I would like to express my sincere gratitude to: the members of the Kwéyòl Donmnik language community who contributed the data analyzed in this research; Sylvia Henderson for transcribing and translating the corpus data; the Dominica Overseas Nationals Association for their help with participant recruitment and recording space; Marlyse Baptista, Sarah G. Thomason, Diana L. Ranson, Susan Gelman, and anonymous reviewers for their constructive comments and suggestions; and audiences at various discussion groups and conferences for their insightful questions and feedback.

1. Introduction

Tab. 1, Selected PMs

KWÉYÒL PMs	FRENCH COUNTERPARTS (ETYMA)	ENGLISH COUNTERPARTS
konsa ‘so’	(ou) comme ça ‘(or) like that’	so
èben ‘well’	(eh) ben ‘well’	well
papa/Bondyé ‘father/God’	bon dieu ‘good God’ and other similar expressions (e.g., mon dieu ‘my God’)	oh my God and other similar expressions (e.g., gosh)
la ‘there’	là ‘there’	here/there

2. Kwéyòl Donmnik & Creole Emergence

2.1. Sociohistorical Context

2As highlighted by researchers like Weinreich (1953), Thomason & Kaufman (1988), and Mufwene (2001, 2008), sociohistorical context is critical to the unfolding of language contact phenomena. The circumstances surrounding the emergence of Kwéyòl Donmnik, a Lesser Antillean French lexifier Creole of the Caribbean island of Dominica, involves a layered history of contact with two CSLs, making the interplay between Kwéyòl, French, and English an intriguing context for exploring mechanisms of language contact and emergence.

3Unlike “the prosperous sugar islands of Barbados, Antigua, St. Kitts, Guadeloupe and Martinique [which] were far more attractive to the [enslavers]” (Honychurch, 1995: 53), Dominica’s small size and mountainous terrain was not conducive to sprawling plantations. Instead, early French settlements were small-scale, allowing for intense contact and facilitating the lexifier’s lasting linguistic influences. By 1745, about half of the 3,032 documented inhabitants of Dominica were enslaved (Honychurch, 1995: 54). Most were Caribbean born and transshipped from other islands (Honychurch, 1995: 53), perhaps bringing with them other Creole varieties. Meanwhile, the indigenous Kalinago, users of an Arawakan language, became increasingly isolated to “the rough, rocky north-east quarter” of the island (Honychurch, 1995: 50) where other people of color are known to have encountered and lived among them (Honychurch, 1995: 64; Taylor, 1977: 25). By 1778, “…the island’s population consisted of 1,574 whites, most of them French, 574 free mulattos and blacks, and 14,309 [enslaved people]” (Honychurch, 1995: 87). More diachronic research is needed, but it is reasonable to presume that Kwéyòl began to develop in earnest over the course of the 18th century as more colonizers and enslaved and escaped people of color arrived on the island.

4Dominica changed hands between the French and the British several times but remained primarily under British rule from 1763 until Dominica’s independence in 1978, a layering of colonial influences that shifted the linguistic landscape. Over time, English has become the foremost colonial language with which Kwéyòl coexists and is the language of Dominica’s government, education, and commerce. Most Kwéyòl users are elders; “the language is losing fluent speakers and is no longer spoken as a first language by the majority of Dominican children; by most measures, then, [Kwéyòl] would be considered an endangered language” (Paugh, 2012: 9). Nearly all remaining users are bilingual (typically in English), and today, many are spread across an English-dominant diaspora, including the United Kingdom and the United States. English influence extends beyond lexical borrowings to changes in the Creole’s sound system, such as the introduction of word-final tch [tʃ] and dj [dʒ] and word-initial r (Christie, 2003: 26). Contact is so intense that Kokoy, an English lexifier Creole, has emerged on the island (Christie, 2003: 30; Aceto, 2010). Given this sociohistorical context, it is unsurprising that while PMs of French origin pervade Kwéyòl speech, they are used alongside English markers like well and so.

2.2. Mechanisms of Language Contact & Emergence: Congruence

5As Section 2.1 illustrates, Creole emergence can involve several languages, and creolists debate how Creole genesis proceeds. As summarized by Baptista (2020: 160), Whinnom (1956, 1965) suggests that all Creoles originate from a single linguistic ancestor, while Bickerton (1981, 1984, 2014) argues that Creoles are rooted in language universals and thus display similar grammatical features regardless of their source languages (SLs). Others propose that a particular SL has the most significant impact on a Creole’s grammar (e.g., Chaudenson, 2001, 2003; Lefebvre, 1998). However, many “would agree that Creoles mix properties of their source languages” (Baptista, 2020: 161), and researchers seek to understand how this process unfolds.

6One prominent proposal is from Mufwene’s (2001, 2008; also see Aboh 2009, 2015) biological model of language evolution: the diverse pool of linguistic features supplied by the various SLs that make up a Creole’s linguistic ecology compete for selection, or inclusion in the emerging grammar. Baptista (2020) suggests that preexisting congruence plays a critical role in this process: “the similarities (congruent features) that speakers perceive between the languages in contact are favored to participate in the emergence and development of a new language” (Baptista, 2020: 161). These perceived similarities may be in form, function, and/or syntactic distribution. For example, the form, functions, and preverbal positioning of ka, the negative marker in Cabo Verdean Creole (CVC) and Guinea-Bissau Creole (GBC), retain traits shared by Portuguese nunca ‘never’ and negators in the Creoles’ African SLs, such as Mandinka buka (Baptista, 2020: 173-174). Given that social factors like language users’ attitudes are “a powerful force in promoting or inhibiting change” (Baptista, 2020: 162 summarizing Thomason, 2001), Baptista’s (2020: 162) “goal is NOT to propose a predictive, deterministic model of language emergence and contact-induced change”, and she “consider[s] congruence to be just one of many mechanisms involved in Creole formation—but an important one”.

7Some linguists may refer to congruence as convergence. However, congruence should not be confused with areal convergence: “long-term coexistence [that] can lead languages to CONVERGE with each other, ultimately leading to the rise of congruent features among them (see Joseph 1983, 2010)” (Baptista, 2020: 163). Both phenomena have been explored by numerous researchers (e.g., Thomason & Kaufman, 1988; Silva-Corvalán, 1994, 2008; Aboh & DeGraff, 2014, 2017); for a detailed discussion, see Baptista (2020: 163-166). To my knowledge, research on congruence and other mechanisms involved in Creole emergence does not examine PMs. However, our understanding of language emergence must also take into account the discourse-pragmatic domain; their dynamic multifunctionality and flexibility with respect to syntactic distribution make PMs a promising site for such research.

2.3. Two Contact Situations

8There is also a need for more research into how contributions from a non-lexifier CSL are integrated into a Creole language. Research on Kwéyòl is well-suited to this as there are two CSL contact situations involved in its development: (a) contact with the French lexifier from the start of Kwéyòl’s emergence and (b) contact with English, a second (and now dominant) CSL, that began early in Kwéyòl’s emergence. Anticipating that mechanisms like congruence likely played a role in both of these overlapping contact situations, I expected to find that congruencies between its CSL counterparts are reflected in a Kwéyòl PM’s properties. Given that Creole languages can also develop innovations that “evolve independently from [their SLs]” (Baptista 2020: 161), I also expected to find facets of a PM’s usage in the Creole that are distinct from its CSL counterparts. Future work will investigate whether the properties of PMs in SLs that are not as well documented, such as the Kalinago language, may have contributed to the traits of Kwéyòl’s PMs.

3. Pragmatic Markers

3.1. Defining Pragmatic Markers

10Researchers have also proposed several theories to capture how PMs are interpreted and integrated into discourse (e.g., Schiffrin, 1987; Blakemore, 2002; see Aijmer, 2013 for an overview). One promising proposal suggests that “parts of a word’s meaning are evoked, activated or materialised, foregrounded or backgrounded, in different ways in the different contexts, in which it is exploited” (Norén & Linell, 2007: 390 cited by Aijmer, 2013: 12). A user’s mental representation of a marker is organized around a limited set of core, conventionalized meanings that permit “less conventionalised (or ad hoc) meanings to be created in the communication situation” (Aijmer, 2013: 13). So long as they are compatible with these meaning potentials (Norén & Linell, 2007), “new functions can be created in the interaction” (Aijmer, 2013: 12).

11This approach aligns with research by Cuenca (2008: 1382-1385), whose work on English well suggests that PMs are radial categories: semantic networks of interrelated senses and within which some functions are more closely or more peripherally related to the PM’s core meanings, or prototype foci. This approach also allows a PM’s functions to expand while still maintaining semantic links to the lexical item from which the PM originates. For example, English well, whose lexical parent is the adverb well, and French ben (often combined with eh to form eh ben, Kwéyòl èben’s etymon), whose lexical parent is the adverb bien ‘well’, retain semantic links to evaluation and acceptance: both introduce undesirable responses, like disagreement, and convey concession or begrudging acceptance (see Section 5.1). Notice, however, that French ben [bɛ̃] does not retain the diphthong present in the adverb bien [bjɛ̃]; like many PMs, it is phonologically “a reduced form” (Barnes, 1995: 816) compared to its item of origin.

3.2. Pragmatic Borrowing

12For work on PMs that centers on language contact, I turned to the literature on borrowing. Though much of this research focuses on the transfer of lexical items, languages in contact may also display pragmatic borrowing: “the incorporation of pragmatic and discourse features of a source language (SL) into a recipient language (RL)” (Andersen, 2014: 17). Andersen’s (2014) approach to analyzing pragmatic borrowings proved a useful framework for examining how discourse-pragmatic contributions from Kwéyòl’s CSLs have been integrated into the Creole. Much like investigating a potential case of congruence requires closely examining the form, function, and distribution of a linguistic feature in a Creole (the RL) and its SLs, Andersen (2014: 18) proposes that researchers determine the “degree of parallelism” (Andersen, 2014: 23) between the properties of a pragmatic item in the SL and the RL. This allows scholars to “detect pragmatic functions that have been transferred, functions that are not transferred and new functions that may have emerged post hoc in the RL” (Andersen, 2014: 23).

13Andersen (2014: 23-24) recommends the researcher (A) assess the distributional features (“discourse-structural and syntagmatic aspects”) of the SL and RL markers, (B) inventory the PMs’ functions to determine whether functional stability (“no observable change in the pragmatic function of the marker”) or adaptation (a narrowing, broadening, or shift in function) has taken place during integration into the RL, and (C) “take into account sociolinguistic aspects and consider relevant demographic predictors and factors such as register and style”. Step C is beyond the scope of this investigation, but Steps A and B form the basis for this study’s methodology (see Section 4.3).

14Other researchers define borrowing differently. For instance, van Coetsem (1988) contrasts borrowing, “the transfer of linguistic materials from [an SL] into [an RL] via the agency of speakers for whom the latter is the linguistically dominant language” (Winford, 2020: 12 summarizing van Coetsem, 1988), with imposition, “a process by which the speaker transfers features of her linguistically dominant language (as SL) into her version of the recipient language (RL)” (Winford, 2020: 8 summarizing van Coetsem, 1988). However, in contact situations like the one investigated here, it is unclear which language those who used Kwéyòl (the RL) were dominant in when material from the CSLs was incorporated. For example, early in Kwéyòl’s history, English would presumably have been a non-dominant language for users of the emerging Creole. Now, most Kwéyòl users are English-dominant. Though use of items like so and well as PMs in English likely preceded contact with Kwéyòl, it is unclear when they developed the functions they now perform or when they began to influence Kwéyòl’s PM inventory (see Section 4.4 for further discussion). That Andersen’s (2014) approach does not place dominance restrictions on borrowing makes it particularly adaptable to the current study.

3.3. Pragmatic Markers in French Lexifier Creoles

15In the corpus data sources used for this study (see Section 4.1 for details), there were several pragmatic items attested: o ‘oh’; a ‘ah’; pis ‘because’; bon ‘well’; mé ‘but’; manman ‘mother’; the invariant tag ennit ‘isn’t it’ (see Pichler 2021 for a discussion of this tag’s development in London English and its potential origins); konsa ‘so’, èben ‘well’; papa ‘father (God)’; Bondyé ‘God’; and la ‘there’. Although Creole PMs are understudied, similar markers are found in other French lexifier varieties. St. Lucian features items like Bondyé ‘good Lord!’, Bondous ‘my God!’, dakò ‘okay, agreed’, bon ‘well’, and awa ‘oh no!’ (Frank, 2020). Colot & Ludwig (2013, citing Ludwig & Pfänder, 2003) describe use of la ‘there’ as a deictic discourse marker in Martinican and Guadeloupean (also members of the Lesser Antillean family), and work by Chady (2021) explores the Mauritian extension particle sipa ki ‘whether’.

16Since there were many, I chose to analyze a subset of the PMs present in the Kwéyòl data sources. First, I chose konsa ‘so’ and èben ‘well’ because of their frequency and because their English counterparts so and well also surface in the data; I coded these English tokens as well, curious whether they were used differently in the Creole. I included Bondyé ‘God’ and papa ‘father (God)’ because of their religious content, which is a cultural artifact of the Lesser Antilles’ colonial history. I analyzed them together due to their similar functions and their potential for cooccurrence as papa Bondyé ‘father God’. While there are no tokens of this collocation in the data, it surfaces in a footnote by the London Corpus transcriber/translator (see Section 4.1 for more on the London Corpus) and was mentioned to me during interviews with Kwéyòl users (Peltier, 2022). Another familial PM, manman ‘mother’, surfaced in the data, but whether it shares similar religious undertones is a topic for future research. Finally, I included la ‘there’ in this study to determine whether it functions as a PM in Kwéyòl.

4. Methodology

4.1. Data Sources

17I analyzed data from three sources:

• The London Corpus (LC), comprised of fieldwork data I collected in 2018, contains 90 minutes of three dyads of Kwéyòl-English bilinguals using Kwéyòl: one man and five women between 58 and 82 years of age. Five were Dominica born; the sixth, born in the UK to Dominican immigrants, spent her formative years in Dominica. Participants completed four tasks: dialogue, responding to a silent film called The Pear Story, narrating Mayer and Mayer’s (1975) wordless picture book One Frog Too Many, and a pattern-building task (Cooperrider et al., 2014, 2018). A Kwéyòl-English bilingual translated/transcribed the data. Participants were issued codes made up of their initials, gender identity (f = female, m = male), and age. For three samples of this corpus data, including audio recordings and transcripts, please visit :

18

19

20

21

22

23

• The Corpus Créole (CC) (Ludwig et al., 2001), a collection of Creole conversations, includes one interview and one radio segment in Kwéyòl, both of which took place in Dominica in 1986.

• The Ma’ Bernard Folktales (MBF), my name for a trio of Kwéyòl folktales contributed by a user named Ma’ Bernard, were documented by Douglas Taylor (1977).

24There are orthographic differences across the data sources. For example, in the MBF transcriptions, which were written before Kwéyòl’s standardized writing system was developed, nasal vowels are indicated by a tilde (e.g., èbẽ ‘well’) rather than an n (e.g., èben). In the CC, there is variation such that words containing a w (e.g., apwézan ‘presently’) are sometimes written with an r (e.g., aprézan).

4.2. Identifying Tokens

25After importing the data into Atlas.ti qualitative data analysis software, I labeled the properties of each PM token by assigning to it searchable tags called Codes. Since èben ‘well’, well, and so do not have lexical counterparts in Kwéyòl, locating them was simple, and I verified that tokens of papa/Bondyé ‘father/God’ were being used secularly. Isolating tokens of konsa ‘so’ required excluding instances of kon sa ‘like that’, its lexical, non-PM counterpart. I verified that each token was indeed functioning as a PM and not as a phrase in which sa ‘that’ indicated an intra- or extralinguistic referent (e.g., yon bagay kon sa ‘a thing like that’).

26To isolate relevant tokens of la, I first excluded instances of the definite determiner, distinguishable based on its post-nominal position (e.g., liv-la ‘the book’) and its allomorphic morphophonological patterns: in Kwéyòl, the determiner /la/ is realized as [la] after a consonant and [a] post-vocalically (e.g., moso papyé-a ‘the piece of paper’). This determiner undergoes similar morphophonological changes in other French lexifier Creoles, such as Haitian (Déprez, 2007: 270 citing Joseph, 1989; Lefebvre, 1998: 79). I also ruled out instances of la ‘there’ that provided locational information or occurred in existential constructions (la ni… ‘there is/are…’).

27These homophonous la morphemes contrast with instances like the one bolded in (1) below. Here, the utterer asks her daughter to take a cardboard box to the house of a man who recently died. No location of death is referenced in the discourse, suggesting that la performs another discourse-pragmatic function. This example resembles instances documented in Guadeloupean and Martinican in which a noun phrase containing definite la (nonm-la ‘the man’) is followed by a relative clause also ending in la (ki mò la ‘who died [there]’. The second la has been called a deictic discourse marker (Colot & Ludwig, 2013), and I hypothesized that this la, which also surfaces in contexts other than relative clauses (see (44) and (45) in Section 5.6), might be a PM in Kwéyòl as well.

(1) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

Dèmen,	O,	ou	sav	sa	mwen	té	vlé	’w	fè?
Tomorrow	oh	2SG	know	what	1SG	ANT	want	2SG	do
‘Tomorrow, oh, you know what I wanted you to do?

 

Pou	’w	té	mennen	an,	an,	an	katon	koté
for	2SG	ANT	bring	INDF	INDF	INDF	carton/cardboard box	by
For you to bring a, a, a carton/cardboard box by

 

nonm-la	ki	mò	la.
man-DEF	who	died	PM
the man who died [there].’

4.3 Analysis

28First, I assigned Codes to each token (e.g., “Placement: Initial”) based on Andersen’s (2014: 23) list of distributional features, or “discourse-structural and syntagmatic aspects”:

Utterance Placement: Does the marker occur utterance-initially/medially/finally?

Scope: Does it take into its scope a full proposition or a propositional constituent?

Orientation: Does it point forward or backward in the discourse, qualifying upcoming or preceding material?

Degree of Syntactic Integration: Is it used as a free-standing device, or is it to some degree syntactically integrated?

Collocational Features: To what extent is the marker a constituent of a fixed or semi-fixed phrase or collocation?

(Adapted from Andersen, 2014: 23)

29Keeping in mind that a single PM token can perform multiple functions, I then took notes (Comments in Atlas.ti) on each token, documenting my initial impressions of how the utterer used the PM to mark junctures in the discourse or to convey attitudinal/cognitive states. I used those Comments to group the tokens into functional categories, assigning each function a Code (e.g., “Function: Self-Repair”). Next, I compared the coding results with descriptions of the markers’ CSL counterparts in the literature to (a) identify instances where congruence may have favored a PM to display certain properties in the Creole and (b) determine whether integration of the CSL PMs into Kwéyòl had resulted in

Functional Stability: No observable change in the pragmatic function of the marker in the SL and RL; the marker is associated with the same type of speech act, user attitude, and/or illocutionary force in both the SL and RL

or in a subtype of Functional Adaptation:

     Functional Narrowing: Loss of some function of the marker in the transition from the SL to the RL, or transfer of only one function of the multifunctional SL marker

     Functional Broadening: Acquisition of a new pragmatic function in the RL not observed in the SL

     Functional Shift: Loss of some function of the marker in the transition from the SL to the RL combined with acquisition of a new pragmatic function in the RL, or modification of an existing pragmatic function in the transition from the SL to the RL

(Adapted from Andersen, 2014: 24)

4.4. Limitations

30No corpus-based analysis is exhaustive. While some of the PMs were more plentiful (30 tokens of èben ‘well’), others were less so (six of papa ‘father (God)’), and the Kwéyòl markers may perform functions that were not attested in the data. Likewise, the CSL markers may perform functions that were not reported in the literature or that have developed since past studies were conducted. Use of the PMs discussed here may also vary geographically or socially in ways that are beyond the scope of this study.

31Another limiting factor is the scarcity of diachronic information, making it unclear when the Kwéyòl and CSL PMs emerged and developed their various functions. For instance, it is possible that Kwéyòl èben ‘well’ was being used as a self-repair device prior to intense contact with English well (which also facilitates self-repairs), having already incorporated this function from its lexifier etymon (eh) ben earlier in the Creole’s emergence. It is even possible, if unlikely, that Kwéyòl èben developed this function independently, as “it is generally difficult to decide whether an innovation is due to pragmatic borrowing [here, integration into the Creole RL] or a parallel development in two or more languages” (Andersen, 2014: 21).

32However, consider that multiple causes may contribute to the linguistic outcomes of language contact (Thomason, 2008: 47); both Kwéyòl-internal innovations and external contributions from one or both CSLs can have led a Kwéyòl PM to develop certain traits. For example, even if èben had already acquired its self-repair function from the lexifier, use of èben to perform this function may have later been reinforced (i.e., increased in prominence and frequency of use) by the growing dominance of English well. Also, though the only relevant historical work of which I am aware is limited to the English PMs, it does suggest that these items were performing some PM functions prior to Kwéyòl’s emergence. Use of well as an emphatic attention-getting device extends back to Old English (Jucker, 1997: 91; Marcus, 2009: 215), and use of so as an “introductory particle” is attested as early as the 1590s (Harper, n.d.). Though explicit use of g-words (expressions containing God) in secular contexts has increased in recent decades, euphemistic versions have been in use since the Middle Ages (Tagliamonte & Jankowski, 2019: 213). Despite its limitations, this study provides rich insights that can inform further research into the emergence and development of Creole varieties at the discourse-pragmatic level, particularly in cases involving linguistic contributions from both a lexifier and a non-lexifier CSL.

5. Results

5.1. Kwéyòl Èben

33Of the 30 tokens of èben, seven collocated with other pragmatic elements like a ‘ah’, o ‘oh,’ so, oké ‘okay’, and apwésa ‘afterwards’. Usually, èben was utterance-initial (25 tokens), as in (2), though there were also -medial (three tokens, see (3)) and free-standing instances (four tokens, see (4)). This aligns with the French etymon (eh) ben ‘well’ and English counterpart well, which can also occur utterance-initially, -medially, or as free-standing utterances.

(2) (Book Narration, LC, EDf82, gloss mine)

Èben	sa	sé,	sa	sé	“story”-la	ki	fèt	la.
PM	DEM	is	DEM	is	story-DEF	that	happened	there
‘Well, that is, that is the story that happened there.’

(3) (Book Narration, LC, EDf82, gloss mine)

I	ka	sanm,	tiwé	yon	“frog”	la	anlè,	èben	yonn,	dé	“frog”,
3SG	PROG	seem,	take.off	INDF	frog	LOC	on.top,	PM	one,	two	frog,
‘It seems, take off a frog there on top, well one, two frogs,

 

sété	dé.
was	two
it was two.’

(4) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

A:	Zò	ka	sizé	la	èvè	bwè?
	2PL	PROG	sit	there	and	drink
	‘You sit there and drink?’

 

B:	Èben!
	PM
	‘Well!’

 

A:	Èvè	tout	sé	moun-la?
	with	all	PL	person-DEF
	‘With all the people?’

34French (eh) ben builds coherence between discourse chunks at points of discontinuity (Barnes, 1995; Bruxelles & Traverso, 2001) and indicates an orientation shift in the discourse’s deictic center (Barnes, 1995). English well, too, is a coherence builder and orientation shifter (Barnes, 1995; Schiffrin, 1987). Thus, these PMs both mark major textual junctures: topic changes and introducing new or unexpected content (Bruxelles & Traverso, 2001: 44 and Barnes, 1995: 817 regarding (eh) ben; Jucker, 1997: 97; Beeching, 2011: 99; and Cuenca, 2008: 1388 about well), relaunching abandoned topics (Cuenca, 2008: 1388 for well; Bruxelles & Traverso, 2001: 44 for (eh) ben), facilitating self-repairs and utterance reformulations (Barnes, 1995: 814 with respect to both markers; Cuenca, 2008: 1388 about well), and introducing quoted utterances or reported discourse (Barnes, 1995: 817 about (eh) ben and Schiffrin, 1987: 685 about well).

35Kwéyòl èben facilitates navigation of textual junctures as well. In (5), èben introduces a new topic: the beginning of the second Ma’ Bernard folktale. In (6), a character in the first folktale has just received advice from a soothsayer about freeing his wife from a zombie’s grasp. Here, èben highlights the transition into the next scene: the husband embarking on his quest. This marker can also introduce surprising, unexpected information. In (7), this function is realized sarcastically. A widower has followed a soothsayer’s instructions to transform his dog into a woman; the marker highlights that it is in fact not a surprise that the lonely man happily accepted her as his partner, an effect underscored by non ‘no, of course’.

(5) (Taylor, 1977: 240, MBF #2, gloss mine)

Ebẽ,	sete	yõ	vye	kô	ki	pa	te	ni	zãfã,	epi
PM	there.was	INDF	old	body	who	NEG	ANT	have	child	and
‘Well, there was an old fellow who was childless, and

 

madam-li	vini	mô.
wife-3SG.POSS	came	dead
his wife died.’

(6) (Taylor, 1977: 236, 238, MBF #1, gloss mine)

Ebẽ,	mahwi-a	pwã	sjimẽ	-y;	i	ale,	i	ale.
PM	husband-DEF	took	way	3SG.POSS	3SG	went	3SG	went
‘Well, the husband set out on his way: he went on and on.’

(7) (Taylor, 1977: 239, 240, MBF #2, gloss mine)

Ebẽ,	nõm-la	pwã-y	pu	fè	mun-li	nõ.
PM	man-DEF	took-3SG	to	make	mate-3SG.POSS	no
‘Well, the man took her as his mate, to be sure!’

36Like (eh) ben and well, èben can be used to revisit abandoned or interrupted topics. In (8), Interlocutor A is interrupted while instructing Interlocutor B to give a widow a cardboard box. After reminding B of the woman’s name, A returns to giving instructions; èben marks this shift. Èben is also similar to its CSL counterparts in that it can facilitate self-repairs and reformulations. This was illustrated in (3) above.

(8) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

A:	Am,	non,	pli	ta	mon	ké	mété	katon-la
	um	no	more	late	1SG	FUT	put	cardboard box-DEF
	‘Um, no, later I will put the cardboard box

 

	andidan	lapòt	madanm-la	ba	’y.
	in	door	woman-DEF	for	3SG
	In the wife’s door for her.’

 

B:	Ki	non	’y	ankò?
	what	name	3SG.POSS	again
	‘What’s her name again?’

 

A:	S.
	S
	‘S.’

 

[…]

A:

Mé

sé

ou

ki

di

ou

pa

sa

“remember”

non,

non

fanm-la.

but

is

2SG

who

said

2SG

NEG

that

remember

name

name

woman-DEF

‘But it’s you who said you cannot remember that woman’s name.’

 

B:	Wi,	mwen	ké	chonjé.
	yes	1SG	FUT	remember
	‘Yes, I will remember.’

 

A:	Oké.
	okay
	‘Okay.’

 

B:	Chonje,	chonjé,	wi.
	remember	remember	yes
	‘Remember, Remember, yes.’

 

A:

Èben

lè

’w

alé

la,

di

S

mon

di,

am,

“my

condolences”.

PM

when

2SG

go

there

tell

S

1SG

said

erm

1SG.POSS

condolences

‘Well when you go there, tell S I said, erm, my condolences.’

37Both well and (eh) ben can soften the impact of undesirable responses, such as replies that are indirect, inadequate, or delayed or that express disagreement (Barnes, 1995: 816; Beeching, 2011: 99; Jucker, 1997: 97; and Cuenca, 2008: 1380 regarding well; Barnes, 1995: 816 and Bruxelles & Traverso, 2001: 45 about (eh) ben). They can also express concession or partial agreement/acceptance (Bruxelles & Traverso, 2001: 47 about (eh) ben; Beeching, 2011: 99 with respect to well). Similarly, èben introduces undesirable responses and indicates begrudging agreement/acceptance. In (9), after Interlocutor A reminds Interlocutor B that B’s car is not in any condition to drive to the widow’s house, èben introduces B’s undesirable response: that A will have to make the delivery herself on foot. A’s next utterance, also introduced by èben, expresses her unhappy acceptance of this.

(9) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

A:	So,	kouman	’w	kè	fè	alé	la?	Ou	pa	sa
	so	how	2SG	FUT	be.able	go	there	2SG	NEG	can
	‘So, how will you be able to go there? You can’t

 

	mété	motoka	’w	asou	chimen-la.
	put	car	2SG.POSS	on	road-DEF
	put your car on the road.’

 

B:	Èben	ou	ké	ni	pou	maché	la	ou	menm.
	PM	2SG	FUT	have	for	walk	there	2SG	self
	‘Well you will have to walk there yourself.’

 

A:	Èben	sé	sa	ki	mon	ni	fè;	mon	menm	ké	alé
	PM	is	that	what	1SG	have	to.do	1SG	self	FUT	go
	‘Well that is what I have to do; I myself will go

 

	la.	Mon	pa	“mind”.
	there	1SG	not	mind
	there. I don’t mind.’

38Though their functions overlap, there are differences between the two CSL markers. Only (eh) ben is associated with underscoring comparisons and contrasts (Hansen, 1998: 288) and with introducing illustrative examples or other elaborative information (Hansen, 1998: 289; Bruxelles & Traverso, 2001: 44). Kwéyòl èben performs both of these functions. In (10), a Kalinago chief uses èben to underscore a comparison between how he and his staff were once paid annually and the monthly pay system that was instituted later in his term. In (11), the same interviewee has been telling the interviewer about the treaty that granted the Kalinago their territory. Èben indicates that he is about to elaborate on a related side-topic: the signing process by which the treaty was approved.

(10) (Interview, CC, gloss and translation mine)

[…] mè	lè	mwen	menm	té	rantré […]	chef	kwayib
but	when	1SG	self	ANT	become	chief	carib
‘…but when I myself became […] Kalinago chief

 

yo	té	ban	nou	on	ti	lajan	pa	(adan)	lanné […]
3PL	ANT	give	2PL	INDF	little	money	per	in	year
they gave us a little money per (in a) year […]

 

èben	dépi	aprésa	yo	désann	yo	té	ka	ban	nou	on
PM	since	after.that	3PL	decided	3PL	ANT	PROG	give	1PL	INDF
well since then they decided they gave us a

 

ti	lamoné	pa	mwa
little	money	per	month
little money per month’

(11) (Interview, CC, gloss and translation mine)

Èben	sé	biten-sala	i	té	ni	menm	WITNESS	té	ka
PM	PL	thing-DEM	there	ANT	have	even	witness	ANT	PROG
‘Well there were even witnesses who were

 

siyé	anba	a	yo	komki	THEN	yo	siyé	on	on	kontra
sign	below	ah	3PL	like	then	3PL	sign	INDF	INDF	contract
signing below on those things like they sign a contract’

39Likewise, only well introduces closing remarks that bring an end to the discourse (Cuenca, 2008: 1388); expresses emotional—often negative—reactions like indignation, disappointment, or sadness (Beeching, 2011: 98 citing Corréard et al., 2007); mitigates face-threatening utterances (Jucker, 1997: 97; Beeching, 2011: 99); and signals a desire to hold or reclaim the floor, such as by “fill[ing] interactional silences” (Jucker, 1997: 97).

40Like well, èben introduces concluding remarks, as shown in (2) above in which a fieldwork participant ends her picture book narration. Example (4) above illustrates how this marker can also express emotional reactions. There, the utterer insists she and her husband do in fact take a bottle of wine to London Fields to enjoy together. Beyond underscoring the surprising nature of this revelation, èben expresses the utterer’s frustration at her interlocutor’s skepticism. Èben can even indicate that the utterer has not yet completed their contribution, as illustrated in (8) above. There, Interlocutor A is not finished giving Interlocutor B instructions but is interrupted when B asks for the widow’s name. To facilitate her return to the contribution she was making and regain control of the floor, A uses èben.

41Kwéyòl èben does perform functions that are not reported in the literature for either CSL counterpart. For instance, when introducing a face-threatening utterance, èben does not have well’s mitigating effect. Rather, it emphasizes the pointedness of an uncomfortable or insulting question, as shown in (12). This is reinforced by the utterance-final tag then.

(12) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

A:	Oké.	“So”,	mwen	ni	pou	kwiyé	yo?
	okay	so	1SG	have	for	call	3PL
	‘OK. So, I have to call them?’

 

B:

Wi.

I

di

mon,

o,

i

di

mon

pou

kwiyé

yo

kon

mwen

èvè

’w.

yes

3SG

said

1SG

or

3SG

said

1SG

for

call

3PL

as

1SG

with

2SG

‘Yes. She told me, or, she told me to call them as I’m with you.’

 

A:	Èben,	poutji	ou	pa	fè	sa	“then”?
	PM	why	2SG	NEG	done	that	then
	‘Well, why haven’t you done that then?’

42Èben also has a result-marking function demonstrated in (13). Here, èben highlights that, because of the gradual arrival of Kalinago individuals to the scene of a conflict with the police, the courtyard became full of people.

(13) (Interview, CC, gloss and translation mine)

[…] yo	komansé	vini	yonn	pa	yonn
3PL	started	come	one	by	one
‘[…] they started to come one by one

43 

èben	lè	yo	vini	yonn	pa	yonn	aprézan	lakou-a	koumansé	plen
PM	when	3PL	came	one	by	one	then	courtyard-DEF	became	full
well when they came one by one then the courtyard became full’

44Finally, èben can signal that the utterer is waiting for their interlocutor to take the floor. In (14), the Kalinago chief has answered the interviewer’s previous question, and èben to indicates that he is awaiting the next.

(14) (Interview, CC, gloss and translation mine)

A:	Èben
	PM
	‘Well’

 

B:	Èskè	aprézan	[…] yo	ka	mandé	w	ti	avi?
	Is.it.that	now	3PL	PROG	ask	2SG	2SG.POSS	opinion
	‘These days, […] do they ask you your opinion?’

45Kwéyòl èben, like well and (eh) ben, can provide discourse-pragmatic information about the utterance that follows it (forward orientation), as it does when introducing a new topic; respond to the preceding discourse chunk (backward orientation), such as when it indicates partial or begrudging acceptance of what was just uttered; or highlight a relationship between preceding and upcoming information, a combined forward/backward orientation it displays when highlighting a contrast or comparison. Also, like its CSL counterparts, èben tends to take within its scope the full propositional content of the utterance(s) it modifies. For example, when introducing a new topic, it indicates that the content conveyed by the entire utterance it precedes constitutes the start of new material. An exception is when èben facilitates a mid-utterance self-repair; then, the PM modifies only the repaired portion of the utterance.

5.2 Well in Kwéyòl

46Two of the five tokens of English well were accompanied by other PMs: o ‘oh’ and an han ‘uh huh’. The features of well in the corpus aligned closely with those reported for well in the literature: it was usually in initial (see (15) below) or medial position (see (16)). However, examples like (17) are best categorized as free-standing; here, well and o ‘oh’ form one stand-alone collocation.

(15) (Silent Film Discussion, LC, SMf59 & PJf58, gloss mine)

Èvè,	“well”	[…] pon	panyen-a.
And	PM	take	basket-DEF
‘And, well […] take the basket.’

(16) (Dialogue, LC, SLm82 & FMLf80, gloss mine)

A:	I	di	i	lè	wè	..., i	lé	palé	ba	’w.
	3SG	said	3SG	wants	see	3SG	wants	talk	to	2SG
	‘She said she wants to see ..., she wants to talk to you.’

 

B:	An	han!	“Well”,	nou	palé.
	ah	ha	PM	1PL	spoke
	‘Ah ha! Well, we spoke.’

(17) (Book Narration, LC, SLm82, gloss mine)

O,	“well”.	I	ka	alé.
oh	PM	3SG	PROG	go
‘Oh, well. He is going.’

47One token’s function was unclear because the utterer was interrupted mid-sentence, but the others displayed known capacities of English well (see Section 5.1). In (17) above, the utterer has reached a point in the picture book’s plot where a frog has been reprimanded by his owner and is depicted walking away. Well, combined with oh, expresses sadness and a resigned acceptance of the frog’s decision to leave. Meanwhile, in (16) above, well introduces an undesirable response that contradicts and corrects the addressee: Interlocutor B tells Interlocutor A that she had already spoken with the person A said wanted to talk to her. The corpus tokens also reflected English well’s tendency to modify the entire propositional content of the utterances it accompanies and its ability to take on forward, backward, or forward/backward orientation. The results of this analysis of èben ‘well’ and well in the Kwéyòl data sources are summarized in the table below.

Tab. 2, Property comparison table for Kwéyòl èben ‘well’ and English well as they arise in the Kwéyòl data sources analyzed, as well as for French (eh) ben ‘well’ and English well as discussed in the literature

	ÈBEN	(EH) BEN	WELL	WELL (IN CORPORA)
Utterance Placement:	Initial	Initial	Initial	Initial
	Medial	Medial	Medial	Medial
			Final
	Free	Free	Free	Free
Degree of Integration:	Integrated	Integrated	Integrated	Integrated
	Free	Free	Free	Free
Orientation:	Forward	Forward	Forward	Forward
	Backward	Backward	Backward	Backward
	Forward/Backward	Forward/Backward	Forward/Backward	Forward/Backward
Scope:	Full Proposition	Full Proposition	Full Proposition	Full Proposition
	Constituent	Constituent	Constituent
Discourse-Pragmatic Functions:	New Topic	New Topic	New Topic
	Abandoned Topic	Abandoned Topic	Abandoned Topic
	Self-Repair	Self-Repair	Self-Repair
		Reported Discourse	Reported Discourse
	Undesirable Response	Undesirable Response	Undesirable Response	Undesirable Response
	Concession	Concession	Concession	Concession
	Contrast	Contrast
	Elaboration	Elaboration
	Concluding Remark		Concluding Remark
	Emotional Reaction		Emotional Reaction	Emotional Reaction
	Face-Threat
			Face-Threat Mitigator
	Floor-Holding		Floor-Holding
	Result
	Floor-Ceding

5.3. Kwéyòl Konsa

48Konsa ‘so’ (30 tokens) is derived from French comme ça ‘like that’; when used as a PM, comme ça often collocates with ou ‘or’ to form ou comme ça ‘or like that’ (see Béguelin & Corminboeuf, 2017 for an analysis of ou comme ça). Only one token surfaced alongside another pragmatic element (English alright) and all but one free-standing example (see (18)) were integrated into an utterance in either initial (24 tokens, see (19)), medial (four tokens, see (20)), or final position (one token, see (21)). While this positioning aligns with English so, which is typically utterance-initial but can also surface as utterance-medial, -final, or free, French PM (ou) comme ça tends not to occur in initial position.

(18) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

A:	Èvè	jòdi	mwen	isit-la	èvè	’w,	èvè	dèmen	mon
	and	today	1SG	here-DEF	with	2SG	and	tomorrow	1SG
	‘And today I’m here with you, and tomorrow I

 

	ni	pou	alé,	am,	fizyo.
	have	for	go	erm	physiotherapy
	have to go, erm, to physio[therapy].’

 

B:	Wi,	èvè...
	yes	and
	‘Yes, and...’

 

A:	Konsa...
	PM
	‘So...’

 

B:	Fizyo-la	ké	wédé	’w.
	Physiotherapist-DEF	FUT	help	2SG
	‘The physio will help you.’

(19) (Book Narration, LC, HMMf63, gloss mine)

Sa	pa	bon.	Konsa	mwen	ka	di	’w:	pa	fè	sa	ankò.
that	not	good	PM	1SG	PROG	say	2SG	not	do	that	again
‘That’s not good. So I’m telling you: don’t do that again.’

(20) (Book Narration, LC, HMMf63, gloss mine)

Mé	toutmoun	ka	gadé,	mé	gwo	“frog”-la	pa	enmen	sa,
but	everyone	PROF	look	but	big	frog-DEF	NEG	like	that
‘But everyone is looking, but the big frog does not like that,

 

é	konsa	mon	doubout	èvè	mon	ka	gadé	yo,	pou	vwè	sa
and	PM	1SG	stood	and	1SG	PROG	look	them	for	see	what
and so I stood and I was looking at them, to see what

 

yo	ka	fè.
3PL	PROG	do
they are doing.’

(21) (Interview, CC, gloss and translation mine)

pis	nonm	té	ka	séré	toupatou	an	bwa	toupatou	konsa.
thus	men	ANT	PROG	hide	everywhere	in	tree	everywhere	PM
‘thus the men hid everywhere, in the trees, everywhere so.’

49(Ou) comme ça ‘(or) like that’, has been most closely examined in Swiss French and functions as a hedge or list extension particle indicating uncertainty or approximation (Corminboeuf, 2016; Béguelin & Corminboeuf, 2017). Meanwhile, so highlights inferential, resultative connections (Schiffrin, 1987; Buysse, 2014; Bolden, 2009; Blakemore, 1988; 2002) and (re)launches topics that are “pending” (Bolden, 2008: 306) or otherwise tied to the utterer’s communicative objective. Despite these differences, the two CSL markers overlap functionally. (Ou) comme ça can introduce reported discourse (Béguelin & Corminboeuf, 2017: 13-15) and facilitate self-repairs (Béguelin & Corminboeuf, 2017: 13-15; Corminboeuf, 2016: 9), and my search of the Corpus of Contemporary American English (COCA) uncovered examples of so performing both of these functions as well. Surprisingly, I did not find examples of konsa performing either function in the data.

50When used to “indicat[e] to the hearer that some kind of inferential connection between…two propositions needs to be made” (Bolden, 2009: 976 summarizing Blakemore, 1988, 2002), so often introduces a result or other conclusion that follows from the preceding discourse. However, the portion following the marker can be elided or implicit, leaving it to the addressee to infer the result or conclusion. Inference and elided information are also central to a function of (ou) comme ça, which can be used to express confidence that the addressee is able to infer the rest of an incomplete enumeration or to acknowledge the existence of other relevant examples that the utterer could mention if they chose to provide a list (Béguelin & Corminboeuf, 2017: 11-12).

51This congruent ability to suggest an inferential connection is reflected in Kwéyòl konsa, whose inference-marking function closely resembles so’s. Notice in (22) how konsa indicates that, because the utterer anticipates many people will attend a Kwéyòl cultural event, she will go as a vendor. Like (ou) comme ça, konsa can also be used to mark the elision or intentional incompleteness of a list, as in (21) above; rather than list every place Kalinago men hid from the police, the chief gives an example, emphasizes the men were toupatou ‘everywhere’, and ends his utterance with konsa.

(22) (Dialogue, LC, SMf59 & PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.9fd968h3, 4:50-4:55]

La	ké	ni	anpil	moun	ka	vann	biten.
there	FUT	have	a.lot	person	PROG	sell	thing
‘There will be a lot of people selling things.

 

Konsa,	mwen	ka	’y	fe	’y.
PM	1SG	PROG	3SG	do	3SG
Therefore, I’m going to do it.’

52So and (ou) comme ça are also similar in their floor-holding capabilities (Buysse, 2014: 83; Bolden, 2009: 97 summarizing Local & Walker, 2005). English users employ so to indicate that, despite “a digression or an aside” (Buysse, 2014: 83), they wish to continue speaking. (Ou) comme ça, meanwhile, is a ponctuant ‘puntuator’ (Corminboeuf, 2016: 5); it indicates that the utterer’s contribution is not yet finished by punctuating or segmenting utterances into digestible portions, such as intonational groupings (Corminboeuf, 2016: 10) or crucial chunks of information (Dostie, 2007: 54).

53Konsa displayed both floor-holding functions. In (23), the Kalinago chief uses konsa to return to his central narrative (how the police illegally arrested the wife of a Kalinago merchant) after digressing into an aside. In (24), konsa punctuates the utterer’s narration of the actions of a farmworker in the silent film; the farmworker is tempted to eat a pear but acts with integrity and returns it to the basket.

(23) (Interview, CC, gloss and translation mine)

A:	[…] madanm-la	tousèl	té	la	yo	HANDCUFF	madanm-la
	wife-DEF	alone	ANT	there	3PL	handcuffed	wife-DEF
	‘[…] only the wife was there. They handcuffed the wife.’

 

B:	wé
	yes
	‘Yes’

54 

A:	épi	yo	[…] monté	jik	anho	koté	mouché	JAMES
	and	3PL	climbed	to	up	by	mister	James
	‘And they […] climbed up to Mister James’ place

55 

ola	yo	ka	vann	sé	biten-la	osi	mouché	JAMES
where	3PL	PROG	sell	PL	thing-DEF	also	Mister	James
where they were also selling the merchandise. Mister James,

56 

mouché	Pyè	tousa	té	ka	vann	sé	biten-la	san	lisans
mister	Pierre	all	ANT	PROG	sell	PL	thing-DEF	without	permit
Mister Pierre, all of them were selling the merchandise without a permit,

 

mé	sété	pli	gran	boutik-la	sété	isi
but	was	most	big	store-DEF	was	here
but the biggest shop was here.

 

konsa	yo	HANDCUFF	madanm-la […]
PM	3PL	handcuffed	wife-DEF
So they handcuffed the wife…’

(24) (Silent Film Discussion, LC, EDf82 & HMMf63, gloss mine)

A:	Konsa,	i	té	vlé	yonn	di	yo.
	PM	3SG	ANT	want	one	of	3PL
	‘So, he wanted one of them.

 

Konsa,	i	té	ka	alé	nétwayé	pou	manjé	’y.
PM	3SG	ANT	PROG	go	clean	for	eat	3SG
So, he was going to clean to eat it.’

 

B:	Oké.
	okay
	‘Okay.’

 

A:	I	fè	konmsidi	i	té…,	moun-la	té	ni	an
	3SG	acted	like	3SG	ANT	person-DEF	ANT	have	INDF
	‘He acted as if he had..., the person had a

 

“conscience”.	Konsa	i	di	i	pa	ka	’y	pwan	’y.
conscience	PM	3SG	said	3SG	NEG	PROG	3SG	take	3SG
conscience. So he said he’s not going to take it.

 

B:	Oké.
	okay
	‘Okay.’

 

A:	Konsa,	i	mété	’y	viwé	andidan	“basket”-la,	èvè	i	viwé	twavay.
	PM	3SG	put	3SG	back	in	basket-DEF	and	3SG	returned	work
	‘So, he put it back in the basket, and he went back to work.’

57(Ou) comme ça also performs functions that are not reported in the literature for so, such as indicating that a quantity or figure is approximate or imprecise (Béguelin & Corminboeuf, 2017: 10) or introducing a request for confirmation (Corminboeuf, 2016: 4). It can even stand alone as an evasive, uncooperative response (Cormbinboeuf, 2016: 5). Of these three functions, I found Kwéyòl konsa to perform the second. In (25), Interlocutor A uses konsa to begin her request for confirmation that her inference is correct and that Interlocutor B is not going to work.

(25) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

A:	Bonjou.
	good.day
	‘Hello’

 

B:	Bonjou	manm’.	Sa	ka	fèt?
	good.day	mother	What	PROG	happen
	‘Hello mother. What’s happening?’

 

A:	Sa	’w	ka	fè	jòdi?
	what	2G	PROG	do	today
	‘What are you doing today?’

 

B:	Mon	menm	pa	menm	konnèt,	non.
	1SG	myself	NEG	even	know,	no
	‘I myself don’t even know, no.’

 

A:	Konsa,	ou	pa	ka,	am,	twavay?
	PM	2SG	NEG	PROG	um	work
	‘So, you’re not going to work?’

 

B:	Wi,	pli	ta.
	yes	more	late
	‘Yes, later.’

 

B:	Mon	menm,	mwen	pa	ka	twavay	jòdi	pis...
	1SG	self	1SG	not	PROG	work	today	because
	‘I myself, I am not working today because…’

 

A:	Ou	pa	byen.
	2SG	NEG	well
	‘You are not well.’

 

B:	Mon	pa	byen.
	1sg	NEG	well
	‘I am not well.’

58A similar but distinct function of so that konsa also performs is acting as a “marker of connection” (Howe, 1991: 93 cited by Bolden, 2008: 306). So suggests familiarity or common ground between interlocutors, introducing some aspect of their shared knowledge that “highlight[s] the speaker’s involvement in the addressee’s life world” (Bolden, 2006 cited by Bolden, 2008: 306). In (25) above, Interlocutor A asking about B’s work plans is rooted in the interlocutors’ shared knowledge that B has not been well lately.

59So also launches new or abandoned topics, particularly ones that are central to the utterer’s agenda (Bolden, 2009: 996; Bolden, 2008: 3012), and can be “deployed as a stand-alone unit to prompt the addressee to produce the next relevant action” (Bolden, 2008: 306 summarizing Raymond, 2004), such as taking their turn to speak (Buysse, 2014: 31). Kwéyòl konsa only resembled so’s agenda-(re)launching uses when employed as a floor-holding device that transitioned the discourse back to the central topic (see (23)). However, konsa does perform the “turn-transition prompt” function reported by Buysse (2014: 30) for so. In (18) above, a Kwéyòl user utters konsa then trails off, inviting the addressee to speak.

60Konsa’s orientation is usually forward/backward given that, like so, it highlights inferred connections between preceding and subsequent content. Possible exceptions are when it indicates the incompleteness of a list like French (ou) comme ça (backward orientation) or prompts a turn transition like so (forward orientation). Except when konsa indicates incompleteness of a list and thus modifies only a portion of the utterance, this marker, like so, takes within its scope the entirety of the propositional content of the utterance(s) adjacent to it. For example, when marking an inferential connection, konsa indicates that the full content of the following utterance (the conclusion, result, outcome, etc.) can be inferred from preceding content.

5.4. So in Kwéyòl

61So (39 tokens) featured prominently in the corpus data. Ten tokens surfaced alongside other pragmatic elements, including bon ‘well’, oké ‘okay’, èben ‘well’, wi ‘yes’, and apwézan ‘presently’. So was almost always integrated into a larger utterance whose entire content it modified (37 tokens) and was either utterance-initial (30 tokens, see (26)), -medial (seven tokens, see (27)), or free-standing (two tokens, see (28)).

(26) (Dialogue, LC, SMF59 & PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.9fd968h3, 3:57-4:00]

“So”	mwen	ni	pou	wèsté	yonn	koté.
PM	1SG	have	for	stay	one	place
‘So I have to stay in one place.’

(27) (Book Narration, LC, EDf82, gloss mine)

Sa	ki	ka	tonbé	èvè	fè	malonnèt-la,	i	ka	swiv	yo
that	which	PROG	fall	and	do	ungrateful-DEF	3SG	PROG	follow	3PL
‘The one that is falling and the ungrateful one, it is following them

 

apwézan,	“so”,	apwézan,	tibway-la	ni	chyen	’y,
now	PM	now	boy-DEF	has	dog	3SG.POSS
now, so, right now the boy has his dog,

 

“tortoise”-li,	èvè	dé	“frog”-la,	èvè	yo	pati.
tortoise-3SG.POSS	and	two	frog-DEF	and	3PL	left
his tortoise, and the two frogs, and they left.’

(28) (Book Narration, LC, PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.04e5ltc7, 2:09-2:21]

I	ni,	i	ka	asiz	kon	i	fè	yon	biten	èvè	yo
3SG	has	3SG	PROG	sit	like	3SG	did	INDF	thing	and	3PL
‘It has, it is sitting like it did something and they are

 

pa	kontan	sa.	“So”.
NEG	happy	that	PM
not happy about that. So.’

62Some of the functions so performed aligned with those reported in the literature on English so (see Section 5.3). It often indicated inferential connections between propositions, as it does in (29) between the fact that both Interlocutor A and her friend D have a background in jewelry-making and their decision to attend an event as vendors. Thus, the tokens of so in the data were usually oriented forward and backward simultaneously. So was also used once with a forward orientation to raise a new topic; see (30), in which Interlocutor A introduces a new subject after a lull.

(29) (Dialogue, LC, SMf59 & PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.9fd968h3, 1:15-1:30]

A:	Am,	mé	zanmi-mwen	D,	i	té	ka	vann,	am,	tibwen
	Ah	but	friend-1SG.POSS	D	3SG	ANT	PROG	sell	erm	some
	‘Ah, but my friend D, she was selling, erm, some

 

A:	bijou	ki	i	té	kwéyé	i	menm	akay-li.
	jewelry	which	3SG	ANT	make	3SG	self	house-3SG.POSS
	jewelry which she made herself at home.

 

A:	I	té	fè	yon	kous	andan	kouman	pou	kwéyé	bijou...
	1SG	ANT	do	INDF	course	on	how	to	make	jewelry
	I did a course on how to create jewelry...’

 

B:	A!	bijou?
	Ah	jewellery
	‘Ah! Jewellery?’

 

A:

Èvè

’y

té

ka

vann

yo,

“so”

nou

té

alé

asanm

pou

fè

sa.

and

3SG

ANT

PROG

sell

3PL

PM

2PL

ANT

go

together

for

do

that

‘And she was selling them, so we went together to do that.’

(30) (Dialogue, LC, EDf82 & HMMf6, gloss mine)

A:	Oké.
	okay
	‘Okay.’

 

B:	Sa	dòt ...
	what	else
	‘What more…’

 

A:	Oké,	“so”	C	lakay?
	okay	PM	C	home
	‘OK, so is C at home?’

 

B:	Wi.	O!	C	ba	mon	yon,	an	“phone number”	pou...
	yes	oh	C	gave	1SG	INDF	INDF	phone number	to
	‘Yes. O! C gave me a, a phone number for...’

 

A:	Pou	ba	mwen.
	for	give	1SG
	‘To give me.’

 

B:	Pou	ba	’w.	“Yeah”,	am,	kont	plas-la	ou	té	vlé.
	for	give	2SG	yeah	erm	about	place-DEF	2SG	ANT	want
	‘To give you. Yeah, erm, about the place you wanted.’

 

A:	Oké.	“So”,	mwen	ni	pou	kwiyé	yo?
	okay	PM	1SG	have	for	call	3PL
	‘OK. So, I have to call them?’

 

B:	Wi. […]
	yes
	‘Yes. […]’

63So was used in contexts that hinged upon shared knowledge and common ground between the interlocutors (a function documented in the literature). However, often these tokens also introduced implicit or explicit requests for confirmation that the utterer and addressee had drawn the same inferential connection (a function of konsa and its etymon (ou) comme ça). In (30) above and in (31), so introduces utterances that index the interlocutors’ shared knowledge: that C was supposed to pass along a phone number to Interlocutor B for Interlocutor A to call about an event venue in (30) and who all was expected to attend an event on Sunday in (31). In (30), so also introduces questions that explicitly request a response; in (31), so introduces a statement, and the request for confirmation is more implicit.

(31) (Dialogue, LC, SLm82 & FMLf80, gloss mine)

A:	Mé	nou	ké	wè	’y	dimanch.
	but	1PL	FUT	see	3SG	Sunday
	‘But we will see him on Sunday.’

 

B:	Nou	ké	wè	’y	dimanch.
	1PL	FUT	see	3SG	Sunday
	‘We will see him on Sunday.’

 

A:	I	ké	vini	dimanch	“because”	lé	dimanch	sa	sé	tan
	3SG	FUT	come	Sunday	because	on	Sunday	that	is	time
	‘He will come on Sunday because Sundays are the times

 

i	ka	vini	ési.
3SG	PROG	come	here
he comes here.’

 

B:	Wi,	wi,	wi.
	yes	yes	yes
	‘Yes, yes, yes.’

 

A:	So,	sé	A	nou	ni	pou	(tèlifonn)	apwézan.
	PM	is	A	1PL	have	for	call	now
	‘So, it’s A we have to telephone now.’

 

B:	A.
	A
	‘A.’

64So also performed another function reported for English so: maintaining control of the floor by relaunching the utterer’s topic of interest after a digression or interruption. However, it also displayed the ‘punctuator’ floor-holding pattern performed by konsa and its etymon (ou) comme ça. Notice in (32) below how the utterer marks each juncture in his account with so (and later èvè ‘and’), indicating that there is more content to come and demarcating crucial chunks of information.

(32) (Dialogue, LC, SLm82 & FMLf80, gloss mine)

I	di	la	ni	fonmi	an	kay-la,
3SG	said	there	are	ant	in	house-DEF
‘She said there are ants in the house,

 

“so”	mon	di	pou	di	R	pou	nétwayé	’y.
PM	1SG	said	for	tell	R	for	clean	3SG
so I said to tell R to clean it.

 

“So”	mon	ka	èspéwé	lè	’y	ké	(tèlifonn)	ankò	pou	sav
PM	1SG	PROG	wait	when	3SG	FUT	telephone	again	for	know
So I am waiting for when she will telephone again to know

 

sa	ki	fèt.	Èvè	yè,	R	(tèlifonn)	nou.	“Is	it?”
that	which	happened	and	yesterday	R	telephoned	2PL	is	it
what has happened. And yesterday, R telephoned us. Is it?’

65The results of this analysis of konsa ‘so’ and so in the Kwéyòl data sources are summarized in the table below.

Tab. 3, Property comparison table for Kwéyòl konsa ‘so’ and English so as they arise in the Kwéyòl data sources analyzed, as well as for French (ou) comme ça ‘(or) like that’ and English so as discussed in the literature

	KONSA	(OU) COMME ÇA	SO	SO (in corpora)
Utterance Placement:	Initial	Initial	Initial	Initial
	Medial	Medial	Medial	Medial
	Final	Final	Final
	Free	Free	Free	Free
Degree of Integration:	Integrated	Integrated	Integrated	Integrated
	Free	Free	Free	Free
Orientation:	Backward	Backward	Backward	Backward
	Forward/Backward	Forward/Backward	Forward/Backward	Forward/Backward
Scope:	Full Proposition	Full Proposition	Full Proposition	Full Proposition
	Constituent	Constituent	Constituent	Constituent
Discourse-Pragmatic Functions:		Reported Discourse	Reported Discourse
		Self-Repair	Self-Repair
	Inference		(Elided) Inference	Inference
	Elided or Incomplete List	Elided or Incomplete List
	Floor-Holding		Floor-Holding	Floor-Holding
	Punctuator	Punctuator		Punctuator
		Approximation
	Confirmation Request	Confirmation Request		Confirmation Request
		Uncooperative Response
			Common Ground	Common Ground
			New Topic	New Topic
	Abandoned Topic		Abandoned Topic
			Agenda Launch
	Turn-Transition Prompt		Turn-Transition Prompt

5.5. Kwéyòl Papa/Bondyé

66Bondyé ‘God’ and papa ‘father/God’ were relatively infrequent: just three tokens of Bondyé ‘God’ and six of papa ‘father/God’. These markers sometimes collocated with other pragmatic elements, such as é(la) ‘and/ah’, a ‘ah’, non ‘no’, and wé ‘yes’, and were either free-standing (see (33)) or utterance-initial (see (34)). This aligns with similar markers in French, like mon Dieu ‘my God’, but contrasts with English g-words like oh my God, which can assume any position.

(33) (Dialogue, LC, SMf59 & PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.9fd968h3, 4:02-4:12]

La	ni	bèl	sòlèy.	Éla	papa!
there	is	beautiful	sun	ah	PM
‘There is beautiful sunshine. Ah papa!

 

Kon	mwen	Donmnik,	èvè	mwen	ni	pou	alé	dèwò,	pou	mété
like	1SG	Dominica	and	1SG	have	for	go	outside	for	put
As if I’m in Dominica, and I have to go outside, to put

 

sòlèy	asi	vijay-mwen.
sun	on	face-1SG.POSS
some sunshine on my face.’

(34) (Book Narration, LC, HMMf63, gloss mine)

A!	Bondyé,	Bondyé,	kité	nou	sòti.
ah	PM	PM	let	1PL	leave
‘Ah! God, God, let us get out.’

67Bondyé ‘God’ is derived from French bon Dieu ‘good God’. Papa is a familiar, informal way of saying father in both French and English, but Dominica was colonized by Roman Catholic and Protestant Christians, and expressions like Father God and Papa God are also used in prayer in many Christian traditions. This is reflected in a footnote by the LC transcriber/translator: “Although the word “papa” means “father”, it is often used as an exclamation in all sorts of situations where it does not mean “father”, to give a phrase more emphasis. Sometimes the word “papa” is used to refer to “God”. For example: “Wi papa”, or “Wi Papa Bondyé” (i.e. “Yes, God the Father”).”

68Phrases like bon Dieu ‘good God’ (Bondyé’s etymon) and nom de Dieu ‘name of God’ may still be perceived as blasphemous by French users in secular contexts (Olivier, 2000: 163). However, PMs of religious origins in both CSLs, including English oh my God and gosh and French mon Dieu ‘my God’, are used to express emotional reactions (frustration, surprise, apprehension, etc.), realization, and the processing of new information, as well as emotional involvement in the form of concern or emotional attachment (Tagliamonte & Jankowski, 2019: 214 for oh my God; Downing & Caro, 2019: 101-3 about gosh; Olivier, 2000: 171 about mon Dieu). These functions were displayed by Bondyé and papa. In (35), papa is repeated, expressing Interlocutor B’s surprise and concern upon realizing that, though a group of children trapped in a cave had been found alive, the rescue process would be arduous.

(35) (Dialogue, LC, SMf59 & PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.9fd968h3, 7:56-8:27]

A:	Yon	lòt	biten	mwen	tann.	É,	yo	tapé,	am,	biten;
	INDF	other	thing	1SG	heard	and	3PL	found	erm	thing
	‘Another thing I heard. And, they found, erm,

 

A:	yo	alé	an	kav,	“cave”-la.	Yo,	yo,	am,	sa	yo	di?
	3PL	went	in	cave	cave-DEF	3PL	3PL	erm	what	3PL	say
	they went into cave, the cave. They, they, erm, what did they say?’

 

B:	O!	Sé	zanfan-a?
	oh	PL	child-DEF
	‘Oh! The children?’

 

A:	Zanfan-a.	Ki	té	dispawèt.
	children-DEF	who	ANT	disappear
	‘The children. Who disappeared.’

 

B:	Wi,	wi.	É	sa	sé	bon!
	yes	yes	and	that	is	good
	‘Yes, yes. And that’s good!’

 

A:	Mé	yo	la	toujou,	en!
	but	3PL	there	still	eh
	‘But they are still there, eh.’

 

B:	Wi,	sé	sa	yo	di.
	yes	is	that	3PL	said
	‘Yes, that what they said.’

 

A:	Yo	ni	pou	enstwi	yo.
	3PL	have	for	instruct	3PL
	‘They have to instruct them.’

 

 

A:	Wi.	Yo	ké	la	pou	tibwen	tan	toujou.
	yes	3PL	FUT	there	for	some	time	still
	‘Yes. They will still be there for some time.

 

A:	Yo	ni	pou	enstwi	yo	pou	plonjé.
	3PL	have	for	instruct	3PL	for	dive
	They have to instruct them to dive.’

69The euphemism gosh has a variety of other functions, all of which I found oh my God to also perform when exploring the COCA: facilitating self-repairs and reformulations (Downing & Caro, 2019: 107), “function[ing] as a narrative device used to punctuate the story, to effect topic continuation, ...[and] to alert the listener that the speaker does not intend to give up the floor”, and introducing reported discourse (Downing & Caro, 2019: 106). However, the only function among these that was performed by Bondyé and papa was expressing emphasis (Downing & Caro, 2019: 104-105). In (36), papa emphasizes that Interlocutor B truly cannot recall doing much of anything on Sunday and introduces the reason why: her broken arm. Similarly, above in (33), papa emphasizes the utterer’s appreciation of the sunshine and the nostalgic memories it brings her of life in Dominica. Like gosh, which emphasizes “evaluatives…[like] copular constructions… rhetorical questions…declaratives with question tags…and ‘literal’ questions” (Downing & Caro, 2019: 104-5), Bondyé in (37) underscores the utterer’s evaluation of one of the frogs in the picture book. She calls it an kalité ‘a type’, by which she seems to be referring to its mischievous, surprising behavior (i.e., ‘a [mischievous/strange] type [of thing]’).

(36) (Dialogue, LC, SMf59 & PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.9fd968h3, 2:50-2:58]

A:	Kisa	ou	té	fè	asou	dimanch?
	what	2PL	ANT	do	on	Sunday
	‘What did you do on Sunday?’

 

B:	Mwen	pa	sa	menm	chonjé;
	1SG	not	can	even	remember
	‘I can’t even remember;

 

non	papa,	“‘cause”	mwen	ni	yon	lanmen	la	ki	kasé	la.
no	PM	because	1SG	have	INDF	hand	there	which	broken	there
no papa, because I have a broken hand there which is broken there

 

Mwen	pa	sa	fè	anyen ...
1SG	not	can	do	anything
I can’t even do anything.’

(37) (Book Narration, LC, EDf82, gloss mine)

A!	Bondyé,	“frog”	sala	menm	sé	an,	an,	an	kalité,	pis
ah	PM	frog	DEM	self	is	INDF	INDF	INDF	type	because
‘Ah! God, that frog itself is a, a, a type, because

 

mi	“frog”-la	tonbé	ankò.	Mi	’y	ka	vòltijé.
look	frog-DEF	fell	again	look	3SG	PROG	fly
look the frog has fallen again. Look it went flying.’

70French markers like mon Dieu ‘my God’ can be used mockingly or sarcastically to suggest that the addressee is sensationalizing something (Olivier, 2000: 170-1), to correct the addressee or suggest that they modify their behavior (Olivier, 2000: 171), and to highlight that the utterer is undergoing an exceptional circumstance (Olivier, 2000: 167). Of these three functions, the Kwéyòl PMs take on the latter two: Correction and Exceptional Circumstance. In (38), Interlocutor A chides Interlocutor B, reminding B that she should not talk to Dominicans about rain. In (34) above, Bondyé highlights an exceptional circumstance that could warrant (but does not literally involve) calling on God. There, the utterer narrates the distress of a little frog in the picture book who is trapped inside a box.

(38) (Dialogue, LC, SMf59 & PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.9fd968h3, 7:23-7:39]

A:	Lè	’w	wè	sòlèy,	sòlèy,	sòlèy,	lapli	dèyè	’y	wi.
	when	2SG	see	sun	sun	sun	rain	behind	3SG	yes
	‘When you see sun, sun, sun, rain is behind it, yes.’

 

B:	Mwen	sav,	mwen	sav.	Èvè	Donmnitjen	pè	lapli,	wi.
	1SG	know	1SG	know	and	Dominican	afraid	rain	yes
	‘I know, I know. And Dominicans are afraid of rain, yes.’

 

A:	É,	papa!
	and	PM
	‘And, papa!’

 

A:	Pa	di	Donmnitjen	kont	lapli,	kon	mwen	menm	pè,
	not	say	Dominicans	about	rain	like	1SG	self	afraid
	‘Don’t tell Dominicans about rain, as I myself am afraid,

 

si	mwen	té	an,	(Donmnitjen).
if	1SG	ANT	in	Dominica
if I was in Dominica.’

71Like their CSL counterparts, papa/Bondyé ‘father/God’ can have a forward orientation (e.g., when introducing a correction of the addressee’s behavior), or a backward orientation (e.g., when expressing an emotional reaction). Their orientation can even be simultaneously forward/ backward; see (37) above where papa emphasizes both the utterer’s evaluation of the sunshine (backward) and the nostalgia it brings her (forward). Regardless, the scope of these markers and their CSL counterparts includes the entire content of the utterance(s) they modify. When expressing an emotional reaction, their scope can even extend extralinguistically to encompass the entire discourse situation. The results of this analysis of papa/Bondyé ‘father/God’ are summarized in the table below.

Tab. 4, Property comparison table for Kwéyòl papa ‘father (God)’ and Bondyé ‘God’ as they arise in the Kwéyòl data sources analyzed, as well as for French religious markers like mon Dieu ‘my God’ and English religious markers like oh my God as discussed in the literature

	PAPA/BONDYÉ	MON DIEU, etc.	OH MY GOD, etc.
Utterance Placement:	Initial	Initial	Initial
			Medial
			Final
	Free	Free	Free
Degree of Integration:	Integrated	Integrated	Integrated
	Free	Free	Free
Orientation:	Forward	Forward	Forward
	Backward	Backward	Backward
	Forward/Backward
Scope:	Full Proposition	Full Proposition	Full Proposition
	Situational Context	Situational Context	Situational Context
Discourse-Pragmatic Functions:	Emotional Reaction	Emotional Reaction	Emotional Reaction
	Emotional Involvement	Emotional Involvement	Emotional Involvement
	Realization	Realization	Realization
			Self-Repair
			Punctuator Floor-Holding
			Reported Discourse
	Emphasis		Emphasis
		Mocking of Overreaction
	Behavior Correction	Behavior Correction
	Exceptional Circumstance	Exceptional Circumstance

5.6. Kwéyòl La

72Locative discourse markers, which I call locative pragmatic markers (LPMs) for consistency with the terminology used in this study, are “locative deictic elements…which under certain contextual conditions…function as discourse markers” (Fellegy, 1998: 31). I analyzed tokens of la ‘there’ to determine whether it functions as an LPM in Kwéyòl as it does in neighboring Creoles (see Section 4.2).

73In her work on Quebec French, Dostie (2007: 50-52) reports that la’s French etymon, là ‘there’, can be used as a PM, sometimes cooccurring with temporal or locative là (see (39)).

(39) (Dostie, 2007: 56, gloss and translation mine)

Vous	allez	vous	asseoir	là	là.
2PL	will	REFL	sit	PM	PM
‘You will sit LOC-there PM-there’

74In English, Schiffrin (1987: 328) notes that the PMs there and here are “often used in narratives to mark surprising outcomes in the complicating action”. The standardized English example below is provided by Fellegy (1998: 61), whose study investigates locative PMs in New Ulm English, a German-influenced variety spoken in Minnesota.

(40) (Fellegy, 1998: 61)

We looked and looked and, here, the dog had buried the keys!

75Examples (41) and (42) are New Ulm examples of the PM dere ‘there’. Like all PMs, dere ‘there’ does not contribute to the utterance’s propositional content. Instead, its contribution is to “draw attention to the speaker’s perceived relationship to the topic and [to] focus attention on the topic itself” (Fellegy, 1998: 62). In (41), the utterer has already specified the farm in question by adding of yours. The LPM provides “meaningful social information” (Fellegy, 1998: 63): the utterer — located in St. Paul — “has positioned himself at a discourse-level inside the town of New Ulm” (Fellegy, 1998: 64), a setting familiar to both interlocutors. In (42), “no place information has been or ever is introduced into the story” (Fellegy, 1998: 64). Rather, dere situates the story within “a private location…which is ultimately kept as the utterer’s personal information” (Fellegy, 1998: 65); it is “a subtle distancing device, used by speakers when they do not wish to take the listener to a specific locale, in these particular cases because the location was personally unpleasant” (Fellegy, 1998: 65). LPMs can also point back to earlier chunks of discourse, usually to correct, contradict, or comment on them, as seen in (43).

(41) (Fellegy, 1998: 63)

It [the road] goes down to Cxx and ya drive past the farm of yours up dere.

(42) (Fellegy, 1998: 64)

It was all grammar and he had a test everyday’n I could get just about a hunnert every day, ya know, and I couldn’t answer that god darn gal down dere.

(43) (COCA, Spoken)

A:	This must be a bar mitzvah outfit here?
B:	No, no, this isn't- That- You're wrong there, Lou. This is a great satin woolback coat that you could wear during the day. Shine is in. It's one of the big important trends.

76Of these, the focus-marking (Forget, 1989: 63-64) and discourse deictic commentary functions (Dostie, 2007: 52) were also reported for the French locative PM là, which often focuses the addressee’s attention on information the utterer deems most critical. Là can also foreground a main topic before the utterer provides additional relevant details or reinforce the importance of an action requested by the utterer (Forget, 1989: 65-66). The latter is illustrated in (39) above, in which the utterer urges the addressee to sit. Là even acts as a floor-holding punctuator (Dostie, 2007: 54; Forget, 1989: 62).

77There were six tokens of Kwéyòl la ‘there’ in the data that were neither post-nominal determiners, nor part of existential constructions, nor contributing locative information. All six were utterance-final or -medial. Like its CSL counterparts, these tokens of la brought the addressee’s attention to information the utterer deemed critical. Sometimes that information was a referent that was part of the interlocutors’ shared knowledge: both interlocutors know the man who had recently died in (11) above, as well as the area surrounding the house the utterer shares with her husband in (44). Note that (44) was not uttered in the said house; la does not reference a location within the situational context.

(44) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

Mé	i	ka	sanm,	ka	wété	pa	dèyè	nou	la.
but	3SG	PROG	seem	PROG	live	by	behind	1PL	PM
‘But he seems, to be living behind us there.’

78La also highlighted the utterer’s relationship with or connection to a critical referent, like the first la in (45) that foregrounds the utterer’s hand. In (45), la even displays its etymon’s punctuating and topic detachment functions, segmenting each piece of critical information: the hand (the topic) and its broken state that prevents the utterer from her usual activities (the additional details). Example (46) shows la expressing an emotional reaction (here, frustration) and referring back to / commenting on something uttered previously.

(45) (Dialogue, LC, SMf59 & PJf58, gloss mine) [https://nakala.fr/​10.34847/​nkl.9fd968h3, 2:50-2:58]

Mwen	pa	sa	menm	chonjé;
I	NEG	can	even	remember
‘I can’t even remember;

 

non	papa,	“‘cause”	mwen	ni	yon	lanmen	la	ki	kasé	la.
no	papa	because	1SG	have	INDF	hand	PM	which	broken	PM
no papa, because I have a broken hand there which is broken there

 

Mwen	pa	sa	fè	anyen...
1SG	not	can	do	anything
I can’t even do anything…’

(46) (Dialogue, LC, EDf82 & HMMf63, gloss mine)

A:	Ou	pa	sa	maché	si	’w	pou	alé	anba	la.
	2SG	not	can	walk	if	2SG	for	go	down	there
	‘You can’t walk if you must go down there.’

 

B:	Non.	Janm-mon	ka	fè	mon	mal.
	no	leg-1SG.POSS	PROG	do	1SG	bad
	‘No. My leg is hurting.’

 

A:	Èvè	sé	sa	menm	mon	ka	di	’w	la,	“ennit”?
	and	is	that	same	1SG	PROG	tell	2SG	PM	ennit
	‘And that same thing that I am telling you there, “ennit”?’

79Finally, there was one token (see (2) above) that the literature on la’s CSL counterparts does not account for. Here, instead of facilitating floor-holding, la cedes the floor as the utterer ends her narration.

80La’s CSL counterparts can be oriented forward or backward; Kwéyòl la, however, consistently modified preceding content (backward orientation). Sometimes the content within la’s scope was an entire proposition, like in (44), but as demonstrated by the punctuating examples in (45), la can also bring the addressee’s attention to specific portions of a proposition. The results of this analysis of la ‘there’ as an LPM are summarized in the table below.

Tab. 5, Property comparison table for Kwéyòl la ‘there’ as it arises in the Kwéyòl data sources analyzed, as well as for French là ‘there’ and English here/there as discussed in the literature

	LA	LÀ	HERE/THERE
Utterance Placement:		Initial	Initial
	Medial	Medial
	Final	Final	Final
Degree of Integration:	Integrated	Integrated	Integrated
	Free	Free	Free
Orientation:	Forward	Forward	Forward
	Backward	Backward	Backward
Scope:	Full Proposition	Full Proposition	Full Proposition
	Constituent	Constituent	Constituent
Discourse-Pragmatic Functions:	Shared Knowledge		Shared Knowledge
	Relationship Focus	Relationship Focus	Relationship Focus
	Emotional Reaction		Emotional Reaction
			Location Distancing
	Discourse Deixis with Commentary	Discourse Deixis with Commentary	Discourse Deixis with Commentary
	Topic Foregrounding	Topic Foregrounding
		Request Reinforcement
	Punctuator	Punctuator
	Floor-Ceding

6. Discussion

6.1. Congruence & Functional Adaptation

81Based on these results, congruence likely played a role in the development of èben ‘well’: it displayed several properties that French (eh) ben ‘well’ shares with English well. Its functional inventory also extends beyond its French etymon, incorporating three more functions of well and two functions not documented for either CSL marker (Result and Floor-Ceding). Overall, èben displays functional broadening: it has gained new functions not performed by its French etymon, perhaps under well’s influence.

82By contrast, while the distributional features of konsa ‘so’ reflect congruent properties shared by its CSL counterparts, neither of the CSL counterparts’ shared functions was performed by the Kwéyòl PM. Instead, konsa’s inventory includes three functions unique to French (ou) comme ça ‘(or) like that’ and four of so’s other functions. Thus, konsa exhibits functional shift: though it may not perform some of its French etymon’s functions, it has gained multiple functions of English so.

83The distributional features of papa/Bondyé ‘father/God’, too, generally aligned with those documented for similar markers of religious origin in Kwéyòl’s CSLs. In addition to the three congruent functions shared by their CSL counterparts and two functions documented for French markers like mon Dieu ‘my God’, these PMs have broadened their inventory to include the Emphasis function of English g-words like oh my God.

84Similarly, most of the distributional features of la ‘there’, which is indeed used as an LPM in Kwéyòl, are congruent across its CSL counterparts. Its functional inventory, too, capitalizes on congruence, encompassing both of the functions shared by its CSL counterparts. Of la’s remaining functions, two are documented for English here/there, two are reported for French là, and one (Floor-Ceding) was not associated with either CSL counterpart. Thus, la, too, exhibits functional broadening compared with its French etymon.

85The properties of the tokens of well that surfaced in the data were all documented for English well in the literature; notably, those properties were all congruent traits also displayed by well’s Kwéyòl counterpart èben. Likewise, the distributional features of the tokens of so in the Kwéyòl data, as well as two of the functions it performed, paralleled both the English literature and the properties of its Kwéyòl counterpart konsa. However, so also performed four more functions in the Creole data, two of which are part of konsa’s inventory but are not documented for English so. Thus, so has a broader functional inventory in the Creole, perhaps due to contact with konsa.

6.2. The Potential Impact of Core Meanings

86Why a Kwéyòl PM may perform some of its CSL counterparts’ functions and not all is a complex question that requires further research. In the case of konsa ‘so’, there was even a pair of congruent functions shared by its CSL counterparts that konsa did not perform in the data. Recall from Sections 3.2 and 2.2 that integration of an SL pragmatic item into an RL often involves adaptation rather than a complete transfer of all the SL item’s traits, and even though congruence may probabilistically favor certain linguistic properties for integration into a Creole, congruence as described by Baptista (2020) is a non-deterministic mechanism operating alongside other linguistic processes and social factors.

87However, future research investigating what the Kwéyòl PMs’ core meanings are and how their semantic networks compare with those of their CSL counterparts may prove enlightening. Recall from Section 3.1 that a PM’s abstract core meanings permit the creation of new functions. However, they may still influence the breadth of functions a PM performs. For example, perhaps a Kwéyòl PM inherited core meanings from its French etymon earlier in the Creole’s development that are not compatible with some of its English counterpart’s functions.

7. Conclusions

88The aim of this study was to gain insight into language contact and Creole emergence at the discourse-pragmatic level by comparing the properties of a selection of PMs in Kwéyòl, an understudied Creole variety, with those of their counterparts in French, Kwéyòl’s lexifier CSL, and English, its non-lexifier CSL. The results revealed a tendency towards functional adaptation (specifically, broadening or shift) as CSL PMs were integrated into the Creole, as well as potential sites of English influence on the properties of Kwéyòl markers with French etyma. The results also highlighted several cases where Kwéyòl PMs of French origin displayed congruent properties shared by their French and English counterparts, and tokens of so and well that surfaced in the data reflected congruencies between how these PMs are used in English and how their Creole counterparts konsa ‘so’ and èben ‘well’ function in Kwéyòl. These outcomes suggest that both lexifier and non-lexifier CSLs influence the properties of a Creole’s PMs and that congruence plays a key role in how their contributions are integrated into a Creole language at the discourse-pragmatic level.