What are symbols in language evolution?
Justin Sulik (University of Edinburgh)
Cognitive semiotics represents the combination of novel methods with ancient categories (e.g. `symbol'). But are traditional definitions of such categories still fit for purpose in modern cognitive science? Is it scientically useful for cognitive semiotics to define symbols as `arbitrary' or `conventional'? Naturally, an assessment of fitness-for-purpose depends on that purpose: a specific research question and the methodologies appropriate to that question. One important topic for cognitive semiotics is the evolution of symbolic communication in our species. In this presentation, I evaluate the fitness-for-purpose of traditional definitions of `symbol' relative to this particular question, rather than `symbol' in general. I provide theoretical and empirical support for an internal definition of `symbol' (i.e. one relative to the cognitive processes needed in interpretation) rather than an external definition (i.e. one relative to features of the sign, such as its being arbitrary or conventional). This allows that traditional denitions may be well suited for other research questions.
First, I explore the notion of arbitrariness in human and animal communication and show that, depending on one's sense of `arbitrary', defining symbols as `arbitrary signs' either results in claims too subjective for scientific progress or focuses unduly on superficial perceptual similarities in a way that runs counter to non-perceptual differences in communicative behaviour: meaning is not ultimately a perceptual issue.
Secondly, I unpack two accounts of `convention', one based on game theory (Lewis, 1969, 1983) and the other on imitation (Millikan, 2005). I argue that, for game-theoretic conventions to play an explanatory role in the evolution of symbolic communication in our species, they require salience-deciding inference (Cubitt and Sugden, 2003; Postema, 2008). For imitative conventions to do the same, they require relevance-deciding inference (Csibra, 2003). I link salience- and relevance-deciding inference to claims about context-deciding inference in pragmatic interpretation (Sperber and Wilson, 1995) to show that what is special about human (or human-like) communication is a matter of this open-ended or comparatively unconstrained form of inference, which I identify as abductive or insightful hypothesis generation.
Finally, I present several cognitive semiotics experiments where participants make guesses about the meaning of novel signs in communicative contexts that are unconstrained in the ways set out above. These show that adopting this inference-based denition of `symbol' has practical, testable implications for language evolution.
Csibra, G. (2003). Teleological and referential understanding of action in infancy. Philosophical Transactions of the Royal Society B, 358:447-458.
Cubitt, R. P. and Sugden, R. (2003). Common knowledge, salience and convention: A reconstruction of David Lewis' game theory. Economics and Philosophy, 19:175-210.
Lewis, D. (1969). Convention: A philosophical study. Harvard University Press, Cambridge, MA.
Lewis, D. (1983). Philosophical papers, volume 1. Oxford University Press, New York.
Millikan, R. G. (2005). Language: A Biological Model. Oxford University Press, Oxford.
Postema, G. J. (2008). Salience reasoning. Topoi, 27:41-55.
Sperber, D. and Wilson, D. (1986/1995). Relevance: Communication and Cognition. Blackwell Publishing, Malden, MA, 2nd edition.