Professional Documents
Culture Documents
1. Introduction
The number and type of different voice1 constructions formed with verbs2 are a
point of great disagreement in Philippine linguistics. Many descriptions of the
1. Voice is defined here as the category that marks correspondences between semantic partici-
pants and the syntactic subject in the sentence. The discussion on whether there is a subject or
whether the properties of the subject are divided between more than one argument in Tagalog is
largely irrelevant for this study. Contrary to the discussion initiated and continued in Schachter
(1976, 1977, 1993), I hold the opinion that Tagalog does have an identifiable subject, which is the
nominative constituent, as evidenced in (Kroeger 1993a, 1993b; Himmelmann 2005: 156–157).
2. Verbs are defined here as lexemes having both aspectual and voice meanings.
https://doi.org/10.1075/sl.17056.kli
Studies in Language 43:1 (2019), pp. 1–43. issn 0378-4177 | e‑issn 1569-9978
© John Benjamins Publishing Company
2 Sergei Klimenko
Tagalog voice system present a simplified version of the real situation. There are
two approaches to establishing the inventory of voices: (i) based on the number
of voice affixes (e.g., Kess 1967; Reid and Liao 2004; Reid 2005, 2008; Himmel-
mann 2008), i.e. only four different voice forms are recognized, as many as there
are voice affixes – <um>, -in, -an, i-, in line with the approach typically used in
the analysis of Formosan languages (e.g. Ross and Teng 2005; Chang 2006); (ii)
based on semantics of constructions (e.g., Schachter and Otanes 1972: 70; Rachkov
1981: 127; Foley 1998; De Vos 2011: 106; Malicsi 2013: 48–55; Gallego 2015; Tanaka
2016: 36–37), which relies on opaque definitions of roles rooted in no or almost no
formal basis. Some authors use a combination of these approaches, where some
voices are established on the basis of the affix used with a verb, while some voice
affixes are said to be markers of a variety of voices, based on semantic roles they
are said to mark (e.g., Starosta, Pawley and Reid 2009). Neither approach is sup-
ported by Tagalog data.
The first approach meets insurmountable difficulties when one considers
voice paradigms consisting of more than four forms. For instance, the following
paradigm of voice forms derived from the root sigaw ‘shout’ contains seven differ-
ent voice forms:
(1) a. S<um>igaw=siya3.
<av>shout[pfv]=3m.nom
‘He shouted.’
b. I-s<in>igaw=ko sakaniya na Saksi ni Jehova=ako.
pv-<pfv>shout=1m.act 3m.nact lk witness pers.sg.act pn=1m.nom
‘I shouted to him that I was a Jehovah’s Witness.’
c. S<in>igaw-an=niya=ako.
<pfv>shout-rv=3m.act=1m.nom
‘He shouted at me.’
d. P<in>ag-sigaw-an=niya ng utos sa mga sundalo ang
<pfv>stem-shout-lv=3m.act gen order obl pl soldier nom
balkonahe=ng ito.
balcony=lk prox
‘He shouted the orders to the soldiers from this balcony.’
e. I-p<in>aN-sigaw=niya sa entablado ang bago=ng
iv-<pfv>ins-shout=3m.act obl stage nom new=lk
mikropono=ng b<in>ili=ko.
microphone=lk <pfv>buy[pv]=1m.act
‘He shouted at the stage using the new microphone that I bought.’
3. Brackets in glossing are used whenever a form expresses a meaning not denoted by any of
its individual morphemes. Consider, for example, the verb s<um>igaw in (1a), which has only
Note that I do not consider presence of a stem-deriving prefix (pag-, paN-, etc.)
in a verb form as a basis to put it into a separate paradigm, unlike the approach
developed in De Guzman (1978). The latter work, for instance, suggests that the
two morphemes: the root sigaw ‘shout’ and the actor voice infix <um>. The verb, however, also
expresses an aspect meaning, which is represented by the whole form as opposed to other forms
in its paradigm, rather than by either of the individual morphemes. Hence the use of pfv inside
the brackets in the gloss.
All examples are given in the traditional orthography, except for two cases. First, the non-
actor personal pronouns, which are usually treated as a combination of the marker sa and a
pronominal form with a space in between are given here as single units. These forms do not
pass the coordination test (it is unacceptable to conjoin two such pronouns in the following
way: *sa akin at kanya ‘obl 1m.act and 3m.act’), hence they score higher than two separate
lexical units on the wordhood scale. Second, lexical units that underwent morphophonological
changes, including nasal assimilation of the stem-deriving prefix paN- with the initial consonant
of the root and subsequent omission of the latter (e.g., paniniwala ← paN-ni~tiwala
‘nmlz-nmlz~belief ’), contraction of the infixes <um> ‘av’ and <in> ‘(I)pfv’ with the stem-
deriving prefixes pag-, paN-, ka-, the prefix paki- ‘COM’, or with the verb stem paging ‘become’
in actor-voice verbs, as well as ma- ‘mod’ (this one combines only with <in>) (e.g., nagluto ←
p<in>ag-luto ‘<pfv>stem-cook[av]’), contraction of the linker =ng with the final /n/ of the pre-
ceding form (e.g., noong ← noon=ng ‘dist.gen=lk’), and omission of vowels in some roots after
being augmented with affixes (e.g., kawalan ← ka-wala-an ‘nmlz-exist.neg-nmlz’, malaman ←
ma-alam-an ‘mod-known-pv[neut]’). All such units are given in their underlying forms.
I follow Wolff and De Guzman’s suggestion to treat the verbal prefixal forms mag-, maN-,
maki-, ma-, and nag-, naN-, naki-, na- as complex (Wolff 1973: 72; De Guzman 1978: 87), con-
sisting of the infix <um> or <in> and a corresponding stem-deriving prefix as it allows a more
adequate generalization of the verbal morphology in Tagalog.
Note that some forms are contracted in the traditional morphology (ng for naŋ ‘gen’, mga
for maŋa ‘pl’), /ŋ/ is represented with the digraph ‘ng’ (e.g., =ng for =ŋ ‘lk’), and the glottal
stop is not indicated (e.g, ang for ʔaŋ ‘nom’), unless it is used after a consonant within a lexical
unit (e.g., mag-araro ← p<um>ag-ʔararo ‘<av>stem-plough[neut]’), or in the initial position
within in a form with an infix (e.g., ʔ<um>uwi ‘<av>go_home[neut]’).
In many examples, the syntactic argument relevant for the discussion is highlighted in bold.
stem bigay is used to derive two voice forms, i-bigay pv and bigy-an DV, both
meaning ‘give’; while the stem pag-bigay is used to derive one voice form with
the same meaning, mag-bigay av, also meaning ‘give’ (De Guzman 1978: 146–147).
The present study, however, considers all voice forms derived from the same root
and denoting the same semantic situation as comprising a single voice paradigm.
Thus, the root bigay forms a paradigm with at least five members: mag-bigay av,
i-bigay pv, bigy-an rv, i-pag-bigay bv, and i-pam-bigay iv. This approach is prefer-
able because verbs of the same semantics are grouped together, rather than being
separated into several different paradigms on the sole basis of presence of stem-
deriving prefixes, whose use in the language is quite irregular. The downside of
De Guzman’s purely formal approach is also seen in the fact that there are forms
b<um>igay av ‘to give up’ and pag-bigy-an ev ‘to give in’, which have the same
stems as i-bigay pv and mag-bigay av, respectively, but describe different situ-
ations and, hence, should not be included in the same paradigm. All the voice
forms in (1) describe the same situation of shouting. Each one, however, requires
a different argument in the subject position.
The paradigm in (1) also shows that the Tagalog voice affixes do not function
alone to mark voice forms, but rather work in conjunction with other devices,
namely stem-deriving prefixes (pag-, paN-, ka-).
On the other hand, linguists adhering to the second approach suggest a list
of semantic participants which is normally defined without any formal basis. For
instance, Theme can be defined as a complement that serves as a stimulus of a
verb of emotion, or a complement which is a result of some change of state, or
a complement which is moved, while formally its marking is claimed to depend
on the semantics and subcategorization of the verb (Gallego 2015: 73). Different
studies based on this approach come up with different inventories of Tagalog
voices: for example, Schachter and Otanes distinguish 10 voices (1972), De Vos dis-
tinguishes eight voices (2011: 106), Malicsi distinguishes 16 voices (or focus con-
structions) (2013: 48–55), while Tanaka (2016: 36) distinguishes five voices. Cross-
linguistically, the same approach leads to very different voice inventories for
different Philippine languages. For instance, Batad Ifugao is said to have nine par-
ticipants (and voices) (Newell 2005: 40), while Tuwali Ifugao is said to have 14
participants (Hohulin and Hohulin 2014: 20–21). Thus, such inventories are based
upon individual linguists’ subjective beliefs on what abstract properties of refer-
ents are significant in a semantic situation (e.g., animacy, volitionality, deliberate-
ness, change of state, exhaustibility, etc.), hence the huge variation in the number
and types of participants in posited inventories. The scientific value of such defin-
itions of participants not rooted in any formal properties of the language is dubi-
ous, since classification of participants based on them is unfalsifiable: the reason
why a particular argument receives a particular interpretation is due to the way
a particular linguist views the situation described by the verb. The lack of con-
sensus regarding the nature and definitions of semantic participants, as well as
the controversies regarding building their inventories, their divisibility and delim-
itation are also recognized in the context of the general semantic theory (Dowty
1991: 547–548, 553–555; Kittilä and Zúñiga 2016).
This study suggests using strictly formal morphosyntactic criteria for estab-
lishing an exhaustive inventory of semantic participants and voices in Tagalog.
Although it is true that morphosyntactic marking of Tagalog arguments has long
been in the focus of the research tradition, studies dealing with it pursued other
research objectives and employed substantially different methods, which resulted
in very different conclusions. For instance, Schachter (1961) and Cena (1971) look
into the semantic ambiguity of argument and voice markers; Kess (1967) focuses
primarily on the compatibility of verbal stems with voice affixes and argument
structures of verbs, while the voice inventory in his study is equated with the inven-
tory of voice affixes; Ramos (1974) approaches the problem of classifying voice
morphology on the basis of an inventory of participants identified with the use of
semantic features; De Guzman (1978) seeks to organize the treatment of Tagalog
verbal morphology in a more adequate way through exploring its inflectional and
derivational properties. None of these studies provide a formal method of estab-
lishing an inventory of participants and voices. Such a formal approach has been
used to establish some of the Tagalog participants and voices in some studies (cf.
Schachter and Otanes’ description of the benefactive, “causative” and “referential”
verbs (1972: 312–314, 317)), and some of the marking strategies established below are
well known (cf. Ramos’ treatment of the agent, instrument, benefactive, and “the
affected noun phrase” (1974: 100, 103–105)), but the method has not been applied to
the whole voice system of Tagalog, which is the aim of this paper.
Section 2 presents the formal criteria suggested in this paper for distinguish-
ing semantic participants and voices, Section 3 describes different strategies of
argument marking, Section 4 presents participants and voices established with the
help of the method suggested here, and Section 5 presents the conclusion.
2. Formal criteria
The following set of formal criteria is suggested here to be used for establishing
classes of semantic participants available in Tagalog:
1. Distinct marking in non-subject positions. Since Philippine languages have
a very limited number of case marker sets, and, while some arguments can
be marked with several markers, the choice is always limited, it is suggested
here that the types and number of case marking strategies in principle can be
exhaustively described. In other words, sets of markers that a participant can
be used with in non-subject positions can be used to distinguish participants.
For Tagalog, such marking includes non-personal and personal analytic case
markers, and demonstrative and personal pronouns.
2. Co-occurring forms in the voice paradigm. If two voice forms with the same
propositional meaning require different arguments present in the situation in
the subject position, their subjects are two different participants.
3. Co-occurring participants. If two arguments co-occur in one construction,
they represent two different participants.
4. Existence of a co-referential4 voice form. Not all participants of verbal con-
structions can be made the subject or the predicate co-referenced with a
corresponding voice form, since the latter might be non-existent. Some par-
ticipants do not have co-referential voice forms, some can have such forms
only with some verbal roots, but not others, and some have co-referential
voice forms almost with all verbs they occur with.
The first criterion is considered to be the most important here, since all partici-
pants that can occupy a non-subject position receive certain marking, while the
other three criteria are employed to distinguish between participants in occasional
cases when different participants are used with the same marking in a non-subject
position.
It should be remembered that whatever names are given to participants estab-
lished with the help of these criteria, they are strictly arbitrary labels, since the
definitions have a purely formal basis. So, such labels could be substituted with
any other without any loss of observational power and, hence, should be treated
as mnemonic devices. Evidently, there is a fundamental difference between such
units and the semantic participants that are distinguished under the strictly seman-
tic approach. The latter are units of another, more abstract level, since they are
established a priori and often are not directly linked to facts of every individual
language. The units established in this paper can be likened to phonemes and mor-
phemes in that they are abstract units, whose inventories are unique for every par-
ticular language, while syntactic arguments in particular constructions are their
particular realizations, like allophones and allomorphs are particular realizations
4. Voice forms are considered here to be co-referential with a certain participant. The use of
this term is due to the fact that any content words in Philippine languages can function as
the predicate, the subject, or any other argument in the sentence. For instance, constructions
with a verb in the subject position and a substantive phrase in the predicate position cannot be
described as belonging to a certain voice, since, strictly speaking, the matrix clause in such con-
structions is non-verbal.
3. Argument markers
the Tagalog case system distinguishes three cases for all types of arguments –
non-personal substantives, personal nouns and pronouns – nom, gen, and obl
(e.g. (Reid and Liao 2004; Gallego 2015: 66)), suggesting that there is a complete
symmetry between the four systems. This is, however, unjustified, since two sys-
tems of case can be distinguished in Tagalog: that of non-personal substantives
and demonstratives with case grammemes nom, gen, and obl; and that of per-
sonal nouns and personal pronouns with case grammemes nom, act, and nact.
nom-markers are used to introduce any type of participants that can be co-
referenced with a verb in Tagalog, as well as the argument that is co-referenced
with the predicate in non-verbal constructions. Markers of other cases can be
used to introduce only some of the types of participants occurring in verbal con-
structions, as will be evident from the discussion below. Although the labels act
and nact might be too revealing at this stage regarding the participants they can
mark, the reasons for labeling them this way will become clear later.
Non-personal case markers include three forms – one for each case:
5. The forms in Table 2 are demonstrative pronouns, that is forms that can substitute other
constituents (nouns, substantivized adjectives and verbs). Tagalog also has another set of
demonstratives, homophonous to the nom forms of the demonstrative pronouns, e.g. iyon in (i):
i. Na-saan ang bahay na iyon?
prd-where nom house lk dist
‘Where is that house?’
Such forms possess only the category of proximity and lack the category of case.
Personal noun case markers comprise a paradigm of nine markers with two
categories – case and number. The pl forms have two variants, the second of
which is used only colloquially.
Traditionally, the markers ni and nina/nila are labeled as gen, while kay and kina/
kila are labeled as obl (Schachter and Otanes 1972: 93). However, this is erroneous,
since there is a clear discrepancy between the functions of the so-called “genitive”
and “oblique” non-personal case markers and “genitive” and “oblique” personal
case markers in many Philippine languages. In Tagalog, some participants that are
marked with ng when realized as a non-personal substantive have to be marked
with kay, rather than ni, when realized as a personal noun. Going ahead, the “gen-
itive” non-personal marker ng introduces Actor, Possessor, and a number of other
non-actor participants, while the “genitive” personal marker ni can be used only
to introduce Actor and Possessor (as well as to conjoin personal nouns in inclu-
sory constructions). Any non-actor participants realized as personal nouns have
to be marked with the “oblique” personal marker kay. Thus, it is justified to rein-
terpret the personal case marker paradigm (and the personal pronoun paradigm)
as including the nominative, actor (also marking possessors), and non-actor cases.
To give a concrete example, in (3a) non-personal Actor has to be marked with the
genitive non-personal marker ng, while personal Actor has to be marked with the
actor personal marker ni. In (3b), non-personal Patient has to be marked with ng,
while personal Actor has to be marked with kay.
(3) a. P<in>atay=siya ng kapatid=niya / ni Pedro.
<pfv>kill[pv]=3m.nom gen sibling=3m.act pers.sg.act pn
‘His brother/Pedro killed him.’
b. Siya ang p<um>atay ng kapatid=niya //*ni/
3m.nom nom <av>kill[pfv] gen sibling=3m.act pers.sg.act
kay Pedro.
pers.sg.nact pn
‘He is the one who killed his brother/Pedro.’
The 2m.nom pronoun has two forms: =ka is an enclitic, while ikaw is an indepen-
dent form that is used in the topic and predicate positions, as well as the subject
position when there are other constituents between it and the predicate.
There is another minimal membership dual pronoun – kita – which is difficult
to accommodate in any tabular form, since it stands for the non-existing combi-
nation =ko+ikaw (i.e. 1m.act+2m.nom).
Traditionally, actor pronouns are treated as counterparts of the genitive non-
personal marker, while non-actor pronouns are treated as counterparts of the
oblique non-personal marker (Schachter and Otanes 1972: 88). However, the same
reinterpretation of this paradigm can be done as in the case of the personal noun
case markers, since, for example, in (4a) non-personal Patient has to be marked
with the genitive marker ng, while personal Patient has to be realized as the non-
actor pronoun sakanya; and in (4b), non-personal Path is marked with ng, while
personal Path has to be realized as the non-actor pronoun saiyo:
(4) a. Ako ang p<um>atay ng tao // *=niya / sakanya6.
1m.nom nom <av>kill[pfv] gen human =3m.act 3m.nact
‘I am the one who killed a person/him.’
b. B<um>alik=ako ng bahay // *=mo / saiyo.
<av>return[pfv]=1m.nom gen house =2m.act 2m.nact
‘I returned to the house/you.’
Using the criteria presented above, 16 participants and 13 voices were found in
Tagalog.
4.1 Actor
Let us start with the participant which occupies the subject position with predicate
verbs with the voice affix <um>:
(5) a. K<um>ain=ako / =ito ng ma-sarap.
<av>eat[pfv]=1m.nom =prox.nom gen adj-deliciousness
‘I/This person ate something delicious.’
6. The “i” in the third person minimal membership pronouns is not pronounced in Manila
Tagalog and is frequently omitted in casual writing.
© 2019. John Benjamins Publishing Company
All rights reserved
12 Sergei Klimenko
In (5a) the subject position is occupied by the personal pronoun =ako ‘1m.nom’ or
demonstrative pronoun =ito ‘dem.prox.nom’, while in (5b) – non-personal noun
nanay ‘mom’ marked with the nom marker ang or personal noun Pedro. In non-
subject positions all these participants always occur with gen or act-marking:
(6) a. K<in>ain=ko/ =nito ang ma-sarap.
<pfv>eat[pv]=1m.act =prox.gen nom adj-deliciousness
‘I/This person ate the delicious (food).’
b. I-lu~luto ng nanay=ko / ni Pedro ang adobo=ng
pv-prsp~cook gen mom=1m.act pers.sg.act pn nom adobo=lk
iyon.
dist
‘My mom/Pedro will cook that adobo.’
b. P<in>ag-bigy-an=siya ni Pedro / ng
<pfv>stem-give-ev=3m.nom pers.sg.act pn gen
ka-ibig-an=niya / =nito.
ca-love-nmlz=3m.act =prox.gen
‘Pedro/His friend/This person gave in to him.’
4.2 Benefactive
Let us consider the following voice paradigm derived from the root luto ‘cooking’.
In addition to the forms above in (5b) and (6b), there at least five other voice forms
from the same root (the subjects are highlighted in bold):
(8) a. Lu~lutu-in=ko ang paborito=ko=ng pagkain.
prsp~cook-pv=1m.act nom favorite=1m.act=lk food
‘I will cook my favorite food.’
b. P<in>ag-lutu-an=ko ng adobo ang kaserola.
<pfv>stem-cook-lv=1m.act gen adobo nom pot
‘I cooked adobo in the pot.’
c. I-p<in>ag-luto=ko ng adobo ang pamilya=ko.
bv-<pfv>stem-cook=1m.act gen adobo nom family=1m.act
‘I cooked adobo for my family.’
d. Ni-lutu-an=niya=ako ng noodles.
pfv-cook-bv=3m.act=1m.nom gen noodles
‘He cooked noodles for me.’
e. I-p<in>aN-luto=ko ng adobo ang suka at toyo.
iv-<pfv>stem-cook=1m.act gen adobo nom vinegar and soy_sauce
‘I cooked adobo using the vinegar and toy sauce.’
Let us have a look at the construction in (8c) with the verb i-p<in>ag-luto.
The participant occupying the subject is introduced using the following strategy
in constructions of other voices – the preposition para ‘for’ combined with the
obl-forms of non-personal case markers and demonstrative pronouns or
nact-forms of personal case markers and personal pronouns:
(9) a. P<in>ag-luto=ako ng adobo para saiyo/ kay
<pfv>stem-cook[av]=1m.nom gen adobo for 2m.nact pers.sg.nact
Pedro/ sa pamilya=ko/ dito.
pn obl family=1m.act prox.obl
‘I cooked adobo for you/ Pedro/ my family/ this person.’
This verb, although different formally from the one in (8c), bears the same seman-
tics and requires the same participant in the subject position – hence it has to be
classified as belonging to the same voice. Besides, the construction in (10) cannot
be used with another Benefactive in a non-subject position:
(11) *Ni-lutu-an=niya=ako ng noodles para sa pamilya=niya.
pfv-cook-bv=3m.act=1m.nom gen for obl family=3m.act
Intended: ‘He cooked me noodles for his family.’
Not all verbs have a voice form that requires Benefactive in the subject position:
(12) a. P<in>alo ni Mario ang bata para sa asawa=ng
<pfv>hit[pv] pers.sg.act pn nom child for obl spouse=lk
si Marie.
pers.sg.nom pn
‘Mario hit the child for his wife Marie.’
b. *I-p<in>ag-palo ni Mario ng bata ang asawa=ng
bv-<pfv>stem-hit pers.sg.act pn gen child nom spouse=lk
si Marie.
pers.sg.nom pn
Intended: ‘Mario hit the child for his wife Marie.’
4.3 Instrument
The participant occupying the subject position in (8e) with the verb i-p<in>aN-
luto in constructions of other voices is marked either with gen-case markers or
the preposition sa pamamagitan, or is used as the subject of a modifier clause with
the predicate gamit ‘used’:
With some verbs this role can be introduced with the genitive case marker ng in
addition to the other strategies. This is impossible for the verbs derived from luto;
however it is possible for “verbs of surface contact” (palo ‘hit’, sampal ‘slap’, sin-
turon ‘belt’, taga ‘hack’, suntok ‘punch’, kurot ‘pinch’, saksak ‘stab’, hagupit ‘strike’,
kagat ‘bite’, pukpok ‘pound’, halo ‘mix’, hugas ‘wash’, punas ‘wipe’, walis ‘sweep’, linis
‘clean’) and “verbs of contact that result in a change in the (internal) structure of
objects” (durog ‘pulverize’, wasak ‘destroy’, hiwa ‘slice’, talop ‘peel’) (Ramos 1974: 35,
109. Cited via (Latroite 2011: 117)):
(14) B<in>asag=niya ng bato ang bintana. (Latrouite 2011: 115)
<pfv>break[pv]=3m.act gen stone nom window
‘He broke the window with a rock.’
Note that Latrouite’s analysis of why some verbs are allowed to occur with
ng-marked participants of this type, while others are not, is, probably, incorrect. She
states that only verbs denoting events “with inherent Instrument arguments”, like
hagupit ‘the act of hitting with a whip’ and halo ‘pestle’, can occur with this strat-
egy (Latrouite 2011: 118). However, some of the roots in the list above do not contain
a specific instrument in the interpretation of their semantic situation: e.g., sampal
‘slap’, linis ‘clean’ etc. do not refer to situations with a specific instrument involved,
since it can be a variety of objects (e.g., a palm, flip-flop, money, etc. in the situation
of slapping), although some instrument has to be involved – it is impossible to per-
form the action without it. On the other hand, the same is true regarding roots like
luto ‘cook’, as well: it is impossible to cook anything using nothing.
Some verbs require a participant which can be marked either with gamit ang
or sa pamamagitan ng, or sa, rather than ng:
(15) a. Ang tawag sa tao=ng p<in>ag-su~sulat *ng / sa / gamit
nom call obl person=lk <ipfv>stem-ipfv~write[av] gen obl used
ang kaliwa=ng kamay ay kaliwete.
nom left=lk hand top left-handed
‘People who write with their left hand are called left-handed.’
Note that the sa-marking can be blocked when there is another overt participant,
like Patient in the following example:
(16) P<in>ag-sulat=ako ng liham *ng / ?sa / gamit ang kaliwa=ng
<pfv>stem-write[av]=1m.nom gen letter gen obl used nom left=lk
kamay.
hand
‘I wrote a letter with my left hand.’
Apart from non-personal substantives, only demonstratives can be used just with
gen-marking as a realization of this participant (17), while personal pronouns (18)
and personal nouns (19) can be used only either with sa pamamagitan ng, or gamit:
(17) a. P<in>alo=niya=ako=nito.
<pfv>hit[pv]=3m.act=1m.nom=prox.gen
‘He hit me with this.’
b. La~laban=tayo sa pamamagitan=niyan.
prsp~fight[av]=12nm.nom obl help=med.gen
‘We will fight with its help.’
c. T<in>akot=niya=ako gamit=iyan.
<pfv>fear[pv]=3m.act=1m.nom used=med.nom
‘He scared me with that.’
(18) a. Sa pamamagitan=mo, m<in>a-la~labas=namin
obl help=2m.act <ipfv>mod-ipfv~exit[pv.take_out]=1nm.act
ang mga bagay na hindi ma-pa-labas ng
7
nom pl thing lk neg mod-caus-exit[pv.neut. take_out] gen
amin=ng ma-hina=ng puso.
1nm.gen=lk adj-weakness=lk heart
‘With your help we express things that cannot be expressed by our weak
hearts.’
b. Gusto=ko=ng ma-kamit yun gamit ikaw.
desired=1m.act=lk mod-achieve[pv.neut] dist.nom used 2m.nom
‘I want to achieve that using you.’
7. Neutral aspect in this paper refers to what is sometimes referred to as infinitive in Tagalog.
4.4 Cause
The participant in the subject position in (1f) with the verb i-k<in>a-sigaw in
other voice constructions has to be introduced either with the oblique marker sa
or the prepositions dahil sa, gawa ng:
(21) S<um>igaw=siya sa / dahil sa / gawa ng pag-tulak ng
<av>shout[pfv]=3m.nom obl because obl because gen nmlz-push gen
ka-tabi=niya sakanya.
ca-side=3m.gen 3m.obl
‘He shouted because of his seatmate’s pushing him.’
4.5 Reference
Same marking strategies are also applied to demonstratives (28), personal nouns
(29), and personal pronouns (30) used for realization of this participant:
(28) a. Ano ang dapat=ko=ng ma-alam-an tungkol=dito?
what nom necessary=1m.act=lk mod-known-pv[neut] about=prox.obl
‘What do I have to know about this?’
b. Wala sa kahit na sino=ng kampo ang p<in>ag-salita
exist.neg obl even lk who=lk camp nom <pfv>stem-word[av]
hinggil=dito.
about=prox.obl
‘Nobody’s camp spoke regarding this.’
(29) a. Alden at Derrick, p<in>ag-usap tungkol kay
Alden and Derrick <pfv>stem-converse[av] about pers.sg.nact
Maine!
pn
‘Alden and Derrick talked about Maine!’
b. P<in>anaginip si Basilio hinggil kay Crispin.
<pfv>dream[av] pers.sg.nom pn about pers.sg.nact pn
‘Basilio had a dream about Crispin.’
4.6 Patient
The constructions with i-s<in>igaw (1b) and with i-lu~luto (6b) and lu~lutu-in
(8a) contain participants in the subject position which manifest the same behav-
ior in a non-subject position:
(32) a. Ako ang s<um>igaw ng / sa pangalan=mo.
1m.nom nom <av>shout[pfv] gen obl name=2m.act
‘I was the one who shouted your name.’
b. S<um>igaw=ako ng / *sa pangalan=mo.
<av>shout[pfv]=1m.nom gen obl name=2m.act
‘I shouted your name.’
(33) a. Ang nanay=ko ang p<in>ag-luto ng / sa adobo=ng iyon.
nom mom=1m.act nom <pfv>stem-cook[av] gen obl adobo=lk dist
‘My mom was the one who cooked that adobo.’
b. P<in>ag-luto ng / *sa adobo ang nanay=ko.
<pfv>stem-cook[av] gen obl adobo nom mom=1m.act
‘My mom cooked adobo.’
This is probably also the reason why the following construction with the partici-
pant in question, here realized as a personal noun, describing a situation where an
infant named Pedro is passed on to somebody else, although recognized as well-
formed by my informants, since it is the only way to directly express the intended
meaning with this verb in Tagalog, is considered by some as somewhat unnatural:
(36) Ako ang p<in>ag-ʔabot kay Pedro sakaniya.
1m.nom nom <pfv>stem-reach[av] pers.sg.nact pn 3m.nact
‘I was the one who passed Pedro on to him.’
9. As Latrouite notes, “[m]ultilinguality is the standard in the Philippines” and in spite of com-
ing from Manila, all of her informants speak some other Philippine language as well (Latrouite
2011: 125). This, however, indicates that they are either immigrants from other areas outside the
Tagalog-speaking region, or children of immigrants from such areas who still use their parents’
language at home. Tagalog monolinguality and Tagalog-English bilinguality is the norm for Fil-
ipinos from the Tagalog-speaking region who have never been immersed in any other Philip-
pine language-speaking environment.
In (38a) and (38b), two different voice forms derived from the root lakad ‘walk’
are used – ni-lakar-an and ni-lakad, respectively – with two different participants
occupying the subject position – ang putik… and ang nais punta-han, respectively.
The semantics of the two forms is different, as can be seen from the examples.
If we try to put the two participants into a non-subject position, we will see
that they behave differently. While the participant in (38a) has to be marked in
the oblique case (sa putik…(39a)), the one in (38b) is acceptable with the oblique
marking only for some speakers and otherwise can occur as an obl-complement
of a path predicate in a modifier clause (pa-tungo sa nais punta-han) (39b):
(39) a. P<in>ag-lakad=ako sa putik na ʔ<in>i~ʔiwas-an ng
<pfv>stem-walk[av]=1m.nom obl mud lk <ipfv>ipfv~avoid-pv gen
ilan.
some
‘I walked on the mud that some people avoid.’
b. P<in>ag-lakad si Fred ?sa / pa-tungo sa nais
<pfv>stem-walk[av] pers.sg.nom pn obl dir-head obl desired
punta-han.
go-PaV[neut]
Intended: ‘Fred walked to where he wanted to go.’
(40a) illustrates the form t<in>akbo from the root takbo ‘run’, while (40b) illus-
trates the form t<um>akbo from the same root. The construction with the latter
can be used with the same participant as the one in the subject position in (40a) –
pinto – marked with the oblique case (sa). The marking with a path form is equally
possible (pa-punta sa).
There is also another difference in the possibilities for marking these two
kinds of participants: the participant like the one occupying the subject position
with the verb t<in>akbo in (40a) can be marked with the case marker ng in a non-
subject position when the participant is expressed with a non-personal substan-
tive in the peripheral position (this strategy is not available for non-personal sub-
stantives in the fronted position, pronouns, personal nouns and abstract nouns –
in all these cases only the oblique case form can be used), while the participant
like the one occupying the subject position in (38a) cannot:
(41) a. L<um>ipad=siya sa / *ng ere.
<av>fly[pfv]=3m.nom obl gen air
‘He flew in the air.’
b. L<um>ipad=siya sa / ng Pilipinas.
<av>fly[pfv]=3m.nom obl gen Philippines
‘He flew to the Philippines.’
When marking participants like the one in (41b), if a verb ends in a vowel the
marker ng can be used in a reduced form, phonetically identical to the form of the
linker =ng used after words ending in vowels, /n/ or the glottal stop:
(42) M<in>a-pag-desisyun-an=niya=ng p<um>unta=ng Maynila.
<pfv>mod-stem-decision-RefV=3m.act=lk <av>go[neut]=gen Manila
‘He decided to go to Manila.’
This reduced form can also occur attached to pronominal subjects, since, as encl-
itics, they have to immediately follow the predicate head and, thus, precede any
complement:
(43) D<um>ayo=siya=ng Maynila noon=ng 1967.
<av>go_from_afar[pfv]=3m.nom=gen Manila dist.gen=lk
‘He visited Manila in 1967.’
Some Tagalog speakers’ grammar seems to have a limitation on the use of ng with
indefinite participants of the kind, while for others this limitation does not exist.10
10. Note that Malicsi (2013: 91) claims that non-personal substantives are marked with ng when
indefinite and sa when definite. This is not confirmed by the data from my informants, as seen
from Example (44).
Note that Latrouite’s observation that only locational proper names are commonly
marked with ng when occurring with motion verbs (2011: 122) is incorrect, since
non-proper nouns can be used with the same strategy as well:
(45) P<um>unta=ako sa / ng tinda-han.
<av>go[pfv]=1m.nom obl gen sell-loc
‘I went to the store.’
The construction in (46) contains two participants marked with the oblique case
marker sa – sa UP and sa damo. Only one of them is similar to the participant sa
putik… in (39a), namely sa damo. Note that it can become the subject in a con-
struction with the form ni-lakar-an in (47a), while UP cannot (47b):
(47) a. Ni-lakar-an=ko ang damo.
pfv-walk-lv=1m.act nom grass
‘I walked on the grass.’
b. *Ni-lakar-an=ko ang UP.
pfv-walk-PaV=1m.act nom
Intended: ‘I walked in UP.’
The participant of the type in (39a) cannot be realized as a personal noun or per-
sonal pronoun in a non-subject position. When realized as a demonstrative, it has
to be used in the oblique case:
Let us call this participant Location, while the other participant in (46) can be
labeled Place. The latter can also occur in non-verbal constructions and cannot be
made the subject, so it will not be included in the discussion below.
The participant of the type in (45) when realized as a demonstrative can be
marked only with the oblique case (49a) or as an oblique complement of a path
predicate in a modifier clause (49b), while for personal nouns and personal pro-
nouns, apart from the latter strategy ((50b) and (51b)), only nact-marking is
available ((50a) and (51a)):
(49) a. P<um>unta=ako doon.
<av>go[pfv]=1m.nom dist.obl
‘I went there.’
b. P<um>unta=ako pa-tungo doon.
<av>go[pfv]=1m.nom dir-head dist.obl
‘I headed there.’
(50) a. P<um>unta=siya kay Hesus.
<av>go[pfv]=3m.nom pers.sg.nact pn
‘He went to Jesus.’
b. T<um>akbo=siya pa-punta kay Nadine.
<av>run[pfv]=3m.nom dir-go pers.sg.nact pn
‘She ran to Nadine.’
(51) a. L<um>ipat=ako sakaniya.
<av>move[pfv]=1m.nom 3m.nact
‘I moved to his place.’
b. T<um>alon=ako pa-layo sakaniya.
<av>jump[pfv]=1m.nom dir-move_away 3m.nact
‘I jumped away from him.’
Location can be confused with another participant. Consider examples (1c) and
(1d), repeated here:
(52) a. S<in>igaw-an=niya=ako.
<pfv>shout-rv=3m.act=1m.nom
‘He shouted at me.’
b. P<in>ag-sigaw-an=niya ng utos sa mga sundalo ang
<pfv>stem-shout-lv=3m.act gen order obl pl soldier nom
balkonahe=ng ito.
balcony=lk prox.nom
‘He shouted the orders to the soldiers from this balcony.’
(52a) and (52b) contain two different voice forms derived from sigaw ‘shout’ –
s<in>igaw-an and p<in>ag-sigaw-an, respectively. The two forms have different
participants in the subject position, since in (52a) the subject is the recipient of the
message (ako), while in (52b) the recipient occurs in the oblique case form (sa mga
sundalo), while the subject position is occupied by a participant which can be rec-
ognized as Location (ang balkonahe=ng ito), since the non-personal substantive
phrase used for its realization has to be marked in the oblique case and cannot be
substituted with a personal noun or personal pronoun:
(53) S<um>igaw=siya sa balkonahe=ng ito.
<av>shout[pfv]=3m.nom obl balcony=lk prox
‘He shouted from this balcony.’
So, both participants have the same marking in a non-subject position when real-
ized as a non-personal substantive.
The participant sa mga sundalo in (52b), however, unlike Location, can also be
realized as a demonstrative in the oblique case (54a), personal noun in the nact
case (54b), and personal pronoun in the nact case (54c):
(54) a. S<um>igaw=siya dito sa bata=ng ito.
<av>shout[pfv]=3m.nom prox.obl obl child=lk prox
‘He shouted at this child.’
b. May s<um>igaw kay Carlo.
exist <av>shout[pfv] pers.sg.nact pn
‘Somebody shouted to Carlo.’
c. Siya ang s<um>igaw saakin.
3m.nom nom <av>shout[pfv] 1m.nact
‘She was the one who shouted at me.’
The participants saiyo and para saiyo here, however, belong to different types,
since they can cooccur within one construction:
(56) P<in>ag-dala=ako ng pagkain sa mga tao doon para
<pfv>stem-carry[av]=1m.nom gen food obl pl human dist.obl for
kay Pedro.
pers.sg.nact pn
‘I took food to people there for Pedro.’
4.9 Measure
In the following two pairs of examples, the participants occupying the predicate
position in (57a) and (58a) with the verb t<in>akbo occupying the subject position
(verb forms of this voice are normally used in the subject position, rather than as
the predicate head) have to be marked with ng when in a non-subject position,
while sa-marking is unacceptable ((57b) and (58b)):
(57) a. Sampu=ng metro ang t<in>akbo=ko.
ten=lk meter nom <pfv>run[MV]=1m.act
‘I ran ten meters.’
b. T<um>akbo=ako ng / *sa sampu=ng metro.
<av>run[pfv]=1m.nom gen obl ten=lk meter
‘I ran ten meters.’
(58) a. Kalahati=ng oras ang t<in>akbo=ko.
half=lk hour nom <pfv>run[MV]=1m.act
‘I ran for half an hour.’
b. T<um>akbo=ako ng / *sa kalahati=ng oras.
<av>run[pfv]=1m.nom gen obl half=lk hour
‘I ran for half an hour.’
4.10 Malefactive
Tagalog has a group of verbs which are affixed with the modal prefix ma- and
whose subject position is occupied by a participant (=ako in (60a)) that in a non-
subject position (=ko in (60b)) has to be expressed as a possessor of Actor (ng bait
in (60a)), and cannot be expressed in any other way (60c):
(60) a. M<in>a-sira-an=ako ng bait.
<pfv>mod-break-MalV=1m.nom gen kindness
‘I got crazy.’
b. K<in>a-sira ang bait=ko.
<pfv>stem-break[av] nom kindness=1m.act
‘I got crazy.’
c. *K<in>a-sira saakin ang bait.
<pfv>stem-break[av] 1m.nact nom kindness
Intended: ‘I got crazy.’
Most verbs in this voice cannot be used without the modal prefix:
(61) *S<in>ira-an=ako ng bait.
<pfv>break-MalV=1m.nom gen kindness
Intended: ‘I got crazy.’
4.11 Experiencer
Some Tagalog verbs marked with voice affixes other than <um> do not have coun-
terparts of any other voices. Note that the following constructions are particularly
difficult to classify within the second, purely semantic approach to establishing the
voice inventory:
(62) a. M<in>a-hirap-an=ako sa math.
<pfv>mod-difficulty-ev=1m.nom obl math
‘I found math difficult.’
Although many of the roots used in the examples above can be the derivation base
for other voice forms, they cannot be grouped into a single paradigm, since they
either have different propositional meanings, or the constituent occupying the
subject position cannot be expressed in a non-subject position with verbs of other
voices. An example of the first case is the verb p<in>ag-sabi-han in (62g), derived
from sabi ‘said’, which also forms other voices, like s<in>abi ‘ <pfv>say[pv]’ and
s<in>abi-han ‘<pfv>say-rv’. However, the meaning of the former differs from
that of the latter two in that it refers to the action of telling off, rather than saying
something, which is common for both s<in>abi and s<in>abi-han. An example
of the second case is the verb i-kain in (62b), derived from kain ‘eat’, which also
forms, for instance, k<um>ain ‘<av>eat[neut]’. The subject constituent ang prob-
lema=mo in (62b), however, cannot be expressed with the latter voice form.
The participant occupying the subject position with such verbs will be labeled
Experiencer.
4.12 Collateral
Some participants can be used only in a non-subject position marked with ng, and
they cannot have a personal reference, that is they are always expressed with a
non-personal substantive or demonstrative pronoun:
(63) a. ʔ<um>ulan ng pungla.
<av>rain[pfv] gen bullet
‘It rained with bullets.’
b. ʔ<in>ulan-an=kami ng pungla.
<pfv>rain-ev=1nm.nom gen bullet
‘We got caught in a rain of bullets.’
c. P<in>ag-ka-roon ng bagyo.
<pfv>stem-inch-dist.obl[av] gen typhoon
‘A typhoon appeared.’
d. L<um>ipat=ako ng pwesto.
<av>move[pfv]=1m.nom gen post
‘I changed my position.’
Note that the verb l<um>ipat in (63d) can also form another construction – with
a participant that can be marked with either sa, ng or units like pa-punta sa, etc.
The meaning of such a construction is different:
(64) L<um>ipat=ako sa / ng / pa-punta=ng France.
<av>move[pfv]=1m.nom obl gen dir-go=gen
‘I moved to France.’
The gen-participant in (63d) denotes the object of change. This position can be
also occupied by such nouns like upuan ‘seat’, tirahan ‘residence’, bansa ‘country’,
trabaho ‘work’, etc., but contains no information on the source and goal of moving,
while the participant in (64), referring to the goal of moving, is Path. Note that
the latter can be made the subject or the predicate with a verb of a corresponding
voice (65a), while the one in (63d) cannot (65b):
(65) a. France ang ni-lipat-an=ko.
nom pfv-move-PaV=1m.act
‘France was where I moved to.’
b. *Pwesto ang ni-lipat-an=ko.
post nom pfv-move-V=1m.act
Intended: ‘Position was what I changed.’
4.13 Counter-agent
Comitative verbs require presence of a participant which has the same marking
as Recipient – the oblique case for non-personal substantives and demonstratives
and non-actor case for personal nouns and personal pronouns. The difference
with Recipient is that there are no voice forms with which this participant can be
made the subject. Thus, we can label it as Counter-Agent. Recipient and Counter-
Agent do not seem to be able to co-occur in one construction, so it might be jus-
tifiable to treat the latter as a subtype of the former.
4.14 Causee
Hence, the voice form in (68) that requires this participant in the subject position
can be labeled Patient voice:
(68) P<in>a-tayo=niya=ako sa harap-an=niya.
<pfv>caus-stand[pv]=3m.act=1m.nom obl front-loc=3m.act
‘He made me stand up in front of him.’
Causative verbs derived from verbs with Patient require an additional argument
that behaves like Recipient in a non-subject position, since it can be realized as a
non-personal substantive with obl-marking (69a), or a personal pronoun (69b):
(69) a. Isa=ng titser, p<in>ag-pa-kain ng alikabok *ng / sa
one=lk teacher <pfv>stem-caus-eat[av] gen dust gen obl
estudyante=niya.
student=3m.act
‘A teacher fed dust to her student.’
b. Puro mangga ang i-pa~pa-kain=niya=saakin.
only mango nom pv-prsp-caus-eat=3m.act=1m.nact
‘It’s only mango that he will feed me.’
Latrouite notes that if the “Undergoer” of the verb has “dative” marking, the other
participants has to be marked in genitive, giving the following example as an evi-
dence (2011: 133):
(70) P<in>ag-pa-tulong=siya ng / *sa bata sa mga guro.
<pfv>stem-caus-help[av]=3m.nom gen obl child obl pl teacher
‘He made/let a/the child help the teachers.’ (Latrouite 2011: 134)
4.15 Identifier
Such verbs do not have voice forms which would require this participant to
occupy the subject position. Let us call this participant Identifier.
At least one verb in Tagalog – maging ‘to become’ – requires an argument
which is adjoined to the verb without any overt markers:
(75) P<in>aging titser=ako.
<pfv>become[av] teacher=1m.nom
‘I became a teacher.’
Although Schachter and Otanes claim that this verb is “similar to the verbal pre-
fixes in being obligatorily non-pre-enclitic” (1972: 396), constructions like (76)
with maging in a pre-enclitic position are acceptable for native speakers as well:
(76) P<in>aging=ako=ng titser.
<pfv>become[av]=1m.nom=lk teacher
‘I became a teacher.’
Still, the wordhood status of this verb is under question, since it is more preferably
used with enclitics following its argument, like in (75).
Table 5 shows the sets of markers available for different participants in non-subject
positions.
All the participants above, except for Measure and Collateral and Recipient,
Counter-agent and Causee, differ in the sets of marker available for them in a
non-subject position. Apart from this criterion, co-occurrence of voice forms in
the paradigm was used to distinguish between Location and Path, and Recipi-
ent and Causee; co-occurrence of participants was used to distinguish between
Recipient and Causee; and existence of a co-referential voice form was used to
distinguish between Location and Place, Measure and Collateral, and Recipient
and Counter-agent.
Among these participants, only 13 can occupy the subject position with a
verb of appropriate voice, as can be seen from the discussion above. Hence, 13
voices can be distinguished in Tagalog: Actor, Benefactive, Instrumental, Causal,
The verb pag-ʔipun-an in (78) has an Actor voice counterpart, which requires an
argument marked with the preposition para (79), hence it is a Benefactive voice
verb. The verb pag-planu-han will be discussed in the conclusion.
(79) P<in>ag-ʔipon=ako para dito.
<pfv>stem-save[av]=1m.nom for prox.obl
‘I saved for this.’
Although in many instances the labels applied to participants here and those used
traditionally coincide (e.g., Actor, Patient, Cause, etc.), the method suggested here
allows to group together on the basis of the same morphosyntactic marking some
participants that are labeled differently in studies that employ the purely seman-
tic approach. For instance, sa gilid in (80) is claimed to be Location by Gallego
(2015: 73):
(80) …pa-simple=ng ʔ<um>ihi sa gilid. (Gallego 2015: 73)
man-simple=lk <av>urine[pfv] obl side
‘…simply peed on the side.’
However, this participant can be also realized as a personal noun or personal pro-
noun (81a) and put into the subject position (81b), hence it is Recipient:
(81) a. Ito ang aso=ng ʔ<um>ihi saakin.
prox.nom nom dog=lk <av>urine[pfv] 1m.nact
‘This is the dog that peed on me.’
b. Lalaki, ʔ<in>ihi-an ang kapwa pasahero dahil
male <pfv>urine-rv nom fellow passanger because
p<in>ag-bawal-an p<um>aN-sigarilyo at
<pfv>stem-forbidden-rv <av>stem-cigarette[neut] and
ʔ<um>inom ng alak.
<av>drink[neut] gen alcohol
‘A man peed on his fellow passenger because he was forbidden to smoke
and drink alcohol.’
When in the subject or predicate position, pakpak has to be used with an Actor
voice verb (83a) and this participant has to be marked as act when realized with
a personal noun or pronoun (83b), hence it is Actor:
(83) a. Pakpak ang t<um>ubo saakin.
wing nom <av>grow[pfv] 1m.nact
‘Wings was what grew on me.’
b. B<um>alik=siya sa lupa=ng t<in>ubu-an=niya.
<av>return[pfv]=3m.nom obl land=lk <pfv>grow-rv=3m.act
‘That’s the land where he grew.’
Two identical verbal forms can be of different voices. Let us consider the following
two constructions with the same verbal form ni-lipar-an:
(84) a. Ni-lipar-an=siya ng palda.
pfv-fly-MalV=3m.nom gen skirt
‘Her skirt got blown up by the wind.’
b. Ni-lipar-an=siya ng ipis.
pfv-fly-PaV=1m.nom gen cockroach
‘A cockroach flew to her.’
This verbal form bears different meanings in these two examples, which is
reflected in the way the subject participants have to be marked in a non-subject
position:
(85) a. Ni-lipad ang palda=niya.
pfv-fly[pv] nom skirt=3m.act
‘Her skirt got blown up by the wind.’
b. L<um>ipad (pa-punta) sakanya ang ipis.
<av>fly[pfv] dir-go 3m.nact nom cockroach
‘The cockroach flew to her.’
The two constructions differ in the number of allowed arguments, as (86b) does
not allow Actor =niya. Moreover, only the verb in (86a) can have a non-modal
form (87a), while the latter is not allowed in construction (87b), hence the verb in
(86b) is an Actor voice form:
(87) a. P<in>a-ihi=niya=ako sa tawa.
<pfv>caus-urine[pv]=3m.act=1m.nom obl laugh
‘He made me pee because of laughing.’
b. *P<in>a-ihi=ako (dahil) sa tubig.
<pfv>caus-urine[pv]=1m.nom because obl water
Intended: ‘I had to pee because of water.’
5. Conclusion
The difference between the units established here and those in other works that
build on purely semantic considerations can be illustrated with Instrument. For
Tagalog, Instrument is defined here as the participant that can be realized in
a non-subject position with the preposition sa pamamagitan gen/act ‘with the
help of ’ (or gamit nom ‘using…’). There are also three subtypes of Instrument
that are established on the basis of the additional possibility of being marked
with gen or obl without a preposition for non-personal substantives and
demonstrative pronouns. All the three subtypes have the same marking possibil-
ities when realized as personal nouns or pronouns. Thus, the label “Instrument”
Thus, there is a question of how the verb pag-planu-han in (78) should be clas-
sified. Such verbs that require arguments that can be marked as different partici-
pants do not seem to be very frequent in Tagalog, however at present it is unclear
how many of them exist in the language.
Acknowledgements
This study would have been impossible without the assistance of the following people: Anne
Austria Abcede, Aureen Austria Abcede, Berna Austria, Jei Ente, Paul Julian Santiago, Ron San
Juan. This research was funded by the Russian Science Foundation (project no. 18–18–00472,
“Causal Constructions in World Languages (Semantics and Typology)”).
Abbreviations
References
Apresjan, Juri D. 1995. Lexical semantics (synonymic means of language) (In Russian:
Лексическая семантика (синонимические средства языка)). Moscow: Vostochnaya
Literatura.
Apresjan, Juri D. 2010. A prospectus of the active dictionary of Russian (In Russian: Проспект
активного словаря русского языка). Moscow: Yazyki slavyanskih kul’tur.
Cena, Resty M. 1971. Case opposition in Tagalog. The Ohio State University Working Papers in
Linguistics 10. 133–148.
Chang, Anna Hsiou-chuan. 2006. A Reference Grammar of Paiwan. Canberra: Australian
National University PhD dissertation.
Cunningham, Margaret C. & Joan E. Goetz. 1963. Pronoun formatives in Amganad Ifugao.
Manila: SIL. Goudswaard, Manila: Linguistic Society of Philippines and SIL.
De Guzman, Videa P. 1978. Syntactic derivation of Tagalog verbs (Oceanic Linguistics Special
Publications). Honolulu: University of Hawaii Press.
De Vos, Fiona. 2011. Essential Tagalog Grammar.
Dowty, David. 1991. Thematic Proto-Roles and Argument Selection. Language 67.3. 547–619.
https://doi.org/10.1353/lan.1991.0021
Foley, William A. 1997. Anthropological Linguistics: an introduction. Oxford: Basil Blackwell.
Foley, William A. 1998. Symmetrical Voice Systems and Precategoricality in Philippine
Languages. In Miriam Butt & Tracy Holloway King (eds.), Proceedings of the LFG 98
Conference: Workshop on voice and grammatical functions in Austronesian languages,
45–59. Stanford: CSLI.
Gallego, Maria Kristina S. 2015. Ang mga Nominal Marker ng Filipino at Ivatan. Daluyan
Journal ng Wikang Filipino, tomo XXI 2015. 65–95.
Himmelmann, Nikolaus P. 2005. The Austronesian languages of Asia and Madagascar. In
K. Alexander Adelaar & Nikolaus Himmelmann (eds.), The Austronesian languages of
Asia and Madagascar, 110–181. London: Routledge.
Himmelmann, Nikolaus P. 2008. Lexical categories and voice in Tagalog. In Simon Musgrave
& Peter K. Austin (eds.), Voice and Grammatical Functions in Austronesian Languages,
247–293. Stanford: CSLI Publications.
Hohulin, Richard M. & Elma Lou Hohulin. 2014. Tuwali Ifugao dictionary ang grammar sketch.
Manila: Linguistic Society of the Philippines.
Kess, Joseph Francis. 1967. Syntactic Features of Tagalog Verbs. University of Hawai’i PhD
dissertation.
Kittilä, Seppo & Fernando Zúñiga. 2016. Recent developments and open questions in the field
of semantic roles. In Seppo Kittilä & Fernando Zúñiga (eds.), Advances in Research on
Semantic Roles (Benjamins Current Topis 88), 1–26. Amsterdam/Philadelphia: John
Benjamins. https://doi.org/10.1075/bct.88.01kit
Kroeger, Paul. 1993a. Another look at subjecthood in Tagalog. Philippine Journal of Linguistics
24.2. 1–16.
Kroeger, Paul. 1993b. Phrase structure and grammatical relations in Tagalog. Stanford: Center
for the Study of Language and Information.
Latrouite, Anja. 2011. Voice and Case in Tagalog: The coding of prominence and orientation.
Dusseldorf: Heinrich-Heine Universität PhD dissertation.
Malicsi, Jonathan C. 2013. Gramar ng Filipino. Quezon City: Sentro ng Wikang Filipino.
Newell, Leonard E. 2005 (1993). Batad Ifugao Dictionary with Ethnographic Notes. Manila:
Linguistic Society of the Philippines.
Rachkov, G. E. 1981. Introduction to morphology of contemporary Tagalog (In Russian: Vvedenie
v morfologiyu sovremennogo tagal’skogo yazyka). Leningrad State University.
Ramos, Teresita V. 1974. The Case System of Tagalog Verbs. Pacific Linguistics, Series B. No. 27.
Canberra: The Australian National University.
Reid, Lawrence A. 2005. Tagalog and Philippine languages. In Philipp Skutch (ed.),
Encyclopedia of linguistics. New York: Routledge.
Reid, Lawrence A. 2008. Tagalog. In Bernard Comrie (ed.), The World’s Major Languages (2nd
edition), chapter 47. London: Routledge.
Reid, Lawrence A. & Hsiu-chuan Liao. 2004. A Brief Syntactic Typology of Philippine
Languages. Language and Linguistics 5(2). 433–490.
Ross, Malcolm & Stacy Fang-ching Teng. 2005. Formosan Languages and Linguistic Typology.
Language and Linguistics 6:4. 739–781.
Schachter, Paul. 1961. Structural ambiguity in Tagalog. Language Learning 11. 135–145.
https://doi.org/10.1111/j.1467‑1770.1961.tb00751.x
Schachter, Paul. 1976. The subject in Philippine languages: topic, actor, actor-topic, or none of
the above. In Charles N. Li (ed.), Subject and Topic, 491–518. New York: Academic Press.
Schachter, Paul. 1977. Reference-related and role-related properties of subjects. In P. Cole &
J. M. Sadock (eds.), Syntax and Semantics, Volume 8: Grammatical Relations, 279–305.
New York: Academic Press.
Schachter, Paul. 1993. The Subject in Tagalog: Still None of the Above. UCLA Occasional Papers
in Linguistics, Number 15. Los Angeles: University of California Press.
Schachter, Paul & Fe Otanes. 1972. Tagalog reference grammar. Berkeley: University of
California Press.
Starosta, Stanley, Andrew Pawley, & Lawrence A. Reid. 2009. The evolution of focus in
Austronesian [Unabridged version of paper published in 1982]. In Elizabeth Zeitoun (ed.),
Formosan linguistics: Stanley Starosta’s contributions (Language and Linguistics
Monograph Series C6), vol. 2, 329–481. Taipei: Academia Sinica.
Tanaka, Nozomi. 2016. An asymmetry in the acquisition of Tagalog relative clauses. PhD
dissertation. Manoa: University of Hawai’i.
Wolff, John U. 1973. Verbal inflection in Proto-Austronesian. In Andrew Gonzalez (ed.),
Parangal kay Cecilio Lopez: Essays in honor of Cecilio Lopez on his seventy-fifth birthday,
71–91. Quezon City: Linguistic Society of the Philippines.
Sergei Klimenko
Typology department
Institut lingvisticeskih issledovanij RAN
9, Tuchkov lane
Saint Petersburg, 199053
Russia
sergeybklimenko@gmail.com
https://orcid.org/0000-0001-8901-1037