Tải bản đầy đủ - 0 (trang)
Chapter Seven. Pragmatic Factors Motivating Order Variation

Chapter Seven. Pragmatic Factors Motivating Order Variation

Tải bản đầy đủ - 0trang

1 9 0 7. Pragmatic Factors Motivating Order Variation

I start with the assumption that assertions are the pragmatically

neutral speech act. This can certainly be challenged but I take it as my

starting point. An assertion commonly presupposes some information

(though this may not be absolutely required). Commonly, an assertion

simply adds or links up new information to this presupposed portion;

normally the new information is expressed in the predicate. Alternatively,

an assertion may attempt to link up in new ways pieces of information

already assumed to be in the hearer's information store. In a simple

assertion, the speaker does not assume he or she will be contradicting any

major expectations set up or "activated" by the hearer's current knowledge network. In other words, non-marked speech acts are those assertions which largely rehearse what the speaker assumes are already-established links, or which attempt to make new links of a fairly incremental


Pragmatically non-neutral speech acts include those in which the

speaker assumes the hearer holds some proposition to be true, or alternatively assumes that the hearer has certain established links in his or her

information network. In fact, in pragmatically marked speech acts, the

major portion of the predication (usually including the predicate) is

presupposed and is not asserted. In addition, however, the speaker

assumes that the hearer either (a) has some piece of information incorrectly linked to some other piece of information, or (b) that there is a link

missing which really must be established. In the first instance (a), the

speaker will perform a marked speech act in an attempt to get the hearer

to unlink the incorrect information. In morphosyntactic terms, this is often

done simply by negation. In an instance like (b) and optionally in (a), the

speaker will perform a marked speech act in an attempt to get the hearer

to link up the correct information. To briefly give an example here, based

on contextually given information the speaker may assume that the hearer

holds a certain proposition f(x) to be the case. The speaker may choose a

construction which conveys single focus contrast (Section 7.4.1) in an

effort to get the hearer to substitute the information y for χ in the

remainder of the proposition f.

Thus, the basic difference between marked and unmarked speech acts

has to do with degree of pragmatic force associated with the speaker's

intent to manipulate the information store of the hearer. In the nonneutral situation, the speaker assumes something about what the hearer

holds to be true (or at least will accept without challenging), but takes

pains to modify in some specific way what the hearer takes for granted.

Another way of putting it is that pragmatic markedness has to do with

assumptions about the hearer's expectations based on their knowledge

structure. When simple assertions are made, the speaker assumes there

7.1. General pragmatic structure of Yagua clauses 191

will be a high degree of matching or continuity with the hearer's expectations. When a highly marked speech act is performed, the speaker is

essentially saying: "I assume this information/new alignment will directly

contradict your current expectations - but you should change your

expectations as they are wrong". When a speech act has an intermediate

degree of markedness, the speaker may be saying "I assume this information/alignment will not match anything in your current expectations - I

have reason to believe it will be surprising to you".

Clearly, there are degrees of pragmatic markedness. Here, however,

we will begin with a contrast between fairly neutral assertions, and those

speech acts made on the assumption that the hearer has something definitely wrong in their knowledge network. Some intermediate cases will be

raised throughout the chapter.

A second major type of markedness involves semantic operations.

Here I will simply claim (and not further justify) that heightening the

degree of an expressed quality is a more marked semantic operation than

simply expressing that quality. In addition, we might view negation as

semantically more marked than positive assertion - though generally there

are marked pragmatic conditions for negation.

Section 7.1 briefly presents the major features of unmarked pragmatic

structure in the clause. In Yagua, neutral predications of either the pragmatic or semantic variety occur overwhelmingly with V(A)(0) or V(S)

orders. Substantiating statistics will be presented in Section 7.5 below.

However, our primary purpose is to investigate pragmatic factors which

motivate constituent order variation. Non-neutral predications are found

to occur with verb non-initial orders. The allowable variations in syntactic

order and the pragmatic conditions under which these occur suggest that

there is a marked pragmatic structure option for the (nuclear) predication

(Sections 7.2 through 7.5). This consists of one preverbal constituent of

any syntactic role plus the "remainder" of the predication. The remainder is possibly followed by, or interrupted by, an echo of the preverbal

constituent. Section 7.6 discusses the frequency distribution of various

orders and aspects of overt noun phrase versus clitic reference to participants. The pragmatic conditions under which constituents occur in

preverbal position, plus the frequency data, together argue that verb

initial order is basic. Section 7.7 discusses pragmatic and some syntactic

factors accounting for relative ordering of object and oblique (including

postpositional) phrases.

192 7. Pragmatic Factors Motivating Order Variation

7.1. General pragmatic structure of Yagua clauses

In pragmatic terms, Yagua clauses (or sentences) can have Connective, Delimiting, Nuclear, and Clarification components. The Nuclear

component or predication is essentially the verb, its direct (selectionally

restricted or subcategorized) and oblique (i.e. non-direct) arguments, and

certain aspectual and modal operators which have scope over the verb

plus its arguments.

Connectives are elements such as conjunctions or sequence phrases

which tie the predication in with the preceding context. Delimiting

components limit the applicability of the nuclear predication to some

restricted area in the addressee's referential field (Dooley 1982:310; cf.

also Chafe 1976:50 on "topic"). Unlike nuclear arguments, a non-nuclear

delimiting element is not necessarily related to the nucleus by the semantic case or subcategorization frame of the verb. The non-nuclear delimiting component can consist of a phrase such as a locative or time expression, or a conditional or adverbial clause which has a delimiting function

relative to the nuclear predication.2 Clarification includes phrases which

further specify the identity of, or further delimit, some element of the

nucleus. In (2) below the clarification phrase jớớryoonu rỗvỗỗt 'bushmaster's

poison' further specifies the identity of the possessor previously indicated

just with the third person singular Set I clitic sa- in sa-rỗvỗỗ 'his poison'.3

Whenever two or more pragmatic components occur in a sentence,

pragmatic function and semantic scope group the components into some

type of constituent structure (Dooley 1982:308). The overall pragmatic

structure of Yagua clauses can be diagramed as follows:

(1) [Connective [Non-Nuclear Delimiting

[Nucleus] Clarification] Connective']

Clause (d) of (2) illustrates all except the non-nuclear delimiting

component. The connective element rámutiy 'therefore' logically refers

back to the situation expressed in (2a) and (2c) in which a careless child

made the bushmaster spill his poison; this explains how other animals in

the jungle came to have poisonous, biting, and stinging characteristics.

The last connective' element váriy in (2d) is more sequential in function

than rámutiy (which indicates a logical relation). Váriy indicates how the

expression in (2d) relates in terms of temporal progression to what

preceded in (2c). Muchojimyiy Bacheenu 'musmuqui-eaten ones orphan'

(i.e. the orphan of the one eaten by the musmuqui monkey) is a proper


7.1. General pragmatic structure of Yagua clauses 193


a. Níútiy





3SG-NIY musmuqui-eaten oφhan






'He., the Musmuqui-eaten-ones-orphan. ruined


b. ljpiMrya



Ί see it

c. níítíy






3SG-TIY spill-CAUS-PAST3 bushmaster-INAN


'that heĂ made the bushmaster. spill his. poison.











3PL-paint.poison-PAST3 fer.de.lance-PL





'the fer-de-lances. painted there his. poison,




194 7. Pragmatic Faeton Motivating Order Variation







'the bushmaster's. poison, then.' (LX048)

r v


9 99

The initial connective and the non-nuclear delimiting component

rarely co-occur in naturally occurring discourse. The following example

(the same as [20] of Chapter 3) illustrates a time phrase in delimiting

function. Locative delimiting phrases also occur.





one.ANIM.SG month-extent.of-AL






'For a whole month he was laid up (in bed) with it.' (KT005)

The following example (the same as (5) of Chapter 3) illustrates a

delimiting element coreferential with the nuclear subject. Delimiting

elements coreferential with nuclear objects and obliques also occur.




s ^ j y o




bite-CL.place 3SG-knee-in







big-CL.NEUT INAN-swell.up

The wasp bite in his knee, it swelled up big.' (KT004)

Within the nucleus it is possible to have marked or unmarked pragmatic structuring. In pragmatically unmarked predications Verb(Subject)-(Object) syntactic role order is employed (though there is

potential variation in relative order of direct and oblique objects; Section

7.7). This is illustrated in the nuclear portion of (2d) above, and will be

argued for in Sections 7.2 and 7.6.

7.2. The pragmatically marked nucleus 195

7.2. The pragmatically marked nucleus

For Yagua I define "subject" as the confluence of "S" and " A " in the

sense of Dixon (1979), and "object" as Dixon's " O " (see discussion in

Section 3.1). In transitive clauses, the orders VAO, AVO, OVA, and

Oblique-VAO can occur when full noun phrases are used. When obliques

occur postverbally, they may either precede or follow the direct object

(this is explored further in Section 7.7). In intransitive clauses, SVOblique and Oblique-VS orders can occur when full noun phrases are

used. In both transitive and intransitive clauses, elements of noun phrases

may occur preverbally, discontinuous from the rest of their postverbal

constituent. Here and in Sections 7.3 and 7.4 we will attempt to discover

the conditions and factors correlating with these different syntactic orders.

Before identifying the specific conditions which correlate with nonverb-initial orders, I will present a general overview of what I conclude is

the marked pragmatic nuclear structure. This consists of a pragmatically

marked (PM) component followed by the remainder (RM) of the nucleus.

Although I talk about a "pragmatically marked component", the entire

clause is a marked structure; the PM component is simply that part which

codes information which the speaker intends the hearer should attend to

most explicitly; this information is possibly not presupposed while the

remainder may be (cf. the discussion of answers to information questions

in Section 7.4.3).

The PM component may be echoed in a final PM' component which

follows, or perhaps very occasionally interrupts, the remainder of the

nucleus. The echo is generally limited to one or two words. Very, very

rarely the PM' component may occur without the PM component. This

echo is characteristic of information questions but also occurs in other

pragmatically marked situations. Though "characteristic", it is not clear

to me how well the echo is integrated into the syntactic structure of the

clause. Order and constituency within the marked pragmatic nucleus is

represented in (5).









As might be suggested by the bracketing in both (1) and (5), the

pragmatic constituent structure closely parallels the syntactic structure

posited in (22) of Chapter 3, which describes order and constituency when

full noun phrases (or free pronouns) are used. Such parallelism should

not be surprising, as syntactic structure is in part the result of grammati-

196 7. Pragmatic Factors Motivating Order Variation

cization (over time) of semantic scope and pragmatic function relations.

That is, as the number of tokens evidencing a particular pragmatic function or semantic scope relation increases in naturally occurring discourse,

such a recurring pattern provides one type of pressure towards actual

grammaticization of a syntactic configuration paralleling the semantic or

pragmatic configuration.

In the following example, clause (b) illustrates the pragmatically

marked nuclear structure:



Sanicyee *tỗcõ, tỗcõ, tỗc. "

Sa-nicyee tỗcõ, tỗcõ, tỗcõ


b. [

PM . . .








frequently-now-EMPH 3SG-speak-EMPH






'(a) He said "tỗcõ, tỗcõ, tỗcõ." (b) Frequently

he spoke, frequently.' (KT086-087)

When arguments which are subcategorized or selectionally restricted

by the case frame of the verb occur in the PM position, they are not

resumptively referenced by Set I and Set clitics (see also Section 3.1.1):


[ ã PM . [










The trap, I'm going to see first.'4

There are nine or more specific pragmatic and semantic conditions

which correlate with preverbal placement of some constituent of the

nuclear predication. These are illustrated in Sections 7.4.1 through 7.4.7.

Though different conditions can be identified, the fact that they all

correlate with placing information in preverbal position, and the fact that

all syntactic roles (subject, object, and oblique) are found there, suggests

that what is emic to the Yagua system may be simply the pragmatically

marked status of the predication or of the information coded in preverbal

7.2. The pragmatically marked nucleus 1 9 7

position. The more specific conditions that can be identified are in a

sense etic, at least with regard to order. Emicization just of pragmatically

marked status in grammar is not universal. In other languages (cf.

Watters 1979 on Aghem) different marked pragmatic conditions may each

have different morphosyntactic encodings.5

7.3. Function of the PM' component

The PM' component gives added cognitive salience to the element occurring in the PM position. The hearer's attention is particularly called to

that item of information by virtue of its repetition.6 TTie PM' component is

commonly employed (though not required) in information questions, as in

(8). (Vocative elements and interjections such as néé 'no!' do not clearly

pertain to any pragmatic constituent.)



[ . . .PM . . .



Nỳtyara musiủa



nỳtyara musiy-na



from-now 3SG-die-PAST3 lSG-father-deceased


Divỗỗ, nỳtyara

dớv?? nỳtyara

Mother how

. PM' . . . ]




'From what now did my father die Mother, from

what?' (LX002)

Example (6b) above involves an adverb in the PM and PM' components and is an instance of added detail restatement based on the assertion made in (6a) (Section 7.4.4). The PM' component not only gives

added salience to the adverb in the PM position, but also iconically

emphasizes frequency.

7.4. Functions of the PM component

The Pragmatically Marked (PM) component may code a subject, an

object, a postpositional or other oblique phrase, an adverb, or a modifier

198 7. Pragmatic Factors Motivating Order Variation

which is discontinuous from the rest of its postverbal noun or adpositional

phrase. Phrases in the PM position generally contain given and/or definite

information. However, new information can be introduced into the

discourse in preverbal position if it is simultaneously in one of the

following pragmatically or semantically marked relations: single focus

contrast, multiple (usually double) focus contrast, counter expectation,

restatement, added detail restatement, questions and answers to information questions, a threat, an assertion which is counter to cultural or situational expectations, negation of the constituent, heightened degree of the

quality expressed by a constituent, and perhaps other non-neutral

communicative intents.

There has been some discussion in the literature as to whether

languages with "flat" syntactic structure easily allow discontinuous

constituents (Hale 1982). As just mentioned, discontinuous subconstituents

of noun and postpositional phrases may occur in the PM position, separated from the rest of their noun phrase as in examples (19), (29), (33),

and (37) below. This is not particularly frequent in discourse given that it

occurs only under marked conditions. I know of no other conditions under

which elements of noun or postpositional phrases may be discontinuous

from the rest of their phrase. (Paratactic clarificatory phrases which come

at the end of a clause are probably best viewed as completely separate

syntactic constituents from the phrase they clarify.)

7.4.1. Single focus contrast and other single focus subtypes

Contrast has been defined by Chafe (1976) as a situation where (a) there

is some propositional background knowledge in the hearer's mind but

some item of information is missing or incorrectly assumed in that proposition; (b) there is a limited set of possibilities in the addressee's mind as

to candidates which could supply the missing information; and (c) the

speaker asserts which candidate is the correct one. It is possible that the

speaker assumes the hearer has the wrong candidate in mind, and he

wishes to correct this misunderstanding. Communicative situations meeting these three criteria are generally termed "single focus contrast" situations.7 Dik, et al. (1981) make a finer distinction between "replacing

focus", which corrects an incorrectly assumed piece of information, and

"selective" and "restricting" focus which do not correct information.

Selective focus selects one item from among a presupposed set of possible

candidates, as in: Presupposed background assumption: Tomás bought rice

or beans; selective focus assertion: Tomás bought RICE. Restrictive focus,

on the other hand restricts an antecedently given presupposed set to one

7.4. Functions of the PM component 199

or more correct values, as in: Presupposed background assumption:

Tomás bought rice, beans, and tortillas; restrictive focus assertion: Tomás

bought (just) rice and beans. There may be other single focus subtypes as

well. In Yagua, candidates in replacive, selective, or restrictive focus are

all coded in the PM position. In actual communication, the presupposed

background assumption need not be stated overtly in the discourse. The

assumption may be cognitively built up out of several previous overt

propositions, or may be assumed on the basis of general cultural knowledge and expectations. It is also possible that (part oí) the remainder of

the proposition may be left implicit in the context.

The following section of text is taken from a tale of Mocayu and two

wasp twins. The twins try unsuccessfully to outsmart Mocayu. In this

particular incident the group has come upon a snake and the two parties

are jockeying as to who will kill it. (From here on I use parentheses to

indicate different constituents in the pragmatic structure. The parentheses

are not meant to indicate hierarchical scope relations. Pragmatic structuring will not be indicated in pragmatically unmarked clauses.)



Syytay jifia


3SG-say JUTA

T h e wasp said,


( . . . PM . . .)

"Née! Naay











"No! We will spear him (the snake).








never anyone spear(verb)-NMLZR.INST-NEG

Never have I seen anyone speared with




Jo. "


junúuy-rá jiy-ruvéé

1SG-MALF see-INAN 2SG-spear yes

your spear. "

200 7. Pragmatic Factors Motivating Order Variation



jaachiy j f f


3SG-MALF spear


He speared at it. Nothing!




( . . PM . .)

"Rỏy j f f










3SG-run-enroute Mocỏyu

Finally Mocỏyu came running.






rỏ! "




"I will spear him! "' (KT058-065)

In this excerpt clauses (b) and (f) both code single focus contrast,

employing otherwise unneeded free pronouns in the PM position. The

background assumption shared by both parties is that someone will spear

the snake. The set of possible candidates consists of the wasp twins (it is

unimportant which wasp twin will throw the spear) and Mocayu. In

clause (b) the wasp asserts that the correct candidate is 'us', and in clause

(f) Mocayu asserts that the correct candidate is himself. The clitic j f f in

clauses (b) and (f) communicatively underlines the marked pragmatic

structure. In both cases j f f occurs in quoted material, making a

'progression' analysis of its function in these clauses unlikely (cf. Section

3.4.3). Use of jffta and j f f in clauses (a) and (d) does mark progression

from one event to another. (Clause [e] is possibly presentative in function,

rather than indicating an event per se.)

In (9b) and (f), subjects occur in preverbal position. However, all

syntactic roles may occur preverbally. An oblique occurs before the verb

in (10b). This example comes from a text describing how two groups

made peace following a time of warfare. One group has approached the

offended party asking to be friends since the population of the groups is

declining. The offended party replies:



( . . PM . .) (




raủiy jỗỗtra






MALF begin-INAN fight-INF

'You began the fighting.

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Chapter Seven. Pragmatic Factors Motivating Order Variation

Tải bản đầy đủ ngay(0 tr)