What Philippine Ergativity Really Means

DRAFT/Comments Welcome/

WHAT PHILIPPINE ERGATIVITY REALLY MEANS Ricardo Ma. Nolasco University of the Philippines (Diliman) ([email protected]) 1.

Introduction:

The following Tagalog sentences illustrate how the “focus” system works in Philippine languages (PLs). Actor (1) Tumakbo NEUT.run

ang

weyter.

PART1

waiter

`The waiter ran.’ Actor (2) `Magtakbo NEUT.bring

ka

nga

ng

tubig sa

table no. 5 ’

you (the waiter) please PART2 water PART3 table no. 5 `Please bring (or rush) some water to table no. 5’

Goal (theme) (3) `Itakbo

mo ang

tubig (na ito) sa

NEUT. bring you PART1 water (this) `Bring (or rush) this water to table no. 5.

Goal (beneficiary) (4) `I(pag)takbo mo nga

ng

table no. 5.’

PART3 table no. 5

Marlboro sa

tindahan ang

table no. 5.’

NEUT. run.get you please PART2 Marlboro PART 3 store PART1 table no. 5 `Please run to the store and get the customer at table no. 5 some Marlboro cigarettes.’

Goal (patient) (5) `Takbuhin

mo siya.’ (siya=customer at table no. 5)

NEUT. run after you he (PART 1) `(You) run and try to reach him.’

___________________________________________________________________________ ABBREVIATIONS: A =Agent or Source of the Action; ABS = Absolutive; ERG = Ergative; FOC = Focus; FUT = Future; GEN=Genitive; INT=Intransitive Affix; LKR= linker ; NEG = Negator; NEUT=Neutral Tense-Aspect; NOM=Nominalization; NPAST = Non past; OBL= Oblique; P = Patient or most affected entity; PER= Personal name; PART= particle; PAST = Past; Q= question; TR= transitive affix; S = only argument of intransitive construction; TA = tense-aspect; 1 = 1st person; 12 = 1st and 2nd person; 2=2nd person; 3=3rd person.

1

Goal (direction or location) (6) Tinakbuhan ka

ng

table no. 5.

PAST. run away from you (the waiter) PART2 table no. 5 `Table No. 5 ran away from you without paying the bill.’

Goal (instrument) (7) Ipinangtakbo

niya sa

maraton ang

aking Reebok.

PAST. used for running he PART3 marathon PART1 my `He used my Reebok shoes for running in the marathon.’

Reebok shoes

The received view is that the so-called “focus” affixes (bold faced) identify the semantic relation (sentence headings) of the `focused item’ (underscored) in the sentences above. 2. How has Philippine morphosyntax been analyzed? Philippine languages have been analyzed to be: a) Nominative-accusative • Active-passive dichotomy: all “goal-focus” constructions are passives derived from active. • “ang” marked entity is “subject” or “what is being talked about.” b) Ergative-absolutive • “goal” constructions are basic. • “agent” marked by ergative case (PART2), “patient” marked by absolutive (PART1) • no. 2 is “antipassive” c) Neither nominative nor ergative, but belong to a separate language type • “actor focus” construction (no.2) and “goal focus” constructions are all transitive. • “ang” marked entity is “topic.” Commonalities in previous analysis (but not all) : • The main function of the “focus” affix is to identify the semantic role of the “focused” element. • “Ang” marks the privileged syntactic argument.

3. Goals of the study In this presentation, I will concur with the characterization that the grammatical patterns in pragmatically simple sentences in PLs are ergative. To me, the question of language type crucially depends on what constitutes a transitive construction in PLs. In identifying a transitive from an intransitive construction, I will invoke a set of parameters similar to those laid down by Hopper and Thompson (1980), but I have made

2

some adjustments to the parameters to suit the Philippine condition. Evidence will be presented from several PLs illustrating how speakers use this set of parameters in assigning case to arguments and in choosing a particular voice form. In sum, I point to the robust manifestation of the ergative-absolutive relation in PLs and affirm in substance the findings in Mithun 1994 for Kapampangan. 4. Data: The data used in this study were taken from: a) Stories in Bannawag, an Ilokano weekly magazine, namely Kaputotan ni Anteb (KA), Dec. 31, 2001, by Rodrigo Manongdo; Kartib (KAR), Feb. 19, 2001, by Emidio R. Rosal; Lakay Igme Maysa nga Ama (LIMA) by Manuel Benosa; Taraddek (TAR), March 15, 199, by Robert A. Fernandez; Ti Aningaas (TA), February 19, 2001, by Jovito F. Amorin; Ti Dalagan (TD), Dec. 24, 2001, by Delfin P. Dumayas; Umaykan Ponso, di ka Agbabati (UPDKA), April 22, 2002, by Juan S.P Hidalgo, Jr.; b) Cebuano Folktales 1, San Carlos Publications, Cebu City and the following stories in Bisaya, a Cebuano weekly magazine, namely: Ang Inyong Sexpert (AIS), Dec. 13, 2000, by Greg Laconsay Kadaogan sa Gugma (KG), Aug. 5, 1998, by Edgar S. Godin; Dahlia (DAH), May 13, 1998, by Dodds Giagonia; Pinaskuhan ni Marilou (PM), Dec. 29, 1993, by Loreto Sardovia; Ugang Dugo sa Kamot ni Manoling (UDKM), Dec. 13, 2000, Edgar Godin; c) Three (3) Tagalog romance novels: Kung Kulang pa ang Lahat (KKPAL), 1996, by Veronica Siasoco Puto at Dinuguan (PD), 1994, by Edgar Reyes; Tukso ba ang Umibig (TBAU), 1982, by Gilda Olvidado d) Donya Dulimanan kan Prinsipe Agtanang (DDPA), an Itneg narrative by Jane P. Martinez; e) Mga Sugilon no Minanubo (MSM), a collection of Agusanon stories by Teofila Bada; f) A Manuvuq oral narrative by Mercedes Allan; g) A Blaan oral narrative by Isidro Pandan; h) Two taped conversations with Narding Solomon and Babes Jetajobe of Sorsogon, Sorsogon.

3

5. Definition of transitivity, case and voice in PLs: Transitivity may be defined in terms of the conceptual statuses of `source of action’ and `most affected entity’, partly following Klaiman (1988). a) A transitive construction is one where the source of action (A) is viewed as distinct from the most affected entity (P). An intransitive construction is one where the source of the action is also viewed as the most affected entity. When the A and the P refer to the same entity, it may also be labeled the S. b) Case, as used in this study, pertains to how the arguments of a predicate are formally encoded to distinguish between S, A, P and obliques. S, A, and P may be referred to as the core cases, while any other argument which is not an A, an S or a P is an oblique. c) Voice as used in this study, refers to the alternations in verbal morphology which encode the distinction in viewing how the action’s effects accrue to the clausal elements. d) In this study, voice is derivational, while tense-aspect (TA) is inflectional. But it is sometimes hard to tease apart TA from voice (see example no. 1). Almost all verbs (except root verbs) can be inflected for TA. The prototypical Tagalog verb has at least four members in its TA paradigm: “past (begun), present (begun and continuing), future (not begun) and nonfinite (neutral). There will be no further discussion of TA in this paper. Table 1. Sarangani Manobo Personal Pronouns Personal Pronouns

Person

Set 1

Set II

Set III

Set IV

Singular 1 (Minimal) 12 2 3

a ki ka 0

ku ta ta din

kanak kenita kemno kandin

siak siaken sikemno sikandin

Plural 1 (Maximal) 12 2 3

kay kidon kaw dan

day tadon niyo dan

kenami kenitadon keniyosa kandan

sikami sikitadon sikiyo sikandan

Source: Dubois, Carl. 1976. Sarangani Manobo: an introductory guide. PJL Special Monograph Issue, 6. Manila: LSP.

_______________________________________________________________________ Table 2. Sorsogonan Demonstratives ABS ERG this near speaker ini sadi that near hearer ina sana that far from both idtu sadtu

OBL didi dida didtu

4

FOC ini ina idtu

Source: Author _______________________________________________________________________ Table 3. Binukid Manobo Case Determiners Topic Personal sg. si pl. say Common non specific sa specific su

Attributive hi hay hu ku

Oblique ki kay hu/ta ku

Source: Otanes, Fe T. and Hazel J. Wrigglesworth, eds. 1992. SIPL/Binukid Dictionary 9:2, Manila: LSP and SIL, p. xxvii.

______________________________________________________________________ Table 4. Personal Pronouns in Ilokano, Tagalog and Sebwano

1 S

12 2 3

1 12 2 3 1 P 12 2 3 1 12 OBL 2 3 A

Ilokano Sg -ak -ta -ka 0 ~ isuna

Pl -kami -ta -kayo -da

Tagalog Sg ako kita ikaw siya

Pl kami tayo kayo sila

Sebwano Sg (a)ko kita ikaw siya

-ku -ta -mo -na -ak -ta -ka 0 ~ isuna kanyak kanyata kenka kanyana

-mi -tayo -yo -da -kami -ta -kayo -da kadakami kadatayo kadakayo kanyada

ko natin mo niya ako kita ikaw siya sa akin sa atin sa iyo sa kaniya

namin natin ninyo nila kami tayo kayo sila sa amin sa atin sa inyo sa kanila

nako nato nimo niya (a)ko kita ikaw siya sa ako sa ato sa imo sa iya

Pl (a)ko kita ikaw sila namo nato ninyo nila (a)ko kita ikaw sila kanamo kanato kaninyo kanila

______________________________________________________________________ Table 5. Case Determiners in Ilokano, Tagalog and Sebwano TAGALOG

PER SON COM MON

ABS ERG ni sg si pl sina nina nspec ng spec ang ng

OBL kay kina ng sa

ILOKANO

SEBWANO

ABS ni da

ERG ni da

OBL kenni kada

ABS si sila

ti

ti

Iti

ang

5

ERG ni nila ug ng

OBL kay kanila ug sa

6. How many voice forms are there? Earlier analyses: -um-, mag-, mag-…-an, mag-…-um-…-an, mang-, maki-, makipag-, magpa-, ma-, ma-…-an. mai-, i-, ika-, ipa-, ipakipag-, ka-…-an, ipag-, ipang-, pa-…-in, etc. Proposed analysis: -um- ~ m-, -in, -an and ia) Morphological classes: roots, stems, affixes and particles. A root is any form to which an affix may be added. A stem is that form to which the last affix is added. An affix is a dependent form which attaches to a root or stem. A particle is any unaffixable form. b) The verb stem can consist of a root, or a root plus a derivational stem forming affix. In (8), the verb shows a simple root rugi with the tense-aspect infix -in- and the voice affix –an. Ilokano (8) Rugi=an=na

ketdi ti

PAST=[start=TR(-an)]=3ERG.SG `She started again to teach.’

nangisuro. (TAR, p. 11)

again ABS teach

c) The most frequent and productive of the stem-forming affixes are the following: (p)ag-, pang-, paka- , paki- and ka-. Stems formed with these derivational affixes frequently end up being affixed by an allomorph of the –um- voice affix, namely the replacive m- affix. The m- affix replaces the first sound of the stems, producing socalled mag-, mang-, maka-, maki- and ma- verbs. The examples below illustrate the layered use of these verb stems. Sorsoganon (9) a. dili

n=abuhay

an

NEG PAST. INT(m-)=(ka=life) `My plants didn’t grow.’

b. waraq ka=buhay

an

NEG ka=buhay ABS `My plants didn’t grow.’

(10)

dili na

ako

NEG LKR 1ABS

mga tanom ko.

ABS PL

plant

1GEN,SG

mga tanom ko. PL

plant

1GEN.SG

n=akasulod,

waraq na

PAST.INT (m-)=[(paka=enter)]

NEG

kami

LKR 1ABS.PL

paka=sulod (sa mina). (Transcribed interview with Narding Solomon) paka=enter

(OBL mine.)

6

`I wasn’t able to enter, we were not able to enter the mine.’ Agusan Manobo (11)

…napoyok si worried

Jun

ABS.PER Jun

su

wada og=pakabulig

kandin

because NEG NPAST=(paka=help) 3OBL.SG

`Jun got worried because there was no one to help him (harvest rice.) (MSM, p. 18)

Sebwano (12) …ayaw na

pakig=dula

nila=ng

Beth…

don’t anymore (paki = g)= play 3OBL.PL=LKR Beth `Don’t play with Beth and her companions anymore.’ (PM,

(13)

“Wa ko

m=akigduwa

nila=ng

p. 3)

Beth…”

not 1ABS NEUT.INT(m-)=(pakig=play) 3OBL.PL=LKR Beth `I didn’t play with Beth and her companions.’ (PM, p. 3)

Ilokano (14)

N=agisagana

a

PAST.INT (m-)=[ag=(i=prepare)]

iti

dagus

ni

baket

LKR immediately ERG.PER wife

pangaldaw mi. (KA, p. 20)

OBL breakfast

our

`My wife went to prepare our breakfast.’

(15)

N=angaldaw=kami PAST.INT (m-)= [pang=morning]=1ABS.PL (KA, p. 20) `We took breakfast.’

d) The claim of the layered composition of the verb in PLs captures a generalization which the other analyses have failed to make. It simplifies the analysis of the voice forms and accounts for the similarities and differences between verbs with the same root or stems but with different voice affixes. It also draws attention to the highly derivational character of the process of affixation in these languages.

7

7.

Transitivity Parameters:

Hopper and Thompson (1980) came out with a set of semantic criteria for determining how transitive or how intransitive a particular construction is in a certain language. They saw transitivity as a cline or continuum and set up two columns of ten features associated with high and low transitivity. They put forward the hypothesis that if a certain construction exhibits one feature on one side of the high-low transitivity table, chances are that that particular construction will most probably exhibit the other features on the same side. The original parameters of Hopper and Thompson are given below: a) Hopper and Thompson (1980) High Low A. Participants 2 or more participants 1 participant B. Kinesis action non-action C. Aspect telic atelic D. Punctuality punctual non-punctual E. Volitionality volitional non-volitional F. Affirmation affirmative negative G. Mode realis irrealis H. Agency A high in potency A low in potency I. Affectedness of O O totally affected O not affected J. Individuation of O O highly individuated O non-individuated b) Proposed Transitivity Parameters for PLs: Nolasco (2003) reformulated Hopper and Thompson’s list to suit Philippine conditions and came out with a language-specific set of features. The original reformulation has been further revised as follows:

A. No. of Arguments B. Kinesis C. Aspect D. Punctuality E. Intentionality F. Particularity G. Directionality H. Effort I. Affectedness of P J. Exclusivity of P

High Distinct A and P action telic punctual deliberate particular external effortful P totally affected exclusive P

Low S state atelic non-punctual volitional general internal effortless P not affected non-exclusive P

From a cursory look at the revised list, one can easily discern the similarities with the original Hopper and Thompson (1980). But there are certain noticeable differences as

8

represented by the bold faced items. and provide examples.

We now turn to a discussion of these differences

A. Distinct A and P versus S. In a prototypical transitive construction, we require not only the existence of a semantic agent and a semantic patient, but also that these participants be GRAMMATICALLY encoded as A and P. In an intransitive construction, the A and the P are encoded grammatically as one argument (S). The S takes absolutive case and receives co-indexing from the verb. In a prototypical transitive construction, the grammatical encoding as A and P entails that: (a) A is assigned the ERG case with the P assuming the ABS case. (b) P is co-indexed by the voice affix. A semantic patient encoded as an OBL cannot be interpreted as receiving the full effects of an action and therefore cannot be a P. A construction with a semantic agent and semantic patient, where the agent takes ABS case and the semantic patient OBL, is semantically transitive but grammatically intransitive. The semantic agent is not a true A but only an S. The S and not the OBL patient is interpreted as the most affected entity in the clause. In addition to a voice system and a case system, some Philippine-type languages also possess a cross-referencing or person marking system, which identifies the core arguments of the clause through the use of pronominals. The geographical distribution of this subtype indicates that this is probably a very old feature of the Philippine ancestral language. It is also probably the origin of the person-marking system which Oceanic languages are known for. Kapampangan (16) Malagu ya pretty

ing anak.

3ABS ABS child

`The child is pretty.’ (17) a. Mamangan ya=ng

bayabas ing

INT.eat 3ABS=LKR guava `The child ate guavas.’

b. Pengan ne

anak.

ABS child

ning pusa ing

TR.eat na (3ERG) + ya (3ABS) ERG cat `The cat ate the fish.’

ABS

asan. fish

The Kapampangan examples above show how the co-referencing system in that language formally identifies core arguments from non-core. It also disambiguates a transitive construction from an intransitive construction. (16) may be considered the prototypical intransitive instruction where the only argument (ing anak `the child’) is coreferenced by the third personal absolutive ya. (17a) is a semantically transitive construction, but is formally and grammatically intransitive. Like (16). only the semantic agent (ing anak `the child’) in (17a) is represented by the co-referent pronoun ya. The semantic patient (bayabas `guava’) in (17a) is not co-referenced by a pronoun. (17a) therefore contains only one core argument, an S. On the other hand, (17b) is not only semantically but also grammatically transitive. The cross-referent pronouns for the core

9

arguments A (ning pusa `the cat’) and the P (ing asan `the fish’) are encoded in the portmanteau pronoun ne.

B. Action vs. State. This parameter distinguishes a state or stative event from a dynamic one which contains more action. Sebwano (18) a. Gi=kaguol

kini

pag-ayo

sa

amahan ug anak.

PAST. TR (i-)=[ka=sad] ABS.this very ERG father and child `The father and daughter took it (the mother’s death) very hard.’ (KG, p. 3)

b. N=aguol

pag-ayo ang amahan ug

PAST. INT (m-)=[ka-=sad] very

ABS father

anak

and daughter

tungod ni=ini. because OBL=this `The Father and daughter were deeply saddened by it (the mother’s death).

Gi-kaguol and naguol both have the same stem (kaguol) which more or less means `sad’ or `sadness.’ “Naguol” which is the past form of maguol refers to the natural emotional state a person finds oneself in when a loved one or someone close to him/her passes away. Gikaguol packs a lot more action and presupposes a greater amount of suffering and of will on the part of the experiencers. In the immediately following section of the story from where (18a) was taken, the daughter was described as always in tears, and the father as frequently having bouts of high blood pressure and chest pains. Tagalog (19) a) Sumalpok

ang alon sa

bangka

PAST.INT (-um-) ABS wave OBL boat `The wave struck the boat.’ (Schachter, 1972, p. 70)

b) Sinalpok

ng

alon ang bangka

PAST=[strike=TR(-in)] ERG wave ABS boat. `The wave struck the boat . (Schachter, 1972, p. 70)

These obviously elicited examples come from Schachter (1972), whose glosses have been retained here. Even so, it must be mentioned that 19b describes the boat as being struck and pounded more forcefully by the waves, compared to the natural striking of the boat by the waves in 19a. Ilokano (20) Dayta

a

lalaki, serrak=n=ak

ABS. DEM LKR man,

ket

kayat=n=ak

PAST. forcibly enter.TR(-en)=3ERG.SG=1ABS.SG

a

gundaway=an (TLD, p. 14)

CONJ wants=3ERG.SG=1ABS.SG LKR NEUT.abuse=(TR-an)

10

`This man, he forcibly entered my house and he wants to abuse me.’

(21) In=serrak=n=ak

ni

Marian iti

maysa a

PAST. TR (i-). enter=3ERG.SG=1ABS.SG ERG Marian OBL a `Marian led me to a room.’ (TD, p. 18)

(22) (idi)

srek

ti

maysa a

kuwarto.

LKR room.

baket.

(when) NEUT.INT (-um-) ABS a LKR woman (TD, p. 18) `(Their conversation was interrupted when) a woman entered the room.’

The transitive examples in (20) and (21) contrast sharply with the simple action of entering a room in (22). C. Telic vs. Atelic. By telicity or completion of an event, it is meant that the action or activity terminates with a clear accomplishment of some kind. In the following Sebwano example, the difference between being acquainted with someone and recognizing someone is foregrounded. Sebwano (23) a. N=akaila

ka

ba

sa

biktima?

PAST. INT (m-)= [paka=know] 2ABS.SG QPRT OBL victim `Were you acquainted with the victim?’ (UDKM, p. 23)

b. N=ailhan

ba

nimo

ang biktima

PAST.INT (m-)={[ka=ila]=TR(-an)} QPRT 2ERG.SG ABS victim `Did you recognize the victim?’

Ilokano (24) “No ania ti

makuna=m, Marian, ulien=ta=nto

a

dua.

If what ABS say=2ERG, Marian, climb.TR(-en)=we=FUT LKR two `What are you implying, Marian, that we two scale it (the mountain).’ (TD, p. 16)

(25) Imm=uli

kami

ti

agdan xxx

PAST.INT (-um-)=climb 1ABS.PL OBL stairway `We went up the stairway…’ (KA, p. 16)

As its gloss clearly indicates, ulien refers to a more causative and more purposeful activity which results in the mountain being scaled and conquered. Umuli is a more internally directed activity which merely moves or transfers the effector theme into a different place or location. Tagalog (26) a. Nang mainit na When hot

ang mantika,

iprito

niya

ang kamote.

already ABS cooking oil TR (i-).fry 3ERG.SG ABS camote

`When the oil was already hot, she fried the camote.’ (KKPL, p. 80)

11

b. Nang mainit na ang mantika, n=agprito siya ng kamote. When hot already ABS cooking oil PAST=INT (m-).[pag=fry] ABS OBL camote

`When the oil was already hot, she fried camote.’ In (26a). the most likely interpretation is that the camote strips ended up being fried, while in (26b), what is being asserted is only the activity of frying camote.

D. Punctual vs. Durative. The next examples come from Shibatani (1988). We have retained the original gloss for (27a), but changed the gloss for (27b). This was done to faithfully capture the difference in meaning between long term knowledge of a secret and its instantaneous discovery. Sebwano (27) a. N=akahibalo

si

Juan

sa

PAST.INT (m-)=[paka=know] ABS.PER Juan OBL ` Juan knows the secret.’ (Shibatani, 1988, p. 104)

b. N=ahibaloan

ni

tinaguan. secret

Juan ang

tinaguan.

PAST.INT (m-)={[ka= know]=TR(-an)} ERG.PER Juan ABS secret (Shibatani, 1988, p.104) `Juan discovered the secret.’

More interesting are the Tagalog examples where biting which is presumably a punctual act comes in two forms. The author begins with a generalized depiction of the victim’s retaliatory action described in nangagat as plural, random, indiscriminate and therefore durative. But when the author focuses however on each and every bite that the victim exacts on her tormentor, the verb used is the punctual kinagat Tagalog (28) Pagsusuntok

niya

PAST.hit repeatedly

naman siya sa leeg. truly 3ABS OBL neck Kahit ano, Even anything

ito.

Bilang ganti, hinalikan

3ERG 3ABS.this

In retaliation, PAST.kiss

N=angagat siya. PAST.INT (m-)=[pang=bite] 3ABS

kagat

niya.

[=biteTR(-in)] 3ERG

`She repeatedly hit him. He retaliated and kissed her in the neck. She bit and bit. Anything that she could bite, she bit.’ (TBAU, p. ) Ilokano (29) Takder=an=na

ti

dadaiten=na…

PAST.stand up=TR(-an)=3ERG ABS sewing=3ERG `He stood up from his sewing.’ (TD, p. 10)

12

(30) a. Bangon

ni Ponso sana

matmat=an

PAST. INT (-um-)=sit upright ABS.PER after.this

ti

gaze=TR(-an)

bola. xxx Takder. (UKPDA, p. 17)

ABS ball. xxx PAST.INT (-um-)=stand up

`Ponso sat upright and gazed at the ball. xxx He stood up.’ b.

Nabayag a long

ti

n=agtakder

iti

LKR PAST.INT (m-)=(ag=stand)

tawa

ti

asideg

OBL near

kuarto=da

ERG window ERG room=ERG.PL (TA, p. 31) `She was standing by the window for a long time.’

The difference between timmakder and nagtakder is one of telicity and punctuality. The action of timmakder requires that the agent starts from a sitting or non-standing position at which point the agent stands up. Nagtakder does not presuppose a non-standing position from which to begin the standing activity. In fact, it is unclear when the standing stance begins or ends, a truly non telic and non-punctual affair. Tinakderan is a different matter. Unlike the first two which are intransitive, takderan is a punctual, telic and transitive verb which contains an ergative A and an absolutive P. E. Deliberate vs. Volitional. An action or activity may either be voluntary or carried out willingly, in which case it is volitional; or it can be more determined or wilful, in which case it is deliberate. Sebwano (31) a. …gi=undang

ko

na

ang

pagpanigarilyo…

PAST=stop. TR (-un) 1ERG.SG already ABS smoking `I already stopped smoking (in the early 1960s).’ (AIS, p. 17)

b. …mi=undang

ko

na

sa

pagpanigarilyo…

INT.PAST=stop 1ABS already OBL smoking `I already stopped smoking (in the early 1960’s)

(31a) comes from a column of Greg Laconsay in the Bisaya magazine. The verb giundang denotes that it took a large amount of discipline on the part the author to stop smoking. 31b is a colorless and bland way of describing the event in a matter-of-fact fashion. The next examples were taken from a scene in the Tagalog novel Puto at Dinuguan by Edgar Reyes. Manila has been hit by a powerful typhoon, and many of the streets are flooded. Andy, who is driving a truck, decides to pick up Fe, the girl he is courting. Fe had warned Andy not to pick her up at her place of work because this might

13

provoke a lot of gossip. She refuses to go with Andy and insists that they take public transportation. They quarrel. To appease Fe, this is what Andy does. Tagalog (32) Pagsapit NOM.reach

nila

sa

abangan

3ERG.PL OBL public transportation stop

hubad

niya

ang kanyang jacket

take. off.TR(-in) 3ERG ABS his

jacket

`When they reached the bus stop, he took off his jacket (and gave it to Fe.) (PD, p.66)

Fe hesitates but eventually takes Andy’s jacket and wears it, signifying the end of the spat. The claim is that Andy’s action in (32) is purposeful and deliberate. Let us now compare this to a similar event, that of taking off one’s socks and shoes. Andy and Fe manage to get a ride on a jeepney. The jeepney stalls in the middle of the flood. A copassenger decides to wade through the flood but before doing so, he naturally and understandably does this. Tagalog (33) N=aghubad

ito

ng sapatos at medyas.

PAST. INT (m-)=[pag=take off] ABS.DEM OBL shoes `He took off his shoes and socks.’ (PD, p. 68)

and socks.

The distinction between a volitional and a deliberate action is nicely portrayed in two clauses in an Itneg tale about Donya Dulimanan kan Prinsipe Agtanang (Lady Dulimanan and Prince Agtanang). In this scene, Kulimbibita, the future Donya Dulimanan, goes swimming with the other children by the river, when a goat appears and proceeds to defecate on their clothes. This purely involuntary bodily function is depicted in two different ways in the story. First, when the goat appears: Itneg (34) …ta uwad ta ni=lumtaw nga ugsa when there. OBL.DET PAST.INT.(-um-).appear LKR deer ket mapan takkiya=n dadin bado= da. and went NEUT.defecate.TR.(-an) 3ABS,PL clothes= 3ERG.PL `There appeared a deer and (it) went on to defecate on their clothes.” After seeing what the animal appears to be doing, the other children begin to drive the animal away. They lead it to Kulimbibita’s clothes instead. (The deer’s droppings turned out to be pure gold. )

(35) “Adi ka takki kanta, Don’t 2ABS.SG NEUT.INT.(-um-)=takki there 14

in=ka kanta=n bado ni Kulimbibita.” go=you OBL.DEM=LKR clothes GEN.DET Kulimbibita `Don’t you defecate (on our clothes), go do it on Kulimbibita’s clothes.’ The verb takkiyan with the voice affix –an attributes a more deliberate and conscious action on the part of the deer in (34) than simply tumakki (35) which is the more volitional, unmarked and natural action. In (34), the storyteller strongly hints that the deer was intentionally leaving its waste on the clothes of those it wanted to make rich. F. Particular vs. General action. It is common to find verbal alternations where the opposition or difference in meaning lie in the generality and particularity of the action to which the same verb stem refers. Ilokano

(36) n=agisagana PAST=[ag=(i=prepare)]

iti

a

dagus

ni

baket

LKR

immediately ERG wife

pangaldaw mi. (KA, p.20)

OBL breakfast

our

`My wife immediately went to prepare our breakfast.’

(37) in=sagana amin=en dagiti PAST=[TR(i-)=prepare] all=already ABS.PL

masapsapul. needs

`She put out all that was needed (for the natural delivery of a child).’ (KA, p. 21)

In (36), the verb nagisagana encompasses all the activities involved in the preparation of breakfast. This would include cooking rice, frying eggs, preparing coffee, setting the table and the like. In (37), the activity is limited to simply bringing out what a mother and her new born baby would be needing at birth. Sebwano (38) a. “Dro,

adto=on

ta

si

Haring Wati …”

St. Peter. NEUT. go.to=TR(-on) 12ERG.SG ABS.PER King Worm `St. Peter, let’s go check up on King Worm.’ (HW, p. )

b. “Dro, mangadto ta kang Haring Wati…” St. Peter, NEUT.INT (m-).pang=go 12ERG OBL.PER King Worm.’ `St. Peter, let’s go to King Worm’s house.’ 38a comes from a story about King Worm who asks God to make him a human being so that he could purportedly serve God’s subjects. God grants the worm’s

15

wish. After some time, God and St. Peter pay King Worm a visit. They dress themselves up as paupers to find out if he would still be able to recognize them. The particular motive for the visit—which is to put King Worm to the test-- is captured by adtoon. The use of mangadto in this context would simply mean that God and St. Peter planned to go to King Worm’s house with no specific intention in mind. Tagalog (39) a. N=agbasa

siya

ng

libro.

PAST.INT(m-).[pag=read] 3ABS.SG OBL book. `He did book-reading.’

b. Basa . read.TR(-en) `He read the book.’

niya

ang

libro.

3ERG.SG ABS book.

In these made up Tagalog examples, reading as a general and leisurely activity (39a) and as a conscious, deliberate and particular act undertaken to affect a book (39b) is illustrated. G. Internal vs. External action. Intransitive actions are inherently internal, transitive actions seek external targets. Tagalog (40) a. Balik=an

niya

si

while

Fe sa Broadway Centrum.

PAST=[return=TR (-an)] 3ERG.SG ABS.PER Fe at Broadway Centrum `He went back to pick Fe up at the Broadway Centrum. ‘ (PD, p. 76)

b. Balik

siya

sa

PAST.INT (-um)=return 3ABS.SG `He went back to the pick-up truck.’

pick-up.

OBL pick up truck

The next examples involve the use of the “causative” pa- construction in Tagalog. Earlier accounts highlight the “addition” of a causee in this clausal type. What is glaringly omitted in those accounts are the reflexive reading in 41a and the purely causative interpretation in 41b. Tagalog (41) a. N=agpaluto

ako

ng

adobo sa

INT (m-)=[pag=(pa=cook)] 1ABS.SG OBL adobo

nanay ko.

OBL mother my

`I asked my mother to cook adobo (for me).’ b.

Paluto

ko

ng adobo ang nanay ko.

PAST.pa=cook)=TR(-in) 1ERG.SG OBL adobo

`I asked my mother to cook adobo.’ Ilokano

16

ABS mother my

(42)

asitgan=na ni Mauricia. PAST=[go.near=TR (-an)]=ERG.SG ABS Mauricia `He (the referee) approached Mauricia.’ (FM, p. 9)

(43)

Sitsit=an ni Diputado Agaton PAST.psst=TR (-an) ERG Congressman Agaton ti guardia. ABS guard

asideg PAST.INT (-um-)= near

daytoy ABS.here

`Congressman said “psst” to the guard. He (the guard) drew closer.’ (TA, p. 8)

In (42), the female boxer Mauricia has just hit the canvas from a powerful blow from her opponent. The referee quickly approaches to start counting her out. In (43), what is predicated is the internally directed movement of the guard drawing closer to the one who has called him. The following example was taken from the King Worm story. King Worm couldn’t recognize God and St. Peter in their pauper’s clothes. When God asks him for a drink of water, he orders a servant to bring him water given to chickens. God decides to reveal himself to King Worm and takes him to task. Sebwano (44)

Unya, gi=atubang

sa

atong Ginoo si

Later turn towards.TR (-un) ERG our

Haring Wati..

Lord ABS.PER King Worm

`Later, the Lord confronted King Worm.’ The intransitive counterpart of giatubang is miatubang which means `to turn one face towards someone.’ The use of miatubang in the above context is patently too weak a depiction of this confrontation scene.

H. Effortful vs. effortless. To me, one of the hallmarks of a transitive construction is the great amount of effort that an agent pours into carrying out an action or activity. Sebwano (45) a. “N=angita

ko

sa

isda=ng

PAST.INT(m-)=(pang=look for) 1ABS.SG OBL fish=LKR

wala=y

bukog.’ (CFT1, p.41)

no=LKR bone.

`I am looking for the fish with no bones.’ b.

`Gi=pangita nako ang isda=ng PAST=[(pang=look for)=TR(-in)] 1ERG.SG ABS.NPER fish=LKR

17

wala=y bukog.’ no=LKR bone. `I am searching for the fish with no bones.’

Tagalog (46) a. Langoy

sila

sa

ilog.

PAST.INT (-um-).swim 3ABS.PL OBL river `They went swimming in the river.’

b. Ni=langoy

nila

ang ilog.

PAST.swim=TR (-in) 3ERG.PL ABS river `They swam the river ( and conquered it).’

The Sebwano and Tagalog examples are fairly clear and straightforward, but the Manuvuq examples need elaboration. Manuvuq (47) a. Id=susu

rin

kon

en

ini

so

togotong.

PAST=[(suck milk)]=TR(-en) 3ERG.SG heresay already ABS.DEM LKR eggplant. `It is said that he (the baby) sucked the eggplant of its milk.’

b. Idsusu

sikandin kon

en

dutut

togotong.

PAST=[INT(-um-)=suck milk] 3ABS.SG heresay already OBL.there eggplant `It is said that he (the baby) nursed at the eggplant.’

(47a) was lifted from a story narrated to me by Mercedes Alan about three children whose mother had died and whose father had remarried. As expected, the stepmother was of the cruel kind who didn’t care for the children. The youngest was still a baby who constantly cried of hunger. One day, an eggplant suddenly emerged from the ground. It so happened that the tip of the eggplant touched the baby’s lips. The hungry baby sucked and sucked into the eggplant until he was able to extract milk from it -doubtless an effortful enterprise. 47b is the effortless alternative to 47a, as if milk is naturally obtainable in an eggplant. The two verb forms in (47a) and (47b) are the same, except that the baby in (47a) is encoded in the ergative and the eggplant in the absolutive, while in (47b) the baby is absolutive and the eggplant oblique. I. Exclusivity and non-exclusivity of the P. Hopper and Thompson (1980) spoke of semantic individuation and how referents may be rendered specific through the use of pronouns, names, pronouns and definitizers. In this article, we distinguish between semantic individuation from grammatical individuation of the P, where voice and case

18

mutually signal exclusivity and non-exclusivity of a semantic patient. The next examples are illustrative: Sebwano (48) a. N=amukaw PAST. INT(m-)=(pang= wake up)

pa gani kami even had-to 1ABS.PL

kang Manang Tacia. (DAH, p. 3) OBL Manang Tacia `We even woke Manang Tacia up.’ b. Gipukaw pa gani namo si Manang Tacia.. gi-PAST.-TR.wake up even had.to 1ERG.PL ABS.DET Manang Tacia. `We even had to wake Manang Tacia up.’ The clause in (48a) was taken from a love story in Bisaya magazine. Cita, while on vacation from school, meets Dading in Cagayan de Oro, a city in the southern Philippines. He courts her, but just as they were getting to really know each other, vacation is over. She has to board a ship back to Oroquieta City to resume her studies. Cita promises to give her answer to Dading’s proposal when he sees her off at the pier. Dading fails to make it to the send-off. Cita leaves him a note in the care of Trining and Charing. They take the note to Dading’s boarding place late at night. But, Manang Tacia, the caretaker, is already fast asleep. The clause in (48a) is what Trining and Charing tell Dading when they meet him. The analysis of 48a and 48b is as follows. 48a means that Manang Tacia may have not been the only one awakened from her sleep. The entire house or even the neighbors may have been, too. 48a also indicates that the rousing may have been accomplished by knocking at the front door or gate or by shouting from the street. This probably explains why others may have been unduly awakened. The verb gipukaw means that it was Manang Tacia who was the only target of the waking event and she was awakened directly at her bedside. This is what we mean when we say that Manang Tacia is more individuated in 48b when it is marked as the absolutive than when it is marked by a mere oblique, as in 48a, In this regard, we claim that there is only one grammatical argument , an S, (kami) in 48a, compared to the distinct A (namo) and distinct P (si Manang Tacia) in 48b. Notice, too, the more direct, more effortful, more immediate, more active, and more intentional character of the action in 48b than in 48a. This is what is meant by grammatical individuation and exclusivity of the P. (49) a. Bili

rin siya

ng

isa pa=ng

pares ng

PAST.INT (-um-).=buy also 3ABS.SG OBL one more=LKR pair OBL

rubber sandals para kay rubber sandals

for

Fe.

OBL.PER Fe

19

`He also bought another pair of rubber sandals for Fe.’

b. Ibili

rin niya

si

Fe ng isa pa=ng

PAST. TR (i-)=buy also 3ERG.SG 1ABS.PER.SG Fe OBL one more=LKR

pares ng rubber sandals. pair

OBL rubber sandals.

`He also bought Fe another pair of sandals.’

(49a) comes from the novel entitled “Puto at Dinuguan” by Edgar Reyes. Andy and Fe have just escaped from a street flood, caused by heavy rains. They decide to wait it out in a shopping mall. Andy thinks of buying new clothes and changing the wet ones he’s wearing. He buys a pair of rubber sandals and buys another pair for Fe. While Fe is individuated and the beneficiary of Andy’s action, it is contended that Fe is not the sole beneficiary of the buying action. She is not singled out as the most affected entity, and therefore this explains why the clause is intransitive and the beneficiary is merely encoded as an oblique. It would have been different, had (49b) been used. 49b speaks of a buying action where Fe is the exclusive semantic beneficiary. In fact, this construction in 49b puts in an added sense in that Fe gains a second pair of sandals. J. Wholly Affected and Partially Affected P. Consider the following Ilokano example in (50a). It comes from another love story in the Bannawag magazine, entitled: “Lakay Igme: Maysa nga Ama” (Old Igme: A Father). The story opens with Mando asking Lakay Igme’s only daughter ‘s (Elsa) hand in marriage. Lakay Igme agrees but only under the following condition: (50) a. “Sipapalubos=ak agree=1ABS.SG

nga ag=kasar=kayo ngem masapul LKR INT.ag=marry=2ABS.PL but necessary

nga i=subli=m ti nagasto=k LKR i-.TR=return=2ERG.SG ABS.DET spent=I

iti panagadal na. Lima ribu!”, OBL.DET studies her five thousand (pesos) `I agree, but you must return (all) the money I spent in sending her to school. Five thousand pesos!’ b. “…mang=i=subli ka iti nagasto=k… …m-(pang=i-.TR=return) 2ABS.SG OBL spent=I The verb in 50a (boldfaced) can be replaced by that in 50b, but not without a corresponding drastic change in meaning. Under the same circumstances, 50b would have a more generalized meaning of giving back something, which may or may not be

20

in the form of money. Pragmatically, the default reading is to return a portion of what was spent, and not the entire amount. In other words, the ti nagastok iti panag-adal na in 50a is a wholly affected entity than the same argument appearing in the oblique in 50b. Moreover, there is also a difference in intensity in these imperative constructions. The more intense, more urgent and more command-like is 50a. 50b is more like a request than anything else, which the addressee is free to comply with or not. Under our analysis, 50a is more transitive while 50b is more intransitive. The Koronadal Blaan examples below also demonstrate the partitive and wholistic reading in the affectedness of a semantic patient. Koronadal Blaan (51) tumu ni, Good this

aluwaq gu

ni alas

gu

ni.

give out 1ERG this poison 1ERG this

“It’s better if I just give out all of my poison.” (52) amti la all

alas

amsut dee na m=inum

ale

dituq di

they appeared then and INT(m-)=drink 3ABS here

OBL

alwaq.

poison gave.out

`All of them came and drank from the poison given out.’ The above examples originated from a story told to me by Isidro Pandan. The python used to be the most poisonous of all the snakes, but was tricked by the crow into giving out all its poison. The snake had just bitten a girl who was bathing by the river. The snake did not know what became of the girl, and so he asked the crow to go to her house and check. The crow learned that the girl had indeed died but reported to the snake that she continued to live. This drove the snake into depression, and so he decided to give out all his poison, as quoted in (51) as him saying. Even if there is no voice affix in the verb, we know this to be a transitive construction because the A is assigned the ergative, while the P is assigned the absolutive. To continue the story, the snake invited all his subjects to partake of the poison that he has just released in a large leaf. (52) tells that many came and drank from the poison. Notice how the verb `drink’ is encoded. It is with an intransitive voice affix. The agent is encoded in the absolutive while the poison is in the oblique. In (52), the poison is only partially affected, compared to (51), where it is fully affected.

8.

Generalizations:

21

It would appear then that transitivity has been fully grammaticized into the case marking strategies and in the voice system of PLs. I interpet this to mean that case and voice are indicators of transitivity in PLs so much so that the following generalizations can be made with a high degree of confidence. First, a transitive construction is one where the A (the source of the action) is encoded in the ergative case and the P (the most affected entity) is encoded in the absolutive case. An intransitive construction is one where the S-- representing the source of the action and the most affected entity at the same time—is assigned the absolutive. Second, a transitive construction is one where the verb contains any of the voice affixes –in, -an and i-, or their equivalents in the other PLs. An intransitive construction is one when the verb contains the voice affix –um- or its allomorph m-. 9. Conclusion: While more linguists have come to accept the ergative nature of Philippine morphosyntax, many more have remained skeptical of, and unconvinced by, this view. The reason for this is understandable considering that most work subscribing to the ergative analysis has tended to focus on the formal aspects of the phenomenon, downplaying its semantic, pragmatic and discourse motivations, with a few notable exceptions like Mithun 1994. The meaning–based and formal evidence all points to the robust manifestation of the ergative-absolutive relation in Philippine-type languages and conversely to the weak grammaticization of the starting point or subject-object relation. This presentation has attempted to deepen our understanding on how transitivity influences speaker’s choice of grammatical structures, especially in the assignment of case to arguments and in the selection of voice affixes. To understand Philippine ergativity, one must understand Philippine transitivity. Our proposal has been for a conceptual (re)formulation of transitivity in PLs in terms of source of action and most affected entity. In sum, what Philippine ergativity really means is that speakers give the highest degree of prominence to the most affected entity. Selected references: Blake, Barry. 2001. Case. Cambridge: Cambridge University Press. 2nd edition. Bybee, Joan. 1985. Morphology. Amsterdam: Benjamins. Comrie, Bernard. 1978. Ergativity. Syntactic typology, ed. by Winfred P. Lehmann, 329394. Austin: University of Texas Press. Comrie, Bernard. 1989. Language universals and linguistic typology. 2nd Edition. Chicago: University of Chicago Press. Cooreman, Ann. 1994. A functional typology of antipassives. In Barbara A. Fox and Paul J. Hopper, eds., Voice: form and function, 49-88. Amsterdam: John Benjamins. Croft, William. 1991. Syntactic categories and grammatical relations. Chicago: University of Chicago Press. DeLancey, Scott. 1982. Aspect, transitivity and viewpoint. Tense aspect: Between semantics and pragmatics, ed. by Paul J. Hopper, 167-183. Amsterdam & Philadelphia: John Benjamins.

22

Dixon, R. M. W. 1979. Ergativity. Lg. 55.59-74. Dryer, Matthew. 1997. Are grammatical relations universal? In J. Bybee, J. Haiman, and S. Thompson, eds., Essays on language function and language type, 115-143. Amsterdam: Benjamins. Dubois, Carl. 1976. Sarangani Manobo: an introductory guide. PJL Special Monograph Issue, 6). Manila: LSP. Du Bois, John W. 1985. Competing motivations. In Haiman, John, ed., Iconicity in syntax, 343-365. Amsterdam: John Benjamins. Du Bois, John W. 1987. The discourse basis of ergativity. Language 63: 805-855. Du Bois, John W., William Ashby, and Lorraine Kumpf, eds. 2003. Preferred argument structure: the next generation. Amsterdam: John Benjamins. Fox, Barbara and Paul J. Hopper, eds. 1994. Grammatical voice: Its form and function. Amsterdam: John Benjamins. Givón, T., ed. 1998. Grammatical relations. Amsterdam: Benjamins. Haspelmath, Martin. 1990. The grammaticization of passive morphology. Studies in language 14:25-72. Hopper, Paul and Sandra Thompson. 1980. “Transitivity in grammar and discourse.” Language 56, 2:251-299. Hopper, Paul J. and Sandra A. Thompson. 1984. The discourse basis for lexical categories in universal grammar. Lg. 60.4: 703-752. Kemmer, Suzanne. 1993. The middle voice. Amsterdam: John Benjamins. Li, Charles N. and Sandra A. Thompson. 1976. Subject and topic: a new typology of language, 457-489. In Charles N. Li, ed., Subject and topic. New York: Academic Press. Klaiman, Miriam H. 1988. “Affectedness and control: A typologicalstudy of voice systems. In Shibatani, 1988. Mallinson, Graham and Barry Blake. 1981. Language Typology. Amsterdam: North-Holland Publishing Co. Mithun, Marianne. 1991. Active/agentive case marking and its motivations. Language 67. 510-546. Mithun, Marianne. 1991. “The role and motivation in the emergence of grammatical categories: The grammaticization of subjects.” In Elizabeth Traugott and Bernd Heine (eds). Approaches to grammaticalization, vol. II, 159-184. Amsterdam and Philadelphia: John Benjamins. Mithun, Marianne. 1994. “The implications of ergativity for a Philippine voice system.” In Barbara Fox and Paul J. Hopper (eds.), Voice Form and Function, 247-277. Amstersam: John Benjamins. Mithun, Marianne and Wallace Chafe. 1999. What are S, A, and O? Studies in Language 23.3: 569-596. Nolasco, Ricardo Ma. 2003. Ang Pagkatransitibo at Ikinergatibo ng mga Wikang Pilipino. Ph.D Dissertation. University of the Philippines- Diliman Otanes, Fe T. and Hazel J. Wrigglesworth, eds. 1992. SIPL/Binukid Dictionary 9:2, Manila: LSP and SIL, Schachter, Paul and Fe Otanes. 1972. Tagalog reference Grammar, Berkeley: University of California Press.

23

Shibatani, Masayoshi. 1985. Passive and related constructions: A prototype analysis. Lg. 61:821-848. Shibatani, Masayoshi, ed. 1988. Passive and voice. Amsterdam: John Benjamins. Shopen, Timothy, ed. 1985. Language typology and syntactic description (3 volumes). Cambridge: Cambridge University Press. Thompson, Sandra A. and Paul J. Hopper. 2001. Transitivity, clause structure, and argument structure: evidence from conversation. In Joan Bybee and Paul J. Hopper, eds., Frequency and the emergence of linguistic structure. Amsterdam: Benjamins. Van Valin, Robert and Randy J. LaPolla. 1997. Syntax. Cambridge: Cambridge University Press. Verhaar, John. 1990. How intransitive is intransitive? Studies in Language 14.1: 93-168.

24