Cultivating Ambiguity: Notes on Issues of Complexity in Creole Discourse

ABSTRACT

This paper is a rejoinder to the dialogue concerning complexity and simplicity in pidgin and creole grammars. My position takes a chronotopic and unfinalized view of grammar, Bakhtinian notions that give crucial importance to temporal and spatial contexts in discussions about linguistic development-expansion and historicity of language use. I take a modular perspective and focus on a pragmatic component of grammar: ambiguity in discourse. Linguistic utterances that contain the phrasal verbs rip off and hot up are analyzed in the present study for their ambiguous¹, double-voicing features. My aim is to underscore the importance of recognizing double-voicing as a complex discursive strategy in Afro-Caribbean creole grammar. This feature and other non-salient, zero-marked constructions remain difficult to account for using current metrics of complexity versus simplicity. I invoke insights from Bakhtin and his theory of dialogism in the hope that it can aid our analysis of plurilingualism and the cultivation of ambiguity in creole discourse.

KEYWORDS:
Creole discourse; Ambiguity; Simplicity and complexity

RESUMO

Este artigo é uma resposta ao diálogo relativo à complexidade e simplicidade em gramáticas pidgin e crioulas. Minha abordagem requer uma visão cronotópica e não finalizada da gramática, bem como noções bakhtinianas que dão importância crucial aos contextos temporais e espaciais nas discussões sobre a expansão, o desenvolvimento linguístico e a historicidade do uso da linguagem. Assumo, nesta pesquisa, uma perspectiva modular, cujo foco recai sobre um componente pragmático da gramática: a ambiguidade no discurso. No presente estudo, enunciados linguísticos que contêm os phrasal verbs¹ "rip off" e "hot up" são analisados por suas características ambíguas² e bivocais. Meu objetivo é ressaltar a importância de se reconhecer a bivocalidade como uma estratégia discursiva complexa na gramática crioula afro-caribenha. Essa característica e outras construções não marcadas, ou marcadas com zero, são difíceis de se justificar utilizando as métricas atuais de complexidade versus simplicidade. Invoco, aqui, ideias de Bakhtin e sua teoria do dialogismo, na esperança de auxiliar em nossa análise do plurilinguismo e do cultivo da ambiguidade no discurso crioulo.

PALAVRAS-CHAVE:
Discurso crioulo; Ambiguidade; Simplicidade e complexidade

Introduction

Pidgins have impoverished morphological systems and simple syntactic properties. The previous statement is meant to have a neutral reading; it cannot be interpreted neutrally, however, for the word impoverished has a resultative, change of state meaning: the language was once rich in morphology, but its current state displays a diminished coding system. Impoverished and simple are words that carry "contextual overtones" and are understood relative to previous states (BAKHTIN, 1981, p.278). When a pidgin is framed in this way, readers are forced to take one of the following positions: (i) pidgin is characterized relative to itself; this means that the language predates the contact scenario from which it emerged and in its previous state it had more complex coding strategies than we find after the contact scenario, or (ii) pidgin is characterized relative to some other set(s) of languages. Position (i) is impossible, since a pidgin does not exist until language contact occurs. The use of impoverished, then, can only be understood relative to languages in contact. Although there is no mention of input languages in the description of pidgins that appears in the first line of this paper, the reader is still forced to understand pidgin relative to some other language(s). The debate is framed for us when we accept characterizations like impoverished, reduced, and simple in descriptions of pidgins and creoles.

Current Discourse on Complexity and Simplicity in Pidgins and Creoles

At the fore of discussions about complexity and simplicity in pidgins and creoles is an obsession with alchemical traditions of analysis, including atomization and stratification. The alchemical approach as a version of divide and rule is still very much a part of the dominant discourse in social sciences research (MERCHANT, 2006; VON WERLHOF, 2001). Even esteemed American anthropologist Edward Hall's term "extension transference" can be traced to alchemical traditions given its structuralist interpretation: extension transference is conceptualized as the representation of knowledge externalized into perceivable units (1976, pp.28-40). The captured utterance, after being stripped of its context and isolated as a distinct unit of a larger system-conceptualized thereafter as a morphophonemic or morphosyntactic constituent-is an example of extension transference.

In the pidgin and creole linguistics context, divide and rule is apparent in attempts to quantify perceivable units of grammar and speech sounds from captured utterances and to group languages based on their overall weights and relative scores. There is a growing tendency to measure complexity and simplicity by examining the number of phonological and morphosyntactic units and rules that are perceived in contact languages with their input languages-what Mufwene (2013) refers to as "bit complexity" analysis (p.162, citing DEGRAFF, 2001, p.268). Recently, the issue of complexity has merged with discussions of creole distinctiveness (BAKKER; DAVAL-MARKUSSEN; PARKVALL; PLAG, 2011; BAKKER, 2016). Complexity is examined comparatively in those studies, and those comparisons are made to examine how un-African and non-European that contact languages are. There are two dominant trends that creolists pursue regarding comparisons of Afro-Caribbean English-lexifier Creoles (AECs) and their European lexifiers and West African substrates and adstrates. First, central to most studies is a focus on surface structural features of the tense, mood, and aspect (TMA) systems. The TMA system reflects the semantics of temporal, emotional, and continuity categories. Creoles like Crucian and Nigerian Pidgin have lexical items that mark functional categories, and creolists cite this strategy as less complex than the coding strategies that exist in metropolitan Englishes, for example, which have inflectional structures that code grammatical features like habituality and completedness. Siegel (2008) summarizes the received wisdom about inflectional morphology: "lexicality corresponds with morphological simplicity while grammaticality corresponds with complexity" (p.43). Second, in many cases this lexical vs. functional dichotomy is coupled with one of two approaches in the search for simplicity or complexity in pidgin and creole grammars: holistic versus modular. Holistic approaches view languages as finalized, monolithic entities. Modular approaches look at specific potentials of language contact that have helped shape a linguistic domain of a contact language (see Essegbey, 2005, and Corum, 2015, for example, for cases in which substratal and adstratal input has influenced the domain of location marking). McWhorter (2011) compares languages based on their ages and whether they meet certain criteria, a kind of "creole litmus test" (pp.1-18), and concludes that creoles have less complex grammars than older languages. Although the author generalizes about whole grammars of creoles, McWhorter's approach is a modular approach in part, focusing on comparative constructions, distal and proximal relations, and the copular system.

What emerges from studies that examine only TMA marking strategies and conceive of a language as a finalized, abstracted system reveals dominant discourses that prevail in social sciences and humanities. Often, conclusions of those studies underscore the superiority of European languages and cultures and European scientific approaches to linguistic analysis, with slight acknowledgment of complex discursive strategies that are embedded in AEC modes of communication-strategies that are largely absent or avoided in Western modes of communication.

The remainder of this paper is a rejoinder to the dialogue concerning complexity and simplicity in pidgin and creole grammars. My position is in line with Siegel (2008, pp.21-22) in that it takes a chronotopic and unfinalized view of grammatical expansion in contact languages, Bakhtinian notions that give crucial importance to temporal and spatial contexts in discussions about linguistic development and historicity of language use. In my analysis, I zoom in on a pragmatic component of creole grammar: ambiguity in discourse. We find ambiguity in pidgins and creoles at both structural and pragmatic levels of grammar. At the structural level, there is zero-marking of number on nouns and gender on pronouns. Ambiguous features like these were not eliminated during contact language formation and, in fact, today they are considered to be typical of noun phrases in both Atlantic and Pacific pidgins and creoles (HOLM, 2000, pp.212-217). Yet, they are still viewed as marginal features that were an "invention" of speakers of contact languages like Hawai'ian Creole (BICKERTON, 1981, p.26). In terms of bit complexity analysis, zero-marking would not be noted because it is not perceivable and, hence, unquantifiable. In sum, ambiguity on both structural and pragmatic levels of grammar remains difficult to account for using current metrics of complexity versus simplicity. My aim in this paper is to underscore the importance of recognizing pragmatic discursive strategies as complex features of AECs. I focus on phrasal verb constructions and their double-voicing features, with particular attention paid to the phrasal verbs rip off and hot up in Crucian. In the following section, I invoke insights from Bakhtin (1981, pp.272-275) and his theory of dialogized heteroglossia in the hope that it can aid our analysis of plurilingualism and the cultivation of ambiguity in creole discourse.

Dialogism and Simultaneity in Pidgin and Creole Grammars

Mikhail Bakhtin was a brilliant Russian thinker² whose ideas about communication and consciousness challenged the Cartesian rationalist and mechanistic view of language and mind. Bakhtin emphasized that the communicative function of language should be conceptualized as a shared dialogic event, as opposed to a monologic activity in which symbols are manipulated and algorithmic operations are performed in isolation by individual speakers. In this dialogic view, a speaker's utterances are understood to be temporal/spatial instantiations of her/his speech events; these instantiations conveyed by utterances are always produced and understood in socio-cultural, historical, and register- or genre-specific contexts (BAKHTIN, 1981, p.272). Even more, the ways in which today we assemble, use, reassemble, and reuse language is inextricably tied up with the ways in which others have assembled, used, reassembled, and reused language in the past given their socio-cultural, historical, and register- or genre-specific contexts. This theory of language is called dialogism, and it almost seems like a quantum mechanical theory of communication, for example, in its description of meaning as an entangled state:

Dialogism argues that all meaning is relative in the sense that it comes about only as a result of the relation between two bodies occupying simultaneous but different space, where bodies may be thought of as ranging from the immediacy of our physical bodies, to political bodies and to bodies of ideas in general (ideologies) (HOLQUIST, 2002, p.19).

Creoles, as instantiations of "nation language" (BRATHWAITE, 1984), are in tune with the theory of dialogism. Nation language is not a single unitary system, but a diverse range of linguistic potentialities. Brathwaite distinguishes the frame of monological unitary, hegemonic concepts of language that are typified by standardized written European languages from dialogical oral traditions from Africa and elsewhere (FARACLAS, 2009). Nation language is not a national language that can be isolated and understood in relation to one people and one space. In a heteroglossic Afro-Caribbean contact setting, speakers have accesses to a range of varieties that include diverse Englishes, mixed codes and entwined languages, and traces of African and indigenous languages. A language is acknowledged in its own right and voiced when needed:

What I am going to talk about this morning is language from the [Afro] Caribbean, the process of using English in a different way from the 'norm'. English in a new sense as I prefer to call it. English in an ancient sense. English in a very traditional sense. And sometimes not English at all, but language (BRATHWAITE, 1984, p.5).

The idea of nation language contradicts the notion of a unitary language, which displays centripetal tendencies imposed by dominant institutions and their normative practices. Nation language is found in heteroglossic practices, where contradictory forces collide and give rise to many truths. In this sense, nation language and creole hybridity find connections with dialogism:

'Both/and' is not a mere wavering between two mutually exclusive possibilities, each of which is in itself logical and consistent, thus insuring the further possibility of truth, since a logic of this restrictive sort is so limiting that only one of the two options can be correct. Dialogic has its own logic, but not of this exclusive kind (HOLQUIST, 2002, p.40).

Nation language does not waver between two mutually exclusive interpretations of one utterance, either. Nation language exists between speakers and listeners, and the voices that can be conveyed by nation language are open to interpretation, at least two of which come from the perspectives of immediate speakers and listeners. But, one interpretation of a single utterance does not cancel second, third, etc. interpretations; those simultaneous interpretations, which are achieved because of the multiple voices that are conveyed through the concrete utterance, carry potentialities of meaning for a given person who may detect a familiar trace of her/his past, present, or future given the register- or genre-specific context of that utterance. This fact of language is conceptualized in this paper as a double-voicing feature of creole discourse, which is based on Bakhtin's notion of double-voicedness.

Similar to particle-wave dualism in quantum mechanics, double-voicedness suggests that utterances can be "internally undecided and two-faced" (BAKHTIN, 1984, p.198). This internal dialogization of an utterance is observed on lexical, syntactic, and pragmatic levels of pidgin and creole grammars. In Ghanaian Student Pidgin, for example, an English-derived form body functions as a grammatical item in locative predication and conveys meanings that mirror uses of similar items in Ghanaian languages of the Kwa group, including Fante, Twi, Ga, and Ewe (CORUM, 2015, pp.117-119); two or more of those Kwa languages form part of the linguistic repertoires of most Ghanaian Student Pidgin speakers. In addition to the multiple meanings that body can have in descriptions of locative relations, the hybrid "two-faced" nature of an utterance containing body becomes more apparent when one considers the syntactic position of the grammatical item. Unlike prepositions in English, body appears in postposition: rope dé tree im body 'The rope is around the tree.' The utterance is internally undecided and double oriented in that the words are English, but the semantics and the syntax are Kwa. The hybrid utterance also carries with it sociolinguistic implications in the Ghanaian context.

Ghanaian Student Pidgin is a refracted variety of the older English-lexifier Pidgin spoken in West Africa. Most speakers of Nigerian Pidgin and Ghanaian Pidgin English, for example, do not use postposed grammatical items in descriptions of locative relations. By placing an item after the noun phrase in Ghanaian Student Pidgin, speakers create a verbal sign that is inextricably bound to an ideological theme. The use of Student Pidgin reflects in-group membership for persons who have been educated formally at secondary and tertiary institutions in Ghana; historically, this has not been a common indexical feature of pidgin in West Africa. Therefore, in Ghana the student variety distorts the very idea of what West African English-lexifier Pidgin represents. An "evaluative orientation" has been made with the reappropriation of a pidgin language in the Ghanaian context (VOLOŠINOV, 1973, p.105): adding novel, distinctive features to the existing contact language allowed speakers to reassign a new value to a language that historically had been stigmatized as broken or corrupt English. In addition, the choice of an historically stigmatized variety over an European-oriented variety as a marker of social indexicality represents a challenge to the normative linguistic habitus of anglophone Ghana.

The important role that members of the Bakhtin Circle placed on the sociological nature of utterances was later echoed by creolists Le Page and Tabouret-Keller (1985), whose work on social-verbal interactions as linguistic acts of identity showed how an "individual's idiosyncratic behavior reflects attitudes towards groups, causes, and traditions but is constrained by certain identifiable factors; and how the identity of a group lies within the projections individuals make of the concepts each has about the group" (p.2). More recently, creolists have begun exploring how linguistic acts of identity can be seen in the light of multiple-voicing phenomena in the Atlantic Creoles. The ideas from Faraclas et al. (2014), for example, are discussed below and set the tone for the remainder of the paper on phrasal verb constructions and their double-voicing features in Crucian.

Complexity and Pluri-Voicing Strategies in Afro-Caribbean English-lexifier Creoles

Drawing on the notions of "double consciousness" from Du Bois (1903, pp.2-3) and "heteroglossia" from Bakhtin (1981, p.263), Faraclas et al. (2014) conceptualize pluri-voicing in the Afro-Atlantic as a means of asserting affiliations with different linguistic identities and cultural heritages, specifically they provide strong support for convergence scenarios among West African substrates, European lexifiers, and proto-Atlantic contact languages. They believe that multifunctional uses of lexical and grammatical items in AECs reveal complexities in discursive strategies that are best characterized as ingenious and creative responses to language in general and language contact in particular. In a dialogic sense, AEC multifunctional words and constructions can be used to engage multiple audiences:

Pluri-voicing has allowed Atlantic Creole speakers to equip themselves with a linguistic repertoire that has enabled them to use what appear at first glance to be the exact same words and structures to simultaneously assert Afro-Atlantic identities, Euro-Atlantic identities, Atlantic Creole identities and other identities to the extremely diverse and often very hostile and dangerous audiences and communities of practice in dialogue with whom they have managed to survive (and even thrive) through slavery, colonialism, and neo-colonialism (FARACLAS et al., 2014, pp.177-178).

A clear example of pluri-voicing as a complex discursive strategy is observed in a speaker's use of lexical items that convey property concepts, that is, words that are considered adjectives in European languages and stative verbs in West African languages (FARACLAS et al., 2014, p.178; MAZZOLI, 2015). In Figure 1 below, a single utterance dì sup swit mì from Nigerian Pidgin is interpreted differently under certain linguistic lenses: European lenses (a), which favor static interpretations of property items, and West African lenses (b), which favor dynamic interpretations of property items.

Figure 1
Double-voicing of the property concept swit in Nigerian Pidgin (adapted from Faraclas et al., 2014, pp.179-180)

There are compelling reasons to acknowledge both voices (a) and (b) in the English-lexified utterance in Figure 1 above. Substrate and adstrate sources were influential in the adoption of multifunctional grammatical items in creole discourse. They impacted not only the motivation for uses of those items in certain constructions, but also the pragmatic strategies that were chosen by speakers during verbal interactions. We still find the impact of substrate and adstrate influence on those three areas in creole discourse of the Afro-Atlantic and Pacific today. Some features of those inherited elements are distinct from elements of Standardized European lexifier languages. First, discursive strategies in West African languages value performance-oriented modes of communication and often employ multifunctional items that promote "cultivation of ambiguity" and indirectness (FARACLAS et al., 2014, p.181; see also Ameka & Breedveld, 2004; Tarr, 1979). Performance-oriented modes of communication result in multiple interpretations of single utterances and other speech acts and intentionally draw listeners into dialogue.

Second, boundaries between lexical and functional items are fuzzier and less rigid in West African languages than they are in Standardized European languages. Lexical items often have overlapping semantic spaces in AECs and, unfortunately, creolists cite this as evidence of underspecification, that is, featurelessness, or non-overt marking in the grammar. Zero-marking, as described earlier in this paper, is another example of what could be termed an underspecified feature of pidgin and creole grammars. In addition to zero-marking of noun phrases, there are also certain copular constructions and verbs in the present tense and in realis mode that are left unmarked. The factative interpretation, for example, does not require overt marking of active verbs. Speakers understand that an event occurred in the past when they encounter the factative; however, this knowledge about non-salient features of grammar has to be accessible to a user so that she/he can interpret the intended meaning of an utterance. Huang (2013) provides an overview of salience in such a discursive context:

Salience is a dynamic and graded notion. It varies at differing stages of processing and in differing discourses. It also reflects the situation where one mental entity has multiple meanings competing for salience at the same time. Importantly, salience functions in a discourse, which is composed of two core elements: the formal linguistic quality of a particular stretch of language, and its individual or group users.... In this sense, salient meanings are subjected to the range of linguistic and extra linguistic information and knowledge accessible to the language users (HUANG, 2013, p.113).

It is also worth noting that AECs and their West African substrates and adstrates have been described as "aspect prominent" in the literature (FARACLAS, 1990, p.105; PARKVALL, 2000, p.87), meaning that the preferred strategy of speakers is to mark aspect rather than tense and, therefore, to report on how an event unfolds rather than when an event occurs. Factative allows speakers to reference time without overtly marking it in discourse. There are ways to make tense and non-realis more salient, however. Speakers can disambiguate for these categories by inserting appropriate markers in an utterance: don for [+completive], de for [-completive], and go for [-realis]. If a marker is not included, then an active verb like go in a Nigerian Pidgin utterance im gó háws would be interpreted multiple ways: 's/he has gone home', or 's/he went home', or 's/he goes home', or 's/he is going home' (FARACLAS, 2009).

Instances of zero-marking and their resulting ambiguous interpretations can be observed in the absence of copulas in locative and equative constructions, as well. Zero-marking in Miskito Creole, for example, occurs when dé is omitted before locative adverbials like prepositional phrases: Ai no nuo if it [dé] in di baibl, 'I don't know if it's [there] in the Bible' (HOLM, 1976, p.266). The absence of the copula dé produces at least two different interpretations in the creole utterance: "[zero-marking] makes de ambiguous as either locative 'be' or 'there' after a zero copula" (HOLM, 2000, p.199). Looking to the Singlish contact ecology, Malay and Chinese languages have zero-copula equative constructions. It makes sense that Singlish would acquire the zero-copula marking strategy, too, since it emerged from a typological matrix in which zero-copula in equative constructions was a dominant type frequency pattern (ANSALDO, 2004, p.135).

In all cases mentioned above, zero-marking contributed to a pragmatic feature of pidgin and creole grammar: cultivation of ambiguity. This linguistic strategy draws speakers into social-verbal interactions. Meanings that emerge from utterances are not fixed, but rather are negotiated by speakers and placed under an evaluative purview that allows for new significances to emerge from formerly recognized meanings (VOLOŠINOV, 1973, p.106). Today, continual uses of zero-marking in pidgin and creole utterances are a testament to the positive evaluations that speakers have towards the flexibility of their grammars; in this paper, the strategy is observed in terms of centrifugal forces that motivate speakers to pull the normative tendencies of their grammars apart.

In the next section, I present another instance of the cultivation of ambiguity in the AEC context by focusing on uses of phrasal verb constructions in Crucian, a creole language of the U.S. Virgin Island St. Croix. The double meanings found in the uses of phrasal verb constructions provide support for the pluri-voicing argument made above in Faraclas et al. (2014). I extracted the instances of phrasal verb constructions from a small corpus of Crucian. The data in the corpus are composed of phone recordings that were transcribed by the St. Croix court system in the early 2000s. I have converted the transcriptions into accessible text (.txt) files to enable a concordancing program, for example AntConc 3.4, to conduct key word in context (KWIC) searches of the files and to identify the most frequent constructions. Although the recordings were made in St. Croix, the corpus is not representative of Crucian in general. The data contain linguistic examples produced by persons from St. Croix and other areas of the Caribbean region who speak varieties of AEC. Also, most of the persons are males between the ages of 18 and 30. Therefore, the conclusions made in this work reflect broad claims about features and uses of AEC³.

Cultivating Ambiguity: Phrasal Verb Constructions and Double-voicing in Crucian

The Crucian corpus is important to the study of creole discourse in general because it provides a large body of natural, heteroglossic language data that consists entirely of dialogic interactions between speakers. Moreover, specificities of interlocutor, time, and space are reflected in the linguistic utterances that make up the corpus in that the data were recorded from phone calls and, therefore, the utterances that were produced by speakers of Crucian were directed towards other speakers of Afro-Caribbean creole varieties. There is little reason to believe that speakers' concrete utterances reflect normative, hegemonic tendencies of English grammar in all cases. In fact, grammatical variation abounds in the Crucian data; it could be argued that speakers use forms that refract features of standardized English to assert linguistic acts of resistance against established institutions and the normative practices that they impose (see Gonzalez-Lopez, 2014, for a discussion of creoles as spirit languages of resistance).

Double-voicing in this paper refers to the complex interaction of linguistic features that permit multiple interpretations of single utterances in creole discourse. Chrono and tope in the term chronotopic utterance that was mentioned at the beginning of this paper are understood in relation to the time frame-or age-of the speakers and the spatial contexts in which they use their variety of AEC -in St. Croix. Specifically, we have a synchronic snapshot of how AEC is employed by young adults between the ages of 18 and 30 in the Northeastern part of the Afro-Caribbean region today. This temporality shapes the character of the utterances that we find in the Crucian data; I would expect data that were extracted from early Surinamese texts to have distinctive and, therefore, chronotopic compositional features, too (ARENDS; PERL, 1995). Nevertheless, there are synthesizing features that leave generations of speakers of AECs with similar strategies of discourse, regardless of their different times and places. These are hybridization features that speak to the "multiaccentuality" of creole discourse (VOLOŠINOV, 1973, p.23), that is, continual convergence patterns of ancestral substrate and present lexifier components of creole grammar. Specific examples of convergence patterns were provided in the immediate section above. Hybrid mixing of features of substrate and lexifier sources in the formation of creole grammars are directly related to Bakhtin's feature of double-voicedness: In the Ghanaian Student Pidgin and Nigerian Pidgin examples above, we saw two or more different linguistic consciousnesses that collided within single concrete utterances, at least two of those consciousnesses can be traced back to African languages and Englishes. In the next two sections, I provide an examination of similar strategies that are achieved through the use of phrasal verb constructions in Crucian Virgin Islands Creole.

Phrasal verb constructions consist of verbs and functional morphemes, for example, bring out and hook up. Phrasal verbs are interesting from a cognitive semantic point of view because they are highly productive constructions that use functional morphemes to add semantic content to verbs. By adding functional morphemes to verbs in phrasal verb constructions, speakers adjust the semantic constraints that are imposed on certain verbs. In the Crucian data, one finds functional morphemes used in places where English prefers affixes, modals, or adverbials. Below, I present instances of phrasal verb constructions from Crucian in which morphemes and their placements result in shifts of meaning that are not observed in U.S. or British varieties of English.

Rip Off

In English, rip undergoes a shift in meaning when it is used transitively and a morpheme is added to the verb. The Oxford English Dictionary defines rip off: 'tear or pull (something) quickly or forcibly away from something or someone else' (emphasis added). When the verb is used with an object, rip provides the 'quick' or 'forcible' meaning and the morpheme provides the 'removal' meaning. In the Crucian corpus, one reads about a description of an event in which an affidavit was stapled to someone's door and it was reported to have been removed:

1) They dou rip it off.

They have ripped it off.

The translation of (1) is a literal reading of the phrasal verb rip off; note the pronoun placement between the verb and the morpheme in AEC and the English translation to render the Oxford English dictionary meaning. In a different part of the Crucian corpus, there is a situation in which two individuals discuss the details of a robbery. In examples (2) and (3), speakers use rip off in a way that is different from the previous instance:

2) If he want thirteen thousand we gon rip off he head.

If he wants thirteen thousand, we will rob him.

3) you got to mek it look like you ripping off my head to.

You have to make it look like you are robbing me, too.

In the conversation above, the speakers employ the phrasal verb rip off to mean rob, that is, cheat. Searches of rip off in two corpora of written English (the Freiburg-LOB (FLOB) corpus of written British English of the 1990s, and Freiburg-Brown (FROWN) corpus of written American English of the 1990s) return instances of the phrasal verb rip off that are similar to the utterances in the Crucian corpus:

4) if you're going to rip someone off when you're ripping off your clothes, who better than Madonna? (FLOB_C16, lines 216-217)

5) You plan to rip me off? (FROWN_C05, line 81)

In both British and U.S. American varieties of English, the pronoun must occur between the verb and the morpheme. If a lexical noun phrase is used instead of a pronoun, then is can appear before or after the morpheme, as it does in ripping off your clothes. If it appears after the morpheme, however, the phrasal verb has a literal interpretation, that is 'removing quickly or forcibly'. Thus, the sentence I will rip off NP's head in U.S. American English cannot mean 'rob/cheat', rather the intended meaning is 'hurt or destroy/kill', either figuratively or literally. The expression rip NP's butt off means 'cheat NP', whereas rip off NP's butt sounds unusual and has some physical connotations. I conclude, then, that the following rules apply to English for the rip off phrasal verb construction⁴: i) if a lexical NP appears between rip and off, the phrasal verb construction has a figurative meaning 'cheat' or literal meaning 'remove'; and, ii) if a lexical NP appears after off, the phrasal verb construction has a literal meaning 'remove', which can be interpreted figuratively as 'kill' or 'render invisible'.

In the Crucian data, I observed that the speakers in examples (2) and (3) used rip off he head and ripping of my head in ways that do not conform to the uses of rip off that were posited in (i) and (ii) above. Even though the NP body part he head appears after the morpheme, the intended meaning is still 'cheat' , according to the translation provided by the court transcriber. At the same time, though, the expected meaning given the placement of the noun phrase posited in (ii) still applies: the act of robbing will result in the removal of the guy from future activities, thereby figuratively killing or eliminating him. Double-voicing of rip off is illustrated in Figure 2 below.

Figure 2
Double-voicing of the phrasal verb rip off in the Crucian corpus

There are two voices that collide in the single utterance containing rip off he head: one voice asserts English grammar, in which lexical noun phrases that appear after morphemes render literal meanings; a second voice asserts Afro-Caribbean creole grammar, in which lexical noun phrases can appear after morphemes and render figurative meanings. Thus, there are two simultaneous but different interpretations of the phrasal verb rip off in Crucian, similar to the double-voicing of swit in Nigerian Pidgin seen in Figure 1. Afro-Caribbean English-lexifier creoles are especially good at this double-voicing phenomenon, which has been framed in terms of cultivating ambiguity in the sections above and which has most recently been explored by Faraclas and The Working Group on the Agency of Marginalized Peoples in the Emergence of the Afro-Atlantic Creoles (2016).

In this brief discussion, I noted that Crucian speakers are able to undermine monologic tendencies of English grammar. In the next section, I focus on another phrasal verb construction that shows potential of meaning that distort the normative grammar of English and contribute to innovative new readings of grammatical constructions in creole grammar.

Hot Up

In previous sections of this paper, I discussed zero-marking of nouns and copulas and argued that this strategy led to ambiguity in discourse. Here, I will look at instances of overt marking in phrasal verb constructions in Crucian. The utterances under consideration include property items + functional morphemes, for example, hot up. Double-voicing emerges from these constructions in that there is a tendency to overtly mark aspect on the functional morpheme regardless of the conceptual interpretation of the property items as static adjectives or dynamic verbs (see Figure 1 for a similar analysis of property concept words and ambiguity that results from their uses in Nigerian Pidgin). Holm (2000) refers to property items of this kind in AECs as "adjectival verbs" (p.200).

In AECs, adjectival verbs take preverbal markers to express tense, mode, and aspect. TMA marking strategies that are used with these grammatical elements have revealed similarities between creoles and African languages, for instance in the coding of an inchoative quality: "In some creoles [adjectival verbs] can also take the marker of progressive aspect [...] indicating that the quality is inchoative (i.e. coming into being) [...] This is also a widespread feature in African languages" (HOLM, 2000, p.201). In the section Cultivating Ambiguity, I noted that functional morphemes add aspectual content to verbs in phrasal verb constructions; they are similar to preverbal markers in this regard in that they can code aspectual information like the inchoative quality. In the section Complexity and Pluri-voicing Strategies, I noted that AECs and their West African substrates and adstrates are aspect prominent. The reader should bear in mind those points from the two sections above as we review the phrasal verb hot up in Crucian. Before I begin the discussion, let us review a remark that Holm (2000) made regarding phrasal verbs in AECs:

Some of the English-based creoles have created a number of new phrasal verbs such as Miskito Coast CE apiir op 'appear, show up', daak op 'turn dark', dronk op 'become intoxicated', hog op 'hug', uol op 'become old', wet op 'soak', wind op 'become flatulent'. However, caution is usually advisable before anything is labelled an innovation.... It is possible that some of the Miskito Coast CE phrasal verbs mentioned above actually represent survivals of archaic or regional British usages.... However, when they include verbs based on English adjectives (e.g. uol op from old plus up), the case for their being true innovations is much stronger (HOLM, 2000, p.131).

The Crucian corpus contains several instances of phrasal verbs like those Holm (2000) mentioned in the quote above, including hide up, search up, and stab up. I focus only on the adjectival verb + functional morpheme construction hot up, as this construction is like those that Holm believes have the most potential for being innovations, that is, instances of grammar that defy centripetal, normalizing tendencies of input languages.

According to the Oxford English Dictionary, British and US American varieties of English have two common meanings associated with uses of the phrasal verb hot up: a literal meaning 'become or make hot', which is used transitively with an object; and a figurative meaning 'become or make more active, lively, or exciting', which usually appears intransitively. In the Crucian corpus, speakers only use hot up in its figurative sense, but both transitive and intransitive uses of the phrasal verb are observed. In example (6), a speaker complains about his companion who used his vehicle and, by doing so, attracted attention from the police. The phrasal verb is used transitively with an object in this case:

6) You hot up me jeep with dem man.

You made my jeep hot with the police. [attracted surveillance]

In example (7), the phrasal verb hot up is used intransitively and co-occurs with the verb get. The phrasal verb references the state or quality of a location that has attracted attention from the police:

7) A telling you eh hot eh geh hot up.

I'm telling you it's hot; it got hot. [became surveilled]

The utterances in (6) and (7) show flexibility of the grammatical uses of property items and functional morphemes in phrasal verb constructions in Crucian. Similar to swit in the Nigerian Pidgin example in Fig. 1, Crucian has dynamic, verbal interpretations of property items. The verbal interpretation of hot in example (6) is confirmed by TMA specification, which appears in the form of a functional morpheme up, as well as the transitive use of the phrasal verb. This interpretation resembles the literal use of hot up that was noted in the Oxford English Dictionary. In example (7), the phrasal verb is used intransitively and, therefore, hot could be analyzed as a static, adjectival property item. The common strategy in English is to treat hot and sweet as static adjectives and, therefore, to code aspectual information on auxiliary verbal items or some other inflectional marking strategy. In those intransitive cases, verbs like become, get, or make or aspectual structures like -ing express an inchoative meaning, for instance, getting hot. In the Crucian data, however, there were several uses of hot up that were used intransitively but did not co-occur with auxiliary items like is, become, or get. According to the translation of the utterance given by the Crucian-speaking court transcriber, those instances of hot up rendered inchoative meanings. The functional morpheme must have provided that aspectual information in those cases. Even when speakers included verbal items, as seen in the use of geh in (7), there was still a tendency to incorporate a functional morpheme as well. This overspecification suggests that a second voice is present in the discourse. The second voice asserts that Crucian is aspect prominent, a characteristic quality of AECs and their West African substrate languages.

Conclusion

It was my aim in this paper to contribute to the discussion concerning complexity in creole grammar by focusing on a polyphonic discursive strategy called double-voicing, also called pluri-voicing in Faraclas et al. (2014), a phenomenon in which multiple voices are engaged through single utterances of lexical and grammatical items in Afro-Caribbean creole languages. I examined instances of phrasal verbs in the Crucian corpus to see whether normative rules of English are refracted to create novel meanings in creole. Multiple voices were found in single concrete utterances of the phrasal verbs hot up and rip off and each offered different potentials of meaning. The analysis of hot up drew attention to a voice in the utterance that overtly marked aspect with a functional item, even in cases in which an auxiliary verb co-occurred with the property concept word. This marking strategy seems like a case of unnecessary overspecification from the point of view of English grammar, but in fact it is a strategy that we might expect from the grammar of an aspect prominent language. In the case of rip off, I observed that one potential meaning followed the U.S. English pattern, in which a postposed lexical noun phrase leads to a literal or figurative meaning 'remove'; and the second potential meaning emerged as a novel Afro-Caribbean creole specific use of the construction in which a postposed lexical noun phrase renders a figurative 'cheat' or 'rob' meaning. In U.S. and British varieties of English, placement of the lexical noun phrase will restrict one or the other intended meanings posited above and generalized in (i) and (ii) in the section Rip Off.

I began this paper with a discussion of dominant discourses that prevail in the study of simplicity and complexity in AECs. I drew from the review of the literature in Siegel (2008) and his discussion about an emerging consensus among creolists over metrics that can satisfactorily measure complexity in creole languages. Although creoles are rightly viewed as natural languages, they are still described as having the simplest grammars because they contain fewer linguistic units, phonetic distinctions, and derivational operations than their input languages. Pragmatic discourse features, like double-voicing, are often ignored in studies of complexity in grammar, since these features do not have grammaticalized expressions that overtly mark their semantic and pragmatic distinctions (MCWHORTER, 2005, p.46). Hopefully, the analysis of hot up in this paper will provide insight on this issue of overt marking and pragmatic distinctions.

Complexity and simplicity have been framed in creole studies as objective notions that are measured by counting linguistic units. My contribution to the discussion takes issue with the strict either/or division that is often made in the study of creole languages, namely that creoles either conform to patterns of the lexifier or carry on patterns of their substrates and adstrates. It is also alarming that discursive strategies are rarely counted as part of the metrics of complexity in creole grammar. Morphosyntax remains the major focus for determining inheritances from input languages, even in situations where creoles continue to be spoken alongside their input languages and have converged typologically with respect to several linguistic domains. Kihm (2011) provides a note about Guinea-Bissau Kriyol in this regard:

The relative scarcity of morphosyntactic influences from the substrate does not mean, however, that Kriyol is not perfectly embedded in its sociolinguistic and cultural environment. Quite the opposite in fact: in terms of lexical semantics, discourse strategies, pragmatics, and language uses in general, Kriyol is just as 'African' as the surrounding languages - scare quotes necessary since the epithet is not susceptible of a precise definition, but it is an impressionistic label for a very complex set of cultural attitudes and practices, shared beliefs, etc. (KIHM, 2011, p.82, emphasis added).

Kihm's impression about complex linguistic and cultural attitudes and practices that creole speakers adopt has been confirmed by the research of Faraclas and The Working Group on the Agency of Marginalized Peoples in the Emergence of the Afro-Atlantic Creoles (2016). That research group's ideas about dialogic interaction and ambiguity in creole discourse have been validated and further supported in this paper with the analyses of phrasal verbs and their double-voicing features in AECs.

• Pontifícia Universidade Católica de São Paulo
• 1
  Cultivating ambiguity is a term borrowed from Faraclas and The Working Group on the Agency of Marginalized Peoples in the Emergence of the Afro-Atlantic Creoles (2016).
• 2
  Editors' Note: Actually, we can assert that there was a Bakhtinian thinking, which was collectively developed by the different members of the Circle.
• 3
  Special thanks are given to the English Department of University of Puerto Rico for providing access to the data (see Vergne, 2008, for a discussion about this data and insightful reflections on ethical uses for linguistic purposes).
• 4
  I ignore rip off to mean 'imitate', as there were no instances of rip off that had that meaning in the Crucian corpus.

REFERÊNCIAS

• AMEKA, F.; BREEDVELD, A. Areal Cultural Scripts for Social Interaction in West African Communities. Intercultural Pragmatics, v.1, n.2, pp.167-187, 2004.
• ANSALDO, U. The Evolution of Singapore English: Finding the Matrix. In: LIM, L. (Ed.). Singapore English: A Grammatical Description. Amsterdam: John Benjamins, 2004, pp.127-150.
• ARENDS, J.; PERL, M. Early Suriname Creole Texts: A Collection of 18th Century Sranan and Saramaccan Documents. Madrid: Bibliotheca Ibero-Americana, 1995.
• BAKHTIN, M. From Discourse in the Novel. In: ______. The Dialogic Imagination: Four Essays Edited by Michael Holquist; translated by Caryl Emerson and Michael Holquist. Austin, TX: University of Texas Press, 1981, pp.259-422.
• ______. Problems of Dostoevsky's Poetics Edited and translated by Caryl Emerson. Minneapolis, MN: University of Minnesota Press, 1984.
• BAKKER, P. You Got Gungbe, but We Got the Numbers: Feature Pools Show that Creoles Are Still Typologically Distinct. Journal of Pidgin and Creole Languages, v.31, n. 2, pp.419-435, 2016.
• BAKKER, P., DAVAL-MARKUSSEN, A.; PARKVALL, M.; PLAG, I. Creoles are Typologically Distinct from Non-Creoles. Journal of Pidgin and Creole Languages, v.26, n.1, pp.5-42, 2011.
• BICKERTON, D. Roots of Language Ann Arbor: Karoma Publishers, Inc, 1981.
• BRATHWAITE, K. Nation Language. In: ______. History of the Voice: The Development of Nation Language in Anglophone Caribbean Poetry. London: New Beacon, 1984, pp.5-19.
• CORUM, M. Substrate and Adstrate: The Origins of Spatial Semantics in West African Pidgincreoles. Berlin: Mouton de Gruyter, 2015.
• CRUCIAN WIRETAPS CORPUS. Compressed Line Sheets [text data files]. Retrieved from the English Department of University of Puerto Rico, Río Piedras, 2006.
• DEGRAFF, M. On the Origin of Creoles: A Cartesian Critique of Neo-Darwinian Linguistics. Linguistic Typology, v.5, n.2/3, pp.213-310, 2001.
• DU BOIS, W. Souls of Black Folk Chicago: A. C. McClurg & Company, 1903.
• ESSEGBEY, J. The Basic Locative Construction in Gbe Languages and Surinamese Creoles. Journal of Pidgin and Creole Linguistics, v.20, n.2, pp.229-267, 2005.
• FARACLAS, N. From Old Guinea to Papua New Guinea: A Comparative Study of Nigerian Pidgin and Tok Pisin. In: VERHAAR, J. (Ed.). Melanesian Pidgin and Tok Pisin: Proceedings of the First International Conference on Pidgins and Creoles in Melanesia. Amsterdam: John Benjamins, 1990, pp.91-169.
• ______. Personal Interview Conducted during an Independent Study at University of Puerto Rico titled Cognitive semantics for creole linguistics, May 15, 2009.
• ______ ; GONZÁLEZ COTTO, L.; CORUM, M.; JOSEPH HAYNES, M.; URSULIN MOPSUS, D.; VERGNE, A.; AVILLÁN LEÓN, P.; CRESCIONI, S.; CRESPO VALEDÓN, D.; DOMINGUEZ ROSADO, B.; LECOMPTE ZEMBRANA, P.; PIERRE, J.; LAO MELENDEZ, H.; AUSTIN, V.; GIBB DEPEZA, H.; JESSURUN, A. Creoles and Acts of Identity: Convergence and Multiple Voicing in the Atlantic Creoles (special issue dedicated to John Holm). PAPIA, v.24, n.1, pp.173-198, 2014.
• ______ ; THE WORKING GROUP ON THE AGENCY OF MARGINALIZED PEOPLES IN THE EMERGENCE OF THE AFRO-ATLANTIC CREOLES. Toward Post-Colonial Frameworks for the Study of the Emergence of Creolized Languages and Societies in the Afro-Atlantic Paper presented at the 3rd Bremen Conference on Language and Literature in Colonial and Postcolonial Contexts, Universität Bremen, Germany, 2016.
• GONZÁLEZ-LÓPEZ, C. Chamu Chamu Ye Tu (Talk Some, Leave Some) A Culture of Resistance and the African Substrate: The Role of 'Spirit' or Secret Languages in the Genesis and Development of Atlantic Creoles. 301 p. Unpublished doctoral dissertation. University of Puerto Rico, Río Piedras, 2014.
• HALL, E. Beyond Culture New York: Anchor Press, 1976.
• HOLQUIST, M. Dialogism: Bakhtin and His World. 2. ed. New York: Routledge, 2002.
• HOLM, J. Copula Variability on the Afro-American Continuum Paper presented to the Society for Caribbean Linguistics. Guyana, 1976.
• ______. An Introduction to Pidgins and Creoles Cambridge, UK: Cambridge University Press, 2000.
• HUANG, M. Solving the Riddle of Metaphor: A Salience-based Model for Metaphorical Interpretation in a Discourse Context. In: EVANS, V.; POURCEL, S. (Eds.). New Directions in Cognitive Linguistics Amsterdam: John Benjamins, 2013, pp.107-126.
• KAPLAN, J. English Grammar: Principles and Facts. New York: Prentice Hall, 1994.
• KIHM, A. Substrate Influences in Kriyol: Guinea-Bissau and Casamance Portuguese-related Creole. In: LEFEBVRE, L. (Ed.). Creoles, Their Substrates, and Language Typology Amsterdam: John Benjamins, 2011, pp.81-104.
• LE PAGE, R.; TABOURET-KELLER, A. Linguistic Acts of Identity: Creole-based Approaches to Language and Ethnicity. Cambridge, UK: Cambridge University Press, 1981.
• MAZZOLI, M. Property Items in Nigerian Pidgin: Verbs or Adjectives? In: CALABRESE, R.; CHAMBERS, J.; LEITNER, G. (Eds.). Variation and Change in Post-Colonial Contexts Cambridge: Cambridge Scholars Publishing, 2015, pp.79-108.
• MCWHORTER, J. Defining Creole Oxford: Oxford University Press, 2005.
• ______. Linguistic Simplicity and Complexity: Why do Languages Undress? Berlin: Mouton de Gruyter, 2011.
• MERCHANT, C. The Scientific Revolution and The Death of Nature Isis, v.97, pp.513-533, 2006.
• MUFWENE, S. Simplicity and Complexity in Creoles and Pidgins. Journal of Language Contact, v.6, pp.161-179, 2013.
• PARKVALL, M. Out of Africa London: Battlebridge, 2000.
• SIEGEL, J. The Emergence of Pidgins and Creoles Oxford: Oxford University Press, 2008.
• TARR, D. H. Indirection and Ambiguity as a Mode of Communication in West Africa: A Descriptive Study. 275 p. Unpublished doctoral dissertation. University of Minnesota, St. Paul, 1979.
• VERGNE, A. Reflections on Ethical Issues in Fieldwork. La Torre, v.13, n.49, pp.511-517, 2008.
• VON WERLHOF, C. Losing Faith in Progress: Capitalist Patriarchy as an 'Alchemical System.' In: BENNHOLDT-THOMSEN, V.; FARACLAS, N.; VON WERLHOF, C. (Eds.). There Is an Alternative: Subsistence and Worldwide Resistance to Corporate Globalization. New York: Zed Books, 2001, pp.15-40.
• VOLOŠINOV, V. N. Marxism and the Philosophy of Language Translated by Ladislav Matejka and I. R Titunik. New York, NY: Seminar Press, 1973.

Publication Dates

• Publication in this collection
  May-Aug 2018

History

• Received
  13 July 2017
• Accepted
  01 Apr 2018