The Concept of Decreolization

PDF Decreolization: A critical review 16 Pages Decreolization: A critical review Jason F Siegel UPLOADED BY Jason Siegel CONNECT TO DOWNLOAD GET PDF Academia.edu Decreolization: A critical review DOWNLOAD IUWPL9: Decreolization: A critical review (2010), edited by J. Clancy Clements, Megan E. Solon, Jason F. Siegel,and B. Devan Steiner, pp. 83–98. Bloomington, IN: IULC Publications. D ECREOLIZATION : A C RITICAL R EVIEW  Jason F. Siegel Abstract: This study looks at the often invoked but frequently vague notion of decreolization in studies of creole languages and language contact. Departing from the notion that scientificterms should have clearly bounded definitions and unique denotata, the notion of decreoliza-tion is examined in terms of its target, internal coherence, and difference with other linguisticprocesses. I conclude with a proposal to abandon the term decreolization in favor of terms ap-plicable to all languages such as debasilectalization . Decreolization remains an insecure notion: insufficiently distinguished from ordinary change processes, possibly conceptually incoherent, and certainly not adequately supported by dia-chronic investigations to date. It will not be utilized in this investigation, nor will it be linked through definition to any of the key concepts; and this can be recommended as a generalrule. Patrick (1999: 19) The concepts of depidginization and decreolization remain insufficiently studied or under-stood. Kaye and Tosco (2001: 94) It has not been rigorously defined what structural process is inverted or what structural properties are removed by this de-creolization process. DeGraff (2005: 553)1. IntroductionThe notion of decreolization is one that is both widely invoked and at the same timeoften misunderstood. One the one hand, it seems that one can see its effects in a large numberof creoles in contact with their lexifiers, but on the other hand it can be difficult to tell whatexactly decreolization is and how it differs from other processes of language change, bothwithin and outside of creoles. This essay then endeavors to explore decreolization not fromthe perspective of its results, but rather from its use and its definition.A scientific term should have a number of characteristics. The first is that it shouldnot be a loose definition, one capable of meaning a wide variety of vaguely similar objects,concepts, or phenomena without clearly established boundaries about what does and does notbelong in the set of things denoted by term X. This is crucial for scientific discovery, becausereal knowledge does not advance when scientists talk past each other, using the same wordswithout the same understanding. This does not mean that all those using the term have to  I would like to thank Kevin J. Rottet as well as the anonymous reviewers for their comments on earlier draftsof this work. Any faults remaining are of course my own. 84 Jason F. Siegel agree on what is a member of the set at the margins, or that there will not be things in theworld that challenge the general understanding of the boundaries of the term; rather, it meansthat scientists know, upon seeing a term in different sources, that the word means the samething in all the sources, that core membership in the set unless otherwise specified.Another important aspect of scientific terminology is that a unique term should repre-sent a unique phenomenon, concept, or thing. The need for precision in everyday language isoften downgraded in favor of expediency, ambiguity resolved by context or left ambiguouswithout major impediment to conversation. In science— and indeed in other fields such aslaw—, where members of the community share a common goal of discovering the truth orthe facts of the matter at hand, precision is indispensable for clear argumentation as well asinternal and external consistency. Being unique does not mean that the term will not be hyp-onymic to a larger concept nor hyperonymic to sub-concepts. A foot represents a certain levelof suprasegmental phonology; that feet can be iambic , trochaic , degenerate , unbounded , etc.,does not mean that foot fails to identify a unique concept, though. It is ‘unique’ in that thereare not other terms (e.g. syllable group ) that identify the same concept or thing as foot . Syn-onymy in scientific discussion is superfluous and invites increased probability that normalsemantic drift will reduce that actual semantic proximity and reduce the precision of usingthe terms interchangeably.To be clear, I am not saying that scientific terms do not undergo the same kinds of semantic evolution as other words; advances in knowledge and theory are responsible forchanging how we understand basic scientific terms like atom and phoneme , both of whichwere once thought to be indivisible and are now widely recognized as being made up of smaller parts (protons, neutrons and electrons for the former and distinctive features for thelatter), such that an atom and a phoneme no longer mean quite the same thing now as theydid at their coinage, even if they denote the same set of items. Semantic evolution is una-voidable, but conceding this fact is not equivalent to inviting an anything-goes attitude. Be-ing descriptivists of language does not obligate linguists to eschew all prescriptivism. At thelevel of discourse, we certainly embrace prescriptivism in the publishing process, discussingwhat constitutes acceptable discursive structure, adhering to style guides, and the like. To beable to get at a fuller understanding of the nature, functions, and realities of language, we canalso embrace the same level of prescriptivism among ourselves concerning terminology.A number of problems come up when considering the scientific term decreolization .The first and perhaps most obvious is the unclear target of decreolization. What decreolizes:the lexicon, a grammatical domain, a sociolect, the language, the speech community, or anyand all of the above? An obvious follow-up question would be to what extent would a com-bination of the above possibilities represent a ‘unique’ process? Another question that comesinto play is, What makes the language less creole? If a researcher is claiming that a creolelanguage is not simply evolving but decreolizing, then s/he has a responsibility to demon-strate that it is specifically creole features that are being lost. A related point is that defini-tions of decreolization (described more completely in the following section) frequently speci-fy that movement is towards and caused by the lexifier. If decreolization cannot occur in thedirection of another language, then there should be some clear linguistic or social differencesbetween those changes (or their outcomes) and the former. Indeed, this begs the question of what makes decreolization a distinct process of language change. This issue has two parts:what makes decreolization different from other language change; in other words, is it a spe-cific type of change with results or processes that are qualitatively different from other types Decreolization: A Critical Review 85 of contact-induced language change, results or processes that can be attributed to the lexifi-er/creole relationship? Another problem addressed here is the relationship of decreolizationto the creole continuum, specifically, does one necessarily beget the other? Lastly, do non-creole languages undergo the same types of changes as decreolizing languages?2. DefinitionsThere are many different definitions of decreolization given, some more precise thanothers. The most common definition is a variant of the following, found in Trask (1997)’s AStudent’s Dictionary of Language and Linguistics : “The process in which a creole changes soas to become more similar to the natural language from which it was originally derived[…]”or this one from Matthews (2005: 93)’s Oxford Concise Dictionary of Linguistics “Historicalprocess by which a *creole is progressively assimilated to a standard language: the assimila-tion of English creoles in the West Indies to standard English[…]”. Kaye and Tosco (2001:94) say, “Basically, […the creoles] become more like the (standard) superstratum languagesince… creoles normally coexist alongside it”. Holm (1988: 53) describes it as “progressive-ly dropping its nonstandard features and adding standard ones.” Thomason and Kaufman(1988: 98) define it as when “the language of a substrate creole-speaking population gradual-ly changes through structural and lexical borrowing from the superstrate vocabulary-baselanguage.” Rickford (1987: 16) says it is “movement away from creole norms and toward thenorms of lexically related standard languages.” Mufwene (2005: 79) says that decreolization“consiste en la perte des traits structurels associés au basilecte” (‘consists of the loss of struc-tural traits associated with the basilect’). Winford (2003: 314) limits it to “contact-inducedchange in a more radical creole due to continuing contact with its lexifier.” However, themost important definition, the one that seemed to set the study of decreolization in motion,has come from Bickerton (1980: 109-110): “In decreolization, speakers progressively changethe basilectal grammar so that its output gradually comes to resemble the output of an acro-lectal grammar… Decreolization then consists in the creole abandoning, one after another,those features which distinguish it from the superstrate, and immediately replacing eachabandoned feature by its superstrate equivalent.”If we wanted to construct a prototype of decreolization according to these definitions,we see that its principal defining features are the following. It involves at least a creole lan-guage as a basilect, and generally the lexifier as the acrolectal target variety, though variousdefinitions allow for other acrolectal varieties (e.g. the mesolect closest to the standard lexifi-er), an important point of disagreement. There is loss of basilectal features and replacementof those features by acrolectal ones. For some it is purely grammatical, and for others there isthe possibility of lexical decreolization. Bickerton’s definition is useful in that it makes itclear that for him the term covers both the individual and community grammars. For the pur-poses of this article, decreolization will be defined broadly as “change in a creole in the di-rection of the lexifier-base language due to contact with some variety of the latter.” Usingthis definition is a descriptive exercise and should not be interpreted a priori as an ac-ceptance of this definition as one that is scientifically adequate; indeed, it is a ‘loose’ defini-tion to be refined later. 86 Jason F. Siegel 3. What decreolizes?The exact target of decreolization is unclear. Some researchers appear to use it tomean that there is a large amount of lexical borrowing by the creole from the lexifier lan-guage. This is the case for Valdman (1986) and especially Zéphir (1993). Indeed, the latter’smain evidence for the decreolization of Haitian Creole comes from the lexical expansion thathas occurred using French borrowings such as “ reforme agraire la ” (‘the agricultural re-form’) and “ relations interpersonnelles ” (‘interpersonal relation(ship)s’) (200). Aitchison(2001: 218) adds, “Massive vocabulary borrowing is the most superficially noticeable aspectof decreolization in Tok Pisin.” Nevertheless, it seems that lexical borrowing, while likely acatalyst, is surely not on its own an example of decreolization, especially in the case of Hai-tian Creole, where creole features are not being lost. Fouse (2003) suggests that if we were touse this definition, we might have to cite the relexification of the Portuguese-based creolesby Spanish as examples of decreolization, as the two languages are so close that it can be dif-ficult to distinguish a great number of words as being from one rather than the other. Withsuch great lexical and phonological overlap, it is at least conceivable that the non-Iberiansspeaking the language in its early days did not distinguish Spanish from Portuguese (i.e. if they did not think they were more than differing dialects of the same language), and thereforethe former could simply have been considered the local variety of the latter by those speakingPapiamentu 1 . We could also ask if for any two closely related varieties, such as Sudanese Ar-abic and Classical Arabic for Juba Arabic Creole, if one dialect (e.g. Classical Arabic) cameto replace the other (Sudanese Arabic) as the acrolect to which the creole assimilated at thelevel of the lexicon only, would this relexification count as decreolization? A clear scientificdefinition of decreolization should be able to answer this question. The loose definition thatwe have adopted would accept this relexification, but it is unclear whether these great dialec-tal differences would constitute different varieties for many of the above researchers.Another possible target of decreolization is the grammar or one of its domains, suchas phonology or morphosyntax. Akers (1981) and Winford (1978) apply the term to phonol-ogy presumably in the context of broader decreolization in the language, and various otherstudies (including DeCamp 1971; Escure 1981; Greene 1999; Mühlhäusler 1997) includephonology as an important part of the decreolization situation. Indeed, it is hard to imagine amovement toward the lexifier-base language that completely ignored phonology. The vastmajority of studies, however, include mostly morphosyntactic changes as the main area of decreolization. These include Bickerton (1975, 1980), DeCamp (1971), Mühlhäusler (1997),Rickford (1986), Schwegler (2001), Washabaugh (1977), and Wood (1971). Given that thisis traditionally thought to be the hardest part of a language to borrow (see Thomason andKaufman 1988: Chapter 1), it should not be surprising that it would also be the most signifi-cant change to the creole. In general, those studies that discuss the decreolization of the cre-ole’s morphosyntax also either assert or imply that the language as a whole is decreolizing, 1 An anonymous reviewer points out that any scholar of the Iberian languages would not consider Spanish andPortuguese to be the same lexifier language because the differences between the two are so numerous. Be thatas it may, there are respected linguists such as Trudgill (1986) that put the two on a dialect continuum. Suchcontinua, as seen below, are a vexing problem for linguists, and while the poles at the time of decreolizationmay have been different enough that we could call them different languages, it is not clear that the acrolectalvarieties in creole-speaking areas were in fact the polar varieties for which we could talk about linguistic ratherthan dialectal differences. For this reason, I have attenuated the language from an earlier draft to reflect that thiswould be a conceivable instance of decreolization rather than an actual one. Decreolization: A Critical Review 87 rather than simply that one domain. That morphosyntactic change or phonological change isan example of decreolization is seemingly uncontroversial among scholars.The last logical possibility, explored by Rickford (1986) and endorsed by Aitchison(2001), does not exclude morphosyntax from its purview, but adds the speech community tothe concept of decreolization. According to him, decreolization could result from the additionof new lects filling in the creole continuum between the basilect creole, an intermediate me-solect, and the acrolect lexifier. Then, as more lects became available, the original basilectwould simply disappear in a manner akin to language death, with fewer and fewer speakersusing it. He refers to this as “quantitative” decreolization, distinguished from “qualitative”decreolization, in which the creole is retransmitted from generation to generation but with anincreasing number of acrolectal features. Aitchison’s discussion of decreolization focuses onquantitative decreolization, coming in her section on language death, not on pidgins and cre-oles, and is preceded by the heading “language suicide.” This gives a sense of agency andintentionality to the decreolization process that is absent from other descriptions of it. Afterall, not all cases of language or dialect shift are fully intentional; sometimes it is just a case of bilinguals who do not separate their languages as completely as possible, bringing aboutdeath by borrowing. Mufwene (1994) warns against the conflation of qualitative and quanti-tative decreolization, as neither one logically implies the other. In this respect, decreolizationsubsumes two very different processes: structural change and community abandonment. Thisin itself is not a reason to consider decreolization to be neither unique nor internally con-sistent. However, it is enough to say that decreolization, subsuming these two processes, is adescription of a result , rather than a process.4. What makes the language less creole?The postulation that a language is de-creolizing indicates that it is becoming less cre-ole. This notion of course presupposes that there is some linguistic definition of a creole fea-ture that can be lost. However, pace McWhorter (1998), who postulates a clustering of threefeatures (roughly, semantically transparent derivation, little use of tone, lack of inflectionalmorphology) as the defining linguistic features of creoles, there is no widely accepted defini-tion of creole without reference to sociohistorical emergence 2 . This point should not be over-looked. With other processes of linguistic change such as grammaticalization or apocope , weneed no other information about languages beyond what they looked like at each relevantstage. With the sufficient, relevant, linguistic data— which need not be recorded attestationsbut may also consist of historical reconstruction—, we can identify these other linguisticchanges or rule them out. Even with the same amount of relevant linguistic data, we canidentify decreolization if and only if we have sociohistorical information about the emer-gence of the language. This calls into question whether decreolization is in fact unique, for if we see similar processes or outcomes to decreolization occurring in languages that are in alexifier/creole relationship long in the past with no sociohistorical documentation, we wouldwrongly conclude that something other than decreolization is taking place. 2 An anonymous reviewer points out that one way to define a creole linguistically is against its lexifier, in thatthe former will be morphologically simpler than the latter. This still seems to assume sociohistorical knowledge.How would we distinguish between sister languages descended from the same source language where one hasbecome significantly more analytic than the other (and possibly undergone other simplifications) and lexifi-er/creole languages where the creolization took place so far in the past that there are no records of it? READ PAPER