistic status of sign language agreement. Theoretical Linguistics
37(3–4). 189–196.
Quer, Josep. 2012. Quantificational strategies across language modalities. In Maria
Aloni, Floris Roelofsen, Galit Weidman Sassoon, Katrin Schulz, Vadim
Kimmelman & Matthijs Westera (eds.), Proceedings of 18th Amsterdam
Colloquium, LNCS 7218, 82–91. Berlin, Heidelberg: Springer-Verlag.
Quer, Josep, Eva Ma. Rondoni, Gemma Barberà, Santiago Frigola, Pilar Iglesias,
Josep Boronat, Marina Martínez, Delfina Aliaga & Joan Gil. 2005. Gramàtica
bàsica LSC. CD-ROM. Barcelona: Federació de Persones Sordes de Catalunya.
Quer, Josep & Rosa Maria Boldú. 2006. Lexical and morphological resources in the
expression of sentential negation in Catalan Sign Language (LSC), Actes del
7è Congrés de Lingüística General, Barcelona: Universitat de Barcelona.
Quer, Josep & GRIN. 2008. Possessive and existential structures in Catalan Sign
Language. In Ulrike Zeshan (ed.), Typological Studies on Possessive and
Existential Constructions in Sign Languages, 33–53. Nijmegen: Ishara Press.
Quer, Josep, Laura Mazzoni & Galini Sapountzaki. 2010. Sign Languages in
Mediterranean Europe. In Diane Brentari (ed.), Sign Languages: Cambridge
Language Surveys, 95–112. Cambridge: Cambridge University Press.
Rathmann, Christian. 2000. Event structure in American Sign Language. Austin,
TX: University of Texas at Austin dissertation.
Rathmann, Christian & Gaurav Mathur. 2002. Is verb agreement the same
crossmodally? In Richard Meier, Kearsy Cormier & David Quinto-Pozos
(eds.), Modality and structure in signed and spoken languages, 370–404.
Cambridge: Cambridge University Press.
Reinhart, Tanya. 1981. Pragmatics and linguistics. An analysis of sentence topics.
Philosophica 27. 53–94.
Recanati, François. 2005. Deixis and anaphora. In Zoltán Gendler Szabó (ed.),
Semantics vs. Pragmatics, 286–316. Oxford: Clarendon Press.
Ribera, Eulàlia. 2007. Projecte de base de dades lexicogràfica per a la LSC (llengua 
de signes catalana). Barcelona: Universitat Autònoma de Barcelona MA
thesis.
Rinfret, Julie. 2009. L’association spatiale du nom en langue des signes québécoise:
formes, fonctions et sens. Montréal: UQAM dissertation.
Roberts, Craige. 1989. Modal subordination and pronominal anaphora in discourse.
Linguistics and Philosophy 12. 683–721.
Roberts, Craige. 1990. Modal subordination, Anaphora, and Distributivity. New
York, NY: Garland Press.
 References 261

Roberts, Craige. 1998. The place of Centering in a general theory of anaphora

resolution. In Marilyn Walker, Aravind Joshi & Ellen Prince (eds.), Centering
Theory in Discourse, 359–399. Oxford: Oxford University Press.
Roberts, Craige. 2002. Demonstratives as definites. In Kees van Deemter & Roger
Kibble (eds.), Information Sharing: Reference and Presupposition in Language
Generation and Interpretation, 89–196. Stanford: CSLI Press.
Roberts, Craige. 2003. Uniqueness in definite noun phrases. Linguistics and
Philosophy 26. 287–350.
Roberts, Craige. 2005. Pronouns as definites. In Marga Reimer & Anne Bezuidenhout
(eds.), Descriptions and Beyond, 503–543. New York, NY: Oxford University
Press.
Rosselló, Joana. 2008. Some arguments against some prevalent ideas on
specificational sentences. Journal of Portuguese Linguistics 7(2). 109–130.
Roy, Cynthia. 1989. Features of Discourse in an American Sign Language Lecture.
In C. Lucas (ed.), The sociolinguistics of the Deaf Community, 231–251. San
Diego, CA: Academic Press.
Russell, Bertrand. 1905. On denoting. Mind 14. 479–493.
Russell, Kevin & Terry Janzen. 2008. The categorical nature of ASL pronoun
locations. In Ronice Müller de Quadros (ed.), Sign Languages: spinning
and unraveling the past, present and future (TISLR9). Petrópolis/RJ, Brazil:
Editora Arara Azul.
Sandler, Wendy. 1989. Phonological Representation of the Sign: Linearity and
Nonlinearity in American Sign Language. Dordrecht: Foris.
Sandler, Wendy & Diane Lillo-Martin. 2006. Sign Language & Linguistic Universals.
Cambridge: Cambridge University Press.
Sandler, Wendy, Irit Meir, Carol Padden & Mark Aronoff. 2005. The emergence of
grammar: Systematic structure in a new language. PNAS 102(7). 2661–2665.
Schembri, Adam & Onno Crasborn. 2010. Issues in creating annotation standards
for sign language description. In Proceedings of the Seventh International
Conference on Language Resources and Evaluation, 212–216. Paris: European
Language Resources Association.
Schlenker, Philippe. 2011a. Donkey anaphora: the view from sign language (ASL
and LSF). Linguistics and Philosophy 34(4). 341–395.
Schlenker, Philippe. 2011b. Quantifiers and Variables: Insights from Sign Language
(ASL and LSF). In Barbara Partee, Michael Glanzberg & Jurgis Skilters (eds.),
Formal Semantics and Pragmatics: Discourse, Context, and Models. The
Baltic International Yearbook of Cognition, Logic and Communication, Vol.
6, 1–22. Manhattan, KS: New Prairie Press.
Schlenker, Philippe. 2011c. Iconic agreement. Theoretical Linguistics 37(3–4).
223–234.
262 References

Schlenker, Philippe & Jonathan Lamberton. 2012. Formal Indices and Iconicity in
ASL. In Maria Aloni, Floris Roelofsen, Galit Weidman Sassoon, Katrin Schulz,
Vadim Kimmelman & Matthijs Westera (eds.), Proceedings of 18th Amsterdam
Colloquium 2011, LNCS 7218, 1–11. Berlin, Heidelberg: Springer-Verlag.
Schlenker, Philippe, Jonathan Lamberton & Mirko Santoro. 2013. Iconic variables.
Linguistics and philosophy 36(2). 91–149.
Scholten, Jolien & Ana Aguilar-Guevara. 2010. Assessing the discourse referential
properties of weak definite NPs. In Rick Nouwen & Jacqueline van Kampen
(eds.), Linguistics in the Netherlands 27, 115–128. Amsterdam: John
Benjamins.
Schulpen, Maartje. 2011. Weak  definites:  Modification,  non-unique  reference  and 
enriched meanings. Utrecht: Universiteit Utrecht MA thesis.
Schwarz, Florian. 2009. Two  Types  of  Definites  in  Natural  Language. Amherst:
University of Massachusetts dissertation.
Segimon, Josep M. & Pilar Fernández-Viader. 2000. Animals. Barcelona: Fundació
privada Illescat.
Senghas, Ann & Marie Coppola. 2001. Children creating language: How Nicaraguan
Sign Language acquired a spatial grammar. Psychological Science 12(4). 323–
328.
Shepard-Kegl, Judy. 1985. Locative relations in ASL word formation, syntax and
discourse. Cambridge, MA: MIT dissertation.
Smith, Carlota. 2003. Modes of discourse. The Local Structure of Texts. Cambridge:
Cambridge University Press.
Stalnaker, Robert. 1974. Pragmatic Presuppositions. In Milton K. Munitz & Peter
K. Unger (eds.), Semantics and Philosophy: Essays, 197–213. New York, NY:
New York University Press.
Stalnaker, Robert. 1978. Assertion. In Peter Cole (ed.), Syntax and Semantics 9:
Pragmatics, 315–332. New York, NY: Academic Press.
Stalnaker, Robert. 1998. On the Representation of Context. Journal of Logic,
Language, and Information 7. 3–19.
Steinbach, Markus & Roland Pfau. 2007. Grammaticalization of auxiliaries in sign
languages. In Pamela Perniss, Roland Pfau & Markus Steinbach (eds.), Visible
variation: Comparative studies on sign language structure, 303–339. Berlin:
de Gruyter Mouton.
Stokoe, William. 1960. Sign language structure. An outline of the visual
communication systems of the American Deaf. Silver Spring, MD: Linstok
Press.
Strawson, Peter. 1950. On referring. Mind 59. 320–344.
 References 263

Suppalla, Ted. 1986. The Classifier System in American Sign Language. In Colette
Craig (ed.), Typological Studies in Language: Noun Classes and Categorization,
181–214. Philadelphia, PA: John Benjamins Publishing Company.
Swabey, Laurie. 2002. The Cognitive Status, Form and Distribution of Referring
Expressions in ASL and English Narratives. Minnesota, MN: University of
Minnesota dissertation.
Talmy, Leonard. 1985. Lexicalization Patterns: Semantic Structure in Lexical Forms.
In Timothy Shopen (ed.), Language Typology and Syntactic Description, Vol.
3, 57–149. Cambridge: Cambridge University Press.
Tang, Gladys & Felix Sze. 2002. Nominal expressions in Hong Kong Sign Language:
Does modality make a difference. In Richard Meier, Kearsy Cormier & David
Quinto-Pozos (eds.), Modality and Structure in Signed and Spoken Languages,
296–320. Cambridge: Cambridge University Press.
Tervoort S. J. 1953. Structurele analyse van visueel taalgebruik binnen een groep
dove kinderen. Amsterdam: Universiteit van Amsterdam dissertation.
Thompson, Robin, Karen Emmorey & Robert Kluender. 2006. The relationship
between eye gaze and verb agreement in American Sign Language: an Eye-
Tracking Study. Natural Language and Linguistic Theory 24. 571–604.
Turan, Ümit Deniz. 1995. Null vs. Overt Subjects in Turkish Discourse: A Centering
Analysis. Pennsylvania, PA: University of Pennsylvania dissertation.
Vallduví, Enric. 1992. The Informational Component. New York, NY: Garland Press.
Vallduví, Enric. 1997. Teoria de dinàmica de centres: les relacions anàforiques locals.
In Rudolf Ortega (ed.), Actes del Segon Taller de Ciències del Llenguatge
(Report GRLMC n/97), Tarragona: Grup de Recerca en Lingüística Matemàtica
i Enginyeria del Llenguatge (URV).
Vallduví, Enric. 2002. L’oració com a unitat informativa. In Joan Solà (ed.),
Gramàtica del català contemporani, 1221–1230. Barcelona: Empúries.
Vallduví, Enric & Maria Vilkuna. 1998. On Rheme and Kontrast. In The Limits of
Syntax, Peter Culicover & Louise McNally (eds.), 79–106. New York, NY:
Academic Press.
De Vriendt, Sera & Max Rasquinet. 1990. The Expression of Genericity in Sign
Language. In Siegmund Prillwitz & Thomas. Vollhaber (eds.), Current Trends
in European Sign Language Research: Proceedings of the 3rd European
Congress on Sign Language Research, 249–255. Hamburg: Signum Verlag.
Vogt-Svendsen, Marit & Brita Bergman. 2007. Point Buoys: The Weak Hand as a
Point of Reference for Time and Space. In Myriam Vermeerbergen, Lorraine
Leeson & Onno Crasborn (eds.), Simultaneity in Signed Languages: Form and
Function, 217–235. Amsterdam/Philadelphia: John Benjamins.
264 References

De Vos, Connie. 2012. Sign-spatiality in Kata Kolok: How a village sign language
in Bali inscribes its signing space. Nijmegen: Radboud University Nijmegen
dissertation.
Walker, Marilyn. 1989. Evaluating discourse processing algorithms. In Association
for Computational Linguistics (eds.) Proceedings of the 27th annual meeting
on Association for Computational Linguistics, 251–261. Vancouver, British
Columbia, Canada.
Walker, Marilyn, Masayo Iida & Sharon Cote. 1994. Japanese discourse and the
process of centering. Computational Linguistics, 20(2). 193–232.
Ward, Gregory & Betty Birner. 2001. Discourse and Information Structure. In
Deborah Schiffrin, Deborah Tannen & Heidi Hamilton (eds.), The Handbook
of Discourse Analysis, 119–137. Malden, MA: Blackwell Publishing.
Webber, Bonnie. 1979. A Formal Approach to Discourse Anaphora. New York, NY:
Garland Press.
Wilbur, Ronnie. 1979. American Sign Language and sign systems: research and
application. Baltimore: University Park Press.
Wilbur, Ronnie. 1994a. Eyeblinks and ASL Phrase Structure. Sign Language Studies
84. 221–240.
Wilbur, Ronnie. 1994b. Foregrounding Structures in American Sign Language.
Journal of Pragmatics 22. 647–672.
Wilbur, Ronnie. 1995. Why so-called “Rhetorical Questions” are neither Rhetorical
nor Questions. In Helen Bos & Trude Schermer (eds.), Sign Language
Research 1994: Proceedings of the 4th European Congress on Sign Language
Research, 149–169. Hamburg: Signum.
Wilbur, Ronnie. 1997. A Prosodic/Pragmatic Explanation for Word Order Variation
in ASL with Typological Implications. In Marjolijn Verspoor, Kee Dong
Lee & Eve Sweetser (eds.), Lexical and syntactical constructions and the
construction of meaning, 89–104. Amsterdam/Philadelphia: John Benjamins.
Wilbur, Ronnie. 2000. Phonological and Prosodic Layering of Nonmanuals in
American Sign Language. In Karen Emmorey & Harlan Lane (eds.), The Signs
of Language Revisited, 213–241. Mahwah, NJ: Lawrence Erlbaum Associates.
Wilbur, Ronnie. 2008. Complex predicates involving events, time and aspect: Is this
why sign languages look so similar? In Josep Quer (ed.), Signs of the time.
Selected papers from TISLR 8, 217–250. Hamburg: Signum Verlag.
Wilbur, Ronnie. 2011. Nonmanuals, semantic operators, domain marking, and the
solution to two outstanding puzzles in ASL. Sign Language & Linguistics
14(1). 148–178.
Wilbur, Ronnie. 2012. Information structure. In Roland Pfau, Markus Steinbach
& Bencie Woll (eds.), Sign Languages (Handbooks of Linguistics and
Communication Science, HSK), 462–488. Berlin: de Gruyter Mouton.
 References 265

Wilbur, Ronnie & Cyntia Patschke. 1998. Body leans and the marking of contrast in
American Sign Language. Journal of Pragmatics 30. 275–303.
Wilbur, Ronnie & Cyntia Patschke. 1999. Syntactic Correlates of Brow Rase in
ASL. Sign Language & Linguistics 2(1). 3–41.
Wilson, Julie. 1996. The Tobacco Story: Narrative Structure in an American Sign
Language Story. In Ceil Lucas (ed.), Multicultural Aspects of Sociolinguistics
in Deaf Communities, 152–182. Washington D.C.: Gallaudet University Press.
Winston, Elisabeth. 1995. Spatial Mapping in Comparative Discourse Frames. In
Karen Emmorey & Judy Reilly (eds.), Language, Gesture and Space, 87–114.
Mahwah, NJ: Lawrence Erlbaum Associates.
Zamparelli, Roberto. 1995. Layers in the Determiner Phrase. Rochester: University
of Rochester dissertation.
Zeshan, Ulrike. 2000. Sign Language in Indo-Pakistan: A description of a signed
language. Amsterdam: John Benjamins.
Zeshan, Ulrike. 2010. The significance of village sign languages for the typological
study of sign languages. Paper Presented at Theoretical Issues in Sign
Language Research (TISLR 10). West-Lafayette, 30 September–2 October.
Zimmer, June & Cynthia Patschke. 1990. A Class of Determiners in ASL. In
Ceil Lucas (ed.), Sign Language Research: Theoretical Issues, 201–210.
Washington D.C.: Gallaudet University Press.
Zucchi, Sandro. 2004. Monsters in the Visual Mode? Milano: Università degli studi
di Milano.
Zucchi, Sandro. 2011. Event descriptions and classifier predicates in sign languages.
Paper Presented at FEAST, Venice, Università ca’Foscari.
Zucchi, Sandro. 2012. Formal semantics of Sign Languages. Language and
Linguistics Compass 6(11). 719–734.
Zwicky, Arnold M. & Geoffrey K. Pullum. 1983. Cliticization vs. inflection: the case
of English n’t. Language 59(3). 502–513.
Zwitserlood, Inge. 2003. Classifying hand configurations in Nederlandse Gebarentaal 
(Sign Language of the Netherlands). Utrecht: UiL OTS, Universiteit Utrecht
dissertation.
Zwitserlood, Inge & Ingeborg van Gijn. 2006. Agreement Phenomena in Sign
Language of the Netherlands. In Peter Ackema, Patrick Brandt, Maaike
Schoorlemmer & Fred Weermann (eds.), Arguments and Agreement, 195–229.
Oxford: Oxford University Press.
Zwitserlood, Inge, Asli Özyürek & Pamela Perniss. 2008. Annotation of sign
and gesture cross-linguistically, In Onno Crasborn, Thomas Hanke, Eleni
Efthimiou, Inge Zwitserlood & Ernst Thoutenhoofd (eds.), Construction and
Exploitation of Sign Language Corpora. 3rd Workshop on the Representation
and Processing of Sign Languages, 185–190. Paris: European Language
Resources Association.
Index
Adamorobe Sign Language Common ground 88–90, 97, 123, 126,
(AdaSL) 20 131–132, 134, 140, 142, 144, 146,
Al-Sayyid Bedouin Sign Language 148, 154–155, 189–190, 198, 228
(ABSL) 45–46 Coreference 9, 17, 53, 58, 65, 94–95,
Ambiguity 36, 96, 150–151, 153, 160, 104, 107–108, 128, 150, 167, 187,
181, 184, 186, 189, 204, 207–208, 191, 196, 216, 219, 227
220 Danish Sign Language (DSL) 28, 55,
American Sign Language (ASL) 17, 69, 135–136
27, 55, 70, 156, 183 Definiteness 6, 10–11, 38, 40–41,
Anaphora 11, 28, 31, 54, 63, 66–67, 122–125, 129–131, 133–136, 141,
79, 82, 90, 92–94, 96, 101, 108, 114, 146–148, 150, 154, 156, 158, 169,
123–124, 131–133, 136, 138, 141, 190–193, 201, 224
148, 155, 186, 195–197, 199–202, definite 11, 38–40, 70–71, 75, 89,
216, 218, 221, 227 93, 96–97, 107, 111, 118, 123–
Assertion 11, 88, 110–111, 123–124, 128, 130–132, 134–136, 139,
126, 130, 136, 141–145, 148, 181, 141, 144, 147–151, 155, 169–
191–192 170, 190, 192, 197, 199, 201
Australian Sign Language (Auslan) 5, Deixis 2, 11, 31, 45–46, 64, 79–80,
9 96, 101–102, 119, 123, 131–134,
136, 138, 140–141, 148, 201, 203,
227–228
Body-anchored location 84–85, 142
Descriptive localisation 22–23, 27
British Sign Language (BSL) 5, 17,
Determiner 38, 40, 77, 104, 112, 134–
27–28
135, 139, 143, 150, 156, 161–162,
180–186, 192–193, 195
Catalan Sign Language (LSC) 1–7, Discourse 1–5, 7, 9–14, 16–18, 22–23,
9–12, 15–17, 23–24, 26, 32, 35, 37, 25–26, 28–29, 35–38, 43–45, 47,
40, 47, 49–50, 52–57, 59–62, 64–78, 49–50, 52–53, 60–61, 63–69, 71,
80–82, 85–87, 90, 99, 101–106, 73, 79–80, 82–84, 86–108, 113,
108–112, 114, 116–124, 127, 131, 115–117, 121–129, 131–133, 136,
133–136, 141–143, 145–149, 157– 138–139, 141–142, 144–146, 148–
164, 166, 168–174, 177–178, 180– 158, 161, 167–168, 175, 186, 188,
186, 189–195, 200–203, 207–208, 190–219, 221–228
210–211, 215–218, 223–229
discourse model 35–37, 88–89, 94,
Classifier 4, 9, 24, 26–27, 84, 115, 142, 97, 99–100, 103–104, 124, 132,
216–217 138–139, 141, 144, 148, 200–
Clitic 47, 49–52, 86, 180, 216 202, 204–208, 216, 227–228
 Index 267

Discourse referent (DR) 9, 11–12, 16, Eye gaze 10, 40, 42, 53–54, 57, 59–60,
23, 43, 52–55, 59, 63–64, 66, 68, 64–65, 76, 83, 85, 112, 134–135,
84–87, 90–93, 96–101, 103–110, 157, 160–163, 165–166, 168, 173,
118, 120–121, 123–127, 129–137, 175–178, 187, 189, 205, 229
139–148, 150, 152, 154–158, 160–
169, 171, 175–177, 181–184, 187–
Familiarity 123–134, 136, 138, 140–
188, 190–191, 193–200, 202–213,
141, 144–148, 174, 196–198
215–229
Frame of reference 25, 69, 73, 137,
Discourse Representation Structure
209, 219
(DRS) 2, 11–12, 63–64, 66–67, 69,
71, 78, 83, 86–88, 90–97, 99–114, French Sign Language (LSF) 6, 36
117, 119–126, 128–129, 132–152,
155–160, 163–164, 166–169, 171– Genericity 11, 40, 85–86, 96, 108,
172, 178–179, 182–184, 188–195, 117–121, 171
197–199, 201–215, 217–221, 223,
225–228 German Sign Language (DGS) 5, 24,
55
Discourse Representation Theory
(DRT) 2, 11, 90–91, 93–95, 101– Gesture 1, 13, 18–19, 21, 33, 39, 41,
103, 108, 110–111, 121, 136, 140, 46–47, 102, 132, 207–208
151, 172, 190–191, 210–211, 225,
227–228 Hong Kong Sign Language (HKSL) 40,
Discourse structure 10, 102, 156, 190, 134–135, 141, 148, 156–157, 192
193–196, 198, 201, 209–210, 213,
217, 224, 226, 228 Identifiability 78, 120, 125, 146, 150–
Distributivity 114, 117 151, 154–156, 164–168, 172, 175,
Donkey sentence 90, 93, 112 177–179, 184, 193, 229
Indefiniteness 11, 40, 75, 124, 134–
135, 141, 146–147, 150, 156
Elicitation 4, 6–8, 57, 178, 228–229
indefinite 38–40, 71, 75, 90, 93–94,
Entity 9, 11, 15–17, 20–24, 29–30, 33–
96–97, 107–108, 118, 123–124,
35, 38, 40–41, 43–44, 46, 50, 56–57,
126, 134–136, 141–143, 146–
59, 63–66, 68, 71–72, 74–75, 77–78,
154, 156, 160–162, 164–166,
80, 82, 84–87, 89–90, 94, 96–100,
171, 175–176, 189–190, 192,
102, 106, 116, 118–119, 122–125,
197, 201–202
129, 132–133, 136, 138–142, 144,
146, 148, 150, 154, 167, 175, 179–
182, 186, 189–192, 195, 197–203,
205–207, 210–211, 215, 218, 220–
222, 229
268 Index

Index sign 33–37, 40, 43, 46, 52–54, Localisation 6, 10–12, 16–17, 22–26,
61, 64–65, 69, 71–72, 78–79, 104, 28, 30, 32, 41, 43, 47, 50–54, 56–57,
106, 112, 118–119, 134, 137, 144– 59–79, 81, 83–86, 104, 109–112,
145, 156–157, 160, 163, 205, 207, 115, 118, 120, 122, 130, 135–136,
216, 221–222, 228 138, 141–149, 157, 159–160, 162–
pointing 13–14, 20–21, 33–35, 37– 163, 165–166, 168–183, 185–186,
41, 45, 47–48, 50–53, 64, 67, 78, 188, 190–194, 202, 204, 209–210,
80, 103, 133–134, 137, 156, 207 219–223, 226, 228–229
Indo-Pakistani Sign Language descriptive localisation 23
(IPSL) 55, 70 non-descriptive localisation 25
Infinity 19, 28, 30–31, 36–37, 39, 41– Location 4, 9–12, 14–16, 18, 22–25,
42, 52, 203–207, 216 27–28, 30, 35–38, 40–41, 43–45,
Information structure 59, 100, 147 49–50, 52–54, 57–60, 63–68, 71–77,
80, 82, 84–86, 103–106, 109, 112–
Interpretation 1, 6, 9, 17, 21, 25, 33– 117, 121–123, 131, 135–136, 142,
35, 51, 68, 76–77, 83, 85, 89–91, 144–146, 149, 157–163, 166, 168,
103, 115–116, 118, 125–127, 131, 170–172, 175–178, 182, 186–190,
137–138, 149–154, 156–157, 162, 192–195, 203–208, 210–211, 215–
166, 172–178, 180–181, 183–184, 222, 224–227, 229
188, 197–198, 204, 208, 221–222
body-anchored location 84–85, 142
reading 85, 91, 96–97, 101, 120,
125, 141, 146, 150–153, 155, spatial location 4, 11–12, 16, 25,
160, 162, 169, 180–181, 184, 35–36, 40–41, 44, 50, 57, 63, 65–
186, 189 67, 71–73, 82, 85–86, 103–104,
106, 109, 112, 115–117, 121–
semantic interpretation 9, 178 123, 135–136, 142, 144–146,
Irish Sign Language (ISL) 5, 36, 56 157–161, 166, 168, 171–172,
Israeli Sign Language (ISL) 5, 36, 56 175–178, 182, 186–188, 192–
195, 203, 206, 208, 210–211,
Italian Sign Language (LIS) 40, 157
215–217, 219–220, 222, 224–
227, 229
Kata Kolok (KK) 20, 25 Locative 20, 39, 46, 58–59, 72–74, 78,
Kind 119–120 82
Locus, loci 16, 29–30, 36, 54

Meaning 14, 18, 32, 38, 45, 56, 64,

77–78, 82–83, 86–87, 89, 147, 173,
208–209, 216–217, 220–222, 225
 Index 269

Modality 1, 7, 11, 13, 18–22, 47, 154 Partitivity 151, 153–154, 156, 162,
audio-vocal modality 1, 19, 21 164, 172–173, 179, 183–186, 193,
229
visual-spatial modality 1, 13, 22, 47
Phonology 1–2, 4–6, 13–14, 18–19,
Modal subordination 114, 155, 167,
25–26, 28, 37, 50, 69, 76–77, 83–84,
171, 173, 175, 183, 186, 188–191,
178, 200
193, 201
Plane 11–12, 14, 16, 18, 23, 25–26, 28,
Morpheme 11, 39, 47, 49–52, 54, 82–
46, 49, 61–81, 83, 86, 105–106, 114,
84, 86, 103, 179, 182–183, 205, 225
122, 134–135, 145–146, 149, 157–
Morphology 5, 17, 19–20, 36, 39, 46, 159, 162–164, 166, 168–193, 195,
51, 71, 99, 115–117, 150, 153, 205 209, 215–218, 221–227, 229
frontal plane 11–12, 14, 49, 64, 68–
Nicaraguan Sign Language (ISN) 45, 78, 80–81, 83, 86, 122, 134–135,
95 146, 149, 157–159, 162–164,
Non-descriptive localisation 23, 25, 166, 168–190, 192–193, 215,
175 226, 229
Nonmanual 10, 14, 42–43, 50, 54, 58– horizontal plane 11–12, 16, 18, 23,
60, 75–76, 85, 110, 124, 134–135, 25, 28, 46, 49, 62–68, 72, 78, 80,
146–147, 156–157, 160, 165, 173– 83, 86, 105–106, 114, 145, 173,
178, 189, 205–206 195, 209, 215–218, 221–224, 227
Non-specificity 12, 40, 49, 71, 74–75, midsaggital plane 11, 64, 78–80,
77–78, 82–83, 86, 96, 108, 114, 135, 193
148–164, 166, 168, 171–177, 179– spatial plane 61
193, 195, 226, 229 vertical plane 26
Noteworthiness 155–156, 186, 190, Pragmatics 3, 10, 92, 98, 101, 105–
193–196, 201–202, 215 106, 115, 123–126, 132, 197, 229
Noun Phrase (NP) 9, 28, 40, 52–54, Presupposition 89, 123, 127, 129, 132–
59, 64, 68, 72, 74, 76–78, 85, 89–94, 133, 136, 141, 144–146, 154, 174,
96–97, 104, 107–112, 115–116, 118, 181–182, 192, 197
120–121, 125–126, 128–130, 132,
Prominence 12, 35, 42, 51, 72, 100–
139, 142–144, 146, 150–155, 160–
102, 106, 125, 128, 132–133, 140,
163, 165–166, 168, 176, 181, 192,
154, 194–202, 208–224, 226–229
196–197, 201–202, 212, 219, 228

Operator 94–95, 107–109, 113–114,

119–122, 126, 148, 151–152, 156–
158, 162, 167, 171, 186, 188–189,
191–193, 199, 211, 214
270 Index

Pronoun 2, 9, 23, 31, 34–35, 39, 42– Semantics 1–3, 9–11, 13, 27, 39, 41,
43, 47, 49–51, 53, 56, 61, 64–67, 43, 47–48, 58, 65, 69, 82, 87, 89–92,
72, 77, 89–90, 92–94, 96, 100–104, 94–95, 100–105, 108–109, 111–113,
107–108, 113, 116, 121, 131–133, 117, 120–121, 124, 126, 130, 136,
137–140, 144, 148, 150, 152–153, 140, 142, 151, 156, 166, 172, 178,
160, 166–167, 169, 171, 186–187, 184, 188, 193, 196, 204–205, 207,
191, 195–200, 202–208, 212, 214, 210, 212–215, 220, 225, 227–229
216, 218–221, 227 Sign Language of the Netherlands
(NGT) 5, 18, 39, 55
Quantification 93–94, 114–115, 117, Sign Language (SL) 1–6, 9–10, 15, 36,
125, 153–154, 162, 178, 184, 201 40–41, 45, 49–50, 52–53, 55–57, 59,
Quebec Sign Language (LSQ) 57, 60, 69–70, 87, 101–102, 123, 134–136,
135 149, 156–157, 183, 195, 203–205,
210

Reference 4, 11, 13, 15–17, 19, 23, 25, Space 1–2, 6, 9–38, 40–55, 57–61, 63,
28–31, 33, 35–36, 38, 40–43, 45, 65–67, 69–70, 72–73, 78, 80, 82–83,
47, 52, 54, 58, 63–64, 69, 72–73, 85–87, 102–104, 106, 109–112,
81, 84–85, 89, 92, 94–97, 100, 104, 114–123, 134–138, 141–149, 156–
107–108, 118–121, 123, 125, 130– 161, 166, 168, 171, 173, 175–177,
133, 135–141, 152, 155, 157, 168– 180, 182, 187, 193–195, 202–207,
170, 196–199, 203–204, 207–210, 209–210, 215–219, 221–225, 227–
218–219, 228–229 229

Referential status 100 real space 29–31, 37–38, 46, 138

Role shift 4, 10, 30, 65, 84, 142, 210, shared space 24–25
223 signing space 1–2, 6, 9–11, 13–14,
16–19, 23–25, 27–28, 31, 37–38,
40–47, 49, 51–54, 57, 59–61, 63,
Scope 122, 151, 169, 192 65–67, 69, 72–73, 78, 82–83, 85–
narrow scope 11–12, 107–109, 87, 102–104, 106, 111–112, 114,
113–114, 117, 121–122, 149, 116–117, 119, 123, 134, 136–138,
151, 155–156, 158–159, 161– 141–144, 147, 149, 157, 159, 166,
162, 164, 167–169, 171–172, 171, 173, 175–176, 182, 187,
179, 184, 186, 189, 191–195, 193–195, 202, 204–205, 215,
199, 201, 213–215, 225–226, 229 217–219, 222, 224–225, 227–229
wide scope 11–12, 87, 107–109, syntactic space 27
121–122, 151, 155–160, 162, topographic space 27
164, 166–168, 172, 179, 184,
186, 192–193, 195, 199, 201, use of space 11, 13, 18–19, 21–23,
209, 211, 215, 225–226, 229 27, 30, 46–47, 49, 57, 60, 73, 80,
114, 203
 Index 271

Spanish Sign Language (LSE) 3, 28 Topicality 40–41, 47, 136, 194, 202,
Spatial location 4, 11–12, 16, 25, 210–211
35–36, 40–41, 44, 50, 57, 63, 65–
67, 71–73, 82, 85–86, 103–104, Underspecification 52, 92, 101, 105,
106, 109, 112, 115–117, 121–123, 195–198, 206, 216–217
135–136, 142, 144–146, 157–161,
Utterance 9–10, 53, 107, 125, 128,
166, 168, 171–172, 175–178, 182,
131–132, 137, 142, 166, 186, 199,
186–188, 192–195, 203, 206, 208,
202, 209, 211, 219, 221–222
210–211, 215–217, 219–220, 222,
224–227, 229
Spatial modification 51, 54–55, 57, 60, Variable 11–12, 87, 91–94, 97, 100,
85, 181 103–105, 108–109, 111–112, 114,
117, 120–123, 125, 139–141, 151–
Specificity 1, 6, 10, 12, 40–41, 47, 152, 155–156, 160, 162–164, 166–
50, 68, 70, 74–75, 77–78, 122, 136, 168, 171–172, 184, 188, 190–193,
148–151, 153–160, 168–169, 172– 195, 198–199, 211–215, 222, 224–
174, 178–179, 181–184, 186, 190– 226, 229
194, 201, 224–226, 229
Verb 9, 15–17, 20–23, 27, 29–32, 42,
Spoken language 1, 7, 9, 18, 36, 42, 68, 44, 49–51, 54, 57–60, 73, 76, 83–85,
101, 103, 208, 216 91, 107, 112–117, 143, 160–161,
Swedish Sign Language (SSL) 5, 55, 165, 175, 179–181, 183, 187, 200,
204, 210 216, 220–221, 223
Syntax 5, 13, 18–19, 22–23, 27, 36, agreement verb 16, 31, 115–116,
38, 44–45, 48, 50, 58, 91–93, 96, 99, 165, 187, 221, 223
109–110, 139, 147, 150–151, 154, plain verb 54, 221
196–197, 212, 229

Topic 4, 9–10, 33, 39, 89, 98–100, 122,

136, 148, 168, 188, 200, 202–203,
211, 215, 221, 228–229
contrastive topic 221
s-topic 98–100