Interactive alignment in dialogue: evidence for parity of representations in production and comprehension.

Simon Garrod
University of Glasgow

Pickering and Garrod (2002) propose a model of language processing in dialogue which assumes that interlocutors align their linguistic representations at many levels. Interactive alignment comes about because of three things: (1) Priming of representations at many levels ranging from the phonological to the semantic and situational, (2) Parity of representations used in comprehension and production at these various levels, and, (3) The reciprocal nature of interactive dialogue.

In this paper I will present some preliminary evidence in support of the second of these assumptions: parity of representations used in production and comprehension. The experiments investigate parity at the semantic, syntactic and phonological level. Parity of semantic and syntactic representations is tested using a comprehension-to-production priming paradigm in which participants produce a previously presented word immediately after hearing a semantically and/or syntactically congruous or incongruous sentence fragment. Both the latency of production of the previously presented word and the error pattern indicate a strong priming effect from comprehension to production. Furthermore, the semantic and syntactic priming effects are both present and do not interact with each other.

Parity at the phonological level is tested using a comprehension to production version of the tongue twister paradigm developed by Wilshire (1999). The idea behind the experiment is to establish whether simply hearing phonologically confusable words is sufficient to produce the tongue twisting effect associated with producing the same sequence of words. Pilot data indicate that such an effect is present.

The results of these experiments will be discussed in relation to the interactive alignment model of dialogue processing.

References

Pickering, M. & Garrod, S. (2002) Toward a mechanistic psychology of dialogue. Manuscript under revision for Behavioral and Brain Sciences.

Wilshire, C.E. The "tongue twister" paradigm as a technique for studying phonological encoding. Language and Speech. 42, 57-82.