Lakoff, George and Mark Johnson. 1980. Metaphors We Live ...wilbur/Wilbur EALing...intonation contour by stress shift to attain togetherness of focus and prominence, while Catalan

88 LENA EKBERG

Lakoff, George and Mark Johnson. 1980. Metaphors We Live By. Chicago and London:The University of Chicago Press.

Langacker, Ronald W. 1987. Foundations of Cognitive Grammar. Vol. 1. TheoreticalPrerequisites. Stanford, CA: Stanford UniversityPress.1991a. Foundations of Cognitive Grammar. Vol. 2. Descriptive Application. Stan­ford, CA: Stanford University Press.1991b. Concept, Image, and Symbol. The Cognitive Basis of Grammar. Berlin:Mouton de Gmyter

Lindner, Susan. 1981.A Lexico-semantic Analysis of English Verb Particle Constructions

with OUT and UP. Ph.D dissertation Universityof California, San DiegoMiller, George A. and Philip N. Johnson-Laird. 1976. Language and Perception. Cam­

bridge: Cambridge UniversityPress.Pokorny, Julius 1959. Indogermanisches etymologisches Wdrterbuch. Vol. I. Bern:

Francke Verlag.Sweetser. Eve 1995 "Coalignment in metaphorical systems" Paper read at the 4th Inter­

national Cognitive Linguistics Association Conference, Albuquerque, NM, July1995.

Svorou, Soteria. 1995. "Iconicity in the Grarnmaticalization of Locative Constmctions".Paper read at the 4th International Cognitive Linguistics Association Conference,Albuquerque, NM, July 1995.

Vandeloise, Claude. 1991. Spatial Prepositions. A Case Studyfrom French. Chicago andLondon: The University of Chicago Press.

A Prosodic / Pragmatic Explanation for Word OrderVariation in A-SLwith Typological Implications

Ronnie WilburPurdue University

This paper will address the issue of word order variation in ASL sentences, whichhas intrigued researchers for over 20 years. I will argue that an informationpackag­ing model provides a natural explanation for ASL word order permutations interms of the identification of information as focus and the prosodic requirementthat focus must receive prominence. This argument was essentially made for Eng­lish in Creider (1979), wherein it is argued that stress and intonation provide theexplanation for movement transformations, and that a typology of discourse organ­izing languages could be developed. There are three parts to this discussion: (1)implications of an information packaging model for surface word order; (2) inter­action of a language's phrasal stress assignment system with information packag­ing; and (3) distinctions that arise among languages in their treatment of informa­tion that is not in focus. I will suggest that a typology can be constructed on thebasis of the interaction of two parameters: Vallduvi's (1991) feature [±plastic]which indicates whether a language allows stress shift, and one that I will call[±GR] , which refers to whether a language gives primacy to Grammatical Rela­tions or to Discourse Roles in the determination of surface word order.

Arguments have been offered that ASL is typologically an SVO language, thatdifferent word orders reflect both syntactic and pragmatic functions, and that suchword order changes are marked prosodically (Aarons et al. 1992; Fischer 1975,1990; Liddell 1978; Petronio 1991, 1992; Romano 1991; Wilbur 1991, 1994a).The role of morphological marking has also been recognized (Kegl 1976). I willprovide an overview of the arguments that lead to the conclusion that even thoughASL basic word order is SVO, its surface word order is primarily determined bywhat information is in focus and the fact that ASL has fixed phrasal stress assign­ment in final position. Catalan provides a language comparison, in that it also hasbasic SVO word order and fixed phrasal stress. However, there are significant dis­tinctions between ASL and Catalan in how they accomplish the goal of puttingfocus in main clause final position: ASL prefers leftward movement (preposing) toremove non-focused material from final position, while Catalan prefers rightwardmovement (right detachment or dislocation) to accomplish the same task. Also,ASL allows considerably greater freedom of constituent ordering in the mainclause, while Catalan preserves as much of its SVO ordering as it can (Vallduvi1991). Russian provides yet another type of comparison, in that it displays deter­mination of surface word order by discourse relations like ASL but it allowsphrasal stress to shift. to different positions in the sentence like English (King

90 RONNIE WILBUR WORD ORDER VARIATION IN ASL 91

1993). Spanish is a spoken language representative of the category contammgASL, with both the absence of stress shift and the preference for Discourse Roledeterminants of word order.

1. Information Packaging Perspective

From an information packaging perspective, focus is the central determinant ofboth surface word order and prosodic structure (Chafe 1976; Lambrecht 1994;Prince 1986; Vallduvi 1992). The presentation of information in a sentence isstructured according to the speaker's belief regarding the hearer's knowledge andattentional state (whether something is in the hearer's mind at the time of utter­ance; variant formulations of this generalization exist, but the distinctions that theyare intended to capture are not relevant here; Chafe 1976; Delin 1992). For infor­mation packaging purposes, focus is defined as the "information the hearer is in­structed to enter into knowledge-store" (Vallduvi 1992). The non-focus informa­tion, or ground, indicates to the hearer where and how the focus informationshould be entered into the knowledge store. Vallduvi argues that ground includesat least two different specifications: link information, which is commonly viewed astopic or theme, indicates where the information should be entered in the hearer'sknowledge store, while tail information, if present, may indicate to the hearer tosubstitute the focus information in place of existing information in the knowledgestore. In this framework, then, the absence of a tail would imply that the focus in­formation was additional information, while the presence of a tail would imply thatthe focus information was replacement information. Typically, the order of infor­mation presentation would be link - focus - tail. There appear to be important dif­ferences between ASL and Catalan in their treatment of tail information.

2. Interaction of Information Flow, Prosody, and Word Order

2.1. Role of stress placement plasticity

In English, phrasal stress may be shifted to different positions for different focusreadings while the syntactic structure remains unchanged. Example (1) illustrates afocused object NP, (2) a focused V, and (3) a focused sentence, which shows alsothat in English the stress prominence is on the last argument, in this case the sub­ject (capital letters indicate the location of the primary stress):

In Catalan intonational prominence is fixed on clause-final position and syn­

tactic operations must be used to make the focus fall together with prominence.Vallduvi (1991) notes that in Catalan, ground information is "forced out of thecore clause". The Catalan focused object NP structure is essentially the same asEnglish (1), as both languages have basic SVO order and the object NP is thereforealready in final position for focus in Catalan. The Catalan focused. V requires thedirect object NP to be moved out of final position in the core clause (indicated by tand a comma), leaving the V in final position for focus (2). The rightward moved

object NP is now in its own clause (at least intonationally) and it must remain un­stressed. The Catalan focused sentence shows that the argument available forstress, namely the subject (as in the English equivalent), appears in final positionfor focus and prominence (3)1

Vallduvi proposes that the difference between English and Catalan is theirsetting on the Plasticity Parameter: English is [+plastic] and allows molding of theintonation contour by stress shift to attain togetherness of focus and prominence,while Catalan is [-plastic] and doesn't allow stress shift. As a consequence, Catalanrequires that the focus information and the prominence in final position be broughttogether by other (syntactic) means. I have argued that ASL is also [-plastic], thatis, the order of words must be adjusted to put the focus item in a position ofprominence rather than adjusting the prosody to make the prominence fall on thefocus item (Wilbur 1994a, c, 1995b, 1996). Like Catalan, ASL prefers to have thefocused item in final position of the main clause (of course, unstressed clitics mayattach to the stressed item; Petronio 1993; Wilbur 1994c). However, differencesbetween the two languages will be a constant thread in the following discussion.

2.2. Effects offixed stress on ASL word order

In general, the most neutral of stress patterns is the one in which the entire sen­tence is in focus (broad focus; Bolinger 1985; Cruttenden 1986; Gussenhoven1983, 1985; Ladd 1980). These are the forms that might occur as a statement fol­lowing a conversational opener like "Hey, know what?,,2

bflmd

(4) TOMORROW GO WORK PRO. 1 NOT HAVE- TO, CAN STAYbflmd

HOME'Tomorrow I don't have to go to work, I can stay home.'

In other contexts, when the item to receive prominence may not be in final

position, at least three strategies are observed in ASL. First, the item may be dou­bled, so that it appears both in its original slot and in final position (Bos 1995,Petronio 1993, Wilbur 1995a):

92 RONNIE WILBUR WORD ORDER VARIATION IN ASL 93

(5) HND-#OUTPRO.l CAN'T STAY HOME CAN'T

'I discovered that I really can't stay home.'

Second, the item may appear in final position3 Examples (6) and (7) illustratesimple cases of modals in final position; (8) shows that such structures can be em­bedded; and (9) shows that simple negation may also participate in this structure:

(6) BUT STAY HOME ALL-DAY EVERYDAY CAN'T

'but I can't stay home all day every day.'

3. How ASL Differs from English and Catalan

The lack of stress shift in ASL represents a major distinction between it and Eng­lish. The result of this difference can be seen in the fact that ASL also lacks exactlythose variants of syntactic structures in which the focus is not in final position. InEnglish, predicate nominals (11) and wh-clefts (12) have so-called plain and re­versed forms (Collins 1991a,b; Heggie 1988; Heycock 1991, 1992; Wilbur inpress):

br

(7) MARY BECOME DOCTOR SHOULD'Mary should become a doctor.'

Plain

(11) My sister is the doctor.(12) Lee's tie is what I don't like.

Reversed

The doctor is my sister.What I don't like is Lee's tie.

br

(8) BILL THINK MARY BECOME DOCTOR SHOULD'Bill thinks Mary SHOULD become a doctor. '

______ b~r __ n(9) JOHN WALK PARK NOT

'John didn't walk to the park.'

The plain form of the predicate nominal may be viewed informationally aslink-focus (cf. also theme-rheme) while the reversed form (with stress on doctor) isfocus-ground. The plain form of the wh-cleft is focus-ground and the reversedform is link-focus. The absence of stress shift in ASL means that the two cases in

which the focus is not in final position, namely the reversed predicate nominal andthe plain wh-cleft, do not occur; only those forms with the focus in final positionare found:

br

(IS) DOCTOR PT WHO, MY SISTER'The one who is the doctor is my sister.'

To illustrate, consider a situation in which people are attending a play in whichone person's sister has the role of a doctor. In English, the person has both optionsin (11), but in ASL, only the plain predicate nominal in (13) One structure thatlooks similar to the reversed predicate nominal is the wh-c1eft (15), but it is mysister that is in focus, not the doctor:6

*LEE pass TIE WHAT PRO.!DISLIKE'Lee's tie is what I don't like'.

Third, there may be overt syntactic focusing. Example (10) illustrates the wh­cleft (or pseudocleft), which is extremely common in ASL usage:4

br

(10) FIND-#OUT WHAT, STAY HOME CAN'T

'What I discovered is/was that 1can't stay home.'

In a series of papers, -I have demonstrated that the wh-cleft in ASL meets allthe criteria for a true focus structure and that the common references to it in the

ASL literature as the rhetorical question structure fails to capture its appropriatestructure and function (Wilbur 1994a, 1995b, 1996). In this structure, the firstclause equivalent to the English what 1discovered represents non-focused presup­posed material containing an open proposition (that is, one containing a variable),and the second clause contains the focus material (which provides the informationfor the variable; Prince 1978, 1986).

Another syntactic focusing structure that has been identified for ASL is the

cleft which is marked with THAT, identified by S Fischer (p.c) as the equivalentof the English it-cleft. I will return to this structure later in Section 35

Plain

br

(13) MY SISTER DOCTOR'My sister is the doctor.'

(14)

Reversed

br

*DOCTOR PT MY SISTER

'The doctor is my sister.'

br

PRO.! DISLIKE WHAT, LEE passTIE'What I don't like is Lee's tie.'

94 RONNIE WILBUR WORD ORDER VARIATION IN ASL 95

ASL

fuA: KAY THAT (DRIVE CAR)

A: Kay was driving her Dad's new sports car and ran into a tree.B: WHO was driving his car?

English

(18)

A: It was Kay who was drivinghis car

Here again, the repeated information who was driving his car is acceptable(but not required) in English and is usually omitted but possible in ASL. The ASLform involves the focuser THAT (Wilbur 1994b) (19) provides a situation in

(17) A Kay and Kim got in a wreck Saturday. I think she wasn't wearingher glasses or something.

B WHO wasn't wearing her glasses?English ASL

_IfA Kay wasn't wearing her (a) KAY

glasses. (b) KAY, THAT WBO, THAT(c) KAYPT, GLASSES, NOT

HAVE

In English, it is normal and acceptable (but not required) to repeat the infor­mation wasn't wearing her glasses as part of the answer to the wh-question. InASL this information is normally deleted. Trailing information, such as THATWHO, THAT in (17b), repeats the focus of the echo question WHO wasn't wear­ing her glasses?, but none of the ground information. The option shown in (17c) isa distant third in preference, but it does show that it is possible to have the groundinformation repeated, and also that modifications in word order and pausing aremade for emphasis; that is, unlike Catalan, which allows all possible variations ofright dislocated constituents, ASL has a preferred sequence for redundant informa­tion when it does occur

A similar situation obtains in (18), this time with only one antecedent for the

subject. The echo question who? now reflects a request for information which theaddressee did not catch (Completion Focus; Ferro 1993):

conversation, complete sentences are expected, at least from children. In ASL, theappropriate answer is much more likely to be just ROCK, and efforts to elicit acomplete sentence from ASL consultants result in frustration on both sides, orworse, signed English versions of the desired complete sentence. ASL stronglyavoids redundancy in conversational turns. If apparently redundant information ispresent, then a purpose is being served.

Consider first (17) which illustrates a response to an echo question (who?)

that requests the proper person (Kay or Kim) be selected as clarification (SelectingFocus; Ferro 1993):

The actual ordering of tail phrases used is discourse dependent, but thesepossibilities illustrate another aspect of Catalan, namely that when the non-focusedmaterial is moved rightward out of the core clause, the strict SVO word order seenin main clauses is no longer maintained. The first tail serves the function of indicat­ing that the focus information should be substituted for what the speaker believesthe hearer has in knowledge store. I will refer to this use of tail information as amild contrast indicator, in the sense that there are much stronger ways that thespeaker could tell the hearer that the information the hearer already has is incor­rect. Subsequent tails are presumably redundant for this function. The fact thatCatalan allows multiple tails when only one is needed to identifY the focus as asubstitute indicates that it has a higher tolerance for repetition of discourse-oldinformation.

ASL conversation generally does not use this type of mild indication of con­trast. If the intended purpose of the focus information is to contrast strongly withthe hearer's belief as assumed by the speaker, ASL uses formal contrastive struc­tures rather than mild tail indications7 If the tail would be merely redundant, it isomitted entirely. This fact became obvious in my attempts to use focus questions toelicit different focused constituents as answers (cf. Kanerva 1990). In English, thesequence "What did the boy throw at the house? The boy threw the ROCK at thehouse." is perfectly acceptable; the answer part even has a stereotypic stress andintonation pattern. Indeed, in some contexts, such as classrooms or polite/formal

(16) [AL CALAIX,} jicara el ganivet el Juli.In-the drawer, put-future the knife JuliJAL CALAIX,} jicara, el ganivet, el Juli.[AL CALAIX,} jicara, el Juli, el ganivet.[AL CALAIX,} el Juli, el ganivet, jicara.[AL CALAIX,} el Juli, jicara, el ganivet.[AL CALAIX,} el ganivet, el Juli,jicara.[AL CALAIX,} el ganivet, jicara, el Juli.'Juli will put the knife [in the DRAWER.]'

In sum, then, the absence of stress shift in ASL results in only two syntacticforms for which English has four; the two forms that do not occur are those inwhich the focus would not be in final position and hence would not receive promi­nence. These examples illustrate the interaction between the ability to shift stress,the location of focus, and the surface word order possibilities in English and ASL.

A comparison between ASL and Catalan reveals other differences, ones whichare particularly interesting because both languages have basic SVO order, do notallow stress shift, and prefer focus in final position. Vallduvi argues that the basicstrategy for achieving focus in final position in Catalan is by multiple right detach­ment, that is, by moving non-focused material rightward. Note that all possibleorderings of tail information can occur:

96 RONNIE WILBUR WORD ORDER VARIATION IN ASL 97

(22) Link Focus (Tail)~MARY JIM LOVE TEASE

'As for Mary, Jim loves to tease her.'

Consideration of this functional distinction provides a means for distinguishingtwo ASL structures with apparently identical syntax, Topicalization for ContrastiveFocus purposes (21), which Lambrecht (1994) cal1s contrastive focus, and Topicfol1owed by traditional Comment/Assertion (22), with the function of the Topic toidentifY or reintroduce who is being talked about:

Focus information in (21) is provided in contrast to other information that thespeaker believes the hearer holds in knowledge store; no link wil1 occur in the samesentence. In (22), the link establishes (or re-establishes) Mary as the topic of dis­cussion and asserts about her that Jim loves to tease her. The focus information is

not intended to replace existing information with respect to either Mary or Jim, andno tail is used.

There are two further implications of this analysis for actual production, onefor the stress marking and the other for the non-manual marking of topics. MARYin (21) is a focus and should receive primary stress for the entire sentence; in (22),the primary stress for the sentence should appear in the clause JIM LOVE TEASE(our data indicate that LOVES is the usual location, this being the head of the finalVP constituent).

Aarons (1994) discusses three types of non-manual marking for topics(phonetic details of which wil1not be discussed here). Of these, tml is used only ontopics which may be considered to have been moved from their original position; inthe case of (21), Mary is presumably the underlying object argument of tease andhence would be predicted to occur only with tm!. In (22) MARY is a plain topicand should have tm2 marking, which Aarons identifies as serving to change dis­course topic (among other functions).

which the repeated information may not be omitted in ASL in response to an echoquestion for an object NP that the addressee did not catch (Completion Focus):

(19) A: I told Kay she should consider going into counseling.B: You told WHO?

English ASL

lliA: Kay. A: KAY THAT, TOLD FINISH

It was Kay who I told. *KAY (THAT)

What we have seen so far is that ASL al10ws but avoids repetition of cross­turn information, while English easily permits such information; Catalan also per­mits such information but it is right dislocated. Repetition of cross-turn redundant

information is more acceptable in ASL when the focused material is an object (19),as opposed to a subject (17, 18). The key to the acceptability of such informationin (18) and (19) is the presence of THAT, an ASL foeuser that must always fol1owits focus associate. Recal1 Fischer's suggestion that the ASL equivalent of the it­cleft is a cleft with THAT. In (18) and (19), the form 'NP THAT, XP' is a cleft itwas NP who XP It was Kay who was driving; It was Kay who I told (Wilbur1994b) Delin (1992) discusses the fact that the function of tail information in clefts

is more to remind the hearer than to assert to the hearer (or to mark a substitute,as in non-cleft structures). Thus, the presence of tail information in these cleft ex­amples is unremarkable.

What remains to be snown is how ASL treats ground information when thefocus is offered as substitute information, that is, that ASL does require tail infor­mation in exactly those situations described by Val1duvi. Consider (20), which isfrom Aarons (1994), and the unacceptable options, which are from my data:

__ b_r

(20) JOHN NOT-LIKE JANE. MARY, HE LOVES*MARY (THAT)

'John doesn't like Jane. It's MARY he loves.'

Here, Mary is the focus which is offered as the correct information as to who

it is that John loves; the tail he loves signals that Mary is a substitute to be put intothe listener's entry of John loves Jane. Note that the English al10ws a cleft for thecorrective clause It's Mary he loves while ASL does not. English can signal thatMary is the focus by shifting stress; ASL puts Mary in final position in its clause bypreposing/topicalizing it out of the core clause (which some dialects of English canalso do) Thus, (20) provides evidence that ASL can use tails as Catalan does, toserve an information packaging function. The two languages differ with respect totheir treatment of tails that do not serve such a function, Catalan permitting multi­ple tails and ASL strongly preferring to omit non-functional tails.

(21) (Link) Focus~MARY,

(Jim doesn't like totease. '

Tail

JIM LOVE TEASE [t]

tease Jane.) 'It's MARY who Jim loves to

98 RONNIE WILBUR WORD ORDER VARIATION IN ASL 99

cused material to move rightward around non-focused material to final position(examples 5-9) while Catalan retains strict SVO order and must move non-focusedmaterial out of the main clause until the focused material remains in final position.Catalan is more rigid in word order than ASL. Outside the main clause, Catalanallows free order among the moved constituents, while ASL moves non-focusedmaterial leftward or omits it.

Preliminary analysis of the facts of Spanish indicate that it is also a member ofthe [-plastic] group. Moreover, Spanish is like ASL and unlike Catalan in the de­termination of surface word order. Jimenez (1995) discusses the information status

of Spanish subjects that appear in pre-verbal, post-verbal, and sentence-final posi­tions, noting that pre-verbal subjects are discourse links and post-verbal and sen­tence-final subjects are in focus, with the sentence-final subjects more specificallyin contrastive focus. As a result, various word orders may occur in the main clause

(SVO, OVS, SOY) and these orders are determined by discourse roles (Ocampo1995). In Catalan, if the subject occurs in the main clause, it must be pre-verbal tomaintain SVO word order. Also, Spanish, like ASL, omits the subject if it is simplyredundant (that is, not a link, not contrastive, or not in focus). Spanish, then, is [­plastic], has various word orders allowed in the main clause, omits redundant ma­terial and does not make heavy use ofR.D. Spanish patterns with ASL more thanwith Catalan.

Further consideration of differences between English and Russian reveal that

parallel distinctions must be made within the [+plastic] group as well. There is aclear contrast between English, which determines word order in accordance with

grammatical relations, and Russian, which determines word order on the basis ofdiscourse roles. Clearly, the single parameter [plastic] is insufficient to capture thedifferences between English and Russian, and between Catalan and Spanish!ASL.These differences are not prosodic in nature, but rather reflect whether a language

prefers to give greater prominence to grammatical relations (which results in morerigid word order) or discourse/informational status (which results in more flexibleword order) There is a need for another parameter to capture how languages dif­fer on the relative priority of grammatical and discourse relations; I will term thisthe Grammatical Relation parameter. [+GR] indicates grammatical relations have

primacy; [-GR] indicates that discourse roles have primacy. The following possi­bilities result:

4. Typological Implications

4.1. Same goal, different strategy

From a typological perspective, it is important to appreciate the implications of thediffering behavior of ASL and Catalan with respect to focus. ASL and Catalandiverge in two ways with respect to how the focused item ends up in final position.

First, Catalan makes extensive use of Right Dislocation (RD.) of ground in­formation, with one or more phrases that, in essence, leave behind the focusedconstituent in final position in the main clause. ASL, however, makes little use ofRD. and frequently omits ground material.

Second, Catalan has strict SVO word order in the main clause, but free wordorder among multiple RD.s (as illustrated in [16];Vallduvi 1991). ASL also hasbasic SVO word order, but more freedom is allowed in the main clause. Fischer(1990) notes the possibility ofminitopicalization - fronting oflexical items particu­larly if the item is definite. Her analysis is consistent with the idea offered here, thata constituent which is not the focus of a sentence is moved out of final position toavoid being construed as focus. This type of movement gives the impression thatASL shuffles word order (hence the long-standing debate over whether it has abasic word order). In actuality the movement possibilities are considerably morelimited than the term shuffling might suggest, and as Fischer observed, are primar­ily leftward (see further discussion in Wilbur 1994a, 1995a).

In sum, Catalan and ASL both prefer stress in final position and ensure thatfocused constituents appear in that position. Catalan accomplishes this by heavyuse of rightward movement while maintaining strict SVO ordering of whateverremains in the main CP; in contrast, ASL tolerates word order movement withinthe main CP and relies on deletion or leftward preposing of non-focused materialto ensure that the focus phrase appears in final position. The critical prosodic re­quirement that both languages must meet is that of final prominence in the mainassertion. The critical information packaging requirement is that focused informa­tion must receive prosodic prominence. The two languages use different strategiesto achieve the same goal.

I will now turn to the typological implications of the above observations.There has been a tendency in linguistics to divide languages into two mutually ex­clusive groups: configurational and non-configurational, or, more recently, dis­course configurational (Kiss 1995) On this basis, I initially separated English fromCatalan, Russian, and ASL. Vallduvi's work on Catalan provided an excellent op­portunity for comparison with ASL, and his typological parameter of Plasticityprovided a prosodic basis for the division of English from Catalan and ASL, and amotivation for surface word order variations in languages with fixed phrasal stress.Yet, within the [-plastic] group, careful examination of the differences betweenASL and Catalan suggested further subdivision. Both languages are SVO, henceconfigurational in the traditional sense. Within the main clause, ASL allows fo-

Table 1.

[+GR][-GR]

[+ Plastic]EnglishRussian

[- Plastic]CatalanSpanish, ASL

100 RONNIE WILBURWORD ORDER VARIATION IN ASL 101

With a language sample of only 5, it is difficult to make generalizations;

whether these categories represent meaningful typological distinctions among lan­

guages remains to be demonstrated. However, the suggestion here is that surface

word order is not fully explained by the location of stress and information packag­

ing, but that languages differ on the extent to which discourse relations that are

reflected in the information packaging interact with within-sentence grammaticalroles.

4.2. Same structure, dijJerentjunction/goal

There is one further observation on the difference between ASL and English that

needs to be made. We saw that English permits plain and reversed predicate nomi­

nals and wh-clefts; this was explained as the result of English allowing stress shift.

If focused information is final, stress and focus are together. If focused material is

other than final, then stress must be shifted to achieve togetherness of focus and

prominence. In ASL, non-focused material that is not omitted must be preposed

(unless in a cleft, which has been identified as a frequent exception; Collins

1991a,b). Clearly, preposing serves different functions in the two languages. In

English, the presence of preposing is primarily a pragmatic issue, occurring in

situations in which the appropriate contexts for inversion are met (Ball 1991, Bir­

ner 1992, 1994, Birner and Ward 1993, Coopmans 1989, Ward 1988, Ward and

Birner 1994). In ASL, preposing is part of syntactic-pros odic-pragmatic conspiracy

to put focused information in final position with stress and is therefore required as

a consequence of the absence of stress shift (Wilbur 1994a, c, 1995a). Put another

way, the absence of stress shift in the prosodic component requires assistance from

the syntactic component to achieve desired pragmatic-informational goals.

Acknowledgments

This research was funded in part from NIH grant ROI-DC00935 from the National Instituteon Deafness and other Communication Disorders. I would like to thank Sandra Wood for her

invaluable assistance with the data and linguistic analysis, Bob Channon for discussion of theparameter analysis applied to Russian, Bob Hammond for discussion of the Spanish data, andKim Bjarkman for her editorial help.

Endnotes

1. Jimenez (1995) offers an information packaging explanation for Spanish that would seem toapply to the Catalan example in (3), namely that the subject appears post-verbally when theentire sentence is in focus. Sentence-initial subjects in Spanish are links, and sentence-finalsubjects are themselves in focus.

2. ASL glosses are typically written in capital letters: TOMORROW. When more than one Eng­lish word is required to gloss a sign, hyphenation is used: HAVE-TO. Non-manual markingswhich mayor may not affect the acceptability of a sentence in a particular context are indi­cated above the gloss line, with a line showing the durational domain of the marker (whichusually corresponds with syntactic structure). The non-manual markers used in this paper are:bf for brow furrow, br for brow raise (discussion of syntactic behavior in Wilbur 1995a), mdfor mouth down, n for negative headshake, If for lean forward, and Ib for lean back (for leans,Wilbur and Patschke 1996). Two others that will be referred to are tml and tm2, topic mark­

ing 1 and 2, from Aarons (1994).3. There are three potential analyses for this structure: (1) the focused item moves rightward to

the end; (2) the item is doubled as illustrated in (5) and the original occurrence is omitted (cfBos 1995); or (3) the non-focused material is preposed. It may be the case that one or more ofthese strategies is involved, depending on such things as whether the structure represents anexample of broad or narrow focus, but the issue is beyond the scope of this paper.

4. Note that in the ASL. not only is STAY HOME CAN'T focused by virtue of the special wh­cleft syntactic structure and by placement at the end of the sentence, but also that the modalCAN'T is in final position within its clause, as in example (6).

5. Fischer (1990) also discusses a definiteness effect which she called "mini-topicalization"; insuch cases the complement precedes the head instead of following it. The result of suchmovement is that non-focused material does not appear in (he final position of the mainclause. ASL also has a variety of other focusing mechanisms (Wilbur 1994b). These includelexical focusing items such as SAME 'even' and ONLY 'only', and body leans (forward!backward for inclusive/exclusive and right/left for parallel focus) (Wilbur and Patschke1996)

6. Another similar looking structure has been offered:

(i) DOCTOR THAT-PT, MY SISTER'That doctor is my sister. '

Its discourse function, proper syntactic analysis, and non-manual markings have not yet beendetermined; hence it is too early to speculate on its relevance to the various categories underdiscussion.

7. It is frequently observed that discourse in thc Deaf community tends to be more direct than isusual for the hearing population. Certainly, the frequent use of strong contrastive indicators,such as the signs WRONG(-YOU) 'you're wrong' and NAW' 'no, that's not right', to overtlyreject previous speaker's propositions or presuppositions is a reflection of this directness.

References

Aarons, Debra. 1994. Aspects of the Syntax of ASL. Boston University. [Dissertation]Aarons, Debra, Ben Bahan, Judy Kegl and Carol Neidle. 1992. "Clausal structure and a

tier for grammatical marking in ASL". Nordic Journal of Linguistics 15: 103-142.Ball, Catherine. 1991. The Historical Development of the It-Cleft. Ph.D. dissertation Uni­

versity of Pemlsylvania.Birner, Betty. 1992. The Discourse Function of Inversion in English. Ph.D. dissertation

Northwestern University.

102 RONNIE WILBUR WORD ORDER VARIATION IN ASL 103

Birner, Betty. 1994. "Information status and word order: An analysis of English inver­sion". Language 70: 233-259.

Birner, Betty and Gregory Ward. 1993. "There-sentences and inversion as distinct con­structions: A functional account". Berkeley Linguistics Society 19: 27-39.

Bolinger, Dwight. 1985. "Two views of accent". Journal of Linguistics 21: 79-123.Bos, Heleen F. 1995. "Pronoun copy in sign language of The Netherlands". In H. Bos and

T. Schenner (eds.), Sign Language Research 1994: Proceedings of the FourthEuropean Congress on Sign Language Research. Hamburg: SIGNUM, pp. 121­147.

Chafe, Wallace. 1976. "Givenness, contrastiveness, definiteness, subjects, topics, andpoint of view". In C. Li (ed.), Subject and Topic. New York: Academic Press, 25­55.

Collins, Peter C. 1991 a. Cleft and Pseudo-Cleft Constructions in English. New York:Routledge.1991 b "Pseudocleft and cleft constmctions: A thematic and informational interpre­tation". Linguistics 29 481-520.

Coopmans, Peter. 1989. "Where stylistic and syntactic processes meet: Locative inversionin English". Language 65: 728-751.

Creider, Chet A. 1979. "On the explanation oftransfornlations". In T. Givan (ed.), Syntaxand Semantics, Vol. 12: Discourse of Syntax. New York: Academic Press, 3-21.

Cruttenden, Alan. 1986. Intonation. Cambridge, UK: Cambridge University Press.Delin, Judy. 1992. "Properties of it-cleft presupposition". Journal of Semantics 9: 289­

306.

E. Kiss, Katalin (ed.). 1995. Discourse Configurational Languages. New York: OxfordUniversity Press.

Ferro, Lisa. 1993. "On 'self as a focus marker". In M. Bernstein (ed.), ESCOL '92: Pro­

ceedings of the Ninth Eastern States Conference on Linguistics. Ithaca, NY:Cornell University, 68-79.

Fischer, Susan. 1975. "Influences on word order change in ASL". In C. Li (ed.), WordOrder and Word Order Change. Austin: University of Texas Press, 1-25.1990. "The head parameter in ASL". In W. Edmondson and F. Karlsson (eds.), SLR'87. Fourth International Symposium on Sign Language Research. Hamburg:SIGNUM, 75-85.

Gussenhoven, Carlos. 1983. "Focus, mode and the nucleus". Journal of Linguistics 19:377-417.

1985. "Two views of accent: A reply". Journal of Linguistics 21: 125-138.Heggie, Lorie. 1988. The Syntax ofCopular Structures. Ph.D. dissertation Southern Cali­

fornia.

Heycock, Caroline. 1991. Layers of Predication: The Non-lexical Syntax of Clauses.Ph.D. dissertation University ofPeIUlsylvania.

Heycock, Caroline. 1992. "Extraction and the syntax of copular constructions". Paperpresented at the annual meeting of the Linguistic Society of America, Philadelphia,PA.

Jimenez, Maria-Luisa. 1995. "Semantic and pragmatic conditions on the position of thesubject in Spanish". In L. Gabriele, D. Hardison and R. Westmoreland (eds.), FLSM

VI: Proceedings of the Sixth Annual Meeting of the Formal Linguistics Society of

Mid-America, Vol. 2. Bloomington, IN: Indiana University Linguistics Club Publi­cations, 25-34.

Kanerva, Jonni M. 1990. Focus and Phrasing in Chichewa Phonology. New York: Gar­land.

Kegl, Judy. 1976. "Relational grammar and American Sign Language". Manuscript, MIT.King, Tracy Holloway. 1993. Configuring Topic and Focus in Russian. Ph.D. dissertation

Stanford.

Ladd, D. Robert, Jr. 1980. The Structure of Intonational Meaning: Evidence from Eng­lish. Bloomington, IN: Indiana University.

Lambrecht, Knud. 1994. Information Structure and Sentence Form: Topic, Focus, andthe Mental Representations of Discourse Referents. Cambridge, UK: CambridgeUniversity Press.

Liddell, Scott. 1978. "Nonmanual signals and relative clauses in American Sign Lan­guage". In P. Siple (ed.), Understanding Language through Sign Language Re­search. New York: Academic Press, 59-90.

Ocampo, Francisco. 1995. "The word order of two-constituent constmctions in spokenSpanish" In P. Downing and M. Noonan (eds.), Word Order in Discourse [Typolo­gical Studies in Language, Vol. 30] Amsterdam/Philadelphia: Benjamins, 425-447.

Petronio, Karen 1991. "A focus position in ASL". In 1. Bobalijik and T. Bures (eds.),

Papers from the Third Student Conference in Linguistics. Cambridge, MA: MITWorking Papers, 211-225.1992. "WHY CAN MUST appear in final position". Paper presented at the Fifth

International Symposium of Sign Language Research, Salamanca, Spain.1993. Clause Structure in American Sign Language. Ph.D. dissertation Universityof Washington.

Prince, Ellen. 1978. "A comparison ofwh-clefts and it-clefts in discourse". Language 54:883-906.

1986. "On the syntactic marking of presupposed open propositions". In A. M. Far­ley, P T. Farley and K.-E. McCullough (eds), Chicago Linguistic Society: Papersfrom the Parasession on Pragmatics and Grammatical Theory, Vol. 22. Chicago,IL: Department of Linguistics, University of Chicago, 208-222.

Romano, Christine. 1991. "Mixed headedness in American Sign Language: Evidence fromfunctional categories". In J. Bobalijik and T. Bures (eds.), Papers from the ThirdStudent Conference in Linguistics. Cambridge, MA: MIT Working Papers.

Vallduvi, Enric. 1991. "The role of plasticity in the association of focus and prominence".In Y. No and M. Libucha (eds), ESCOL '90: Proceedings of the Seventh Eastern

States Conference on Linguistics. Columbus, OH: Ohio State University Press,295-306.

1992. The Informational Component. New York: Garland.Ward, Gregory. 1988. The Semantics and Pragmatics of Preposing. New York: Garland.Ward, Gregory and Betty Birner. 1994. "Constituents out in left field: The functions of

fronting in English". Paper presented at the annual meeting of the Linguistics Soci­ety of America, Boston, MA.

Wilbur, Ronnie. 1991. "Intonation and focus in American Sign Language". In Y. No andM. Libucha (eds.), ESCOL '90: Proceedings of the Seventh Eastern States Confer­

ence on Linguistics. Columbus, OH: Ohio State University Press, 320-331.

104 RONNIE WILBUR

Wilbur, Ronnie. 1994a. "Foregrounding stnrctures in ASL". Journal 0/ Pragmatics 22:647-672

1994b. "The pragmatics of focus stnrctures in ASL". Paper presented at theGeorgetown University Conference on Pragmatics of ASL.1994c. "Stress, focus, and extrametricality in ASL". Paper presented at the annualmeeting of the Linguistic Society of America, Boston, MA.1995a. "What the morphology of operators looks like: A formal analysis of ASLbrow-raise". In L. Gabriele, D. Hardison and R. Westmoreland (eds.), FLSM VI:

Proceedings of the Sixth Annual Meeting o/the Formal Linguistics Society o/Mid­America, Vol. 2. Bloomington, IN: Indiana University Linguistics Club Publica­tions, 67-78

1995b. "Why so-called 'rhetorical questions' (RHQs) are neither rhetorical norquestions". In H. Bos and T. Schermer (eds.), Sign Language Research 1994: Pro­ceedings of the Fourth European Congress on Sign Language Research Hamburg:SIGNUM, 149-169.1996. "Evidence for the function and stnrcture of wh-clefts in ASL". International

Review o/Sign Linguistics 1: 209-256.Wilbur, Rormie and Cynthia Patschke. 1996 "Body leans and marking contrast in ASL".

Journal of Pragmatics. In review.

Part II:

Grammatical Morphemes versus Lexical Units