Title: Variationist Sociolinguistics -syntax

The variationist sociolinguistics model assumes that the human language faculty accommodates and generates language variation, and that the workings of grammar may have a quantitative and non-categorical component. By means of the linguistic variable, one describes and analyzes orderly heterogeneity in language use. The linguistic variable as a structural unit depends on the model of syntax one employs. It is argued that there exists a fairly direct connection between grammar and language use. A challenge for interdisciplinary accounts between the variationist sociolinguistic framework and formal syntactic theory is how to account for intraand interspeaker variation. Biography Leonie Cornips wrote her dissertation (University of Amsterdam, 1994) on syntactic variation in the speech community of Heerlen, a former coalmining town of 90,000 inhabitants in the southern part of The Netherlands situated near the Belgian and 2 German border. Her study of Heerlen Dutch in the principles and Parameters approach shows that when the theoretical primitive, i.e. lexical properties of functional categories causing syntactic variation is located in a group of speakers' grammars, these speakers show similar correlations with respect to their social variables. Since 1994 she works at the departmentment of Language Variation at the Meertens Institute of the Royal Netherlands Academy of Arts and Sciences. At this moment, she investigates syntactic variation in new developing regional variaties due to induced language contact situations between standard Dutch and local dialects and emerging ethnic varieties (Moroccan-, Turkish-, Surinamese-Dutch). She co-edited two volumes of The Linguistics in The Netherlands (John Benjamins 2003 & 2004). She co-edited (with Dr. Karen Corrigan, University of Newcastle upon Tyne) an interdisciplinary volume of formal syntax and sociolinguistics e.g Biolinguistic and Sociolinguistic Accounts of Syntactic Variation: Doubts about Dichotomy ( John Benjamins CILT). Variable rule/Linguistic variable In the 1960's both sociolinguistic and formal syntax models contained formal rules that could be applied obligatorily or optionally. Formal rules in the earliest Chomskyan transformational model were transformations that connected 'deep' structures with 'surface' structures on the basis of rewrite rules. Optional rules, for example, derived passive, negative or question sentences from declarative sentences. Labov introduced the concept of variable rule as an extension of this optional rule to include social and stylistic e.g. external dimensions of language use along with 3 linguistic e.g. internal dimensions. There is some confusion about the notion of the variable rule since in the earliest literature it refers to both a theoretical model how to analyze and account for language variation as to a method of statistical analysis as well (Cedergren and Sankoff 1974). However, both paradigms soon followed their own avenues, that is: the successive transformational models assume the existence of categorical rules only while variationist sociolinguistics has maintained the notion of the optional rule. The two perspectives on the nature of formal rules reflect deepseated differences between the two models. The variationist ideas are that the output of a linguistic rule can be probabilistic rather than discrete, and that a linguistic constraint can have a quantitative rather than deterministic effect on the outcome of the process. Formal syntax, however, postulates a blind, deterministic application of a series of procedures given a certain starting point (Chomsky 1995). Labov has modified the generative model with the variable rule as a means to accommodate interspeaker and intraspeaker variation. The variationist sociolinguistic practice that has evolved from studies of language variation and change since then takes the principle of accountability as basic (cf. Sankoff 1990: 296). This principle states that the variants belonging to the same syntactic variable must be specified by the total number of occurrences and the potential occurrences or non-occurrences, in the variable environment, i.e. it ranges between 0% and 100%. It guarantees that the entire range of variability present in the data will be dealt with. The principle of accountability inevitably follows the synonymy principle. This principle is the prerequisite for variants to be assigned to the same linguistic variable; that is, only syntactic variants that are equivocal with regard to referential meaning i.e. variants that are “alternate ways of saying ‘the same’ thing” belong to the same variable. In 4 practice, the assignment of meaning or function of syntactic variants is considered as problematic (Lavandera 1978). The indeterminacy of synonymy/functional equivalence is considered not so problematic when examining the phenomenon of verbal agreement, which is present in many vernacular dialects of English all around the world (cf. King 1994 voor French varieties in Canada), as illustrated in (1): (1) a. We parch it (cf. Poplack & Tagliamonte 1989:49) b. We parches the coffee and copula variability as in African-American Vernacular English (AAVE) (Rickford et al 1991) or creole varieties as illustrated in (2): (2) a. I am just telling the boys (cf. Weldon 2003:43) b. I'm gonna get me a blue tag c. I feel like I Ø fourteen The examples above illustrate that two or more forms of functional categories such as number agreement in (i) and tense in (2) are considered to belong to one linguistic variable. Other linguistic variables that can be found in variationist sociolinguistic practice in the last fifteen years in journals such as Language Variation and Change and Journal of Sociolinguistics are, for example, null versus overt subjects in varieties of Spanish (Cameron 1993) and Bislama (Meyerhoff 2000), particles in Japanese (Takano 1998) and Mandarin Chinese (Ziqiang Shi 1989), null versus overt objects in Hungarian (Kontra 2001), double objects in Dutch (Cornips 1998), 5 relativization strategies in English varieties and AAVE (Tottie & Rey 1997), and negation in Scots English (Smith 2001). Furthermore, these studies examine non-standard rather than standard varieties and they may differ considerably in the frequency of the tokens collected. It may range from 29 (Cornips 1998), 164 (Tottie and Rey 1997), 670 (Meechan and Foley 1994) to 6809 (Tagliamonte 1998). It appears that both a low and high frequency of (morpho)syntactic variables may reveal significant correlations between the use of linguistic variants and social dimensions in the speech community. Structural representation of the syntactic variants In addition to the synonymy principle, the notion of the linguistic variable as a structural unit was also based on the assumption that the variants have an identical underlying structure or representation, which is subject to variable surface realizations (Winford 1996: 177). So, the alternation between active and passive examples was considered to be instances of different surface manifestations of the same underlying or deep structure. Santorini (1993) and Pintzuk (1995) are examples of studies within the Government and Binding framework where it is argued that different word orders such as SOV and VSO are derived by movement. However, in later Generative models, the idea of a derivational model was abandoned in favor of a configurational model (most recent Minimalism), in which a single representation is subject to various constraints. An example is the study of Meechan and Foley (1994) where variable verbal number agreement in existentials is analysed as a reflection of different configurational positions of the postverbal NP (a second example is Meyerhoff (2000)). 6 As a consequence, the original notion of the linguistic variable as a structural unit such that syntactic variants are different surface realizations of an underlying structure has been lost too. The consequences are that the syntactic variable as the structural unit have to be defined again depending on the particular model of syntax one employs. Other studies that can be mentioned in the older Parameters and Principle framework are Henry (1995) and Cornips (1998) where variants constituting the linguistic variable are brought about by different properties of functional categories. Since the consecutive models in Generative Grammar follow each other at rather high speed different make-ups of the structural unit can be detected in the last decades. An overall theory of grammar and the locus of variability Many researchers claim that a bridge between variation and theory remains possible; that is, it is assumed that there exists a fairly direct connection between grammar and language usage. The organization of the grammar may be reflected in the patterns of usage (Taylor 1994) or quantitative results may lend strong support to structural analysis (Pintzuk 1995). Or to putit difefrently, through the use of variationist methodology and knowledge about the factors contributing to syntactic variation, one catches "a glimpse of grammatical structure" (Meechan and Foley 1995:82). Wilson and Henry (1998:8) phrase the advantages of an overall theory of grammar including variation and formal theory as follows: "...we may be able better to understand language variation and change as they are driven by social factors but constrained (at one level) by the nature of possible grammars." One problematic issue in this socially realistic account (Henry and Wilson 1998) is how to account for the fact that individual speakers can use several variants of the syntactic variable when 7 maintaining the same style level. This issue is related to the questions within the successive Generative models about the locus of syntactic variation, its restrictions and predictions. In the literature, two alternative approaches to this 'choice' are suggested (Muysken 2005). Either, the 'choice' is put outside the grammatical mechanisms (Adger and Smith 2005, Kroch 1989, or it is put inside the grammar by re-introducing optional rules (Henry 2002, Wilson and Henry 1998). The first option is advocated by Kroch and his associates who claim that the grammar is a blind, autonomous system and the notion of 'choice' (optionality, variability) is not part of it. Instead, the individual speaker revealing variability has separate or competing grammars. Adger and Smith (2005) also argue that the notion of 'choice' cannot be accounted for within the autonomous grammar. In contrast with Kroch's vision, this doesn't imply that individual speakers "have different grammars, per se, but rather a range of lexical items open to them, some of which will have syntactic effects". In their analysis, the notion of 'choice' concerns the level that serves as the input for the autonomous, grammatical system. However, Henry (1995) argues that individual grammars include variability and, consequently, the speaker has a real choice in terms of syntactic operations, for instance, optional verb movement. However, all agree that an analysis of language use (quantitative analysis, intraand interspeaker variation) in addition to formal theory e.g. the study of both internal and external constraints has the potential not only to assess but also to refine theoretical conceptions of linguistic structure. 8 *I like to thank Karen Corrigan, Frans Hinskens, Pieter Muysken and Jennifer Smith for their valuable comments on a draft version of this paper.