Methodology in Cognitive Linguistics
Why Cognitive Linguistics could (easily) be an empirical science
If we assume that linguistics is indeed in need of an empirical revolution that will eventually bring about an intertheoretical level of comparability that is similar to what we find in experimental psychology, what are the chances for Cognitive Linguistics of being part of this methodological transformation of linguistics? Before we look at the actual state of affairs, we should establish whether the specific theoretical stance of Cognitive Linguistics is conducive to the empirical turn. If we look at the way in which Cognitive Linguistics describes and defines itself, does its specific form of linguistics have features that would either encourage or impede an empirical approach?
Both are in fact the case. On the one hand, at least three features of Cognitive Linguistics would seem to create favorable circumstances for an empirical approach: its very cognitive nature, the usage-based perspective it takes, and the importance it attaches to contextualizing linguistic structures. Conversely, on the other hand, there is one dominant feature of Cognitive Linguistics that could work in the other direction: the emphasis on the analysis of linguistic meaning over linguistic form might go hand in hand with the view that only introspection is a reliable method of direct access to semantic phenomena. Let us now have a closer look at each of these factors.
Cognitive Linguistics, if we may believe the name, is a cognitive science, i.e. it is one of those scientific disciplines that study the mind (which, by the way, includes the totality of the mental life of human beings, and not just knowledge in the restricted sense). It would seem obvious then, that the methods that have proved their value in the cognitive sciences at large have a strong position in Cognitive Linguistics: the experimental techniques of psychology, computer modelling, and neurophysiologic research.
We will discuss below whether this is indeed the actual state of affairs (it is not entirely, as will become clear), but let us perhaps establish first that the conclusion as such is not unproblematic: a self-conception as a cognitive science does not necessarily entail the application of an empirical methodology. Generative grammar, in fact, is a clear case in point: as part of the so-called cognitive revolution of the 1950s, its mentalist self-conception definitely means that it considers itself to be a cognitive science - but the methodology of generative grammar is predominantly introspective. The argumentation is well-known: the grammaticality of sentences is determined by (or at least restricted by) innate principles of grammar, and native speakers have immediate access to the relevant grammaticality judgements. Given the assumption of a completely homogeneous language community, this epistemological starting-point does not invite empirical research. (When the assumption of a completely homogeneous community is relaxed, attempts at statistical analysis of grammaticality judgements may enter the picture. This is a recent, albeit marginal trend in generative grammar: see e.g. Keller and Sorace 2003.)
But if, then, generative grammar is both a cognitive form of linguistics and a largely introspective one, why would Cognitive Linguistics be more prone to empirical research than generative grammar? Is there a sense in which Cognitive Linguistics is a different type of cognitive enterprise than generative grammar? And does this difference suggest a different attitude with regard to empirical research? The specific way in which Cognitive Linguistics conceives of its “cognitive” character would indeed seem to favor a more empirical approach to linguistic methodology. In generative grammar, the ultimate foundation for an introspective methodology is the autonomy of language and linguistics, whereas Cognitive Linguistics conceives of language as integrated with the other cognitive capacities of man: giving up the autonomy assumption motivates a broader range of methodological techniques, in line with the practice used by the cognitive sciences.
Turning to a second relevant feature of Cognitive Linguistics, the appeal of empirical methods within the cognitive approach is boosted by the growing tendency of Cognitive Linguistics to stress its essential nature as a usage-based linguistics– a form of linguistic analysis, that is, that takes into account not just grammatical structure, but that sees this structure as arising from and interacting with actual language use. The central notions of usage-based linguistics have been programmatically outlined in different publications (specifically, see Barlow and Kemmer 2000), and an increasing number of recent volumes show how the program can be put into practice. The link between the self-awareness of Cognitive Linguistics as a usage-based form of linguistic investigation and the deployment of empirical methods is straightforward: you cannot have a usage-based linguistics unless you study actual usage– as it appears in corpora in the form of spontaneous, non-elicited language data, or as it appears in an online and elicited form in experimental settings.
As a third theoretical feature of Cognitive Linguistics promoting an empirical methodology, we may again point to its contextualized conception of language. The relevant context, however, does not just consist of the other human cognitive capacities (everything that links up with the embodied nature of the mind), it also involves society and culture. In this sense, an interest in language variation forms an integral part of Cognitive Linguistics. Now, language variation has been studied in Cognitive Linguistics and adjacent approaches primarily from three points of view: from a diachronic perspective, including grammaticalization research (see e.g. Geeraerts 1997; Bybee 2001; Hopper and Traugott 1993; Heine 1991), from a comparative and anthropological point of view (see e.g. Pederson 1998; Kövecses 2000; Levinson 2003), and from a developmental point of view (see e.g. Tomasello 2003). Language-internal variation and sociolinguistic diversity has been much less studied, but still, we may note a number of developments within Cognitive Linguistics that are likely to contribute to an increased interest in sociolinguistic research.
First, there is the interest in cultural models and the way in which they may compete within a community: see e.g. many of the papers collected in Dirven, Frank, and Pütz (2003). Specifically, it has been pointed out (Geeraerts 2003) that such models may also characterize the beliefs that language users entertain regarding language and language varieties. In this way, Cognitive Linguistics may link up with existing sociolinguistic research about language attitudes.
Second, there is a growing tendency in the theoretical conception of language entertained by Cognitive Linguistics to stress the social nature of language. A number of researchers have started to investigate social variation from the perspective of Cognitive Linguistics: see e.g. the work by Kristiansen on phonetic variation (2003). In the realm of lexicological investigations, the research tradition started in Geeraerts, Grondelaers, and Bakema (1994) and continued in works such as Geeraerts, Grondelaers, and Speelman (1999), or Speelman, Grondelaers, and Geeraerts (2003), integrates theoretical ideas from Cognitive Linguistics with a variational perspective and a corpus-based methodology.
Third, the usage data that Cognitive Linguistics, as a self-styled usage-based approach, is confronted with, automatically include sociolinguistic variation. Assuming that the primary empirical source for a usage-based approach to linguistics consists of corpus materials, then more often than not, the corpus will not be internally homogeneous: because the texts collected for the corpus come from various sources, it will not be known in advance whether the variation that may be observed in the corpus is due to lectal factors or not. (The term lectal is used here to refer to all types of language varieties or lects: dialects, regiolects, national varieties, registers, styles, idiolects, whatever). As such, filtering out the effects of such factors will be necessary for any cognitive linguistic attempt to analyze the usage data– even if the analysis is not a priori interested in lectal variation.
The importance of socially and culturally determined variation for Cognitive Linguistics has an obvious methodological correlate: these socially oriented forms of Cognitive Linguistics would naturally support the adoption of the quantitative, empirical methodology that is dominant in sociolinguistic research at large. We may conclude, in short, that using the main types of empirical methodology is motivated by the theoretical stance of Cognitive Linguistics: its very cognitive nature, its usage-based perspective, and its contextualizing approach may lead in various ways to the use of experimental data, corpus material, survey techniques, and advanced quantitative analysis.
