Australian Linguistics Society
1998

Simon Musgrave , Department of Linguistics and Applied Linguistics, University of Melbourne, Melbourne, s.musgrave@linguistics.unimelb.edu.au

FULL PAPERS

All languages have strategies which can be used to increase the communicative salience of some portion of a message [1]. This paper describes a highly specific strategy used in one dialect of the Sasak language of Lombok which allows a class of adjuncts to be moved to the front of the clause, a salient position in the utterance. The following example illustrates a normal sentence pattern for the verb béng 'give', followed by an example of the fronting construction coding the same information [2]:

1a

Mun

ebéngk

hadiya

no

isiq

Ali

muq - n

béng - k

hadiya

no

isiq

Ali

PAST-3

give-1sg

present

DEM

by

Ali

'Ali gave me the present'

b

aku

lain

ebéng

hadiya

no

isiq

Ali

aku

lai - n

béng

hadiya

no

isiq

Ali

1sg

lai -3

give

present

DEM

by

Ali

'To me Ali gave the present'

The word lai is used to front a noun phrase which must meet semantic and syntactic restrictions. In section 1, I argue that the fronted NP has the pragmatic function of Focus. In section 2 I discuss the constraints on the construction: semantically, there must be motion towards the referent of the focussed noun phrase, although this motion can be abstract or only implied; and syntactically, the noun phrase which is focussed must be an oblique argument of the clause. Section 3 argues that lai is a verb of some sort, but with a limited distribution and defective argument structure, and section 4 briefly mentions an implication of the construction for language typology.

1 Pragmatic function of lai

The construction with lai is a marked sentence type: example 1b is marked with respect to example 1a. The natural assumption is that the lai construction has some pragmatic function, in particular, that it makes the fronted noun phrase salient. In this section, I will try to make this characterisation more specific.

The intonation pattern associated with the lai construction shows that the fronted noun phrase is still treated as a part of the body of the clause [3]. The noun phrase and lai form an intonation contour together, but no pause occurs between lai and the remainder of the clause. This contrasts with an example such as:

2

Ali (...)

muk

ebéng

hadiya

no

Ali

muq - k

béng

hadiya

no

Ali

PAST - 1sg

give

present

DEM

'Ali, I gave the present to'

where a distinct pause between the first two words is required to make sense of the sentence. This restriction on the intonation possibilities shows that the fronted noun in a lai construction is not set apart from the main predication. The pragmatic function of lai then is not to "present a domain or universe of discourse with respect to which it is relevant to pronounce the following Predication" (Dik 1981: 130).

In several discussions, an attribute of the Topic function is definiteness (Givon 1976, Li and Thompson 1976, Davies 1993), and the fronted noun of the lai construction does not entirely fit this pattern. Sasak does not obligatorily mark definiteness; absence of the demonstratives no/ni does not mean that a noun phrase is interpreted as indefinite. Nouns not marked as definite frequently occur in the lai construction (see examples 8b, 11, 18b and 24), but an unmarked noun, fronted with lai can be interpreted as indefinite and may be modified to fix its reference:

3a

Peken

laik

beli

sayuran

peken

lai - k

beli

sayuran

market

lai - 1sg

buy

vegetable

'From a market I bought vegetables'

b

Peken

Ampenan

laik

beli

sayuran

peken

Ampenan

lai - k

beli

sayuran

market

Ampenan

lai - 1sg

buy

vegetable

'From Ampenan market I bought vegetables'

If definiteness is criterial for Topics, clearly these nouns do not have that function.

There is strong evidence available which suggests that the location in a lai construction has the pragmatic function Focus. Bresnan and Mchombo take it as a postulate that "in interrogative clauses the interrogative pronoun or questioned constituent universally bears the FOC[us] function" (1987: 24-25), and a location can be questioned with a lai construction:

4

Sai

laik

ebéng

kelambi

no

sai

lai - k

béng

kelambi

no

who

lai - 1sg

give

shirt

DEM

'Who did I give a shirt to?'

A typical use of the construction is in replying to a question which is phrased in the same way:

5a

Sai

lain

ebéng

sepéde

no

isiq

Ali

sai

lai - n

béng

sepéde

no

isiq

Ali

who

lai - 3

give

bicycle

DEM

by

Ali

'Who did Ali give the bicycle to?'

b

Aku

lain

ebéng

sepéde

no

isiq

Ali

aku

lai - n

béng

sepéde

no

isiq

Ali

1sg

lai - 3

give

bicycle

DEM

by

Ali

'To me Ali gave the bicycle'

Here both the question word and the new information are placed at the front of the sentence, further evidence both for the location noun being Focus, and for it not being Topic. Another use of this construction is to make a noun phrase contrastive. The following example would be a felicitous reply if it was suggested to me that some third party had bought eggs from another person:

6

Aku

lain

bebait

teteloq

nani

aku

lai - n

be [4] - bait

teteloq

nani

1sg

lai - 3

take

egg

now

'It was me he got the eggs from just now'

Contrastiveness is closely related to Focus (Lambrecht 1994: 286-291), and such an example is further evidence for the position argued above. In the absence of data from natural discourse, the conclusion that the lai construction functions to make a noun denoting a location or goal the Focus of the clause seems the most that can be said.

2 Constraints on the use of lai

3.1 Semantic constraints

The lai construction occurs with eleven verbs in my database, which fall into two groups as follows:

7a

Verbs taking the role Goal:

becerite

tell a story

bekuwiq

shout

béng

give

ceritaq

tell a story

lalo

go

ngéngat

face

pegitaq

show

tayek

board, get on

b

Verbs taking the role Source

bait

take

beli

buy

jauq

take

The verbs taking a goal argument are of three semantic types. Firstly, there are three verbs denoting physical motion: lalo, ngéngat and tayek. These all involve the motion of a body or a body part to a new location and the Goal argument could be coded as a prepositional phrase headed by jok 'to' [5]:

8a

Lalok

jok

alang

bait

paré

lalo - k

jok

alang

bait

paré

go - 1sg

to

rice barn

take

rice

'I went to the rice barn and got rice'

b

Alang

laik

lalo

bait

paré

alang

lai - k

lalo

bait

paré

rice barn

lai - 1sg

go

take

rice

'To the rice barn I went to get rice'

Secondly, there is the single verb béng 'give' denoting a transfer of physical goods. This verb has already been exemplified, see example 1 above. Again, the Goal argument can be coded by a prepositional phrase:

9

Muk

ebéng

hadiya

no

jok

Ali

muq - k

béng

hadiya

no

jok

Ali

PAST - 1sg

give

present

DEM

to

Ali

'I gave the present to Ali'

The verbs in the third group denote the transfer of information. For these verbs, no physical motion is involved in the action, but the information does move in a particular direction. Again, if lai is not used, the preposition jok can be used:

10

Aku

becerite

jok

anak

aku

be - cerite

jok

anak

1sg

V - story

to

child

'I told a story to the child'

These verbs then have in common the semantic element that motion directed to a particular goal is part of the action denoted.

With the second group of verbs, lai fronts a noun which refers to the Source of some physical goods:

11

Peken

laik

beli

sayuran

peken

lai - k

beli

sayuran

market

lai - 1sg

buy

vegetable

'From the market I bought vegetables'

At first sight, the movement in question here would seem to be opposite to that which characterises the first group. There is reason to think that the movement which lai focuses on with these verbs is that of the Actor to the place of the action. One piece of evidence for this is that enton 'get off, disembark', which is the antonym of tayek, cannot be used with lai. On this account, the most natural English translation of example 11 would be I went to the market to buy vegetables, but this should not be taken to mean that lai constructions are biclausal. Note the contrast between example 11 and examples 8a and 8b, which have two verbs.

For a lai construction to be possible with a verb such as bait 'take', it is necessary for the goods transferred to be movable as can be seen from the following contrast:

12a

Ali

bait

balén

Ahmat

Ali

bait

balé - n

Ahmat

Ali

take

house - 3

Ahmat

'Ali took Ahmat's house'

b

*Ahmat

lain

bait

balé

isiq

Ali

c

Balén

Ahmat

lain

Ali

bait

kursi

balé - n

Ahmat

lai - n

Ali

bait

kursi

house - 3

Ahmat

lai - 3

Ali

take

chair

'Ali took chairs from Ahmat's house'

While such examples show that the semantic justification of the lai construction with such verbs is complex, they do not contradict the point made previously. It is clear on the basis of real-world knowledge, that Ali must move to Ahmat's house before he can take anything from that location.

I suggested above that the lai construction cannot target direct arguments of a verb. Many of the relevant verbs have variable subcategorization frames, with the goal or recipient occurring either as a direct argument or as an oblique prepositional phrase. Crucial evidence is provided by derived verb forms which have a different valency from the basic verb. Valency-changing systematically affects the possibility of using the lai construction

Firstly, there is the prefix consisting of a nasal segment, realised as ng- before vowels (Thoir, Reoni and Karyawan 1985/86 [1997]: 10). Nasal verbs are not common in Meno-Mené Sasak, and often have the effect of reducing the valency of the verb:

13a

Muk

éngat

dengan

mame

no

muq - k

éngat

dengan

mame

no

PAST - 1sg

face

person

male

DEM

'I faced the man'

b

Dengan

mame

no

ngéngat

jok

aku

dengan

mame

no

N - éngat

jok

aku

person

male

DEM

face

to

1sg

'The man faced me'

The crucial fact in the current discussion is that example 13a has no corresponding sentence with lai but 13b does [6]:

14a

*Dengan mame no laik éngat

b

Dengan

mame

no

laik

ngéngat

dengan

mame

no

lai - k

N -éngat

person

male

DEM

lai - 1sg

face

'Towards the man I turned;

A similar contrast can be seen in with the applicative morpheme -an which normally in this dialect of Sasak can only promote Beneficiaries to direct argument status. But in some cases, the applicative can be used with a Recipient:

15a

Ali

kerim

surat

jok

ariq

ninen

Ali

kerim

surat

jok

ariq

nine-n

Ali

send

letter

to

sibling

female-3

'Ali sent a letter to his sister'

b

Ali

keriman

ariq

ninen

surat

Ali

kerim-an

ariq

nine-n

surat

Ali

send-APPL

sibling

female - 3

letter

'Ali sent his sister a letter' or 'Ali sent a letter on behalf of his sister'

Again, the focus possibilities are restricted as would be predicted if lai can only front obliques:

16a

Ariq

ninen

lain

kerim

surat

isiq

Ali

ariq

nine-n

lai-n

kerim

surat

isiq

Ali

sibling

female-3

lai-3

send

letter

by

Ali

'His sister, Ali sent a letter to'

b

*Ariq

ninen

lain

keriman

surat

isiq

Ali

The only verb of those listed above, which cannot at least optionally take a Goal argument introduced by the preposition jok is bait 'take'. That verb takes a locational argument with the preposition éléq 'from, and it can therefore be maintained that lai is used to focus non-direct arguments of the verb.

3 The lexical category of lai

Focus construction similar to the one discussed here are possible in Meno-mené Sasak, for example:

17

Tunjan

no

isiqek

empuk

basong

tunjan

no

isiq-k

empuk

basong

stick

DEM

with-1sg

hit

dog

'With the stick, I hit the dog'

Austin (1998) discusses various possibilities, such as the prepositions isiq and éléq and the noun taoq 'place'. These words all occur elsewhere in the language, unlike lai which is restricted to the construction discussed here and one other construction:

18a

Mbé

yam

lai

mbé

yaq - m

lai

where

FUT - 2

lai

'Where are you going?'

b

Peken

yaq

laik

peken

yaq

lai - k

market

FUT

lai - 1sg

'I'm going to market'

c

Yak

lalo

jok

peken

yaq - k

lalo

jok

peken

FUT - 1sg

go

to

market

'I'm going to market'

Example 18a is a greeting formula, example 18b and 18c are two possible replies. This limited distribution raises the question of what lexical category lai should be assigned to.

It seems clear that lai is not a preposition. Although prepositions can follow their complements in focus constructions, a preposition can actually co-occur with lai:

19

Jok

peken

laik

jauq

sayuran

no

jok

peken

lai - k

jauq

sayuran

no

to

market

lai - 1sg

take

vegetable

DEM

'To the market I took the vegetables'

Unlike taoq, lai never occurs as a noun and it would be implausible semantically to claim that it was one. Nor can it be treated as a post-head modifier of the noun which it focuses, because a constituent which is clearly part of the main body of the clause can intervene:

20

Peken

wah

laik

beli

sayuran

peken

wah

lai - k

beli

sayuran

market

PERF

lai - 1sg

buy

vegetable

'At the market, I have bought vegetables'

and a negative particle with the focussed noun has scope over only the noun phrase, suggesting that lai higher in the structure than the NP:

21a

Ndéqek

tayek

jok

bis

ndéq - k

tayek

jok

bis

NEG - 1sg

board

to

bus

'I didn't get on the bus'

b

Ndéqen

bis

laik

tayek

ndéq - n

bis

lai - k

tayek

NEG - LNK

bus

lai - 1sg

board

'It wasn't a bus I got onto'

In example 20, lai is in a position typically associated with an auxiliary verb in Sasak, whilst in examples 18a and c, it is in the position of a verb. These examples suggest that lai is a verb of some sort, but not one that normally takes any arguments. This last point can be seen in an example where lai focuses a location with the verb bait in its applicativised form. The clause has three core arguments in it besides the focussed noun, making it highly unlikely that focussing has changed the syntactic status of the Location:

22

Yosef

lain

yaq

baitangk

jaje

sekéq

Yosef

lai - n

yaq

bait - ang - k

jaje

sekéq

Yosef

lai - 3

FUT

take - APPL - 1sg

cake

one

isiq

Simon

isiq

Simon

by

Simon

'Simon will take a cake from Yosef for me'

Examples 18a and c would then either be instances in which it is used as an intransitive verb, or examples of ellipsis of the main verb in a formulaic greeting. Lai can certainly co-occur with the verb lalo 'go':

23

Mbé

lain

lalo

wiq

mbé

lai - n

lalo

wiq

where

lai - 3

go

yesterday

'Where did he leave for yesterday?'

This analysis is not fully satisfactory, but the only alternative is to claim that lai is a particle, which is to make no claim at all. The analysis is perhaps more plausible if the construction is seen as a fossilised serial verb construction which has been retained only to allow the language to meet certain functional pressures [7].

4 Conclusion

I have claimed that lai is used to focus a location (usually a Goal) in Meno-mené Sasak, and that a similar construction is possible with a variety of other words. It is a very striking fact that the focussed noun in such constructions can be the head of a relative clause, and that this is the only way of relativising on many semantic roles in this dialect of Sasak:

24

Masjid

saq

laik

tetep

lalo

léq

masjid

saq

lai - k

tetep

lalo

léq

mosque

REL

lai - 1sg

always

go

LOC

perempatan

perempatan

junction

'The mosque that I always go to is by the junction'

Austin (1998) discusses these possibilities in more detail; here I only draw attention to the fact. In common with most Austronesian languages, Sasak offers limited possibilities for relativisation and does not permit obliques to head relative clauses normally. But this restriction can be overcome by focussing the oblique noun phrase. This results in the typologically unusual feature of the pragmatic status of a noun phrase affecting its accessibility to relativisation. Further research into other dialects of the Sasak language is needed; as these dialects have alternative strategies for making such noun phrases accessible to syntactic processes, it is uncertain whether similar constructions will exist in them.

References

• Austin, Peter K. (1998) Relativisation in Sasak. Paper presented to the ALS Conference, Brisbane
• Bresnan, Joan and Mchombo, Sam (1987) Topic, Pronoun and Agreement in Chichewa (CSLI Report 87-78) Stanford CA:CSLI Publications
• Davies, William D. (1993) Javanese subjects and topics and psych verbs Linguistics 31:238-277
• Dik, Simon K. (1981 ) Functional Grammar Dordrecht:Foris Publications
• Givon, Talmy (1976) Topic, pronoun and grammatical agreement. In Li (1976) pp149-188.

• Lambrecht, Knud (1994) Information structure and sentence form Cambridge UK: Cambridge University Press
• Li, Charles N. (ed.) (1976) Subject and Topic New York:Academic Press
• Li, Charles N. and Thompson, Sandra A. (1976) Subject and topic: a new typology of language. In Li (1976) pp457-489
• Thoir, Nazir, Reoni, Ketut and Karywan, I Ketut (1985/86 [1997]) Tata Bahasa Bahasa Sasak Jakarta:Pusat Pembinan dan Pengembangan Bahasa. English translation by Dewi Novi Djenar, edited by Peter Austin:Department of Linguistics and Applied Linguistics, University of Melbourne

Notes

1. This paper could not have been written without Herman Suheri and the generosity and enthusiasm with which he shared his knowledge of his native language. I am grateful also to Yosep Kroon and Yusuf Eades who worked with me in sessions with Herman and shared their insights into our data with me, to Judith Bishop for her help with the intonation of some of my data, and to Peter Austin for comments on a previous version of this paper. (return to text)

2. The following abbreviations are used in glossing examples:PAST = past tense, PERF = perfect aspect, FUT = future tense, PASS = passive, N = nasal prefix, CAUS = causative APPL - applicative. 1,2,3 = 1^st person etc., sg = singular, pl = plural, NOM = nominalising affix, DEM = demonstrative. NEG = negative marker, REL = relativizer, LOC = locative preposition, LNK = linker. lai is not glossed in examples. This dialect of Sasak does not distinguish singular and plural in 2nd and 3rd person pronouns and clitics. e indicates a central vowel, and é a mid-front vowel. Sasak avoids consonant clusters, and epenthetic vowels frequently occur across word boundaries. In particular, when the verb béng follows a consonant-final word, it is pronounced ebéng. This is indicated in the examples by the difference between the first, text, line and the second, morphemic, line. (return to text)

3. This paragraph is based on a discussion with Judith Bishop. (return to text)

4. The function of the verbal prefix is not clear to me in this example. (return to text)

5. The preposition is in fact optional in the case of tayek. (return to text)

6. Example 13b and example 14b were elicited on separate occasions and are not parallel in the positioning of nouns, however the essential point is clear when the two examples are compared. (return to text)

7. I owe this idea to John Bowden. (return to text)