On the Development of the Aspect System in Some Philippine Languages

Lawrence A. Reid

Oceanic Linguistics, Vol. 31, No. 1. (Summer, 1992), pp. 65-91.

Stable URL:
http://links.jstor.org/sici?sici=0029-8115%28199222%2931%3A1%3C65%3AOTDOTA%3E2.0.CO%3B2-J

Oceanic Linguistics is currently published by University of Hawai'i Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/uhp.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org
Sat Feb 16 23:50:08 2008
 ON THE DEVELOPMENT OF
THE ASPECT SYSTEM IN SOME
PHILIPPINE LANGUAGES 1
LAWRENCE A. REID
UNIVERSITY OF HAWAII

The verbal morphology of Philippine languages, typically complex, usu-

ally includes distinctions of voice, mode, and aspect that are problematic
with respect to whether they are inflectional or derivational. This paper
attempts to give an account of the development of the aspect systems of
Tagalog and Ilokano. Their voice systems, closely intertwined with the
aspect systems, are also considered. The aspect system posited for Proto-
Extra Formosan closely resembles that of Ilokano. The Ilokano system
resulted from a switch in the order of the two infixes <urn) and <in),
whereby the <in) completive aspect infix came to precede rather than
follow the <urn) focus marker, and from consequent phonological
changes. The development of the Tagalog system, typical of the Central
Philippine languages, was considerably more complex, involving a series
of phonological changes that in turn set the stage for a series of re-
structurings, all of which are retraced in detail in the paper.

1. INTRODUCTION. Numerous descriptions of the verbal morphology

of Philippine languages exist in the literature, frequently drawing attention
to the considerable complexity of these systems.? In addition to a wide
range of forms that are clearly derivational, there are typically three sets of
formal distinctions that affect almost every verb and are problematic with
respect to whether they are inflectional or derivational: those marking
voice (sometimes called focus), mode (sometimes called mood), and as-
pect. 3 The voice distinctions are given this label because they have typi-
cally been analyzed as signaling the case relation of the subject. 4 The
modes that are distinguished include indicative (or general), potential (or
abilitative), and imperative (dependent, or irrealis), while the aspects mark
the verb for features of frequency, duration, completion, and so forth."
A comparison of these systems in Tagalog and Ilokano shows that
Tagalog has a morphologically simpler voice system than Ilokano, but it
has a considerably more complex aspect system." The purpose of this

Oceanic Linguistics, Volume 31,no. 1 (Summer 1992)

© by University of Hawaii Press. All rights reserved.
66 OCEANIC LINGUISTICS, VOL. 3], NO. I

paper is to provide an account of the development of the aspect systems

in these two languages. Modal affixation will not enter into the discussion.
All forms are cited in the indicative or general mode of the languages being
compared. However, because the voice and aspect systems are closely
intertwined, especially in Tagalog, it is necessary to include some discus-
sion of the voice systems as well. Although Ilokano is seen as having
maintained, with relatively few changes, a system that is reconstructible to
Proto-Extra Formosan," Tagalog is shown as having undergone consid-
erable restructuring of its system. The main cause of this complexity is
found in a single phonological change that appears to have taken place at
some time early in the development of the Central Philippine languages,
the subgroup of which Tagalog is a member (Zorc 1977).

2. THE TAGALOG SYSTEM. Finite verbs in Tagalog generally have

three forms marking aspect; these forms are described by Schachter and
Otanes (1972) as contemplated, imperfective, and perfective." There is, in
addition, a set of non-finite forms, unmarked for aspect. De Guzman
(1978: 140-144) uses the features finite [fin], begun [beg] and completed
[comp] (adopted from Otanes 1966:54,1970:36-37) to capture the rele-
vant categories.
Table 1 (De Guzman 1978, slightly modified) illustrates the intersection

TABLE 1. TAGALOG ASPECT PARADIGMS

[+V]

[ -fin] [+ fin]

[- beg]
~ [ + beg]

[ -comp]
.>-. [+comp]

(a) kumuha ku.kuha kumu.kuha kumuha

'to take' 'will take' 'takes' 'took'

(b) magbigay magbi: bigay nagbi.bigay nagbigay

'to give' 'will give' 'gives' 'gave'

(c) pulutin pu.pulutin pinu.pulot pinulot

'to pick up' 'will pick up' 'picks up' 'picked up'
(d) punasan pu.punasan pinu.punasan pinunasan
'to mop' 'will mop' 'mops' 'mopped'
(e) Pitapon Pita.tapon Pitina.tapon Pitinapon
'to throw' 'will throw' 'throws' 'threw'
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 67

of the aspect categories in Tagalog with voice affixation. The forms in rows
(a) and (b) are marked for agentive voice, while those in rows (c) through
(e) are marked for one of the non-agentive voices, namely, objective (c),
locative (d), and instrumental (e).9 The nonfinite verb forms in the first
column are usually considered to be the basic forms. They consist of a root
form with a single voice-marking affix. In row (a), the infix (um), a reflex
of Proto-Extra Formosan (PEF) *(um), precedes the first vowel of the
root. In row (b), the prefix mag-, a reflex of PEF *maR-, is attached to the
root.'" In row (c), the suffix -in, a reflex ofPEF *-en, is attached to the root
in the nonfinite [-fin] and future [- beg] form. In row (d), the prefix i-, a
reflex of PEF *?iSi-, is attached to the root. 11
The forms in columns 2 and 3 all show a root reduplication of the form
CV:-. Vowel length on this increment is not normally written, but is
invariably present, regardless of whether or not the first root vowel carries
length. Thus kumukuha is pronounced [kumu:ku:ha], with vowel length on
the first root vowel, whereas magbibigay is pronounced [magbi:bigay] with
no vowel length on the first root vowel.
Forms in rows (c) through (e) of columns 3 and 4, the perfective
[+comp] and imperfective [-comp] non-agentive voice forms, all show an
infix (in), a reflex of PEF *(in), while the corresponding forms in row
(b) show nag-, a reflex of PEF *minaR-, that is *maR- + *(in) (Reid
1987: 47-48), with loss of the first two segments of the (infixed) prefix.
Row (a) appears irregular in several ways. First, neither of the 'nonfuture'
[ + beg] forms shows any reflex of the *(in) affix, with the result that the
nonfinite [-fin] and the perfective forms [+comp] are identical; secondly,
the 'future' [- beg] form carries no voice affix.
The aspect system is considered by some to be inflectional, in that
there is a clearly definable set of forms that commute with each other in
a paradigm within which most verbs can appear. It is also productive,
affecting, for example, newly borrowed forms from English such as mag-
basketbol 'to play basketball', and so on.
The voice system on the other hand has been treated by some as in-
flectional (De Guzman 1978, 1991) and by others as derivational (Bender
1988, Starosta 1986, 1988). The system seems to be inflectional in that all
verbs typically carry one of the voice affixes, whose form and meaning is
generally predictable based on the semantic features of the verb root, and
whose occurrence can be linked to syntactic features in the case frame of
the verb. The system is viewed by others as derivational because probably
no verb may take the complete set of voice affixes; that is, they do not
freely commute within a paradigm. The choice of one affix rather than
another usually marks a semantic distinction between the forms, and this
meaning change is not always predictable; 12 and the agentive affixes derive
intransitive verbs, 13 while the non-agentive affixes always derive transitive
68 OCEANIC LINGUISTICS, VOL. 31, NO.1

verbs. Traditionally, inflectional affixes do not carryover in derivation,

but each of the voice affixes occurs as a noun-deriving affix, in addition to
its verbal functions. Furthermore, the disparate nature of the set, with one
infix, two suffixes, and several prefixes suggests derivation (or at least a
derivational origin) rather than inflection.
Since derivational processes are by definition word-forming, that is,
they create new lexical entries that are potentially inflectable, the relative
order of affix application in Tagalog would seem to require that, for
Tagalog at least, aspect and voice must be either both derivational, or
both inflectional. With a form such as Pitinapon, it is possible to claim
that it is first derived for voice iPi-tapon 'to throw X'), and then inflected
<
for completed aspect ("insert in) before the first vowel of the root"
Pi-t ciny apon 'threw X'). It is also possible to claim that with a form such
as Pitina.tapon, it is first derived for voice (as above) and then inflected,
first by reduplication ("reduplicate the first two segments of the root,
and lengthen a short reduplicated vowel"): Pi-ta.tapon, and then by <in)
infixation: i-t Cin ra.tapon 'throws X'. However these rules cannot be ap-
plied in parallel fashion to produce a form such as kumu.kuha 'takes'.
Derivation for voice ("insert <urn) before the first vowel of the stem")
would produce the form kumuha, reduplication would then result in the
ungrammatical "kukumuha.t " It appears that the root must first be redu-
plicated before <urn) is infixed. In other words, what has been character-
ized as inflection must take place prior to derivation.l " An alternate
analysis, and one that is used below in discussing the development of the
Tagalog aspect system, treats the language as having two <urn) infixes.
The first, <urn 1 ) , occurs in the nonfinite form, k cum ; )uha 'to get'. The
second, <urn 2 ) , occurs in the [+ beg] forms, k<urn2)uha 'got' and
k<urn2)u:kuha 'getting'. In these forms it is an allomorph of <in).
A related problem occurs with a form such as pinulot. The voice affix on
this form is absent, although it is present on forms in the same paradigm
that do not have an <in) infix. This has been treated as a case of haplology
(Bender 1988). However, there is no Philippine language in which these
two affixes co-occur, even though the vowel of the suffix is reflected
differently from that of the infix in some languages. They act as though
they belong to the same inflectional paradigm. Historically, though, it is
apparent that the completive aspect affix (in) and the objective voice affix
-in were derivational alternatives. The former was a derivational affix
*(in) that derived nouns that were the result of the action of the verb, as
in Tagalog b(in)ili 'something that was bought' (bili 'buy'). The latter was
a derivational affix *-en that derived nouns that would receive the action
of the verb, as in bilihin 'something to be bought'. They could not co-occur
(Starosta, Pawley, and Reid 1982: 162-163).
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 69

TABLE 2. ILOKANO ASPECT PARADIGMS

[+V]

[-fin] [+fin]

[-comp]
~ [+comp]
[-cont]
»<. [+cont]
(a) tumakder tumakder tumaktakder timmakder
'to stand' 'stands' 'is standing' 'stood'
(b) Pagbuggu ?agbuggu ?agbugbuggu nagbuggu
'to wash' 'washes' 'is washing' 'washed'
(c) lutuen lutuen lutlutuen linutu
'to cook' 'cooks' 'is cooking' 'cooked'
(d) sagadan sagadan sagsagadan sinagadan
'to sweep' 'sweeps' 'is sweeping' 'swept'
(e) Pibelleng Pibelleng Pibelbelleng Pimbelleng
'to throw out' 'throws out' 'is throwing out' 'threw out'

3. THE ILOKANO SYSTEM. Ilokano verbs, like those of Tagalog,

can be described as having three aspect forms. However, there is little
systematic correspondence between the two languages. Only one of the
features that define the Tagalog system can be used for Ilokano. In
Ilokano, all verbs are either completive or not [± comp]. Completive verbs
typically have past time reference and are perfective. Noncompletive verbs
are typically used with either present or future time reference, although
they may readily occur with past reference as well (examples 4 and 5
below.) They may be either continuative or not [± cont]. Continuative
verb forms are imperfective. In Ilokano, there is no distinct nonfinite verb
form. The basic form, that is, the one that occurs only with a voice affix, is
marked in the same way as the noncontinuative, noncompletive verb form.
The continuative [+ cont] form is marked by reduplication of the first
three segments of the root, CVC-, with the general sense of 'distributed
referent'. Completive verbs are marked with the infix <in).
Table 2 illustrates the intersection of the Ilokano aspect categories with
voice affixation.!" As in Table 1, the forms in rows (a) and (b) are affixed
for agentive voice, while those in rows (c) through (e) are affixed for one
of the non-agentive voices: objective (c), locative (d), and instrumental (e).
In row (a), the infix <um) (or one of its variants) precedes the first vowel
70 OCEANIC LINGUISTICS, VOL. 31, NO.1

of the root. In row (b), the prefix ?ag-, a probable development of PEF
*maR-, is attached to the root. In rows (c) and (d), the suffixes -en and -an
are attached to the root in the noncompletive [-camp] forms. In row (e),
the prefix ?i- is attached to the root. The completive infix <in) combines
with Pi- as in-(?i- + <in) -+ Pini- -+ ?in-). In other words, the voice affix is
attached to the verb first, with the infix being inserted into it. The same
sequence is apparent in row (a). The form t Cimm'rakder is the result of first
infixing the voice affix <um) and then inserting <in), with subsequent
vowel deletion and nasal assimilation «inum) -+ <inm) -+ <imm»).
Although CVC-reduplication may also co-occur with verbs that have an
<in) infix and are sometimes translated as past imperfective, 1 7 they do not
carry an imperfective sense, but are either repetitive (example 1), or dis-
tributive (example 2), depending on the verb stem, and the context. The
action is always understood as completed. Compare example 3 (com-
pleted, distributive), with 4 and 5 (noncompletive, continuative). In a
negative construction, CVC-reduplication with an <in) affix is interpreted
as 'ever, at any time, at all', as in example 6. 1 8

1. Immin-inum diay aso.

<in) <um)CVC-drink DET dog.
'The dog drank and drank.'
2. Gimmatgatangda ti sapatos.
<in) <um)CVC-buy-they DET shoes
'They kept on buying shoes.'
3. Nagbasbasaak ti libro.
<in)ag-CVC-read-I DET book
'I kept on reading a book (or books).'
4. Agbasbasaak ti libro.
ag-CVC-read-I DET book
'I am reading a book.'
5. Agbasbasaak ti libro idi simrekda.
ag-CVC-read-I DET book when <in) <um)enter-they
'I was reading a book when they entered.'
6. Saanak a gimmatgatang ti sapatos.
not-I LIG <in) <um)CVC-buy DET shoes
'I never bought any shoes.'

The CVC-reduplicative pattern occurs with various other functions in

Ilokano, including pluralization of some nouns, and the comparative
degree of adjectives. Since the number of nouns that can be pluralized
in this way is highly restricted, reduplication here is clearly not to be
interpreted as an inflectional category, but rather as a special instance of
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 71

the more general pattern described above meaning 'distributed referent'.

There are also other reduplicative patterns. One is CV-, which derives
words that interpret their patients as distributed in time and/or space. It
optionally occurs on ag-derived verbs whose subjects are plural. Another
reduplicative pattern is CV(C)CV(C)- (i.e., all of a disyllabic root); it
derives forms that express multiply repetitive action.
One other fact about Ilokano reduplication needs to be mentioned here.
The CVC-pattern (regardless of its function) has a variant containing a
long vowel, CV:-, occurring with roots that have a medial glottal stop,
regardless of the stress pattern of the root, as in examples 7 and 8 (inter-
vocalic glottal stop is not written in standard Ilokano orthography). This
can be seen as resulting from a phonotactic constraint that does not permit
consonant clusters with initial glottal stop. Vowel length occurs instead of
glottal stop in the pattern.

7. Agsasao. [?agsa:sa?6]
ag-CVC-speak
'He is speaking.'
8. Agdadait. [Pagda.da.Pit]
ag-CVC-sew
'He is sewing.'

4. PROTO-EXTRA FORMOSAN ASPECT. In order to account for

the development of the two systems characterized above, an attempt
must be made to reconstruct an aspect system for their immediate proto-
language. In order to do this, several questions need to be answered.
They include the following. (I) Was PEF more like Ilokano in having
a completed/noncompleted feature set, or like Tagalog, with a begun/
nonbegun set? (2) Was PEF more like Ilokano in requiring voice affixation
before infixation of (in), or like Tagalog in which (in) appears to be
infixed before voice affixation? (3) What was the form of the 'distributed
referent' reduplicative pattern that marked continuative verb forms in
PEF?

4.1 An examination of the aspect systems of Philippine languages from

various subgroups suggests that PEF was probably more like Ilokano in
having a distinction between completed and noncompleted verb forms,
with the former marked by the infix (in), than like Tagalog, in which
(in) marks an action that has begun, as against one that is still in the
future.
In Ivatan, in the far north of the Philippines, Hidalgo and Hidalgo
(1971 : 53, 70) describe tense as either past or nonpast, with past tense
72 OCEANIC LINGUISTICS, VOL. 31, NO. 1

marked with a reflex of *<in). Subject focus (i.e., agentive voice) verbs,
which use reduplication to mark repetitive action, have m-initial forms,
not n-initial forms as do the Tagalog equivalents.

9. 1vatan (Hidalgo and Hidalgo 1971 : 72)

mayvayvahay 'building a house'
manutunutung 'cooking'
tumatada 'dancing repeatedly'

Sambalic languages similarly restrict the use of *<in) reflexes to perfective

forms.

10. Kapampangan (Gonzalez 1972: 30)

muran 'it will rain'
mumuran 'it is raining'
minuran 'it has rained'

11. Botolan SambaI (Antworth 1979: 26-27)

itanem 'to plant x (contemplated)'
an-itanem 'to plant x (imperfective)'
intanem 'to plant x (perfective)'

All Cordilleran languages are like Ilokano in restricting the use of *<in)
reflexes to perfective forms.

12. Ibanag (Brandes and Scheerer 1927-28: 20): "in the formation of
the perfect, -in- is added to infix -um-."
lumakar 'walks'
lirninakar' 'walked'
umulur' 'descends'
minulu? 'descended'

13. Balangaw (Shetler 1976: 78): "Tense is either past or nonpast.

Nonpast tense is indicated by m- Past tense is marked by
addition of -in- or simply n-"
mansoblak-ayu 'you wash'
nansoblak-ayu 'you washed'
tombala 'I will answer'
tenumbala 'I answered'
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 73

14. Pangasinan (Benton 1971: 123): "Tense is determined by means

of a formal contrast between a pair of affixes ... that can be
described as 'past' and 'non-past' ... on- (actual, incomplete),
-inm- (actual, complete)"
oninom '(he) will drink'
oniinom '(he) is drinking'
inminom '(he) was drinking, (he) drank'
onbatik 'he will run'
onbabatik 'he is running'
binmatik 'he ran'

Manobo languages in Mindanao likewise use reflexes of *(in) only in

perfective verb forms. 19

15. Ilianen Manobo (Shand 1964: 54): "The markers for tense are
ed- 'non-past tense', mid- 'past tense', and med- 'unreal tense."
edtibas 'will slash'
midtibas 'slashed'
wara'medtibas 'did not slash'

16. Ata Manobo (Morey 1964: 71): "The ... morpheme og-means
that the action expressed by the verb is incomplete. Whether that
action has begun or has not begun is irrelevant. ... The meaning
of the aspect nig- is 'complete'; the action has been completed or
is no longer happening.... The 'unreal' aspect, marked by m-,
expresses the unreality of an action"
qogqogot 'scolds, will scold'
nigqogot 'scolded'
qogot/" 'scold'

The Danao languages of Mindanao likewise use reflexes of *(in) only

in perfective verb forms.

17. Maranao (McKaughan 1958: 28): "There are four tenses: neutral,
past, immediate future, and present progressive.... The past tense
is marked by the infix -i-, and indicates that the action has been
completed."
tabasan 'slash'
tiabasan 'slashed'
tebasan 'will slash'
petebasan 'is slashing'
74 OCEANIC LINGUISTICS, VOL. 31, NO.1

18. Agutaynen of Northern Palawan (Quakenbush, pers. comm.),

possibly an isolate in the Philippine language family (Blust 1991),
restricts the infix <in) to completive forms. It, like other lan-
guages ofPalawan (see note 21), shows influence from Tagalog in
that it distinguishes future from present. Quakenbush suggests
that the tense-aspect system in Agutaynen is structured as shown
in Table 3.

The only languages that are like Tagalog in using <in) to mark a
distinction between begun and not-begun verb forms are the other lan-
guages of the Central Philippines, including Bikol, Bisayan, Mansaka, and
languages such as Palawano (where Tagalog is the trade language) and
possibly Kagayanen Manobo, which has been heavily influenced by one
of the Bisayan languages.j ' Zorc (1977: 118-121, 133-138) provides verb
paradigms for each of the Bisayan speech varieties, including Tausug, all
of which pattern like Tagalog in distinguishing "actual" [+ beg] from
"contingent" [- beg], and perfective [ + comp] from imperfective [- comp]
forms.:" Actual perfective aspect is used for past and perfect; actual
imperfective is used to convey present, progressive, and habitual mean-
ings. All "general mode" actual forms show a reflex of *<in).
In discussing the morphological innovations that characterize the Cen-
tral Philippine subroup, Zorc (1977: 232) provides a reconstruction of
some of the Proto-Malayo-Polynesian (PMP) verb affixation. He recon-
structs PMP *nagCV- for "progressive durative active," and *magCV- for

TABLE 3. AGUTAYNEN ASPECT PARADIGMS

[+V]

[-real] [+ real]

[-fin]
~ [+ fin] [-comp]
~ [+comp]
(a) kumuran kumuran kukuran kiminuran
"to rain' "will rain' "rains' "rained'
(b) magturul magturul pagturul nagturul
"to give' "will give' "gives' "gave'
(c) lutukun lutukun aglutukun linutuk
'to cook' 'will cook' 'cooks' 'cooked'
(d) telekan telekan agtelekan sinelekan
'to sweep' 'will sweep' 'sweeps' "swept'
(e) Pibur'lug Piburlug agbubur'lug binurlug
"to throw out' 'will throw out' 'throws out' "threw out'
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 75

"future durative active," thereby implying that the basic aspectual distinc-
tion between begun and not-begun found in the Central Philippine lan-
guages is inherited rather than innovated. The evidence presented above,
however, suggests the opposite, that is, the immediate parent language
of all of the Philippine languages (PMP in Zorc's terms, PEF in mine)
made a basic aspectual distinction between completed [ + comp] and non-
completed [-comp] forms, with the former marked by *<in), and that the
Central Philippine languages have innovated their present systems. In fact,
for a large number of Bisayan dialects distributed across the various
Bisayan subgroups (as reported by Zorc 1977: 130), a basic distinction
between completed and noncompleted forms is still maintained in the
"potential mode" of the verb forms. The innovations are typically found
only in the "general mode" of the verb.

4.2 The second question that needs to be answered is whether PEF was
more like Ilokano in requiring voice affixation before infixation of <in),
or like Tagalog in which the insertion of <in) appears to occur prior to
voice affixation.
Verbs with suffixal voice affixation tell us nothing about the relative
order of affixation, since either infixation or suffixation could take place
prior to the other, without any difference in the output. Verbs that are
prefixed, as with Pi-, or infixed, as with <um), are revealing, in that if <in)
infixation takes place before voice affixation, the results are Pi-Ccin ) V,
and Ccum'; <in) V- respectively, whereas if <in) infixation takes place
after voice affixation, the results are ?<in)i-CV, and C<in) <um) V- re-
spectively. Ilokano reflects the latter type, with earlier *?<in)i- becoming
in- before consonant-initial stems (with nasal assimilation), but iny- before
vowel-initial stems.P as in 18a, and with *<in) <urn) becoming <imm) in
stems without vowel syncopation, but <im) in stems with vowel syncopa-
tion, as in 18b.

18a. belleng ibelleng imbelleng

lako ilako inlako
uli iuli inyuli
agas iagas inyagas

18b. dakkel dumakkel dimmakkel

gatang gumatang gimmatang
lako lumako limmako
uli umuli immuli
dekket dumket dimket
serrek sumrek simrek
ebbal umbal imbal
76 OCEANIC LINGUISTICS, VOL. 31, NO. 1

An examination of languages from each of the major subgroups in the

Philippines suggests that the Tagalog order of affixation, in which <in)
appears to be infixed prior to the affixation of voice affixes, is historically
prior to the order found in Ilokano.r" This assertion is made despite the
fact that there are a number of languages in different subgroups that show
<in) being infixed subsequent to <um) affixation. For example, Bolinao
linmibed (from earlier l<in) <um)ibed) 'went around' (but see Kapam-
pangan minuran, example 10 above, in which the reverse order occurs),
and Mansaka kyomaan (from earlier k<in) <um)aan) 'finished eating'
(Svelmoe and Svelmoe 1974: 31), but mindateng 'arrived', and minlabay
'passed by' (Miller and Miller 1976: 48, 50), which reflect the reverse order.
Similarly Y ogad ginumalit 'departed', dinumanga 'arrived, and dinumama
'walked' (Reid 1978), but Ibanag liminaka? 'walked' (with vowel har-
mony, example 12 above) show opposite orders of affixation. Itawis
imminum 'drank', and limmabbet 'returned home' (Reid 1978) could reflect
the same order as Ibanag with vowel harmony and progressive nasal
assimilation, or the Y ogad order with regressive nasal assimilation.
Although most of the languages of the Northerm Philippines are similar
to Ilokano in requiring voice affixation prior to infixation of <in), a
number of languages show frozen forms in which the order is reversed, and
in some languages (e.g., Kalinga) the older order of affixation is retained
for verbs that are affixed with <um), while the innovated order affects only
verbs that are affixed with Pi-,

19. In Kalinga as spoken in Guinaang (Gieser 1963 : 18, 19, 32) <um)
is noncompleted, and Cummv (from <um) <in») is completed
(19a), whereas ?i- is noncompleted, but Pin- (from ?<in)i-) is
completed (19b).
19a. dumakol dummakol
tumulung tummulung
19b. Pitar'od Pintar'od
Pibaga Pimbaga

20. In Balangaw (Shetler 1976: 43) the pattern described above for
<um) verbs in Kalinga applies only to verbs in which the initial
vowel is not syncopated (20a). If the vowel is syncopated, the
order of affixing follows the innovated pattern of <um) first,
then <in) (20b).25 Affixation of ?i- verbs is similar to that de-
scribed for Kalinga and Ilokano, with infixation occurring fol-
lowing prefixation (20c).
20a. tale Pumale Pummale
tudu, tumudu tummudu
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 77

20b. Pedah Pumdah Pinumdah

tebal tombal tenumbal
20c. tale Piyale Pinyale
baga Pibaga Pimbaga

In a number of Northern Philippine languages, including Northern and

Southern Alta, and Casiguran Dumagat (Reid 1988), affixation of the
completed aspect of actor focus, pag- and pang- derived verbs, is minag-,
or minang-, requiring reconstruction of Proto-Cordilleran *m(in)aR- and
*m(in)aN-,26 each of which implies infixation of *(in) prior to voice
affixation. In numerous other languages, these forms have now been
reduced to n-initial prefixes, such as Ilokano nag- and nang-.
In Sarangani Manobo (Du Bois n.d.: 24, 47) the order of affixation for
both (um) verbs and Pi- verbs with (in) (22a, b) is the same as that
described above for Ilokano (18a, b), but for verbs with either an initial
labial stop or a glottal stop, the "past tense" of (um) verbs is min- with
deletion of the initial consonant (22c).

22a. dateng domateng dimateng

goyod gomoyod gimoyod
kaen komaen kimaen
22b. data idata indata
ligbin iligbin inligbin
agbel i-agbel in-agbel
22c. bayad mayad minayad
panggas manggas mmanggas
adek madek minadek

Languages that reflect the reconstructed *(um) (in) order without

change include Ivatan and Itbayaten, as in tominbay 'answered', tominwaw
'appeared' (Larson 1986: 160, 165); Agutaynen and Aborlan Tagbanwa
(imin) (with vowel harmony), as in Agutaynen kiminuran 'rained', see
Table 3); and Sorsogon (umin) (Zorc 1977: 138). Zorc reconstructs
*(umin), *(im), and *(in) for Proto-Bisayan (1977: 247).
I have attempted to show that even though both orders are found
throughout the Philippines, in languages that show both orders, it is
the older (um) (in) order that is found in frozen forms. Now, if so
many languages have switched the order affixation from (um) (in) to
(in) (um), apparently independently, is it possible to determine what the
motivation could have been for the change? In Section 2 of this paper, the
problem of characterizing Tagalog voice and aspect affixes as either deri-
vational or inflectional was discussed. It was noted that both affix sets
78 OCEANIC LINGUISTICS, VOL. 3], NO. ]

have been reconstructed as derivational, noun deriving affixes in Proto-

Austronesian, and that voice affixation, for the reasons given above, is
probably still derivational not only in Tagalog, but in all Philippine lan-
guages. It may be, however, that as <in) has become more and more
productive as an aspect affix on verbs, it has moved from being a deriva-
tional affix to an inflectional affix, and would therefore require that verbs
first be derived with voice-marking affixes such as <um) and Pi- before
being inflected with <in).

4.3 The third question that must be answered in order to reconstruct the
PEF aspect paradigm is whether the reduplication that marked contin-
uative verb forms was CVC-, as in Ilokano, or CV:- as in Tagalog. The
Cordilleran languages generally agree with Ilokano in showing CVC-,
whereas Central Philippine languages agree with Tagalog in having either
CV:- or CV- without length. There are two considerations that suggest
that it is Ilokano that directly reflects the PEF reduplicative forms. The
first is that it is necessary to reconstruct for PEF *CV- reduplication to
mark noun plurality. This form is reflected widely throughout the family.
Ilokano and other Cordilleran languages distinguish between this form
and CVC- as the marker of continuative verb forms. It seems more likely
that languages that do not distinguish the two reduplicative patterns have
collapsed them, than that those languages that do distinguish the patterns
have innovated a distinctive pattern. The second is that Ilokano and other
Cordilleran languages typically have a phonologically conditioned variant
(CV:- in Ilokano), occurring with verbs that have a medial glottal stop,
since the ?C sequence that would result from the CVC- reduplication of
these forms is disallowed in these languages. In Ilokano, the glottal stop
is deleted with compensatory vowel lengthening. Other languages have
developed different strategies to cope with such disallowed sequences.
Bontok, for example, assimilates the glottal stop to the following conso-
nant, so that CVC- + darit 'sew' becomes daddarit 'sewing'. This pattern
is generalized in Bontok (with some modifications) to verb stems that
have medial glides and liquids, even though consonant clusters with
initial glides and liquids are not otherwise disallowed in the language. In
Tagalog, as in other Central Philippine languages that also show CV:-,
consonant clusters with initial glottal stops are also disallowed. It is proba-
ble that the same process of loss of glottal stop in this position with
compensatory vowel lengthening and subsequent generalization to all
consonants in this position has brought about the change to CV:- in these
languages.
Table 4 presents the aspect paradigm for Proto-Extra Formosan
*(um) and *?i- verbs. The verb roots that appear in the chart are not
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 79

TABLE 4. PROTO-EXTRA FORMOSAN ASPECT PARADIGMS

[+V]

[-fin] [+ fin]

[-comp]
.:'>; [+comp]

[-conij
~ [+conij
(a) tumakder tumakder tumaktakder tuminakder
'to stand' 'stands' 'is standing' 'stood'
(e) Pibeleng Pibeleng Pibelbeleng Pibineleng
'to throw out' 'throws out' 'is throwing out' 'threw out'

TABLE 5. PRE-ILOKANO ASPECT PARADIGMS, STAGE 1

[+V]

[-fin] [+ fin]

[-comp]
~ [+comp]

[-cont]
~ [+cont]
(a) tumakder tumakder tumaktakder tinumakder
'to stand' 'stands' 'is standing' 'stood'
(e) Pibeleng Pibeleng Pibelbeleng Pinibeleng
'to throw out' 'throws out' 'is throwing out' 'threw out'

necessarily PEF reconstructions. They are pre-Ilokano forms that are used
in the following section to demonstrate the development of the system into
modern Ilokano.

5. DEVELOPMENT OF THE ILOKANO ASPECT SYSTEM. Given

the Proto-Extra Formosan aspect system described in the previous sec-
tion, only two stages need be postulated to account for the development
of the modern Ilokano aspect system. The first was an affix-order change
by which focus affixation preceded, rather than followed, the infixation of
the <in) completive aspect marker. The results of this change are presented
in Table 5.
80 OCEANIC LINGUISTICS, VOL. 31, NO. 1

TABLE 6. MODERN ILOKANO ASPECT PARADIGMS, STAGE 2

[+V]

[-fin] [+fin]

[-comp]
~ [+comp]

[-cont]
~ [+cont]
(a) tumakder tumakder tumaktakder timrnakder
'to stand' 'stands' 'is standing' 'stood'
(e) Pibelleng Pibelleng Pibclbelleng Pimbelleng
'to throw out' 'throws out' 'is throwing out' 'threw out'

Stage I. Affix order change.

a. <urn-in> became <in-urn>
b. ?i-(in> became ?(in>i-

The second stage brought about various phonological changes that were
the result of the affix-order switch of the first stage. Vowel syncopation
and nasal assimilation as described below produced the forms given in
Table 6. The gemination rule described below is not pertinent to the
development of the aspect system, but is included to account for the
difference is root forms between the pre-Ilokano and modern Ilokano
stages of tables 5 and 6. This change could have taken place prior to the
switch in affix order. One other phonological change took place, result-
ing in the change of CVC- to CV:- when the final consonant of the
reduplicative pattern was glottal stop, as described above in Section 3
(see examples 7 and 8). This change was the result of a general loss of
glottal stop in preconsonantal position in Ilokano, with associated com-
pensatory vowel-lengthening, and was not limited to reduplicative prefixes
with final glottal stop. It may have preceded or followed the switch in affix
order.

Stage 2. Phonological changes.

a. Vowel loss. An affix vowel immediately following (in) was lost.
<
in-urn) became <in-m )
?(in)i- became ?(in)-
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 81

TABLE 7. PROTO-EXTRA FORMOSAN ASPECT PARADIGMS

[+V]

[-fin] [+fin]

[-comp]
~ [+comp]
[- cont]
~ [ + cont]
(a) kumuha kumuha kumuhkuha kuminuha
"to take' "takes' "is taking' "took'
(e) Pitapun Pitapun Pitaptapun Pitinapun
"to throw out' 'throws out' 'is throwing out' 'threw out'

b. Assimilation. The nasal of <in) was assimilated to the point of

articulation of an adjacent consonant.
<in-m) became <im-m )
?<in)- became ?<im)- before labials
c. Gemination. Medial root consonants following *e were geminated.
be/eng became bel/eng

6. DEVELOPMENT OF THE TAGALOG ASPECT SYSTEM. The

development of the Tagalog aspect system was considerably more com-
plex than that described above for Ilokano, involving not only several
phonological changes but also a complete restructuring of the system
itself. As a starting point to describe these changes, Table 4 is repeated
here as Table 7, but with pre-Tagalog verb roots inserted.

6.1 Stage 1. The phonological changes that apparently resulted in the

restructuring of the Tagalog system must have occurred at an early stage
in the development of the Central Philippine languages, since the Tagalog
aspect system is typical of the Central Philippine languages (see Section
4.1 above). Vowel syncopation and assimilation occurred first, as de-
scribed below and illustrated in Table 8. These changes resulted in gemi-
nate nasal clusters in the completive aspect form of <um) verbs. Although
this stage is not attested in any Central Philippine language, identical
changes have brought about the forms found in other languages of the
Philippines, such as Kalinga (see example 19a).
82 OCEANIC LINGUISTICS, VOL. 31, NO.1

Stage 1. Phonological changes.

a. Vowel loss. The vowel of <in) was lost following <um).
Cum-in') became <um-n)
b. Assimilation. The nasal of <in) assimilated to an adjacent conso-
nant.
<um-n) became cum-mv

6.2 Stage 2. The major phonological change that took place in the second
stage (illustrated in Table 9) was the simplification of geminate consonant

TABLE 8. PRE-TAGALOG ASPECT PARADIGMS, STAGE 1

[+V]

[-fin] [+fin]

[-comp]
~ [+comp]

[-cont]
~ [+cont]
(a) kumuha kumuha kumuhkuha kummuha
"to take' "takes' "is taking' "took'
(e) Pitapun Pitapun Pitaptapun Pitinapun
"to throw out' "throws out' "is throwing out' "threw out'

TABLE 9. PRE-TAGALOG ASPECT PARADIGMS, STAGE 2

[+V]

[-fin] [+fin]

[-comp]
~ [+comp]
[-cont]
.r>: [+cont]

[-beg]
/\ [+beg] [-beg]
/\ [+beg]
kumuha kumuha kumuha kumu.kuha kurnu.kuha kumuha
"to take' "will take' "takes' "willbe taking' "is taking' "took'
Pitapun Pitapun Pitapun Pita.tapun Pita.tapun Pitinapun
"to throw "willthrow "throws out' "will be throw- "is throwing "threw out'
out' out' ing out' out'
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 83

clusters. No instances of geminate consonants are presently found in

Central Philippine languages, except for those resulting from prefixation
as described for Tagbanwa (example 21).

21. Aborlan Tagbanwa (Hussey 1965: 8-9) "No sequences of two

identical consonants were found in the data apart from those
formed by prefixation of a stem."
mag-gurgur 'one will weed"
mag-gated 'one will be hurt'

Loss of gemination must have resulted in considerable formal ambigu-

ity. The nonfinite, noncontinuative, and completive forms would all have
been formally identical. There were, in addition, other areas of formal
ambiguity. No distinction was made in the verb between future and pre-
sent forms. Both the [-cont] and [+cont] forms could be interpreted as
either [ - begun], that is, future, or [ + begun], that is, present. It is proba-
ble, though, that such distinctions would have been marked in other ways,
for example with a postclitic (22), or an "auxiliary" verb (23), as is found
in Northern Philippine languages that have not restructured their aspect
systems.

22. Ilokano:
uminom 'drinks'
uminom to 'will drink'
23. Bontok:
omey 'goes'
asomey 'will go'

One other phonological change that must have taken place fairly early
in the history of the Central Philippine group, since it is shared by all the
languages, is the change of CYC- to CY:-. As in Ilokano, it was probably
the result of the loss of glottal stop in preconsonantal position, with
compensatory vowel lengthening, but unlike Ilokano, the change was
generalized to affect all consonants in this position.

Stage 2. Phonological changes.

a. Loss of consonant gemination
b. CYC- became CY:-

The phonological changes that resulted in the loss of <in) as part of

the completive aspect form of (um) verbs in Tagalog may also have
taken place early in the development of other Central Philippine languages
84 OCEANIC LINGUISTICS, VOL. 31, NO. 1

TABLE 10. PRE-TAGALOG ASPECT PARADIGMS, STAGE 3a: (um) verbs

[+V]

[ -fin] [+fin]

[-camp]
~ [+comp]
[- beg]
.r-: [ + beg]
(a) kumuha kumu.kuha kumu.kuha kumuha
'to take' 'will take' 'takes' 'took'

such as Cebuano, which are like Tagalog in marking completive aspect

with <um) alone (or a subsequent development of Cum ), such as Cebuano
mu-). However there are a number of Central Philippine dialects that still
reflect the full form <umin), such as Sorsogon, or that have apparently
switched the order of affixation, such as Samar-Leyte and Waray (in which
the form <inm) reflects an earlier <inum»), and Hiligaynon (in which the
form <imn) shows a metathesis of the bilabial and alveolar nasals, a
process common throughout the Bisayan dialects) (Zorc 1977: 58, 182,
138, 290). The presence of <in) in these forms makes it clear that the
changes being described here for Tagalog must have taken place after the
initial dispersal of Central Philippine languages.

6.31 Stage 3a. The first step in the restructuring of the system would have
affected only <um) verbs. Resolving the ambiguity between the [+comp]
and [-cont] forms of these verbs was accomplished by eliminating the
distinction between [- cont] and [ + cont] in favor of the [ + cont] forms.
This would still however have left ambiguity between the present [ + be-
gun] and future [- begun] forms, as seen in Table 10. Completive verbs
now fall into two classes, those marked morphologically by <um2 ) , and
those by <in). The affix <um!) marks nonfinite verbs, and with reduplica-
tion, noncompleted finite verbs.

6.32 Stage 3b. The second step in the restructuring of the system resulted
from the association of <um2 ) with completed aspect, and its presence
with reduplication in [+ begun] forms. This allowed the morphologization,
and thus the disambiguation, of the features [± begun]. The affix <um2 )
became the marker of [ + begun], and its absence the marker of [- begun]
(see Table 11).

>
6.33 Stage 3c. With <um2 now functioning as the marker of [+ begun],
analogical pressure must soon have resulted in its corresponding mor-
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 85

TABLE 11. PRE-TAGALOG ASPECT PARADIGMS, STAGE 3b: (urn) verbs

[+V]

[-fin] [+fin]

[-beg] [+ beg]

[-comp]
.r>. [+comp]
kumuha ku.kuha kumu.kuha kumuha
'to take' 'will take' 'takes' 'took'

TABLE 12. TAGALOG ASPECT PARADIGMS, STAGE 3c

[+V]

[-fin] [+fin]

[- beg]
»<:>: [ + beg]

[-comp]
~ [+comp]
(a) kumuha ku.kuha kumu.kuha kumuha
'to take' 'will take' 'takes' 'took'
(e) Pitapun Pita.tapun Pitina.tapun Pitinapun
'to throw out' 'will throw out' 'throws out' 'threw out'

pheme (in) beginning to function in the same way. The system for (in)
verbs was restructured to match that of the (um) verbs, and the modern
Tagalog system came into being (Table 12).

6.4 The changes that have been discussed here, which resulted in the
modern Tagalog aspect system, can be plausibly characterized as abduc-
tive innovations, in the sense described by Anderson (1973). Anderson's
concern was primarily with providing explanations for diachronic phono-
logical changes. However, abductive inference is just as applicable as an
explanation for innovations that have taken place in morphology and
syntax as it is for those that have occurred in phonology (see Starosta
1991b, for a discussion of abductive change in syntax). As described by
Anderson, abductive inference occurs in the process of a child's formula-
tion of his/her grammar. "Abduction proceeds from an observed result,
86 OCEANIC LINGUISTICS, VOL. 31, NO.1

invokes a law, and infers that something may be the case" (Anderson
1972: 775). The "observed result" is the output of the grammars of the
adults who provide language input to the child. The "laws" that are
invoked are the universal principles of syntactic organization that children
are presumably born with. The inferences that children make are the
morphological and syntactic rules that they develop to account for the
input. Where ambiguity exists in the language of the older generation, the
stage is set (Anderson's [1972: 789] "necessary condition") for abductive
innovative change. We have seen how loss of gemination in Tagalog
resulted in considerable formal ambiguity in the language. Children, faced
with providing a syntactic explanation as they developed their own inter-
nalized grammars for the fuzzy data that they were receiving, restructured
the aspectual system, favoring the use of some forms, eliminating the use
of others, and redefining the functions of some morphemes. The sum of
all of these abductive innovations eventually resulted in an aspectual
system considerably different from the one that must be reconstructed for
the language of their PEF ancestors.

NOTES

1. This paper was presented to the 6th International Conference on Austronesian

Linguistics, Honolulu, May 1991. An early version was presented to a meeting
of the Linguistics Society of the Philippines, Manila, March 1990. I wish to
thank Byron Bender and Stanley Starosta for reading and commenting on
various versions of this paper. I also wish to acknowledge support from the
Social Science Research Institute at the University of Hawaii.
2. Gonzalez (1972: 72) notes, "In actual usage the number of semantic deriva-
tional units attached to the root is usually only one or two, three at the most.
Still these monstrous accretions are possible: ... makapakipagpakapabayu 'to
be able to participate [with some group] in exerting [oneself] in causing [some-
body] to make [something] new'." Hidalgo and Hidalgo (1971: 124) cite forms
such as kapakapachipangadadaw 'has just been able to participate in showing
affection'. Vanoverbergh (1955: 225-233) cites around seven hundred possible
forms for the verb sural 'write', the great majority of which can also occur with
the root reduplicated in at least two different ways.
3. McKaughan (1959) considers Maranao to have four inflectional categories:
voice, mode, aspect, and tense.
4. Their derivational characteristics are a result of their historical origin as noun-
deriving processes in Proto-Austronesian (Starosta, Pawley, and Reid 1982).
5. It should be noted that although these features are actually aspectual, they
have often been described in the literature as though they were features of
tense, because the distinctions often carryover in translation as differences in
tense. In the sources that are cited in this paper, terms such as "past tense"
(for completive aspect) and "future tense" (for action not begun) are fre-
quently used, and it is for this reason that the label "tense-aspect" instead of
simply "aspect" commonly appears in descriptions of Philippine languages.
Wolff (1973) refers to these affixes as "tense-mode."
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 87

6. Ilokano distinguishes the morphological marking of beneficiary from instru-

mental voice (Ilokano i- -an, i-; Tagalog i-), and temporal from locative voice
(Ilokano panag-, -an; Tagalog -an). (See De Guzman 1978: 139.)
7. The use of this label for the parent of all of the Austronesian languages outside
of Taiwan avoids a commitment to the validity of a Malayo-Polynesian, which
contains all the Philippine languages as a single subgroup within its western
branch.
8. Ramos (1971: 23) characterizes the same forms as contemplated (future), in-
completed (progressive), and completed (past). Similarly Llamzon (1976: 88)
refers to the contemplated aspect as "future aspect."
9. There is a wide range of analyses and associated terminology connected with
voice distinctions in Philippine languages (see Constantino 1971, Kerr 1965,
Kess 1967, McKaughan 1958, Reid 1966). I am more concerned in this paper
with the development of the verb forms than with their syntactic functions,
and so will not attempt any discussion of these analyses. The terminology used
here is simply a set of convenient labels with which to refer to verb forms that
have the same phonological characteristics, some of which have widely varying
syntactic functions in the language.
10. De Guzman (1978: 150) analyzes the mag- prefix as a combination of a deriva-
tional prefix pag- plus an m- inflectional prefix, an analysis which neatly
captures the historical development of the prefix. Wolff (1973: 74, 84) also
notes the similarity in form and function of (um) and mag- and considers
mag- to be the surface form of a deep structure pag- + (um).
11. For a discussion of the historical development of these affixes and the focus
system which they mark, see Pawley and Reid 1979; Starosta, Pawley, and
Reid 1982; and, from a different point of view, Wolff 1973.
12. Pittman (1966) discusses the differences represented by the choice between
Tagalog (um) and mag-. He notes that with identical verb stems they may be
either reflexive or nonreflexive, causative or noncausative, centripetal or non-
centripetal, dual (and reciprocal) or nondual (and nonreciprocal), repetitive or
nonrepetitive, or 'having intrinsic color change' or 'external color application'.
13. From the point of view of the now widely accepted ergative character of
Tagalog syntax (Starosta 1988), verbs with "agentive voice" affixation that
have two nominal arguments are antipassive and therefore intransitive.
14. For a discussion of the problems associated with an adequate characterization
of reduplication in Tagalog, see Carrier 1979 and Carrier-Duncan 1984.
15. If, as Starosta (1991a) claims, any rule which refers to a word-internal mor-
phological boundary is wrong, then having to refer to the first vowel of a root
in order to characterize Tagalog infixation for completed aspect is evidence
that aspect-marking in Tagalog cannot be inflectional.
16. Only enough of the voice-affix system is shown here to illustrate the possible
affix combinations.
17. See, for example, Vanoverbergh 1955: 222.
18. In these and subsequent examples, either commonly used orthographic con-
ventions or the orthography of the source is used.
19. Harmon (1977: 105)states that Kagayanen Manobo has two aspects, [- begun]
encompassing contemplated actions and commands, and [+ begun] encom-
passing those actions which have begun and mayor may not be completed.
The only data given, however, are all translated perfectively. If in fact these
are the relevant features for Kagayanen verbs, it is probably the result of
areal influence, in that this language has been heavily influenced by languages
in the Central Philippines which have the same aspect system as Tagalog.
88 OCEANIC LINGUISTICS, YOLo 31, NO.1

"Kagayanen is a Manobo language with a heavy overlay of Bisayan vocabu-

lary" (Harmon 1977: vii).
20. Unreal aspect m- has a zero variant on "modelcss" verbs (Morey 1964: 70).
21. That Palawano has changed as a result of areal influence from a Central
Philippine language, rather than sharing in the innovations that probably
brought about this type of system, is clear from Revel-MacDonald's (1979)
description. In the non-agentive voices (example I below), the aspect system
is structurally identical to that described above for Tagalog, with perfect,
present progressive, future, and imperative (i.e., basic) forms. The affix <in) is
used to distinguish the begun (perfect and present progressive) from the not-
begun (future and imperative) forms. However the agentive voice forms (ex-
amples 2 and 3 below) have only partly changed. Verbs that take the <urn)
affix are like the non-agentive voice forms in terms of the structure of the
system, but they restrict <in) to the perfective form. Verbs that are derived
with pag- distinguish only perfective, present progressive, and future forms,
with <in) being used only in the perfective forms, and with reduplication
serving to mark the progressive forms, similar to what has been described for
Ilokano.
I. Tense-aspect of i- verbs (non-agentive)
ilinamuq 'cooked (rice) with X'
ilinalamuq 'is cooking (rice) with X'
ilalamuq 'will cook (rice) with X'
ilamuq 'cook (rice) with X!'
2. Tense-aspect of verbs with <urn)
suminurung 'went'
sumusurung 'is, was going'
susurung 'will go'
sumurung 'go!'
3. Tense-aspect of pag-derived verbs (agentive)
nagdarak 'ran'
magdadarak 'is running'
magdarak 'will run; run!'
Green (1979) reports that Tagbanwa in Palawan is apparently undergoing
a shift from "incomplete versus complete" aspect marking on its verbs to
"begun versus not-yet-begun" action. He suggests that "it may be an attempt
by the speakers, ... to harmonize Tagbanwa with Tagalog, the lingua franca
of Palawan and a language in which most of the speakers of Tagbanwa
are functionally fluent" (Green 1979: 85). The morphological devices that
Tagbanwa uses to distinguish the begun from the not-yet-begun aspect do not
include a reflex of *<in). In Tagbanwa this. affix is restricted to the completed
action aspect.
22. According to Pallesen (1977), Sarna traders from the Sulu archipelago in the
south of the Philippines took back Bisayan-speaking wives from the Butuan
area of Northern Mindanao some 700 years ago. The language that has
developed as a result is Tausug.
23. That is, stems that originally had initial glottal stops.
24. Contrary to Wolff (1973: 74), who reconstructs for Proto-Austronesian
*-inum- "past independent active affix." Formosan languages generally reflect
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 89

*<umin), e.g. Atayal mnlax "gave up" (alax + mn-) (Wolff 1973), although
Tsouic shows the reverse order: t-in-sm-ani-ula?» 'maltreated' (Tsuchida
1976: 43).
25. In Shetler's (1976: 43) terms, "When Past Tense -in- cooccurs with Subject
Focus -um- in a stem which begins with a CV syllable, the i of Tense is reduced;
the n of Tense and u of Focus metathesize ... and the n of Tense assimilates to
m."
26. From an earlier *p<um) <in)a- sequence. In Proto-Extra Formosan (if not
in Proto-Austronesian), when <urn) was infixed into words with initial bilabial
stops, the first two segments of the infixed word were deleted. This rule is still
present in languages such as Tagbanwa, Palawano, some Manobo languages,
and Blaan (see also Wolff 1973: 84). Frozen forms such as matay die (from
p <um) atay) occur in various other languages, such as Tagalog.

REFERENCES
Allen, Lawrence A. 1977. Reduplication and cyclical rule ordering in Kankanaey
morphophonemics. Studies in Philippine Linguistics 1(2): 280-295.
Anderson, Henning. 1973. Abductive and deductive change. Language 49: 765-
793.
Antworth, Evan L. 1979.A grammatical sketch of Botolan Sambal. Special Mono-
graph Issue Number 8. Philippine Journal of Linguistics.
Bender, Byron W. 1988. Morphology Course Notes. Typescript.
Benton, Richard A. 1971. Pangasinan reference grammar. PALl Language texts:
Philippines. Honolulu: University of Hawaii Press.
Brandes, Carlos and Otto Scheerer. 1927-28. On sandhi in the Ibanag language.
The Archive, Paper Number 6. Manila: University of the Philippines.
Carrier, Jill. 1979. The interaction of morphological and phonological rules in
Tagalog. Ph.D. dissertation, MIT.
Carrier-Duncan, Jill. 1984. Some problems with prosodic accounts of reduplica-
tion. In Language Sound Structure, ed. by Mark Aranoff and Richard T.
Oehrle, pp. 260-286. Cambridge: MIT Press.
Constantino, Ernesto. 1971. Ilokano reference grammar. PALl Language Texts:
Philippines. Honolulu: University of Hawaii Press.
De Guzman, Videa P. 1978. Syntactic derivation of Tagalog verbs. Oceanic Lin-
guistics Special Publication No. 16. Honolulu: University of Hawaii Press.
- - - . 1991. Inflectional morphology in the lexicon: Evidence from Tagalog.
Oceanic Linguistics 30: 33-48.
Du Bois, Carl. n.d. Sarangani Manobo: An introductory guide. Unpublished
typescript. Nasuli, Malaybalay, Bukidnon: Summer Institute of Linguistics.
Gieser, C. Richard. 1963. A grammatical sketch of Kalinga. Mimeo. Nasuli,
Malaybalay, Bukidnon: Summer Institute of Linguistics.
Gonzalez, Andrew, F. S. C. 1972. Pampangan: Outline of a generative semantic
description. Manila: Research Council, De La Salle College.
Green, Peter. 1979. Co-existent aspect-marking phenomena in Tagbanwa of Pala-
wan Island. Pacific Linguistics A-50, Papers in Philippine Linguistics No.9.
Canberra: The Australian National University.
Harmon, Carol. 1977. Kagayanen and the Manobo subfamily of Philippine lan-
guages. Ph.D. dissertation, University of Hawaii.
90 OCEANIC LINGUISTICS, VOL. 31, NO.1

Hidalgo, Cesar A., and Araceli C. Hidalgo. 1971. A tagmemic grammar of Ivatan.
Manila: Linguistic Society of the Philippines.
Hussey, Stewart. 1965. Aborlan Tagbanwa: Verbal system and related topics.
M. A. thesis, Hartford Seminary Foundation.
Kerr, Harland B. 1965. The case-marking and classifying function of Cotabato
Manobo voice affixes. Oceanic Linguistics 4: 15-47.
Kess, Joseph F. 1967. Syntactic features of Tagalog verbs. Ph.D. dissertation,
University of Hawaii.
Larson, Virginia. 1986. Ivatan texts. Manila: Summer Institute of Linguistics.
Llamzon, Teodoro A. 1976. Modern Tagalog: A functional-structural description.
The Hague: Mouton.
Li, Paul Jen-Kuei. 1973. Rukai structure. Institute of History and Philology Special
Publication Number 64. Nankang, Taipei: Academia Sinica.
Miller, Jeanne. 1973. Semantic structure of Mamanwa verbs. Linguistics 110: 74-
81.
Miller, Jeanne, and Helen Miller. 1976. Mamanwa grammar. Language Data,
Asian-Pacific Series, Number 8. Huntington Beach, California: Summer Insti-
tute of Linguistics.
McKaughan, Howard P. 1958. The inflection and syntax of Maranao verbs.
Manila: Institute of National Language.
Morey, Virginia. 1964. Distributional restrictions on co-occurrence of aspect and
focus morphemes in Ata verbs. Oceanic Linguistics 31 : 69-86.
Otanes, Fe T. 1966. A contrastive analysis of English and Tagalog verb com-
plementation. Ph. D. dissertation, University of California, Los Angeles.
- - - . 1970. Some transformation rules for noun derivation in Tagalog. Philip-
pine Journal of Linguistics I : 33-73.
Pallesen, A. Kemp. 1977. Culture contact and language convergence. Ph.D.
dissertation, University of California, Berkeley.
Pawley, Andrew K., and Lawrence A. Reid. 1979. The evolution of transitive
constructions in Austronesian. In Austronesian Studies: Papers from the Sec-
ond Eastern Conference on Austronesian Languages, ed. by Paz B. Naylor,
pp. 103-130. Ann Arbor: Center for South and Southeast Asian Studies,
University of Michigan.
Pittman, Richard. 1966. Tagalog -um- and mag-, an interim report. Linguistic
Circle of Canberra Publications A-8. Canberra: The Australian National
University.
Ramos, Teresita V. 1971. Tagalog structures. PALl Language Texts: Philippines.
Honolulu: University of Hawaii Press.
Reid, Lawrence A. 1966. An Ivatan syntax. Oceanic Linguistics Special Publication
No.2. Honolulu: University of Hawaii Press.
- - - . 1978. Yogad and Itawis Fieldnotes.
- - - . 1987. The early switch hypothesis: Linguistic evidence for contact between
Negritos and Austronesians. Man and Culture in Oceania 3, Special Issue:
41-59.
- - - . 1988. The Alta languages of the Philippines. Paper presented to the
Fifth International Conference on Austronesian Languages, Auckland, New
Zealand.
Revel-Macdonald, Nicole. 1979. Le Palawan (Philippines). Langues et Civilisa-
tions de l' Asie du Sud-Est et du Monde Insulindien Monograph # 4. Paris:
Centre National de la Recherche Scientifique.
Schachter, Paul, and Fe T. Otanes. 1972. Tagalog reference grammar. Berkeley,
Los Angeles, and London: University of California Press.
THE DEVELOPMENT OF ASPECT IN PHILIPPINE LANGUAGES 91

Shand, Jean. 1964. Categories and markers of tense, focus, and mode in Ilianen
Manobo. Oceanic Linguistics 3: 58-68.
Shetler, Joanne. 1976. Notes on Balangao grammar. Language Data, Asian-Pacific
Series, Number 9. Huntington Beach, California: Summer Institute of Lin-
guistics.
Starosta, Stanley. 1986. Focus as recentralization. In FOCAL I: Papers from
the Fourth International Conference on Austronesian Linguistics, ed. by Paul
Geraghty, Lois Carrington, and S. A. Wurm, pp. 73-95. Pacific Linguistics
C-93. Canberra: The Australian National University.
- - - . 1988. The case for lexicase. London: Pinter Publishers.
- - - . 1991a. No more boundaries, no more phemes. Paper presented to the
Tuesday Seminar, Department of Linguistics, University of Hawaii.
- - - . 1991b. The great Aux cataclysm: Diachronic justification for a synchronic
analysis. In Studies in Dravidian and General Linguistics: A Festschrift for
Bh. Krishnamurti, ed. by B. Lakshmi Bai and B. Ramakrishna Reddy, pp. 505-
515. Osmania University Publications in Linguistics No.6.
Starosta, Stanley, Andrew K. Pawley, and Lawrence A. Reid. 1982. The evolution
of focus in Austronesian. In Papers from the Third International Conference
on Austronesian Linguistics, vol. 2, Tracking the travellers, ed. by Amran
Halim, Lois Carrington, and S. A. Wurm, pp. 145-170. Pacific Linguistics
C-75. Canberra: The Australian National University.
Svelmoe, Gordon, and Thelma M. Svelmoe. 1974. Notes on Mansaka grammar.
Language Data, Asian-Pacific Series, Number 6. Huntington Beach: Summer
Institute of Linguistics.
Tsuchida, Shigeru. 1976. Reconstruction ofProto-Tsouic phonology. Monograph
Number 5. Tokyo: Institute for the Study of Languages and Cultures of Asia
and Africa.
Vanoverbergh, Morice. 1955. Iloko grammar. Baguio City: Catholic School Press.
West, Anne. 1973. The semantics of focus in Amganad Ifugao. Linguistics 110: 98-
121.
Wolff, John U. 1973. Verbal inflection in Proto-Austronesian. In Parangal Kay
Cecilio Lopez, ed. by Andrew B. Gonzalez, pp. 71-91. Philippine Journal of
Linguistics Special Monograph Issue No.4.
Zorc, R. David. 1977. The Bisayan dialects of the Philippines: Subgrouping and
reconstruction. Pacific Linguistics C-44. Canberra. The Australian National
University.
http://www.jstor.org

LINKED CITATIONS
- Page 1 of 1 -

You have printed the following article:

On the Development of the Aspect System in Some Philippine Languages
Lawrence A. Reid
Oceanic Linguistics, Vol. 31, No. 1. (Summer, 1992), pp. 65-91.
Stable URL:
http://links.jstor.org/sici?sici=0029-8115%28199222%2931%3A1%3C65%3AOTDOTA%3E2.0.CO%3B2-J

This article references the following linked citations. If you are trying to access articles from an
off-campus location, you may be required to first logon via your library web site to access JSTOR. Please
visit your library's website or contact a librarian to learn about options for remote access to JSTOR.

References

Inflectional Morphology in the Lexicon: Evidence from Tagalog

Videa P. De Guzman
Oceanic Linguistics, Vol. 30, No. 1. (Summer, 1991), pp. 33-48.
Stable URL:
http://links.jstor.org/sici?sici=0029-8115%28199122%2930%3A1%3C33%3AIMITLE%3E2.0.CO%3B2-0

The Case-Marking and Classifying Function of Cotabato Manobo Voice Affixes

Harland B. Kerr
Oceanic Linguistics, Vol. 4, No. 1/2. (1965), pp. 15-47.
Stable URL:
http://links.jstor.org/sici?sici=0029-8115%281965%294%3A1%2F2%3C15%3ATCACFO%3E2.0.CO%3B2-L

Distributional Restrictions on Co-Occurrence of Aspect and Focus Morphemes in Ata Verbs

Virginia Morey
Oceanic Linguistics, Vol. 3, No. 1, Papers in Philippine Linguistics. (Summer, 1964), pp. 69-86.
Stable URL:
http://links.jstor.org/sici?sici=0029-8115%28196422%293%3A1%3C69%3ADROCOA%3E2.0.CO%3B2-W

Categories and Markers of Tense, Focus and Mode in Ilianen Manobo

Jean Shand
Oceanic Linguistics, Vol. 3, No. 1, Papers in Philippine Linguistics. (Summer, 1964), pp. 58-68.
Stable URL:
http://links.jstor.org/sici?sici=0029-8115%28196422%293%3A1%3C58%3ACAMOTF%3E2.0.CO%3B2-Y