Wh-questions in the Trinidad and Tobago Signing Community

(1)

wh-questions in the Trinidad and Tobago Signing

Community

Felicia Bisnath 11310588

First reader: dhr. dr. R. Pfau

Second reader: dhr. prof. dr. E. O. Aboh Date: 24th July 2018

(2)

Acknowledgements

This thesis would not have been possible without the time and insights of my informants in Trinidad — Ian Dhanoolal, Joel Jaggernauth and Bryan Rodrigues I would also like to thank Ben Braithwaite of the University of the West Indies, St. Augustine for conducting the data collection and interpreting for me, my supervisor Roland Pfau for supporting my interest in studying the signing varieties in Trinidad and Tobago, and my second reader, Enoch Aboh for his comments.  

(3)

List of Abbreviations

ASL - American Sign Language BSL - British Sign Language CODA - Child of Deaf Adults CSD - Cascade School for the Deaf cu - chin up

FC - face

GI - General Interrogative ISL - Israeli Sign Language InSL - Indian Sign Language LIS - Italian Sign Language LH - left hand

LSB - Brazilian Sign Language MO - mouthing

NM(M) - Non-manual Marker NZSL - New Zealand Sign Language RH - right hand

SEE - Signing Exact English T&T - Trinidad and Tobago TID - Turkish Sign Language tp - topic

TTSL - Trinidad and Tobago Sign Language YSL - Yolngu Sign Language

(4)

Abstract

Wh-questions are important typologically since they are expected in all languages. Their syntax is also important for understanding the effects of modality on language (Cecchetto et al. 2009). In this study wh-questions are examined in a previously unstudied context, that of Trinidad and Tobago (T&T). In T&T two varieties that emerged out of changes in deaf education (Oralism to Total Communication) exist. One variety is indigenous, Trinidad and Tobago Sign Language (TTSL), and the other is a local dialect of American Sign Language, Trinidad and Tobago American Sign Language (TTASL). They are both routinely mixed and are therefore difficult to separate in practice. Data was collected from 3 members of the deaf community in T&T using elicitation and interviews. Two distinct

wh-paradigms exist for TTSL and TTASL, while the same form of non-manual markers (NMM) and a preference for clause-final positioning of the wh-word, is found for both varieties. The form of the TTSL wh-paradigm appears to be influenced by spoken language and co-speech gestures, which is likely since it developed under Oralism; however, its syntax deviates from spoken language patterns. The TTASL wh-paradigm is identical to that found in ASL, apart from the addition of a sign for (WHAT) TIME (also found in TTSL). Its syntax is more variable than that of TTSL, which could be due to a variety of reasons, one of which being contact with TTSL. It does not appear that wh-NMM is used to mark wh-dependency in either TTSL or TTASL.  

(5)

1. Introduction

Wh-questions or content questions are typologically-valuable constructions because they can be assumed to exist in all languages and are therefore a good construction type for cross-linguistic comparison. One commonly-studied aspect of questions is wh-movement, a syntactic transformation existing within Generative Grammar. Wh-movement refers to the displacement of a wh-word from its original position (in its underlying declarative), to a different position in the sentence that is produced, which marks a dependency between the old and new positions. An example of leftward displacement in English is shown below in (1). t marks the original position of the wh-word in the declarative.

(1) a. “John bought the cat” [declarative]

b. “What did John buy tWHAT?” [movement to the left] In the vast majority of spoken languages, wh-movement is overwhelmingly to the left while rightward movement is almost unattested. This has led Cecchetto, Geraci & Zucchi (2009) to identify wh-questions as a possible point of modality-based typological variation between spoken and sign languages.

In order to investigate this possibility, wh-questions must be studied in as many sign languages as possible. Zeshan (2006) has begun this, by undertaking a cross-linguistic comparison of wh-questions in 37 sign languages; however, the majority of these sign languages are European with just five, three and eleven coming from the Americas, Africa and Asia respectively (p. 33). Caribbean sign languages are not represented, and in general, descriptions of them are rare. Cumberbatch (2012) is the only description of a Caribbean signing community— that found in urban Jamaica — and does not contain a description of interrogatives at all. Parks & Parks (2012) provides sociolinguistic sketches of the deaf communities in Trinidad and Tobago (T&T), St. Vincent, and Grenada, and Parks & Williams (2011) provides sociolinguistic sketches of 24 deaf communities in the Americas; but again, there is no description of wh-questions. This motivates the present study’s focus on a description of wh-questions in the signing community in T&T.

The signing community in T&T is characterised by multi-modal language contact and multilingualism arising from the history of deaf education in the country. This type of situation is not unusual across the world; for example, the case of New Zealand Sign Language (NZSL) almost exactly mirrors that of TTSL (McKee 2006). There are two main varieties that emerged out of changes in education in T&T — Trinidad and Tobago Sign Language (TTSL) is the older variety influenced by the American Sign Language (ASL) manual alphabet, and Trinidadian English/Creole while Trinidad and Tobago American Sign Language (TTASL) is the younger variety influenced by Trinidadian English/Creole, ASL and Signed Exact English (SEE). TTSL and TTASL are routinely mixed to different degrees, so it is difficult to disentangle them from each other in the

(7)

minds of some speakers, and also in practice; therefore, a high level of variation is expected within the community. Wh-questions are a good sentence type to observe this variation and multi-modal language contact because they are associated with a distinct lexicon, prosody (non-manual marking), and syntax.

Studying wh-questions in the T&T context has three scientific benefits. First, it adds to typological studies of wh-movement in languages in general, and in sign languages specifically. Second, it adds a description of wh-questions in a Caribbean sign language, which does not yet exist. Third, it helps to characterise multi-modal language contact, which is often found in deaf communities. Studying wh-questions also has social value since the T&T deaf community is interested in preserving TTSL, and contributes to language maintenance since a description has pedagogical value. The primary aim of this study is therefore to provide an overview of the form of wh-questions found in T&T that will (i) act as a foundation for a more detailed study of their syntax and (ii) situate T&T signing varieties typologically. To do this, three descriptive components are required: the

wh-paradigm, non-manual marking (NMM) and its scope, and positions where wh-words are and are not allowed. Analysing these components together will contribute to determining the direction of wh-movement.

The organisation of this thesis is as follows: chapter 2 describes the signing situation in T&T, chapter 3 summarises previous literature on wh-questions in general and in sign languages specifically, chapter 4 describes the research design, chapter 5 presents a description of wh-questions in T&T signing varieties, chapter 6 discusses this description in relation to other sign languages, and chapter 7 concludes by identifying strengths and weaknesses and possibilities for future research.

(8)

2. Sign Languages in Trinidad and Tobago

The deaf community in T&T is quite small, with just 2000 people reported in 2008 (Simons & Fennig 2018) out of a total population of 1.3 million. It is not known how many deaf people there are currently, but this number is not expected to be much different now since there have not been any major events in the last ten years causing deafness. The sign language situation in T&T is characterised by multilingualism, language contact, and variation. Two main varieties are said to exist — an older indigenous variety which I refer to as TTSL following Braithwaite (2018), and a local variety of ASL which, again following Braithwaite (2018), I refer to as TTASL. These two varieties exist on a continuum due to the conditions of their emergence and use. I first detail the history of their emergence, which entails a history of deaf education. I then present what is know about the grammars of TTSL and TTASL, and the distinctions between them.

2.1 History of TTSL and TTASL

As is the case with many sign languages, the development of TTSL and TTASL is tied to the history of deaf education in the country (Braithwaite et al. 2011, Braithwaite 2018). TTSL emerged at the first school for the deaf opened in Trinidad, the Cascade School for the Deaf (CSD). This school was first opened in 1943 in the north of Trinidad by a hearing signer of British Sign Language (BSL). It was then moved to a larger location, also in the north of Trinidad, in 1945, and accommodated both day pupils and boarders. This was the first time that deaf people from across T&T were able to meet daily. When the CSD was established, the educational philosophy in use was Oralism, which prohibited the use of manual means of communication (therefore, BSL was never formally taught) and instead taught students to speak and lipread. However, Braithwaite (2018) reports anecdotal evidence that the ASL manual alphabet was used during the Oralist period in the 1960s to support the teaching of English reading, writing and speaking. Braithwaite (2018) also notes that a manual alphabet brought by a teacher from the Manitoba School for the Deaf was probably also in use before 1975. It was during the period between the second opening of the CSD and the introduction of Total Communication (1947 to 1975) that TTSL emerged. It would have developed with considerable spoken language input because of the influence of Oralism, and possibly with some input from a North American manual alphabet. This situation described for TTSL is not unusual. Nicaraguan Sign Language (NSL) emerged when a deaf community was formed due to the establishment of several deaf schools, and it, along with New Zealand Sign Language (NZSL) developed under the influence of Oralism (Senghas & Coppola 2001; McKee 2006).

In 1975 a switch to Total Communication began when a deaf signer was invited to give ASL classes at the CSD. Total Communication does not emphasise the use of spoken language strategies, but instead teaches deaf students to communicate with all the means they have at their disposal, including their hands and bodies. After this, ASL and SEE

(9)

were introduced, and are still used at the CSD and at other schools where deaf people are educated. SEE is an artificial sign system and therefore not a natural sign language like ASL. Instead, it attempts to represent all aspects of English using sign and therefore has components, like articles, that are not found in natural sign languages. The two deaf-only schools in T&T are the CSD and the Audrey Jeffers School for the Deaf (AJSD) (opened in 1971). Deaf students have also been taught alongside hearing students in some mainstream primary and secondary schools since 1980, and there is also the Point-a-Pierre Special Education School in Trinidad and vocational schools that accommodate deaf students (Parks & Parks 2012: p. 6). Apart from formal education, it is possible to learn sign language from various religious organisations operating in T&T (Parks & Parks 2012: p. 7-8). People who learn to sign at the Jehovah’s Witnesses Organisation learn ASL and those with the Touch of Christ Catholic ministry and the Open Bible Deaf Church learn a combination of TTSL and ASL. Those who learn to sign just by interacting with the deaf community learn a mix of TTSL and ASL, the proportions of which depend on social factors.

Because of these changes in deaf education, and the presence of ASL in formal education and other social settings, it is expected that TTSL would best be known by people who went to the CSD between 1947 and 1975, or those who are at least 50 years old now. However, Parks & Parks (2012) report that TTSL is being used by younger generations now because of increased interest in its preservation (p. 9).

2.2 Teasing Apart TTSL and TTASL

From the previous discussion, we know that TTSL was in use before the introduction of ASL and that it emerged under contact with spoken and written English and/or Trinidadian English/Creole, a North American sign language, and possibly BSL. From anecdotes reported in Braithwaite (2018), it appears that TTSL is distinct from ASL, at least to some speakers. For example, one signer says:

“At my school, Cascade School, we didn’t use American Sign Language, we only used Trinidadian Sign Language. It was completely different. For example, there were difference

[sic] signs for ‘mother’, ‘father’, ‘woman’, ‘bus’ and ‘home’.” (Braithwaite 2018: p. 6).

However, for speakers who were educated after the introduction of ASL, this distinction between TTSL and TTASL is not the same. They are likely to call what they sign TTSL, but to signers educated before the introduction of ASL, they sign TTASL. This is because they were only exposed to one variety and because of the combined use of TTSL and TTASL within the community.

Regarding grammar, little is known about TTSL and TTASL or their differences. From the quote above we can assume that there are lexical differences, though the extent of these differences is not known. Regarding TTASL specifically, Parks & Parks (2012) describe it as something that “looks like Signed English with heavy initialization of signs” (p. 9),

(10)

and Braithwaite (2018) describes it as a local dialect of ASL. As for TTSL phonology, mouthings are found, which are distinct from ASL mouthings. For example, where ASL would use the mouthing ‘“cha”, meaning “large”, TTSL uses “pow” (Braithwaite 2018). Initialisation is a common word-formation strategy, as well as fingerspelling, which are also found in ASL. As for syntax, the basic word order is SVO, though SOV is also possible (Kwok 2015). ASL is also SVO, but this does not indicate that TTSL may have inherited its word order from ASL since many spoken and signed languages have SVO order.

The main point to be taken from this account is that though TTSL and TTASL may be distinct languages for some speakers, separating them is likely to be difficult in practice. This is because of the extent of contact it has had, and continues to have, with ASL — the ASL manual alphabet was in use when TTSL was emerging, and both TTSL and TTASL are used together in contexts where deaf people socialise, like church. Informants would therefore need to have high metalinguistic knowledge to be able to inform on TTSL and TTASL separately.  

(11)

3. Wh-questions in Sign Languages

Wh-questions, or content questions, are those that ask for a specific kind of information. They contrast with polar questions which only require a yes or no answer. They are typologically valuable since all languages can be assumed to have ways of eliciting

“information pertaining to first-order individuals, locations, times, manners, quantities, and reasons” (Mackenzie 2009: p. 1133). When describing wh-questions in sign languages, there are three main elements to consider: the wh-paradigm (the inventory of wh-words in a language), the syntax of wh-words, and non-manual markers (NMMs). NMMs are involved in both the lexicon and syntax of wh-questions, so in addition to a brief discussion of their form, they are integrated into discussions of the other components.

3.1 Non-manual markers

Non-manual markers (NMMs), as the name implies, do not employ the hands but are articulations of parts of the face that accompany manual signing to mark sentence and clause type, information structure, and prosody. In wh-questions, they are an instance of typological variation between signed and spoken languages since in the latter, articulations of the face are not grammaticalised. Zeshan (2006), in a study of 37 sign languages, found that the most common NMM associated with content questions in Western sign languages is the brow furrow (p. 41). However other patterns exist: in Indian Sign Language (InSL), wh-questions are marked by brow raise + backward head tilt; in Turkish Sign Language (TID) they are marked by a headshake (Zeshan 2006: p. 41) and in Hong Kong Sign Language (HKSL) they are commonly marked with brow raise + eye gaze (Tang 2006: p. 212-213). wh-NMMs are always found on the wh-word and can be considered part of its citation form. They also tend to spread from the wh-word to other parts of the clause and in some cases, as in Italian Sign Language (LIS), this marks a syntactic dependency (see 3.3 for more on this). In some cases, wh-NMM can occur without a manual sign, as in Nihon Shuwa (Japan) below:

(2) ____wh’ COLOUR LIKE

“What colour do you like?”

(Fischer & Osgui 1998, in Zeshan (2006: p. 41)) Mouthings can also be considered part of NMM, though they are not often discussed in this context. In this paper they are defined as “mouth movements resembling a spoken English

mouth movement, associated with a gloss or a translation of the sign” (Nadolske & Rosenstock 2007: p. 40). In the case of manual signs for wh-words, they are usually accompanied by the spoken language wh-word. In some languages, like Yolngu Sign Language (YSL) found in Australia (Bauer 2014: p. 110), and Israeli Sign Language (ISL) (Meir 2004: p. 101), mouthings are used to disambiguate wh-words.

(12)

3.2 Wh-paradigms

Wh-questions can be defined as those that question using words corresponding to English

who, what, when, where, why, which, how, how many/much, which is called the wh-paradigm in a language. In many sign languages there is a general interrogative, with a basic meaning corresponding to what, that can represent the wh-paradigm to different extents:

(i) Type A (minimal): the general interrogative accounts for the entire paradigm e.g. Indian Sign Language (InSL)

(ii) Type B (intermediate): the general interrogative accounts for part of the paradigm e.g. Brazilian Sign Language (LSB), Nihon Shuwa (Japan)

(iii)Type C (large): the general interrogative exists alongside a complex wh-paradigm e.g. American Sign Language (ASL)

In a type A language like InSL, there is only the general interrogative which can be used for the entire wh-paradigm (in such cases the exact wh-word is recoverable from context). In addition to this, the general interrogative can combine with non-interrogatives to form more specific wh-words. For example, in figure 1, when is formed by combing the sign for TIME with the general interrogative, G-WH. Other combinations are FACE G-WH (who), PLACE G-WH (where) and NUMBER G-WH (how many).

In a type B language, there are idiosyncratic signs for some wh-words (e.g. how, why, how

many in LSB) and the remainder are covered by the general interrogative. In a type C language like ASL, there are idiosyncratic signs for all wh-words, and a general interrogative.

wh-questions may also include question particles such as PALM-UP, as in Finnish Sign Language, though this is rare (Zeshan 2013) (see figure 2 for an example).

TIME G-WH “when”

Figure 1: Compositional sign in Indian Sign Language (Aboh, Pfau & Zeshan 2005: p. 16)

(13)

Question particles mark an utterance as a question, are not obligatory in sign languages, and are found in the same prosodic unit as the question (Zeshan 2004: p. 32). They are important to consider when describing wh-paradigms since they could have a similar form to other signs such as WHAT and/or general interrogatives, making it necessary to disentangle these signs from each other — for example, if a language had a sign for WHAT like ASL (shown in figure 3), a general interrogative like that in InSL (as in figure 1) and also used PALM-UP (as shown in panel 3 of figure 2).

3.3 Syntax of Wh-questions

Wh-questions are valuable in the Generative tradition because they provide support for the existence of transformational operations in syntax (given the assumption that the initial form of all sentence types is a declarative). One such operation is overt movement, which

“connect[s] two positions, typically an argument position and some higher position to which the argument is related [(the complementiser)]” (Cecchetto et al. 2009: p. 296). In wh-questions,

wh-movement occurs. An example from English (example (1) repeated as (3)) is shown below.

(3) a. “John bought the cat” [declarative]

b. “What did John buy tWHAT?” [movement to the left] Given that English is an SVO language, in (3a) the element to be questioned (the object) appears in final position of the declarative i.e. its underlying position. In (3b), the question,

_______________________________________lowered brows _____head tilt PAPER WHERE PALM-UP ‘Where is the paper?’

Figure 2: Use of PALM-UP in a wh-question in Finnish Sign Language (Zeshan 2004: p. 33)

(14)

the wh-word that asks for the object (what) is not in the same position as in the declarative but is in the left periphery, suggesting that between the initial formation of the question in the mind (its underlying form) and its production (the surface form), the object was moved. The original position of the wh-word is indicated by the trace, tWHAT in (3b). Wh-movement represents a syntactic dependency between the trace and the new position of the wh-word in the left periphery, and identifies questions as questions by syntactically differentiating them from declaratives.

In some cases, wh-movement is not required to mark questions. For example, in Portuguese and Japanese the wh-word does not have to be moved but remains in the same place as it would be in the declarative, as shown in (4) from Portuguese:

(4) a. O João comprou o gato. [declarative]

“João bought the cat.”

b. O João comprou o quê? [in situ]

“What did João buy?”

In the great majority of spoken languages, the wh-word is moved to the left, not moved at all (left in situ), with only one of 1200 languages in the World Atlas of Language Structures (WALS) — Tennet spoken in south Sudan— showing a preference for the right periphery (Dryer 2013). In sign languages however, more options are available: wh-words are allowed in situ, in the right periphery, doubled in the left and right peripheries, and there is “no attested case of a sign language that would allow only clause-initial

interrogatives” (Zeshan 2004: p. 64) . Examples from ASL are shown below in (5). (5a) 1

shows the basic word order in ASL (SVO) and also the positioning of temporal adjuncts, which is necessary to distinguish rightward movement from in situ. (5b) shows in situ, with the wh-word in the same position as the object in the declarative, before the temporal adjunct. (5c) shows the wh-word in the right periphery, and crucially after the temporal adjunct. (5d) shows the wh-word in the left and right peripheries. Its position in the left periphery suggests movement, since this is not its normal position in declaratives (only subject elements, which correspond to who, are found at the beginning of declaratives in SVO languages); however, there are different analyses of the wh-word in the left periphery, which are discussed below.

These generalisations are based on a sample of 37 sign languages, of which the great majority are

1

(15)

(5) a. JOHN BUY BOOKYESTERDAY [declarative] “Yesterday John bought a book.”

b. _______________________wh [in situ]

JOHN BUY WHATYESTERDAY “What did John buy yesterday?”

c. ________________________wh [right periphery] JOHN BUY YESTERDAYWHAT

“What did John buy yesterday?”

d. _____________________________wh [left + right periphery] WHATJOHNBUYYESTERDAYWHAT

“What did John buy yesterday?”

(Neidle, Kegl, MacLaughlin, Bahan & Lee 2000) This pattern suggests that wh-questions could be a point of “macrotypological variation between

spoken and sign languages” (Cecchetto 2012: p. 296), possibly attributable to modality differences. However, because more variation in the position of the wh-word is possible in sign languages than in spoken languages, there is room for different movement analyses, even within the same language. The most influential debate of this kind occurs in ASL, with disagreement even occurring over the data itself, possibly because of dialectal differences. The only uncontroversial position is in situ (5b) since many languages allow it to exist alongside movement. I discuss the differences between the leftward and rightward analyses by making specific reference to (5c and 5d) above, and (6) and (7) below. After discussing ASL, I summarise the rightward analysis of LIS and its proposal that movement to the right occurs because of the availability of multiple channels for expressing syntactic dependencies in sign languages, namely wh-NMM (Cecchetto et al. 2009). Understanding the debate in ASL and Cecchetto et al.’s (2009) proposal is important because it informs method.

3.3.1 Movement Analyses in American Sign Language (ASL)

The leftward analysis (Petronio & Lillo-Martin 1997) aligns ASL with spoken languages suggesting that movement is to the left. For Petronio & Lillo-Martin (1997) it is not necessary that rightward movement, which is unattested in the vast majority of other languages, be proposed, since other explanations are available. The rightward analysis (Neidle et al. 2000) takes a more neutral starting point, and simply claims that wh-words are in the right periphery because there is movement to the right.

For sentences such as (5d) , the leftward analysis claims that the wh-word found in 2

the right periphery does not arise out of wh-movement but is used for focus or emphasis. They support this by referring to modals, quantifiers and verbs, for which the same pattern is observed. The rightward analysis instead claims that the wh-word in the left

These questions should not have a significant pause before the wh-word.

(16)

periphery is a wh-topic. The evidence they provide is that wh-topics have the same distribution as base-generated topics and that their NMM is a mixture of wh-NMM and topic-NMM.

Sentences like (5c) are predicted by the rightward analysis but are a problem for the leftward. Petronio & Lillo-Martin (1997) claim that such sentences are actually instances of (5d), but with a covert wh-word in the left periphery. Null arguments are allowed in ASL, given context, so Petronio & Lillo-Martin (1997) argue that their presence in wh-questions is not unusual. Support for this comes from the finding that signers who initially reject instances like (5c), presented without context, accept them when presented with context. Therefore, the leftward analysis requires a difference in grammaticality associated with presence or absence of context for questions like (5d) while a rightward analysis does not.

For seemingly straightforward sentences like (6), the leftward analysis has problems, as pointed out by Cecchetto (2012: p. 298). The leftward analysis predicts that this sentence should be grammatical, while the rightward predicts that it should not. Petronio & Lillo-Martin (1997) report varying judgements, and claim that they arise due to stylistic reasons, with signers preferring to have overt material in the position of the final complementiser. Neidle et al. (2000) report that (6) is always ungrammatical, which is in line with their predictions.

(6) ____________________wh #WHO JOHN HATE “Who does John hate?”

(Neidle et al. 2000: p. 127) Finally, sentence (7) is said to be ungrammatical by Petronio & Lillo-Martin (1997) but grammatical by Neidle et al. (2000). The ungrammaticality of (7) is predicted by the leftward analysis since complex wh-phrases cannot be in the position of heads.

(7) ______________________________________________________________wh #WHICH COMPUTER JOHN BUY WHICH COMPUTER “Which computer did John buy?”

(Petronio & Lillo-Martin 1997: p. 33) In summary, the possibility of having wh-words in different positions in ASL leads to different movement analyses. The leftward analysis aligns ASL with other languages and attributes the presence of wh-words in the right periphery to information structure, specifically focus and emphasis. It also makes analogy with other comparable elements in ASL which are found in the left periphery. The rightward analysis takes a neutral starting point and explains the presence of wh-words in the left periphery by information structure as well, namely topic. The main issue with this debate is that the basic facts are disputed,

(17)

which makes it difficult to propose a coherent movement analysis. However, there are several aspects of the debate that are instructive. First, it shows that even within establishment sign languages like ASL, there is grammatical variation. This suggests that this variation is likely to exist in T&T varieties since they have been in contact with ASL for most of their lives. Second, the debate shows that information structure and style affect the position of wh-words and should therefore be accounted for when designing elicitation stimuli.

3.3.2 Non-manual Marking in Italian Sign Language (LIS)

Cecchetto et al. (2009) propose that wh-movement in LIS is to the right and goes one step further in suggesting why this may be due to modality. Their major claim is that wh-NMM is a means of marking a wh-dependency, like wh-movement, and that wh-NMM spreading forces rightward alignment in LIS. I first briefly summarise some basic properties of LIS and the pattern of wh-NMM spreading, and then Cecchetto et al.’s (2009) proposal.

LIS is a head-final language with SOV being the unmarked word order. Wh-words may appear in situ or in the right periphery, and they are banned from the left periphery for the signers that worked with Cecchetto et al. (2009) . Regarding wh-NMM, the most 3

common is furrowing of the eyebrows. In all cases there must be wh-NMM on wh-words, and when the phrase is in the right periphery NMM tends to be restricted to the wh-word, as shown in (8a) and (8b). When the direct object is questioned, NMM may spread onto the verb (indicated by dashed lines in 8c), and when the subject is questioned it may spread over the entire clause (8d). When the wh-word is in situ (8e), NMM must spread and spreading is to the right.

Geraci et al. (2015) found that wh-words are allowed in the left periphery but that this property is

3

(18)

(8) a. ________________wh PAOLO STEAL BOOKiWHICHi

“Which book did Paolo steal?” b. _______wh

BOOK STEAL BOYiWHICHi

“Which boy stole the book?” c. _ _ _ _ _ _________________wh PAOLO STEAL BOOKiWHICHi

“Which book did Paolo steal?” d. _ _ _ _ _ _ _ _ _ _ _ _______________wh

BOOK STEAL BOYiWHICHi

“Which boy stole the book?” e. ________________________wh

PAOLO BOOKiWHICHiSTEAL

“Which book did Paolo steal?”

(Cecchetto et al. 2009: p. 285-286) Based on the optional nature of spreading when there is wh-movement and the mandatory nature of it when there is no wh-movement (in situ), Cecchetto et al. (2009) propose that

wh-NMM is a modality-specific strategy used to connect the same argument positions that

wh-movement does. They then show that leftward movement in LIS produces wrong patterns of wh-NMM. Finally, based on Richards’ (2006) proposal that “a prosodic domain

that marks the wh-dependency can be created only if the prosodic boundary is located on the opposite

side of the complementizer” (in Cecchetto et al. 2009: p. 310) they argue that the natural position for wh-words in LIS, and sign languages in general, is the right periphery. This is because sign languages have two means of marking wh-dependency (syntactic and prosodic), that exist in two different channels (non-manual and manual), while in spoken languages, these two strategies exist in the same channel (oral); therefore, it is cognitively easier for sign languages to place elements in the right periphery and to combine multiple means of marking wh-dependency than it is for spoken languages.

The most relevant part of this discussion of LIS to us is that it demonstrates that

wh-NMM can have important syntactic consequences in a sign language and therefore any description of wh-questions requires a description of wh-NMM. The case of LIS also shows that in addition to dialectal differences causing possible differences in the position of wh-words (as in ASL), sociolinguistic factors like age and education are relevant too; therefore, an ideal description of wh-questions in TTSL should account for the variation that exists. Finally, Cecchetto et al.’s (2009) final proposal that the natural position for wh-words in sign languages is the right periphery could imply that leftward movement may be due to the influence of spoken languages. Therefore, in TTSL, one could expect to see rightward movement because it is a sign language, and leftward movement because of

(19)

contact with ASL and SEE, and because Oralism was the dominant educational philosophy during the time it developed.

3.4 Wh-questions in Trinidad and Tobago Sign Language

From the previous discussion, it should be clear that a description of wh-questions requires several pieces of information, namely:

(i) the wh-paradigm

(ii) word order in declaratives (iii)wh-NMM and its spread

(iv) where wh-words can and cannot appear

Regarding word order, Kwok (2015) found that for intransitive constructions the basic surface word order is SV, while for transitive ones, SVO (also found in ASL) and SOV occurred. Other word orders were also found but they are rare. What is clear from this is that the subject precedes the verb, which makes it possible to observe rightward movement of WHO. For the purposes of studying movement of other wh-words, I assume SVO. An example is shown in (9).

(9)M-A-N WORK HOUSE BUILD BRICKS BUILD+REPT×5BUILD+REPT

[ S V O ] [ V O (V) (O) ] ‘The man works on the house, he builds bricks.’

(Kwok 2015: p. 17)

Regarding interrogatives in general, Lewis (2016) studied word order and NMM in polar and content interrogatives. This study has its limitations but is a useful starting point. Data was collected using an elicitation task that involved role playing a job interview. It was done with four signers of varying ages (42, 32 and 28 ). As we mentioned previously, 4

TTSL is associated with signers who attended the CSD between 1947 and 1975; therefore, these signers are too young to have been around when TTSL developed and their signing is unlikely to represent it. One signer did report that he learned TTSL informally while attending the CSD. The wh-paradigm was not explicitly identified, so we do not know exactly what wh-words exist and what their forms are, but evidence of a sign meaning “question” was used by signer 3 and signer 4 (a married couple). This sign was not used by the other participants, but was understood by them, which suggests that (i) its form is iconic enough to be understood and/or (ii) it is used in the wider signing community.

As for NMM, the following were found: backward/forward tilted head, raised/ furrowed eyebrows, widened/closed/squinted eyes, open/closed mouth, tongue protrusion/ side-to-side movement, spread/protruded lips, and puffed/pulled in cheeks (Lewis 2016: 25). The most common type of marking was furrowed eyebrows but there was

The age of signer 3 is not reported in Lewis (2016).

(20)

participant variation with signer 03 using the most eyebrow furrowing (15 times) and signer 4 using the least (4 times); however, signer 4 is different from the other signers in that she was not born deaf and did not attend a deaf school. The use of furrowing by signers 01 and 02 is comparable to that of signer 03, which suggests that it may be used systematically to mark questions. wh-questions followed the ASL pattern of squint + furrowed eyebrows + forward tilt of the head and body (Lewis 2016: p. 22-23); however, the scope of this NMM was not identified.

With respect to syntax, no specific position for the wh-word was identified, though the majority of question words were found clause initially. Evidence of them in clause final position and “sentence medially” was reported, though no examples were given (Lewis 2016: p. 23). The question marker mentioned earlier was always found clause finally; however, it was only used between a married couple and its position is likely to be due to the influence of spoken language. One signer also showed syntactic variation based on the

wh-word itself— what was always found clause initially, while where and how much were found clause initially, clause finally, and both clause initially and clause finally (Lewis 2016: 23). Duplication of the wh-word could have been emphatic.

Clearly there is no definitive description of wh-questions in TTSL. Previous studies suggest that TTSL may have a Type C wh-paradigm, that eyebrow furrowing will be part of wh-NMM, and that wh-words will be found clause initially and possibly in situ. What is not documented is:

(i) the form of wh-signs (ii) the spread of wh-NMM

(iii)where wh-words are and are not allowed

In order to answer these questions, the research design outlined in chapter 4 was developed.

(21)

4. Research Design

Following Van Herreweghe & Vermeerbergen (2012: p. 1033), an integrated approach to data collection is adopted. This combines different techniques and is useful since it makes it possible to collect different kinds of data, and to compensate for the weaknesses of individual techniques. I combine formal and informal elicitation and grammaticality judgements and apply them in the following order: formal elicitation task > informal elicitation > grammaticality judgements. The formal task is done first to elicit as much information as possible without explicitly asking for question forms. Informal elicitation is then done to fill any remaining gaps in the wh-paradigm. Both elicitation tasks are used to make hypotheses about the wh-paradigm, NMM, and syntax which are then tested using grammaticality judgements. All of the data was collected remotely. The elicitation was conducted in Trinidad by Ben Braithwaite, and I conducted the grammaticality judgements and interviews, with Ben Braithwaite as interpreter, over video calling services. I describe the tasks in detail after describing the participants.

4. 1 Participants

Three signers of TTSL, of various ages, participated in this study. The only selection criterion was that they be members of the signing community in Trinidad. It would have been better to have a larger sample and to include signers over the age of 55. However, this was not achievable because not many signers in T&T are trained to provide linguistic intuitions, and because of the remote nature of the data collection. It was more beneficial to work with a few representative signers who were familiar with providing linguistic intuitions, and to use informal elicitation and interviews (described in the next section). Zeshan (2006) also found this in her study of negatives in TID. She restricted the collection of a specific negative sign to “the three informants who had a basic understanding of the

aims and methods of sign language research due to their involvement as research assistants”, and also avoided less direct techniques (which would include picture elicitation) since “when working

with untrained informants…the outcomes can be unpredictable” (p. 131). Regarding the small sample size, it is not unusual in studies such as this one. For example, Zeshan’s (2006) study of negatives and interrogatives in TID used five signers (p. 130), Tang’s (2006) study of negatives and interrogatives in HKSL used two signers (p. 200) and Antzakas’ (2006) study of negative head movements in Greek Sign Language used three signers (p. 261).

The informants in this study are unusual in that none of them were part of the generation that was present during the emergence of TTSL; however, they are all active members of the Deaf community and are likely to accurately represent the variation that exists within it. Some of their characteristics can be found in table 1. Signers 01 and 03 both learned TTSL from signers who were present during its emergence, and signer 03 is particularly proficient since he has several deaf relatives and his mother and father attended the CSD during the development of TTSL. A description of all the participants

(22)

is given below in table 1. In terms of proficiency, signers 01 and 03 are more proficient in TTSL than signer 02 and are better able to distinguish between TTSL and TTASL.

Table 1: Description of participants

CODA = child of deaf adults

4.2 Elicitation

To determine the wh-paradigm and NMM, a formal elicitation task combined with informal elicitation was used. The formal task is based on an adaption of Geraci et al. (2015) and Friedmann & Szterman (2010). Elicited data cannot reflect natural use but are useful since “wh-constructions are used relatively infrequently in spontaneous conversations” (Geraci et al. 2015: p. 136). Geraci et al. (2015) studied variation in wh-signs in LIS with data collected using a picture elicitation task (see figure 6 for an example ). The task was done 5

in pairs in a role play context. One signer held the car accident scene and the other an insurance form. The signer with the insurance form had to ask the signer with the accident scene various wh-questions to complete the form. Geraci et al. (2015) do not say exactly which wh-words were elicited, and do not comment on how effective the individual prompts were, but from figure 4, it seems likely that the method could not elicit the entire

wh-paradigm — for example, it is not clear to me if any of the stimuli would have easily elicited why. This issue may have been negligible to Geraci et al. since they collected data from 165 signers, but in the case of a significantly smaller study such as the present one, it was not profitable to directly replicate this method.

Signer Age Sex Status Job Education Represents Family Where they

learned sign language

1 43 M Deaf Teacher CSD TTSL, ASL Hearing CSD from

maintenance staff

2 23 M CODA Interpreter Mainstream

hearing school

TTASL Deaf

parents

From parents

3 38 M Deaf Salesperson CSD TTSL Deaf

parents and family members

From parents and family

Geraci et al. (2015) also used a second context involving a domestic accident scene and domestic

5

(23)

Friedmann & Szterman (2010) used a similar task to elicit subject and object who-questions from Deaf/Hard-of-Hearing children. An example of their stimuli is shown in figure 5. In this task, the entity being questioned is hidden. The child is shown the picture by the researcher and is required to ask a question to find out who was being concealed (Friedmann & Szterman 2010: p. 219).

Figure 4: Insurance form task Geraci et al. (2015: p. 137)

Figure 5: Pictures used to elicit a subject who-question (Friedmann & Szterman 2010: p. 220)

(24)

In the task used in the present study, I combine and adapt Geraci et al. (2015) and Friedmann & Sztermann (2010)’s approaches. The task is done with pairs of signers (signer 1 and signer 2). A sample of the stimulus to elicit a which-question is shown in figure 6 (see Appendix I for all the stimuli). Signer 1 holds the card with all the information (shown in figure 6a) while signer 2 holds the card with some information missing (shown in figure 6b). Signer 2 is instructed to find out from signer 1 what is missing from his card. Signer 1 is instructed to reply with a full sentence, and both signers are instructed to not communicate with each other otherwise during the task. The signers were positioned in such a way that they were unable to see each other’s cards, but were still in view of each other’s faces. This second point is important to ensure that non-manual markers are used normally. In figure 6, the target question is a which-question, specifically “Which dress did the woman buy?”.

a. which-question stimulus (Signer 1)

b. which-question stimulus (Signer 2)

(25)

Each signer is presented with stimuli designed to elicit the full wh-paradigm (who, what,

when, where, why, which, how (instrument), how many/how much) twice; therefore, each signer receives 16 items in total.

After the formal task, the signers were asked to produce questions with wh-words that did not emerge in an informal elicitation session. In this session I took on the role of someone learning TTSL and asked the signers to give me at least two examples of questions with the missing wh-words. This was necessary because there are different strategies of questioning elements that do not necessarily involve idiosyncratic wh-words. For example, signer 01 produced (10) when presented with a stimulus for how (instrument).

(10) NM: ___________________________________________________________________________________wh1

LH: IX3 WHAT-TTASL1 KNIFE WHAT-TTASL1

RH: FISH CL:FLAT.CUT CL:FLAT3WHAT-TTASL1 KNIFECUTWHAT-TTASL1

“There is a fish that was cut with something. What? Cut with a knife? What?”

(Signer 01)

Instead of using the lexical item for how, signer 01 uses a flat classifier handshape to represent the instrument, points to it and uses the sign for what to ask for the instrument used to cut the fish.

At the end of elicitation, 47 sentences from three signers were collected. From this data it was possible to identify the wh-words used in TTSL, wh-NMM and its scope, and possible positions of the wh-word.

4.3 Interview and (Comparative) Grammaticality Judgements

With the elicitation data and the observations made by Lewis (2016), the following research questions were formulated to be answered with grammaticality judgements and in a follow-up interview.

(i) Can the signs for FACE/PERSON, TIME, PLACE, NUMBER combine with WHAT/ GI to form WHO, WHEN, WHERE and HOW MUCH respectively?

(ii) Can wh-questions only be marked with NMMs?

(iii) Are questions with WHO in clause-final position only grammatical? (iv) What is the preferred position for wh-words?

(v) Do different wh-words have a preference for a particular position? (vi) How do wh-words interact with temporal adjuncts?

(vii) If there is movement, is spreading of wh-NMM mandatory?

Regarding the syntactic grammaticality judgements, they were done in the following way. Sentences were constructed with wh-words in clause-initial, clause-final, in situ, and

(26)

duplicated clause-initially and clause-finally. They were first presented individually, and then the signers were asked to rank them based on preference. This was done because of the variation that exists in the community — because of the ubiquitous mixing of TTSL and TTASL, and the influence of spoken language, I expected that signers would be able to accept wh-words in any position. If all positions were judged as grammatical, this would say nothing about what position is most natural; therefore, I also decided to include ranking. 

(27)

5. A Description of Wh-questions in the Trinidad and Tobago Signing

Community

In this section, a description of wh-words, wh-NMM and the syntax of wh-questions is provided . Some variation is present, most notably in wh-words, that is likely due to the 6

influence of ASL. Both TTSL and TTASL signers appear to use the same wh-NMM, and also generally agree on where wh-words are preferred in questions, though there is some variation.

Because of the history of contact with ASL, TTASL and TTSL are discussed in relation to ASL. This is done to avoid a “monolithic approach” that decontextualises a language from its use (Labov 1998: p. 100). Regarding phonology, similarity is characterised based on handshape, location relative to the body, movement in space (Stokoe et al. 1965), and orientation (Battison 1974). These assessments are made on my own, as a hearing non-signer, but I note that such judgements differ based on hearing/deaf status, and language experience (Hildebrandt & Corina 2002). Lapiak (2018) is the reference for ASL signs. The data collected from the elicitation task can be found in Appendix I. Additional data, collected during interviews, is marked with ^. So, additional data collected from signer 01 is represented as (Signer 01^).

5.1 Wh-paradigm

Two wh-paradigms are observed within the signing community in T&T — one that does not resemble ASL (produced by signer 03), that I refer to as TTSL, and one that closely resembles ASL (produced by signers 01 and 02), that I refer to as TTASL.

5.1.1 TTASL Wh-paradigm

This paradigm closely resembles that found in ASL (see figure 7). It contains idiosyncratic signs for each wh-word, and an additional wh-word, (WHAT) TIME. Variations of individual signs are also found. Each sign is accompanied by the relevant Trinidadian English/Creole mouthing and wh-NMM.

See Appendix I for all sentences collected in the elicitation task.

(28)

WHO

Two signs for WHO were found, which are shown in figure 7a and b. WHO-TTASL1, produced by signer 01, is identical to an old variety of WHO in ASL.

WHAT

Two signs were found for WHAT, shown in figure 7. Signer 01 produced a sign that is phonologically identical to that in ASL, WHAT-TTASL1 (figure 7c). Signer 02 produced a sign that is also identical to ASL, but an older variety, WHAT-TTASL2 (figure 7d).

WHEN

One sign for WHEN is found, WHEN-TTASL (figure 7e). In this sign, the index finger of the non-dominant hand is oriented upwards with the index finger of the dominant hand

c. WHAT-TTASL1

a. WHO-TTASL1

d. WHAT-TTASL2 e. WHEN-TTASL

b. WHO-TTASL2

g. WHY-TTASL

h. WHICH-TTASL i. HOW-TTASL j. HOW.MUCH-TTASL

Figure 7: TTASL wh-paradigm

f. WHERE-TTASL

(29)

tracing a circle counterclockwise on the horizontal plane above the non-dominant hand, and ending with both index fingers touching. This sign resembles the ASL sign for WHEN in handshape, location and movement, but is different in orientation — the index fingers of the ASL sign are oriented toward the sagittal plane.

WHERE

One sign for WHERE (figure 7f) is found and is phonologically identical to that found in ASL. This sign was also identified as belonging to ASL by signer 03.

WHY

One sign for WHY is found (figure 7g) and is similar to the ASL sign in that it involves the Y-handshape. However, the ASL sign only uses the Y-handshape and is articulated at the side of the cheek, while the TTASL variant has its beginning at the forehead with a flat handshape oriented toward the body and ends in neutral space with the Y-handshape. There is also forward movement of the body.

WHICH

One sign for WHICH (figure 7h) is found which is phonologically identical to the ASL sign.

HOW

One sign for HOW (figure 7i) is found and has the same phonological parameters as a variation of ASL HOW. The sign produced by signer 01 shown in figure 13a is slightly different from the variant identified by Lapiak (2018) in that the final position of the hands is not close to each other at the middle of the chest, but apart; however, signer 02 produced the same variant shown by Lapiak (2018).

HOW MANY/MUCH

One sign for HOW MANY/MUCH (figure 7j) is found, which, like WHICH, is phonologically identical to the ASL sign.

(WHAT) TIME

This sign (figure 7k) is a combination of the manual sign for TIME, which is identical to the ASL sign, with the mouthing “what time”. A similar sign exists in ASL, but is only accompanied by NMM (furrowed eyebrows + squint+ chin up).

5.1.2 TTSL Wh-paradigm

In this paradigm (see figure 8), there are two idiosyncratic signs for WHO and (WHAT) TIME, and a general interrogative that covers the remainder of the paradigm. All the wh-words are accompanied by Trinidadian English/Creole mouthings, and in the case of those

wh-words covered by the general interrogative, these mouthings are used for disambiguation.

(30)

WHO

The sign for WHO (see figure 8a) is a flat palm oriented down at shoulder height with up and down movement. It resembles WHO-TTASL2 in handshape, orientation and movement; however, the TTASL variant has an additional initial point to the mouth.

(WHAT) TIME

This sign is the same as that described in the TTASL wh-paradigm above. General Interrogative (GI)

The general interrogative (GI) can be articulated with either one or two hands. It covers WHAT, HOW, HOW MUCH, WHERE, WHICH, WHY, and WHEN, which are differentiated from each other mainly by Trinidadian English/Creole mouthings of the corresponding wh-words. WHY can also be produced by mouthing “for what” while articulating the general interrogative. This general interrogative is also used by signers 01 and 02, and in some cases signers 01 and 02 use it along with mouthings of wh-words instead of using the idiosyncratic signs shown in figure 7.

In some cases, NMMs and bodily movements are also used to identify the wh-word. In the case of WHERE, the sign is articulated with head movements from side-to-side, as if something was being searched for; these head movements were identified as crucial to the form by signer 03. In WHY, there is forward movement of the entire body, which is also seen in the production of signers 01 and 02.

In a paradigm that makes extensive use of a GI one would expect that combinations such as GI + TIME or GI + PLACE could represent “what time?” and “where” respectively, as in InSL, but this is not the case. What is preferred is to combine a manual sign with the mouthing “what”, as shown in (11); however, this appears to be a strategy restricted to TTSL since it is only possible with TTSL lexical items, and not TTASL items (compare (11) and (12)).

a. WHO-TTSL b. GI

Figure 8: TTSL wh-paradigm

(31)

(11) ________what BOY-TTSL “Which boy?” (12) *_______what BOY-ASL (Signer 01^) 5.2 Non-manual Marking

Two main patterns of wh-NMM are found — wh1 and wh2 — that are consistent across TTASL and TTSL. wh1 is the most commonly used marker across signers. wh1 is a combination of furrowed eyebrows and squint (shown in figure 9a), and often occurs with the chin up (shown in figure 9b), though it is not mandatory. wh2 is a combination of raised eyebrows and widened eyes , as shown in figure 9c. 7

wh-NMM can occur independently of a manual sign, as shown in (13) and (14), given enough context.

(13) ___wh1 OLD

“How old (are you?)” (14) _____wh1

NAME

“What is (your) name?”

(Signer 03^) However, this strategy for asking questions is not productive, as shown in (15). This suggests that examples (13) and (14) are probably conventionalised.

wh2 also appears to be the NMM for topics.

7

a. wh1 b. wh1 + chin up

Figure 9: wh-NMM in TTSL and TTASL

(32)

(15) *__wh1 BOY

“Which boy?”

(Signer 01^) In cases such as (15), mouthing of a wh-word is required for the question to be grammatical.

wh-words are always accompanied by wh-NMM. When the wh-word is an object and in situ, wh-NMM can be on the wh-word only (16), or on the entire clause as shown in (17) and (18).

(16) ___________________tp__________________________________________wh1+cu L-O-T-T-O WINHOW.MUCH-TTASLHOW.MUCH-TTASL “How much did she win in the Lotto?”

(Signer 01) (17) NM: _______________________________wh2

MO: which WOMAN BUY DRESS GI

“Which dress did the woman buy?”

(Signer 03) (18) ____wh1___________________wh1+cu

PULL WHAT-TTASL1PULL “What is being pulled?”

(Signer 02)

When the wh-word is clause-final, wh-NMM can be found on the wh-word only, as shown in (19), and on the entire clause, as in (20).

(19) NM: _________________________________________tp_______________wh2

MO: who

SCOOTER CL:PERSON.CL:VEHICLE WHO-TTSL PU “Who is on the scooter?”

(Signer 03) (20) NM: ___________________________wh2

COOK PU WHO-TTSL PU “Who is cooking?”

(33)

When the word is clause-initial, NMM can be found on the word (21) and wh-phrase (22) only. When the wh-word is in situ, wh-NMM can be found on the entire clause (23).

(21) ______________wh1

WHO-TTASL2 SCOOTERWHO-TTASL2 SCOOTER “Who is scooting?”

(Signer 02) (22) ________________wh1

WHICH DRESS BUY WHICH WOMAN BUY DRESS “Which dress did the woman buy?”

(Signer 02) (23) __________________wh1

WHO-TTASL1 COOKIX2

“Who (do you think) is cooking?”

(Signer 01)

5.3 Syntax

Wh-words were found in clause-initial, clause-final and duplicated in clause-initial and clause-final position, as shown in (24) to (26).

(24) WHO EAT FISH [clause initial]

“Who ate fish?”

(Signer 01^)

(25) FISH EAT WHO [clause final]

“Who ate fish?”

(Signer 01^) (26) ________________________________________wh1 [doubled]

WHY-TTASLGIRLCRYWHY-TTASL “Why is the girl crying?”

(Signer 02)

However, the most favoured position for the wh-word was at the end of the sentence for all signers. This is best illustrated in example (27a) below, where rightward movement is observed. In TTSL, the subject always precedes the verb, but in (27a), it is found at the end of the question. Signer 03 had a strong preference for this configuration and also said that questions with the wh-word at the beginning, as in (27c), were more likely to be

(34)

interpreted as answers, or rhetorical questions. (27c), which is a case of leftward movement, is the least preferred of the sentences shown in (27). In TTSL and TTASL, temporal adjuncts like YESTERDAY, when they appear in a sentence, appear sentence initially. In (27c), WHO is moved out of its in situ position and into the left periphery, before YESTERDAY.

(27) a. YESTERDAYEATFISHWHO [right movement]

b. YESTERDAYWHOEATFISH [in situ]

c. WHOYESTERDAYEATFISH [left movement]

“Who ate fish yesterday?”

(all signers^) An additional example is shown in (28) where an object is questioned. In this case it is not possible to distinguish in situ from rightward movement, since we do not have a diagnostic test for the right periphery. However, all the signers prefer (28a) over (28b), showing that clause-final position is preferred over clause-initial position.

(28) a. IX1EATWHAT [clause-final]

b. WHATIX1EAT [left movement]

“What did I eat?”

(all signers^)

5.4 Summary

This description gives an indication of the scope of variation in the form of wh-questions that exists in the signing community in T&T. The most variation is observed within wh-words, where two distinct wh-paradigms are observed — one which is highly articulated with an idiosyncratic sign for each wh-word (TTASL), and one in which there are two idiosyncratic wh-words and a general interrogative covering the remainder of the paradigm. Regarding wh-NMM, two main patterns are observed — wh1 (furrowed eyebrows + squint) and wh2 (eyebrow raise + wide eyes). These groupings were also commonly combined with a head tilt. wh-NMM always appears on wh-words, and can appear on some non-interrogative signs to make them interrogative. Regarding wh-NMM scope, there does not appear to be a clear pattern since wh-NMM is found on arguments that have and have not been moved, and can also spread onto the verb and the entire clause. It can also spread to temporal adjuncts. As for syntax, wh-words can be found clause-intially, clause-finally, and doubled clause-finally and clause-initially. Despite these possibilities, the favoured position for the wh-word is clause-final. The least preferred position is clause-initial, though it is still possible.

(35)

6. Discussion

The description presented in chapter 5 shows variation in the form of wh-questions at the lexical, prosodic and syntactic levels. This is discussed in relation to other sign languages, and language contact (spoken and signed) in 6.1. In 6.2 issues of method are discussed.

6.1 Wh-questions in T&T Sign Varieties and Other Sign Languages

6.1.1 Wh-paradigm(s)

The most salient variation occurs in wh-words, where two wh-paradigms of varying complexity exist — one belonging to TTSL and one to TTASL. An almost identical pattern is also found in New Zealand Sign Language (NZSL) in which “specific signs for

different types of content questions have developed relatively recently (influenced by Signed English) in contrast to an earlier paradigm in which one generic interrogative sign covered most information seeking functions.” (McKee 2006: p. 77). The TTSL paradigm is a Type A, or minimal paradigm while the TTASL paradigm is Type C. In TTSL there are just two idiosyncratic wh-words (WHO and (WHAT) TIME), and one GI that covers the remainder of the paradigm. wh-words are distinguished by Trinidadian English/Creole mouthings. In contrast to this, the TTASL paradigm consists of idiosyncratic signs for WHO, WHAT, WHEN, WHERE, WHY, WHICH, HOW, HOW MANY/HOW MUCH and (WHAT) TIME. Mouthings also accompany these signs, but they are not necessary for disambiguation like they are in the TTSL paradigm.

The use of mouthings to distinguish signs covered by the GI in TTASL is found among older signers of NZSL (McKee 2006), and also in YSL (Bauer 2014). NZSL has a similar history to TTSL in that in both contexts there was a period of Oralist education followed by the introduction of Total Communication using a manually coded form of English (Australasian Signed English in New Zealand and SEE in T&T) (McKee 2006: p. 72-73). YSL is not similar in the same sense; however it is a shared sign language used by hearing and deaf people, and therefore would have developed with considerable spoken language input (Bauer 2014). The form of the GI in TTSL and NZSL is similar in that the palms are flat and oriented up, and movement is from side-to-side. This resembles the palm up co-speech gesture used by hearing non-signers to indicate uncertainty. The similarities between the wh-paradigms in the signing communities in T&T and New Zealand in particular are likely to be due to their strikingly similar histories. It seems that the form of

wh-paradigms in sign languages could be influenced by the developmental history of a sign language. So, in languages that developed with more than usual contact with spoken language, one could expect a minimal/intermediate paradigm with a GI resembling the spoken language gesture that accompanies questions that is disambiguated by mouthings.

6.1.2 Wh-NMM

Regarding wh-NMM, there is little variation in form, but much variation in scope. With respect to form, the furrowed eyebrow component of wh1 (furrowed eyebrows + squint) is

(36)

commonly found in Western sign languages, so its use by T&T signers is not unusual (Zeshan 2006: p. 41). Wh1 itself is also found in ASL and NZSL (McKee 2006). Wh2 (brow raise + wide eyes) is similar to the wh-NMM found in InSL which also includes a brow raise (Zeshan 2006: p. 41). Wh2 also appears to be identical in form to the NMM for topics (glossed as tp). I conclude this based on the patterns exemplified in (29) and (30) (in these examples I gloss for function rather than form). In (29) and (30) “girl” and “Lotto” are topics. They are both in clause-initial position and are marked with eyebrow raise + wide eyes. Also, in (30), wh1 spreads over the entire clause apart from “Lotto”. It is unlikely that two different wh-NMMs would be used together and in such close proximity, so I conclude that the NMM on “Lotto” is for topic.

(29) NM: _____tp ______wh1

MO: where

GIRL HIDE GI

“Where is the girl hiding?”

(Signer 03) (30) ___________________tp__________________________________________wh1+cu

L-O-T-T-O WINHOW.MUCH-TTASLHOW.MUCH-TTASL “How much did she win in the Lotto?”

(Signer 01)

Regarding scope, wh-NMM is always found on wh-words and wh-phrases, but in terms of spreading there is no clear pattern that could be associated with wh-dependency, like that described for LIS. Spreading can be on the verb; the subject and the verb; and on the subject, verb and temporal adjunct, regardless of the position of the wh-word and if movement had occurred. However, in cases like (29), where spreading occurs from the wh-phrase to the verb only, spreading to the entire clause is interrupted by topic marking on the subject. In ASL, spreading must be on the entire clause, apart from topics, if the wh-phrase is in sentence initial position (Fischer 2006: p. 172). However, as we see in example (22), repeated below as (31), this is not mandatory in TTSL. Notably, this sentence was produced by signer 02, who represents TTASL best; therefore, the spread of wh-NMM in TTASL is different from that in ASL.

(31) ________________wh1

WHICH DRESS BUY WHICH WOMAN BUY DRESS “Which dress did the woman buy?”

(Signer 02)

This pattern, or lack thereof in TTSL, could be due to several reasons. First, as mentioned previously, the spread of wh-NMM may be interrupted because of interaction with other

Wh-questions in the Trinidad and Tobago Signing Community