2000 East Nusantara Linguistics Workshop

Deixis Workshop
Saturday 22 July 2000
2.00pm - 5.30pm

Co-ordinator: Aone van Engelenhoven engelenhoven@rullet.leidenuniv.nl

This web page contains a sample version of answers to the questionnaire for Aone van Engelenhoven's workshop on deixis and location at the 2000 East Nusantara Linguistics Workshop. This sample was prepared by Aone van Engelenhoven and relates to Leti, spoken in South-West Maluku. The questionnaire itself can be viewed on line by clicking on the following link:

The questionnaire is also available as a rich text format file by clicking on the following link:

This sample answer to the language contact questionnaire is also available to be downloaded as an rtf file, as are some other sample responses. These can be downloaded by clicking on the links below. (More sample answers to be added to this list as they become available).

If you have any trouble accessing any of these things please email John Bowden at john.bowden@anu.edu.au or Aone van Engelenhoven at engelenhoven@rullet.leidenuniv.nl, and one of us will send you a copy by other means.

Deixis and Location

Deixis

It is generally acknowledged that perception of and orientation in space are determinant factors in human action and interaction. As such, speech heavily depends on knowledge of the context: Where and When is a sentence uttered, and by Whom. These three dimensions are traditionally seen as the so-called deictic centre of all linguistic events, without which no linguistic expression can be properly interpreted. Svorou's (1993) observation, that social and psychological conditions are also relevant factors in the deictic anchorage of language, fully applies to the East-Nusantara Region.

With deixis we mean here all cues provided by a language that localise a speech event and its participants (Speaker, Hearer and narrated participant) in space and time. Anderson and Keenan (1985) distinguish three major categories of deixis: person deixis, spatial deixis and temporal deixis. The category linking social and psychological factors is tentatively labelled 'psychological' deixis at the workshop.

Person deixis usually localises an entity in relation to the position of the Speaker and/or Hearer (a so-called 'positional' system). First and second person pronouns typically refer to the speaking and hearing speech-participant(s), whereas third person pronouns designate the non-speech or narrated participant. Many Austronesian languages in East-Nusantara encode additional information about the referent, for example the number of individuals referred to (singular, dualis, trialis, plural), its classification (male, female, animate, inanimate, edible) or social status (impersonals, deferential pronouns).

Spatial deixis localises both the Speech participants and narrated participants in space. Some languages in East-Nusantara, as for example Ewaw (SE Maluku) only have one term, for which it is very difficult to define its meaning. Many languages, however, may display a two-term or three-term positional system, designating locations in space with reference to the position of the Speaker (cq. 'this'=near the Speaker, versus 'that'=near the Hearer and/or 'yonder'=not-near any of the Speech participants). Not only the relative distance between the object referred to and the Speaker/Hearer may be encoded. Many languages in East-Nusantara are reported to also indicate the level on which the referent is located relative to the Speaker's/Hearer's position. Most non-Austronesian languages on Alor (NTT), for example obligatorily signal whether the object is located above or below the Speaker, or on the same level.

Temporal deixis as it proposed by Anderson and Keenan (1985) localises the speech event in time by means of adverbs ('now', 'then') or nouns ('Tuesday', 'April'). Tense inflection on verbs can also be analysed as temporal deixis in this respect. It is suggested in the literature, that temporal deixis by means of adverbs or demonstratives is least common in the languages of the world. This workshop intends to investigate this claim for East-Nusantara. Leti (Southwest M aluku), for example, has a set of three determiners clearly originating from a spatial deictic set that locates the narrated event in time with reference to the moment of speech.

'Psychological' deixis encodes information about the referent that is related to the psychological framework of the Speech participants rather than to the localisation of the Speech event in space and time. Leti (Southwest Maluku), for example, has a separate set of person pronouns that signals the Speaker's attitude (acceptance versus rejection versus unacquaintance) toward the referent. In Taba (North Maluku), on the other hand, the acquaintance or unacquaintance of the Speech participants is a semantic extension of the directionals meaning 'upward' and 'downward', respectively. In Blagar (Nusa Tenggara Timur), the Speaker's acquaintance is implied in all deictic morphs connoting 'close to the Speaker', whereas the Speaker's unacquaintance is implied by all morphs meaning 'close to neither the Speaker nor the Hearer'. This workshop intends to investigate and compare the patterns of 'psychological' deixis in East-Nusantara.

Location

Whereas not really implied yet in the types of deixis mentioned above, encoding the location of entities requires that the Speaker has a good orientation and understanding of his/her environment (a so-called 'dimensional' system). At this point, languages may diverge significantly from each other, depending on how the spatial arrangement is construed by their speakers. East and West are cardinal points of orientation for most languages in Southeast Asia, obviously connected with the rising and setting of the sun. The oblong form of Kei Besar (Southeast Maluku) along the North-South axis, however, induces that at this island North and South are rather perceived as the main directions, even though the language does not provide separate words for them. In Southwest Maluku on the other hand, most languages do have a word for South, but not for North. Does this mean then, that, for example on Leti, North is not emic?

In the paragraph on deixis it was suggested, that one of the Speech participants (or both) and the Speech event function as the deictic centre. In other words, they are the ground or landmark against which the referred entity, the figure or trajector is profiled. However, when it comes to locating objects, many languages in East-Nusantara rather prefer an entity in the environment instead of the Speech participants/event. An important directional axis in Austronesian languages is seaward-landward, which has 'overridden' the sunrise-sunset or East-West axis on Kei Besar. Here, a location (f.e. 'in front of the house') is principally perceived in relation to the sea or the mainland ('at the seaside of the house'). On big islands downstream-upstream is another major directional axis. Rivers themselves may often be perceived as landmarks themselves, as for example in Nuaulu (Seram, Central Maluku): this side versus opposite side of the river (comparable to Latin Cisalpina -Transalpina).

Many languages in East-Nusantara compulsorily encode the direction of a verbal act in relation to one of the landmarks that the language-speaker acknowledges. In this workshop we intend to compare the different reference frames (Svorou 1993) that are found in the languages.

Sources and suggested background reading

Anderson, Stephen R. & Edward L. Keenan, 1985, 'Deixis', in: Timothy Shopen (ed.) Language Typology and Syntactic Description III: Grammatical categories and the lexicon, 259-308, Cambridge University Press.

Dixon, R.M.W. 1986, 'Noun classes and noun classification in typological perspective' in: Collette Craig (ed.) Noun Classes and Categorization: proceedings of a symposium on categorization and noun classification, Eugene, Oregon, October 1983, 105-112, Typological Studies in Language, Amsterdam/Philadelphia: John Benjamins.

Himmelmann, Nikolaus, 1996, 'Demonstratives in narrative discourse: a taxonomy of universal uses', in: Barbara Fox (ed.) Studies in Anaphora, 205-254, Amsterdam/Philadelphia: John Benjamins.

Langacker, Ronald W., 1999, Foundations of Cognitive Grammar, Volume I: Theoretical Prerequisites, Stanford University Press.

Levinson, Stephen C., 1996, 'Language and space', Annual Review of Anthropology 25:353-383. Senft, Gunter (ed.), Referring to Space, Studies in Austronesian and Papuan Languages, Oxford Studies in Anthropological Linguistics 11, Oxford: Clarendon Press.

Svorou, Soteria, 1993, The Grammar of Space, Typological Studies in Language 25, Amsterdam: John Benjamins.

Talmy, Leonard, 1983, 'How language structures space', in: Herbert L. Picket & Linda P. Acredolo (eds), Spatial Orientation: Theory, Research and Application, 225-320, New York: Plenum Press.

The questionnaire

1.Which language is this questionnaire on and where is it spoken?

Leti
The Leti language, spoken on the west-side of Leti Island in Southwest Maluku.

2. Are this language and its speech community indigenous to the region? If not, where do they originally come from?

Leti
Leti is indigenous to the region.

3. Describe the pronominal system of the language. Does it distinguish singular from plural (cq Leti does, but Maccassarese does not). Does it have separate forms for dualis or trialis? Does it display an exclusive-inclusive distinction (as in Malay kami-kita)? Are there separate deferential pronouns (as for example Malay Beliau). If not, does one of the pronouns have a deferential or honorific function (f.e. ita 'we inclusive' in Tetun for 'you'). Does it use special nouns as deferentials (f.e. Leti aanmu 'your child' instead of 'I', or Classical Malay hamba 'slave' instead of 'I'). Does it use lexical parallelism, for example for honorific address (see the section on parallelism in the questionnaire on oral traditions)? Does it have a special set of possessive pronouns relating to the categorisation of the possession noun (f.e. alienable nouns have a preposed pronoun, whereas inalienable nouns are suffixed; in some languages, f.e. Buli (North Maluku) and Selaru (SE Maluku), nouns referring to food have separate possessive pronouns).

Leti
GLOSS full pronoun proclitics   possessive enclitic
    Subject agreement  
1 sg au/ïau a u- -ku
2 sg oa o mu- -mu
3 sg ea e na- -ne, -nu, -ni
1 inclusive pl ita i ta- -ne, -nu, -ni
1 exclusive pl ami a ma- -ne, -nu, -ni
2 pl mia mi mi- -mi
3 pl ira i ra- -ne, -nu, -ni

4. Is the language's deictic system related to the environment in which this language is spoken (coastal, mountainous, riverain)? Is the system maintained or modified when the language is spoken outside its indigenous location. Motion verbs in Ewaw on Kei (SE Maluku) require a directional indicating whether motion is seaward (=downward) or landward (upward). This is not encoded by Ewaw speakers in Zwolle (The Netherlands) where the environment lacks sea and mountains. Meher speakers in The Netherlands rely on the deictic terms meaning 'left' and 'right'. In the original setting on Kisar Island (SW Maluku) they prefer the seaside-land-side axis, using 'left' and 'right' exclusively for left-handedness and right-handedness, respectively.

Leti
On Leti Island, speakers are compelled to encode the direction of motion. On the west shores (where the Leti language is traditionally spoken), motion towards or away from the shore is encoded by the terms lïòra (seaside) or rïaa (land-side), respectively:

(1)	N-tikli		ba:le	la	lïòr-o/rïa-o.
	3sg-kick	ball	go	seaside-IND/land-side-IND
	"He kicks the ball in the direction of the sea/land."

Except for Tutuwaru, all villages are located on capes. The beach, which runs from the west-side to the east-side via the North is the major route for travel between the villages. The southern shores can only be travelled on horse-back and are seldomly used. Travelling is always encoded as either 'Eastward' or 'Westward' between the the villages of Nuwewang and Luhuleli, respectively in the extreme West and East. For example:

(2)	N-delo		Tomra=po	n-our=tipru		l-la	Vatumïa-o.
	3sg-from	Tomra=then	3sg-travel=east	3sg-go	Batumiau-IND
	"He travelled from Tomra (west-coast) to Batumiau (north-east coast).

5. How many sets of deictic terms does the language have (1, 2, 3 or more?). Do they encode number and/or noun class? Are the deictic categories (person, spatial, temporal and 'psychological') encoded by separate sets (as in Leti), or are they combined (f.e. Tetun ne'e (East-Timor) indicates the referent's proximity to the Speaker in space and time (discourse) and its being known to the Speaker).

Leti
Leti has three sets of three determiners as given in the following table.

GlosstermGlossExample: asu 'dog'
PsychologicalSai
So
Se		'the one I know/like'
'the one I dislike'
'the one I do not know'		As~s~ü~ai
As~s~ü~o
As~s~ü~e
Spatial (stress-bearing)Dí
Dó
Dé		'the one here'
'the one not here'
'the one over there'		As=dí
As=dó
As=dé
TemporalDi
Do
De		'the one discussed now'
'the one discussed once by us'
'the one discussed once by others'		As=di
As=do
As=de
6. If the deictic categories are encoded by means of separate sets, do they co-occur in stacks of separate morphemes (f.e. Leti Kus-dó-di 'cat-there-discussed now' versus Kus-dó and Kus-di.) Is this feature preserved when the speakers switch to local Malay (f.e. SW Malukan Malay itu kucing ni nya 'that cat here once discussed', see the questionnaire on language contact)?

Leti
In Leti, the psychological and spatial terms are complementarily distributed but can occur in combination as stacks with temporal deictics. The contrastive temporal term do only combines with the contrastive spatial term dó. In combinations of temporal deictics, de 'discussed once by others' cannot precede di 'discussed now'. Extreme deictic stacking is attested in topicalised NPs. This feature is preserved in local Malay (Melayu Tenggara Jauh), although it appears to be more elaborate than in the indigenous language. The following table gives some examples.

'the dog. . .'Malay calque
Psychological + temporalAs~s~ü~ai di
As~s~ü~o di
As~s~ü~e di		'I know discussed now'
'I dislike discussed now'
'I don't know discussed now'		Anjing ni ni
Anjing tu ni
Anjing nya ni
Spatial + temporalAs=dí de
As=dó de
As=dé de		'here discussed once by others'
'not here discussed once by others'
'over there discussed once by others'		Ini anjing nya
-
Itu anjing nya
Combinations of temporal deictic stacks occur in for example subject topics, where the first deictic locates the introduction of the referent in the discourse and the second locates the event in time with reference to the speech event:

LetiLocal MalayGLOSS
As=di de n-to:ru.
As=do de n-to:ru.
As=de de n-to:ru		Anjing ni nya gonggong.
Anjing tu ni gonggong
Anjing nya nya gonggong.		'The dog discussed now (once) barked.'
'The dog we discussed once barks now.'
'The dog once discussed by others barked.'

7. How is location encoded, by means of prepositions, postpositions or so-called 'circumpositions': prepositions and postnominal location nouns (as in Ewaw and Leti). What is the origin of the 'spatial grams'? Some may be linked to body-parts (f.e. Leti üò:ne 'its face' = 'in front'), others may not (f.e. Ewaw ratan 'top' = 'on' from the verb rat 'to go up').

Leti
Leti has one neutral preposition indicating location: lo 'at'. For directions it uses the transitive allomorphs of la:va 'to go' for direction away from the Speech participants, of ma:va or 'to come' for direction towards the Speaker and of ti:va for direction towards the Hearer. Leti has a special ablative verb delo, which is a combination of dena 'to stay (at)' and the locative marker lo. The following tables provide the spatial grams of specified location plus their source.

glosstermSource
On1 (same level as Speaker)
On2 (level above the Speaker)
Behind
In front
Under, in
Below
Up
Beside
Inside
In between		Vavna
Vuvnu
Tukra
Üòò-ne
Na:ni
Ïawa
Sïanni
Seri
Rïarma
Letvaru		—
'skull'
'back(bone)'
'face' (obligatory possessive inflection)
—
—
—
side
inner side (plus archaic locative infix)
—

8. What landmarks does the language prefer in its spatial expressions. In how far is it deictically anchored to the Speaker/Hearer? Does it also use landmarks in the environment, for example a mountain or a mountain ridge (Kedang, Lembata, NTT), rivers (Paulohi, Seram, Central Maluku).

Leti
In Leti, spatial expressions are never deictically anchored to the Speech participants. On land, speakers use the sea-land axis, as in example (1) above. On the beach and on the reef the Speaker uses the reef too as a landmark, as in example (3).

(3) N-va:lu   vatu  ma   meti po   ma:nu ra-mta:tu.
    3sg-throw stone come reef then bird  3pl-afraid
    'He throws in the direction of the reef (where I stand) to 
    frighten the birds.'

'Front' (üò:-ne) and 'back' (tukra) are linked to 'East' (Tipru) and 'West' (Varta), respectively. The front of the traditional Leti house (the main entrance) faces the East. Most islands in SW Maluku have a front-side facing the East. These landmarks are confined to islands belonging to the same Alliance (all islands except for Wetar, Damar and the islands East of Babar). Outside this region, the East-West axis = front-back axis is no longer used. In a noncoastal environment Letinese use the left-right axis, rather than seaward-land-ward

9. Does the language distinguish separate levels or dimensions (f.e. Leti uses vavna for the notion 'on' if the referent is on the same level with the Speaker, but uses vuvnu (actually meaning 'skull') if the referent is located above the Speaker)?

Leti
Leti uses vavna 'top' to signal that the referent is on the same level as the Speaker. Above the Speaker the notion on is indicated with 'skull' (thus: lo lïanti vuvan=nu 'in the Heaven' (lit. in the skull of the sky) but lo sère vavan 'on the beach').

10. Which cardinal directions (North, South etc.) exist in the language? How does it encode the direction if it does not exist in the lexicon (f.e. on Leti one says eastward or westward depending on one's position on the island; on Pura (NTT) the island of Alor to its east is 'up', whereas Ujungpandang (Sulawesi) is 'down'). Is it encoded in the language franca?

Leti
Leti has two main cardinal directions, East and West, respectively 'tipru' and 'varta'. South has a separate lexeme: tranna. There is no word for 'North'. On the north-coast 'North' is indicated by 'seaward' (lïòra); on the south-side of the island 'land-ward' (rïaa) is used instead. In the lingua franca the Malay terms, also the one meaning 'North' is used.

10a. Are the cardinal directions linked to other axes? (f.e. In Leti (SW Maluku): East=front - West=back, Ewaw (SE Maluku): South=up - North=down, Paulohi (C. Maluku): towards Ambon (w-sw)=up, Buli (N. Maluku) towards Ternate (w-nw)=up.

Leti
Yes, see also 8.

11. Does the language have a deictic and/or an inherent reference frame? Is the front or back region of an object always determined with reference to a landmark in the environment (f.e. the eastside of a house on Leti is always perceived as the front) or with reference to the Speaker (f.e. the front of a tree in Dutch is the side the Speaker looks at). Do all or some objects have a front and back region of their own that is not related to the Speaker and/or an environmental landmark. Are there objects that lack a front or back region (f.e. speakers of Indonesian, but not necessarily speakers of Dutch, perceive the labelled side of a bottle as its frontside). For more details, see Levinson 1996.

Leti
On Leti, all objects have an intrinsic front region (and consequently too a back region), which is determined by an characteristic irregularity (the label of a bottle, a major dent in a barrel). However, on the island itself the 'East-West' axis is preferred over the object's own front region.

On LetiIn Jakarta
Lo pòtle viel~v~ï~arta/vïel-tipru
At bottle side~west/side~east
'West/east of the bottle'		 Lo pòtle üò-ne/tukar-ne
At bottle face-POS/back-POS
'In front/behind the bottle'

12. How does the language encode direction in motion events: by means of serial verb constructions (f.e. Ewaw (SE Maluku): Noit in lek watuk kokat 'The wind blew away (lit. blow throw) the rice), deverbal or denominal adverbs (f.e. Leti (SW Maluku): N-vaul-seri vatu 'He threw (lit. throw-side) the stone aside') or other devices (f.e. prepositions or postpositions)? Which axes must be encoded (f.e. on Pura (NTT) both seaward-landward and up-down seem obligatory: qana hu met ma bakung da 'he lifted (lit. take come.on.same.level rise come.upward) his spoon'). Which axes are complementarily distributed (f.e. on the Leti coast seaward-landward is preferred over the East-West axis, whereas outside the island or on sea the latter is preferred).

Leti
Leti has deverbal or demoninal adverbs added to motion verbs as is shown in the following table.

Adverb Source/ related lexemeExample
Ala
È:ra
Etu
Ernu
Lòla
Li:ru
Nosri
Pa:sa
Peli
Seri
Taru
Ulu
Vutu		 'towards'
'asunder'
'off, loose'
'down' 'through, via'
'afterward'
'following'
'through'
'away'
'aside'
'behind'
'ahead'
'together'		 -
Vaïè:ra 'to tear'
Ketu 'to cut off'
Kernu 'to descend'
Lòla 'to pass'
Li:ru 'back'
Osri 'to follow'
Pa:sa 'to explode'
Vedi 'to dismiss'
Seri 'side'
-
nïaulu 'first'
Vutu 'to bind'		 l-lèr-ala 'he goes toward'
n-rèï-è:ra 'he pulls apart'
l-lèr-etu 'he goes across'
l-lèr-ernu 'he goes downward'
n-dait-lòla 'he passes alongside'
n-sòpal-li:ru 'he sails in the rear'
l-lèr-nosri 'he goes after s.o.'
n-dait-pa:sa 'he passes through'
n-vaul-peli 'he throws away'
n-vaul-seri 'he throws aside'
n-den-taru 'he stays behind'
n-sòpl-ulu 'he sails in front'
ra-mtïètan-vutu 'they sit together'

13. Are the deictic and locative expressions in the first language copied into the contact language? Is the 'social' or temporal deictic function of lexical parallelism, if any in the indigenous language, pursued maintained in the contact language?

Leti
Melayu Tenggara Jauh structures motion events as the in the indigenous languages, yielding serial verb constructions. These constructions seem in most cases to copy the constructions found in the indigenous languages, as is exemplified below:

(4) Leti:  Mü-ele   tpaku   m~ü~a-o.
           2sg-give tobacco 2sg-come-IND
           
    Malay: Kasi tembakau datang.
           Give tobacco  come
           
           'Give me tobacco.'

If any of these questions needs elaboration for the language you work on, or if its deictic and directional system is radically different from what is proposed here, do not hesitate to contact me on engelenhoven@rullet.leidenuniv.nl

Created and maintained by John Bowden: John.Bowden@anu.edu.au
Last modified: 15 May 2000