Laboratory experiments on the linguistic consequences of communicative interaction
If enough people take the same shortcut across a lawn, their footsteps will eventually create a path marking the route. While such a path certainly results from human action, it is not deliberately designed in the way that paved roads and highways are. Could this be a useful analogy for understanding the design of language? I will present experimental evidence indicating that — if we take a cultural-evolutionary approach to language — the answer is likely to be yes. In particular I will present a set of experiments investigating the emergence, through repeated communicative interaction, of two different kinds of structure: combinatoriality (the recombination of a small set of basic forms, such as letters or phonemes, into more complex forms) and dialectal variation (the use of particular linguistic variants by members of particular social groups). In both cases structure emerges that resembles the structure of existing languages and does so in an environment in which variables can be directly manipulated that are hard to isolate outside the laboratory. Finally, I briefly illustrate how such methods can be applied to non-linguistic tasks, shedding light on how structure may arise in similar ways in different kinds of cultural behavior.