Emotion and affect in the syntax of Formosan languages --- Formosan solution to the lack of prepositions SWL Aug.5,2004 Shuanfan Huang and Fuhui Hsieh
National Taiwan University
1. Posing the Problem: What are the grammatical strategies used by Formosan
languages to express emotion and affect? The research problem makes sense since (a) Emotional cause and emotional affect in English are typically framed by the spatial properties of the prepositions ( at, about, over); physiological reactions caused by emotion events are conceptualized by the image schema of a container ( in, out) , a companion ( with) or frontness ( for).In other words,
emotion events are grounded in our spatio-physical interaction with the world. (b) And yet In many Western Austonesian lanuguages the inventory of prepositions or locative particles is either null ( TSou, Saisiyat, Atayal, Seediq) or at most one ( Saisiyat, Kavalan , Tagalog, Cebuano).
2. Blust (l999) divides Formosan languges into the following nine primary branches: l. Atayalic; 2. East Formosan; 3. Puyuma; 4. Paiwan; 5. Rukai; 6. Bunun; 7. Tsouic; 8. West plains; 9. Northwest Formosan, with Saisiyat and Pazeh-Kulon. All A languages outside Taiwan belong to a single subgroup, called
Malayo-Polynesian (Blust 1999), as shown below:
2.l.
2.2. Saisiyat: a moderately endangered language spoken on the highlands of NW Taiwan by a population of about 45000. It is a strongly subject-initial language.Like many other languages of Taiwan and the Philippines, Saisiyat has a system of voice marking that goes under the general name of ‘ focus’.
3. Focus system: Formosan solution to the poverty or absence of adpositions. 3.l. Many of the Formosan languages have a Philippine style ‘ focus’ system, a system of verbal affixation which allows different arguments t o be placed in subject position and which signals the presence of a particular semantic role associated with the subject. The focus system has also been referred to in the literature as voice system or a trigger system ( Shibatani 1988, Cumming et al. 1987, Wouk 1996) . 3.2 The focus markers are categorized into four types; agent focus(AF), patient focus(PF), locative focus(LF) and referential focus(RF). A fifth type of focus,
Compound focus, is recognized for Saisiyat, as summarized in Table 1. Focus I II Agent focus (AF) m-, -om-, ma-, O Patient focus (PF) -en(-in, -on,-un) -i Locative focus (LF) -an Referential focus (RF) si-/sik- -ani Compound focus (CF)
si-/sik- … -en; si-/sik-…-an
3.3 There are two sets of focus markers in Saisiyat. One set (I) is used with declarative sentences , or with negative sentences containing the negators
kayni’ / okik, while the other set(II) is used in the imperative or in
negative sentences containing the negators ’okay/’izi’/’in’ini’ '
3.4. A corpus count yields the following percentages for the various verbal clause types:
AF: 77%; PF: 19.2%; LF: 0%; RF: 3.8%
( cf. the percentages for Squliq based on a corpus of 1451 clauses: AF: 47.9%; PF: 29.9%; LF: 17.7%; RF: 4.5%)
3.5. The Case Marking System
The case markers in Saisiyat are categorized into six cases and each can be divided into two sets: personal nouns and common
nouns. Table 2 is taken from Yeh (2000:73) and Huang, Su & Sung (2003: Chapter 3).
Table 2 The Case Marking System in Saisiyat
Noun Nominative Accusative Genitive Possessive Dative Locative
Personal name O
hi
hi ni ‘an-a ‘ini’ kan
kala
Common noun O
ka
ka noka ‘an noka-a no ray
3.6 Saisiyat as a split-ergative language: Saisiyat has both accusative and genitive markers and shows symptoms of a split-ergative language. Analysis of the corpus
data shows that its transitive AF clauses tend to occur with the progressive ( ‘am ima, mam) or future auxiliary markers ( iya), while PF clauses show no such tendencies.
3. PF and affect : Note that PF clauses in WA languages are usually semantically and pragmatically active transitive clauses, while AF clauses are usually
intransitives, subject to language-specific exceptions ( Huang 2002) .Most of the normally intransitive verbs may occur in PF form, esp. in Saisiyat. When they do, they may not occur alone, but must occur subordinate to a main clause, where the event denoted by the subordinate clause (SC) indicates temporal
simultaneity/contiguity with that denoted by the main clause(MC). Such
contiguity often pragmatically implies a cause-affect interpretation. As noted by Shibatani (l996) and others, this is a common though not necessary feature of many passives.
A. [ V PF NP GEN ] SC [ NPnom VP]MC
6. Sae’hae’en noka boway, ya’o ima m-in-tani Fall-PF GEN fruit I PROG stop-AF-PERF
“ When the fruit dropped down, I (happened to) stop by (there).” 7. noka kinaat in-sawi-in, ya’o kayni’ ba:iw
GEN book torn-PERF-PF I not want buy ‘ (Because)the book is torn, I don’t want to buy (it).’ 8. asasay-en noka boway, ma’iaeh somi’iael ila
ripe-PF GEN fruit people eat-AF PERF ‘When the fruits are ripe, people eat (them).’
9. ‘obay abisbis-in oka’ ila nanaw PN in pain-PF no PERF still
‘ Obay is in such great pain that he can no longer take it.’ B. NP nom [ VPF (NP) NPGEN ]sc [ NP nom VP] MC 6a. Ya’o sae’hae’en noka boway, (ya’o) ima m-in-tani
“ When the fruit dropped down, I ( happened to) stop by (there)”
4.1. PF construction: Unlike passives in many languages, there is no demotion of agent in Saisiyat, and thus no promotion of patient, even in cases where they focus on the state resulting from an action. And the PFconstruction must occur together with another main clause. Semantically, the intransitive-based PF construction discussed above is used to indicate temporal contiguity and by implication a
cause-affect interpretation. The PF construction is thus a grammatical construction in the sense of Construction Grammar.
5. Restriction on PF verbs
Since any event or state of affairs could have an effect on someone or something, one would expect there to be no restriction on the occurrence of any intransitive in PF form in the (A) or (B) pattern. Our data suggest that this is indeed the case. 10. yami rawas-en noka ra:an, yami lobih ila
we far-PF GEN road we return PERF ‘ The road was too long for us and we returned.’ 11. ya’o oka’en ila noka walo’, pataas ila I have not-PF PERF GEN sugar stop-AF PERF ‘(since) there was no sugar, I stopped (cooking).’ 12. noka kapapama’an alikah-en atomalan, obay motae’ GEN KA-ride-LF fast-PF much PN vomit ‘ The caw was going too fast and Obay vomited.’
6. LF and RF are applicative constructions, which involve verbal derivations that have the effect of encoding as subjects noun phrases that are normally encoded as obliques. This category also refers to verbal derived forms in other accusative languages that have the effect of coding as direct objects noun phrases that are normally encoded as obliques. (Since WA languges allow relativization of subjects only, it is not surprising that these languages also have subject-creating verbal derivations like LF or RF. The interesting correlation between relativization and ‘promotion’ strategies
can be observed in a language like Saisiyat.)
7. Function of LF : A major function of LF is marking location as the subject of a LF c onstruction. In many WA languages, such function is now restricted to nominalization. Examples from Saisiyat are
13.a. ka-patol-an hini’ NMZ-sing-LF here
‘ This is the place for singing.’ b.. kayzaeh ka ka-pahlem-an
good NOM NMZ-sleep-LF “ This is a good bed to sleep in.’
8. Functions of RF: The RF construction is a valency-changing construction. In the RF construction, the genitive argument is the agent, while the nominative argument can express the transported theme, instrument, benefactee, location, or the cause of that action.
14. a. noka korkoring ‘oya’ si-haengih GEN child mother RF-cry ‘ The child cried for her mother.’
b. hini korkoring si-paSay ila ni yaba’ ki ‘oya’ this child RF-die PERF GEN father and mother ‘ The parents of the child died ( on him).’
c. koko:ol si-in-pongah , kayzaeh kita’en mountain RF-PERF-bloom good see-PF
“ The flowers on the mountain are in full bloom and look
nice.’ d. kahoey’ si-Sebet ni ‘oya’ ka ahoe’
tree RF-hit GEN mother ACC dog ‘ Mother hit the dog with a tree stick.’ e. Obay si-Sebet ni ‘oya’ ka ahoe’ PN RF-hit GEN mother ACC dog ‘ Mother hit the dog for Obay.’
f. nisia tatpo’ si-tirotol hi yakin 3S.GEN hat RF-sell ACC 1S.ACC ‘He sold me his hat.’
9.Emotion events and the si-construction in Saisiyat: the nominative NP is either an ( indirect )cause triggering certain emotional experience in the experiencer, which is marked by the genitive case, or an affectee caused by the action performed by the agent in the genitive case . The nominative case in Saisiyat is rarely realized. In situations where there is a direct cause, the pak-construction is used.
9.l. Nominative NP as a cause
15. ’obay sik-boe’oe ma’an PN RF- be.angry IS. GEN
‘ I was angry at/with Obay ( when he was nearby and did something bad).’
16. ‘obay sik-sawa noka korkoring PN RF-laugh GEN child
‘ Thechild laughed at Obay ( when Obay was nearby and did something funny)
17. ‘obay pak-boe’oe iyakin PN cause-be.angry IS.ACC
18. ya’o pak-boe’oe-en ni ‘obay I cause-be angry.-PF GEN PN ‘ I got angered by what ‘Obay did’ 9.2. Nominative NP as an affectee
19. si-osa ni ‘obay, korkoring homangih RF-go GEN PN child cry
‘ The child cried (because) Obay went away.’
20 sik-a-osa nasia, korkoring homangih RF-?-go 3PL.GEN child cry
‘ The child cried )because) they went away.’ 21. si-Sowab ma’an, sia ma-Sowab
RF-yawn 1s.Gen 3s. AF-yawn
‘ I yawned and (affected by me), s/he yawed too.’ 22. si-hila ila yao rima’ ila ray ka-kista:at-an
RF-sun pfv I go pfv Loc school
‘ The sun came out and I was able to go to school.’
9.3. Emotion events in PF vs. si- construction: Saisiayt makes an interesting distinction between emotion events in which the participants involved are just two ( regular PF clauses) and those in which the participants involved are three ( si-construction).
23.
a. ‘obay sahoewis-in ni ataw PN disgust-PF GEN PN ‘ Ataw is disgusted with ‘obay.’ b. ‘obay si-sahoewis ni ataw PN RF-disgust GEN PN
‘ Ataw is disgusted with someone/something and ‘obay is adversely affected ( because of his close association with that someone or something.)
24.
a. ‘obay sai:i’in noka ma’iaeh PN jealou GEN man
‘ Someone (tried to) harm/frustrate ‘Obay out of spite.’ b. ‘obay si-sai:i’ noka ma’iaeh
PN RF-jealous GEN man
c.’obay sik-sai:i’ noka ma’iaeh
‘ Someone (tried to) harm ‘obay’s kin/friend in order to thwart ‘obay.’
9.4. Direction perception and the sik-construction
25. …in’aray ray ‘inoehaes sahae’ ila…(3.4) korkoring from LOC window fall PERF child komita’ ka ahoe’ sik-sahae’
see-AF Acc dog RF-fall
‘The dog falls out the window…The child sees the dog falling.’ 26. yao ‘okik ra:am ka ahoe’ sik-sahae’
1S.Nom NEG know Nom dog RF-fall ‘I did not know the dog fell down.’
10. Emotion events and the compound focus sik----en construction: The RF marker in the compound focus must be sik- and the sik-…-en construction is used only when someone witnesses the event associated with the verb.
27. a. sik-kayzaeh-en ni ‘ataw, iban hoemakas RF-good-PF Gen PN PN envy-AF ‘(Since) Ataw is good at everything, Iban is jealous.’ b. sik-kayzaeh ni ‘ataw, iban hoemakas
‘(Since) Ataw is good at everything, Iban is jealous.’
28. a. sik-hayza’-en ila, ‘obay kayzaeh ka inazazem RF-exist-PF PERF PN good NOM mood ‘(Since) there was someone around, Obay felt good.’
b. sik-hayza’ ila, ‘obay kayzaeh ka inazazem ‘(Since) there was someone around, Obay felt good.’
29. a. yao ‘am rima’ sik-‘oka’-en ila noka ra:an I AUX go RF-no-PF PERF GEN road ‘I wanted to go, (but) there was hardly any road left to go on.’ b. yao ‘am rima’ sik-‘oka’ ila noka ra:an ‘I wanted to go, (but) there was hardly any road left to go on.’
11. A comparison of PF, RF and compound focus sik-…-en constructions: 28 PF: ‘obay oka’-en hi/?ni iban oemahoewis ka PN Neg-PF ACC/GEN PN bad NOM Inazazem
mood
‘Iban was not around (and) Obay felt sad.’ 29. RF: ‘obay sik-‘oka’ ni iban ba:in ila PN RF-Neg GEN PN lazy PERF ‘(Since) Iban was not around, Obay became lazy.’ 30. CF: ‘obay sik-‘oka’-en ni iban sia nonak ila PN RF-Neg-PF GEN PN 3S.NOM self PERF ‘(Since) Iban was not around, Obay was alone by himself.’
31. CF: ‘obay sik-a-‘oka’-en noka ma’iaeh, ba:in ila ‘(Since) nobody was around, Obay became lazy.’
32. AF: ‘obay oka’ hi iban, ba:in ila ‘Obay is without Iban and is lazy.’
11.1 How to feel cold in Saisiyat? 33. yao maskes atomalan I cold-AF very ‘I am cold.’
34. yao kaskes-en ‘oka’ ila nanaw I cold-PF Neg PERF still ‘I am so cold I cannot take it.’
35. yao si-kaskes rini raremoean, ‘ayaeh ila I RF-cold here PN sick PERF
‘I ran into cold weather (when I was) here in Raremoean, and got sick as a result.’
12. The compound focus sik-…-an construction: appears to be very productive and is formed by reduplicating the first syllable of an intransitive verb root and attaching sik-…-an to the product.
36. sik-‘ayaeh noka korkoring, ‘oya’ oema-hoewis ka inazazem RF-sick Gen child mother worried NOM mood ‘The child got sick and mother was worried.’
37. sik-‘ay’ayaeh-an noka korkoring, ‘oya’ oema-hoewis ka RF-sick-LF Gen child mother worried NOM
Inazazem mood
‘The child has been sick and mother is worried.’ 38. ni ‘ataw sik-hayhayza’-an, ‘obay siya’ ila Gen PN RF-exist-LF PN happy PERF ‘Ataw seems to have (it) and Obay is glad.’
39. sik-alialikah-an manra:an, boway ‘am sahae’ RF-fast-LF walk-AF fruit AUX fall ‘If you walk too fast, the fruit will fall down.’
13. Three cultural modes for the discourse on emotion (Huang 2002)
a) The metaphorical model: the conceptual structure of emotions in English depends to a considerable extent on the use of metaphor models and to a lesser extent on the use of metonymic models.
b) The metonymic model: there are extensive conceptual metonymies used in structuring emotion concepts in Chinese.
c) The grammatical mode:
c1: Tsou: prefers to use grammatical prefixes for bodily actions as part of lexicalized verbal expressions that indicate antecedent trigger events leading to emotional onset or behavioral consequences.
cause of an emotional event or its affectee or both. Recent research has shown that there are differences in emotional life in different cultures, in the elicitors and interpretations of emotion, in display rules and in the social functions of emotion and in differential emphasis on the metaphoric and metonymic understanding of a given emotion. In some cultures, some emotions are hyoercognized while others are hypocognized ( see, for example, Kutz 1988; Mesquita and Frijda 1992; Wierzbicka 1994). A thick description of the cultural model of emotion in Saisiyat and in any other language would be a project worth undertaking by researchers in the coming decade.
Notes
l. Clauses with a compound focus are also a frequently encountered pattern in Squliq Atayal. They come in at least two types. In one type, the RF marker s- is attached to an active verb(a) or a stative root (b); in another type,the RF marker s- and the morphological causative morpheme p- are attached to a stative root, as in (c) below:
(a)
Q: s-tmaq-an isu’ qu’ niq-un na’ laqi qa ha. RF-mash-LF 2s.g Nom eat-NMZ Gen child Det Hort ‘ Would you pls mash the child’s food?’
A: Aw. s-tmaq-un-mu kira’ qu’ niq-un-nya’ yes RF-mash-PF-ls.g. later Nom eat-NMZ-3s.g. ‘ Yes, I’ll mash the child’s food later.’
(b) cyux s-
Asp RF-good-LF Gen PN Nom woman-3S ‘ Yumin is amusing his wife.’
(c) s-p-qih-an-maku’ qu’ yumin RF-Cau-bad-LF-1S.Gen Nom PN ‘ I made Yumin angry.’
References cited
Blust, R. 1999. Subgrouping, circularity and extinction: some issues in Austronesian comparative linguistics. In E. Zeitound and P. Li (eds.) Selected papers from the Eighth International conference on Austronesian linguistics. Academia Sinica: Taipei.
Himmelmann, N. 2002. Voice in western Austronesian: an update. In F. Wouk and M. Ross (eds.) The history and typology of western Austronesian system, 7-18. Pacific Linguistics, Australian National University.
Huang, Shuanfan. 2002. Tsou is different: a cognitive perspective on language, emotion and body. Cognitive Linguistics 13.167-186.
Lutz, Catherine. 1988. Unnatural emotions: everyday sentiments on a Micronesian atoll. University of Chicago Press.
Mesquita, B and N. Friida. 1992. Cultural variations in emotions: A review. Psychological Bulletin 112,176-204.
Reid, L. and Hsiu-chuan Liao. 2001. A syntactic typology of the Philippine languages. Paper presented at the International Symposium on Austronesian Cultures. December 10-12. Academia Sinica, Taipei.
Shibatani, M. 1996. Applicatives and benefactives. In Shibatani and Thompson (eds.) Grammatical constructions: their form and meaning. Oxford.
Wierzbicka, A. 1994. Emotion, language and c ultural scripts. In Kitayama, Shinobu and H. Markus (eds.) Emotion and Culture, 133-196. Washington: American Psychological Association.
Wouk, F. 1996. Voice in Indonesian discourse and its implications for theories of the development of ergativity. Studies in Language 20.361-410.
