1 Analyzing language variation: Where sociolinguistics and linguistic typology meet

1 Introduction

This book offers a collection of chapters that explore the interplay between cross-­linguistic and intra-linguistic perspectives to the study of language variation.1

Variation is an inherent property of natural languages and is pervasive across several layers: in the performance of individual speakers, within speech communities, across languages distant in time and space. As a general definition, variation can be understood as the co-existence of formally distinct linguistic forms to express the same content. Nevertheless, the nature of variation itself remains constant irrespective of its locus of manifestation. As Croft (2022: 27) observes "the patterns of variation and change found in [. . .] a particular language are in many cases simply instances of patterns of variation and change found across languages".

Language variation has constituted the core of the research agenda of at least two disciplines of linguistics, that is, variationist sociolinguistics and linguistic typology.2 At a superficial glance, the two appear to deal with quite distinct domains. Typology "concerns itself with the study of structural differences and similarities between languages" (Velupillai 2012: 15), and is based on the study of linguistic phenomena in more or less large samples of languages representing the genetic and geographic diversity of the languages of the world.3 By contrast, sociolinguistics focuses on "the correlation of dependent linguistic variables with independent social variables" (Chambers 2003: ix), and does so by typically focusing on speakers of a single language.4

As a matter of fact, typology and sociolinguistics approach variation with distinct premises and goals. Broadly speaking, typology is interested in crosslinguistic variation because it seeks to explore what the limits of such variation are, and, based on the empirical large-scale study of which linguistic structures are attested and which are not in the languages of the world, it aims at understanding the universal properties of human language.5 Conversely, sociolinguistics takes a keen interest in those linguistic phenomena that have a social meaning, so as to unveil "the mechanisms which link extra linguistic phenomena (the social and cultural) with patterned linguistic heterogeneity (the internal, variable, system of language)" (Tagliamonte 2012: XIV).

What the two fields have in common is the effort to show that language variation, in all its shapes, is not random, but systematically takes place within well-­defined boundaries. As a result, sociolinguistics and typology have developed over time a number of theoretical models and tools that share remarkable similarities. Nevertheless, while a number of studies attempting to address at least aspects of this relationship have appeared in recent years (see e.g. Kortmann 2004a, Trudgill 2011a), a principled discussion on how the two disciplines may interact has not yet been carried out in a programmatic way. This volume aims to fill this gap and to provide a venue for chapters discussing the bridging between sociolinguistic and typological research from various angles, with the ultimate goal of laying out the methodological and conceptual foundations of an integrated research agenda for the study of linguistic variation.

In this introductory chapter we take a closer look at the main intersections that exist between sociolinguistics and typology, with a particular focus on their theoretical and methodological apparatus. The chapter is structured as follows. Section 2 focuses on similarities among the conceptual tools developed by sociolinguistics and typology, while Section 3 deals with shared methodologies and practices. Section 4 addresses the role that diachrony plays in the study of cross- and intra-linguistic variation. Section 5 discusses a number of recent studies that showcase the mutual benefits between sociolinguistic and typological approaches, and introduces the chapters featured in this volume. Section 6 features some conclusive remarks.

2 Analyzing variation: Theoretical models

In addressing the study of variation, sociolinguistics and typology have been confronted with a number of strikingly similar methodological issues, to address which the two fields have, quite independently, come up with comparable theoretical toolkits. In this section, we focus on three key issues, that is (i) the individuation of the linguistic trait to be taken into account when analyzing variation, that is, the notions of variables and comparative concepts (ii) the study of the aggregation of linguistic traits, be it either sociolinguistic varieties or linguistic types (iii) the relationship among linguistic traits and their systematization in terms of implicational hierarchies and universals.

2.1 How to describe variation: Sociolinguistic variables and comparative concepts

The first step in describing variation concerns the need to set the boundaries of the linguistic phenomenon under investigation, that is, the boundaries of the domain of variation. To this end, variationist sociolinguistics relies on the notion of sociolinguistic variable, whose definition is "is the first and also the last step in the analysis of variation" (Labov 2004: 7). A sociolinguistic variable can be defined as the set of "alternative ways of saying "the same" thing" (Labov 1972: 188). In addition to "semantic equivalence", variants as understood by variationist sociolinguistics also correlate with social factors (such as the formality of the context, the geographic origin of the speaker, etc.).

The debate around "semantic equivalence" among variants started in the '70s (see Berruto 2007 [1995]: 139-145 and Tagliamonte 2006: 70-76). Specifically, it soon became clear that while this methodology is particularly apt to describe phonetic variation, its application to other domains of linguistic analysis is far from straightforward. This is a consequence of the fact that more complex linguistic structures necessarily carry a meaning component, which poses a challenge to the individuation of semantic equivalence. This is why Sankoff (1972: 58) broadens his definition of variable and argues that "whenever there are options open to a speaker, we can infer from his or her behaviour an underlying set of probabilities". Similarly, to account for variation outside of the phonetic domain, Lavandera (1978: 181) invokes the principle of functional comparability of linguistic forms in individual contexts. Labov, turning to morphosyntactic variation, offers a narrower interpretation of "semantic equivalence" as pertaining to the referential level only (Labov 1972: 271; 1978; Weiner and Labov 1983).

Over the decades, the boundaries of the notion of variable have been extended so as to include variation at other levels, including morphology, syntax and the lexicon. The most problematic level undoubtedly remains that of pragmatics, and for obvious reasons, if one considers the level of abstractness of the categories used to describe pragmatic phenomena (see Pichler 2013: 6-9; on the variationist approach to pragmatic phenomena see also Tagliamonte 2005, Schneider and Barron 2008, Cameron and Schwenter 2013, and Sansò 2020: 81-104).

Linguistic typology is essentially a comparative branch of linguistics, which deals with the categorization of linguistic traits across languages (Moravcsik 2016). This means that delimiting the domain of variation is a necessary prequisite of the typological inquiry, so as to ensure the comparability and the correct categorization of linguistic traits. As Koptjevskaja-Tamm et al. (2015: 436) put it "cross-linguistic identification of studied phenomena presupposes a procedure which ensures that we compare like with like". This is a relatively straightforward task in the case of the most basic domains of lexical typology (Koptjevskaja-Tamm 2012). Things become more difficult when one tries to compare morphosyntactic entities across languages - which constitutes the core of the typological enterprise (Evans 2020; on the comparability issue see in particular two special issues of Linguistic Typology 20(2) [2016] and 24(3) [2020] and Alfieri et al. 2021) - since most of the notions developed in the Western grammatical thought (and on the basis of Classical languages) cannot be straightforwardly applied across languages: this concerns basic notions such as subject and object (cf. Bickel 2010; Witzlack-­Makarevich 2019), subordinate clause (Cristofaro 2003) and even the notion of 'word' itself (Haspelmath 2011).

To overcome these difficulties, and achieve a meaningful comparison disregarding the structural and formal peculiarities of...