Methodology in Cognitive Linguistics
Why linguistic needs an empirical revolution
Let us now turn to linguistics: why could we say that more empirical research of the type illustrated and demarcated above would be beneficial for the development of linguistics? Theoretical linguistics is characterized by the existence of a multitude of theories: Universal Grammar in the generativist sense, a family of models claiming the label of Functional Grammar, smaller trends like Role and Reference Grammar or HPSG– and of course Cognitive Linguistics, in a number of versions that may each branch off into subversions, and then into alternative realizations of subversions. There is Cognitive Linguistics Langacker-style (Cognitive Grammar in the narrow sense), and there is Cognitive Linguistics Talmy-style– to some extent using the same vocabulary, involving notions like the figure/ground opposition, but still sufficiently different to be easily recognizable by anyone familiar with the relevant literature. And then there is Construction Grammar, which may or may not include Cognitive Grammar (Langacker-style), but which certainly encompasses different approaches like Fillmore and Goldberg-style Construction Grammar, Radical Construction Grammar, and even Embodied Construction Grammar.
Superficially, it might seem that the existence of different “theories” is not really typical of linguistics, as the foregoing passage may seem to suggest. In psychology (to take a discipline that is not too far removed from linguistics), one may come across different theories about how categories are stored in the mind: you get exemplar models that assume that the bulk of the work is done via the knowledge we have of typical members of a category, and you get concept-based models that rather start from the assumption that there is an abstract, schematic definition of a concept, and then you get mixed models that explore different combinations of exemplar- and concept-based models. Is there no difference with linguistics, then?
The difference seems to be that there is a common, commonly accepted way in psycholinguistics of settling theoretical disputes: experimentation. Given a number of conditions, experimental results decide between competing analyses, and psycholinguists predominantly accept the experimental paradigm as the cornerstone of their discipline. The conditions that need to be fulfilled to make the paradigm work are in principle simple: the experiment has to be adequately carried out, and it has to be properly designed in order to be distinctive with regard to the competing theories. That is to say, you need good experimental training (knowledge of techniques and analytical tools), and you need the ability to define a relevant experimental design. The bulk of the effort in psycholinguistic research, in other words, involves attending to these two conditions: setting up adequate designs, and carrying out the design while paying due caution to experimental validity. The contrast with linguistics will be clear: theoretical multiplicity in linguistics is only weakly constrained by a common methodological basis. This observation may be specified in various ways.
First, systematic theory comparison focusing on specific empirical phenomena is not a dominant activity in linguistics. Even within the realm of Cognitive Linguistics, for instance, an analytical and critical comparison of different approaches (as in Langacker 2005, Nuyts 2005 or Croft and Wood 2000) is exceptional.
Second, as a sociological corollary, linguists flock together with their own theoretical crowd. Each of the approaches has a number of founding fathers and leading figures, of the Chomsky, Halliday, Labov type, and these authorities are the core members of sociological networks that include specialized journals, conferences with a specific orientation, and local curricula that determine how a novice learns to do the job (which would probably mean: learn to work within a given theoretical framework). Breaking through the boundaries of a given approach is not obvious.
And third, linguistics at large does not possess a common empirical ground, in the form of a set of observations derived through a generally accepted method, that plays the same role that experimentation does in psycholinguistics. The latter factor is the crucial one: attempts to compare theories (the first point) require data with regard to which to compare them, and bridging distances between theories (the second point) would have to be built on a systematic comparison of theories. It is probably justified to say that linguistics does not invest a lot in the empirical operationalization of its hypotheses. For instance, in the domain of Cognitive Linguistics, what exactly would it mean that a certain construction is a “reference point construction” in the terms of Langacker (1999)? What are the measurable correlates of analyzing a particular construction in such terms? As it stands, a concept like “reference point construction” is primarily a concept that makes sense (and probably good sense at that) of a number of grammatical phenomena. That is to say, it provides an interpretation for those phenomena, but as we have seen, such an interpretation is merely a hypothesis from the point of view of empirical research, and what we would really like to see are operationalizations of such interpretative hypotheses that allow for testing of the kind discussed before. (Incidentally, readers who might be interested in ways of operationalizing the notion of reference point might have a look at Grondelaers, Speelman, and Geeraerts 2002.)
It would seem, then, that (to use the terminology of a traditional theory of science, see Nagel 1961), linguistics should primarily develop an independent observational language that the different theoretical languages of linguistics can be mapped onto: in the same way in which a psychological theory is translated into a set of statements about the behavior of test subjects under certain experimental circumstances, linguistic theories should be translatable into statements about linguistic behavior that are independent of the theory. Even if we take into account that the demarcation between theoretical and observational languages may not be totally rigid (this is a point made by the theory of science that developed in the wake of Kuhn 1962: observational statements may themselves be couched in a theoretical vocabulary), a lot of progress is possible in the operationalization of linguistics, i.e. in the attempt to turn theoretical concepts and statements into independently testable and theoretically distinctive empirical statements.
Such a systematic attempt at operationalization of linguistic hypotheses would indeed constitute an empirical revolution in linguistics. The observational basis for such a revolution would to some extent overlap with that of psychology: experimental methods and neurophysiological evidence would obviously be part of it. But so would a number of other approaches: survey techniques as used in sociolinguistic research, and corpus linguistics. Generalizing somewhat, one might say that a truly empirical form of linguistics would share experimental techniques with psychology, and survey and field techniques with sociology, whereas corpus analysis would be the most typically linguistic method.
