It's only a model: Discussion note on ‘Small model languages as tools for reflection’

Abstract

Rastall’s paper ‘Small Model Languages as Tools for Reflection’ observes that when linguistic models fall short of an accurate representation of language, this offers a rare opportunity to question assumptions and observe linguistic facts. The only reason a model’s proponents might be distressed when a model is proven inadequate is if they harboured the illusion that the model was an accurate representation of language production, or even of cognition. As Rastall points out, no model can make this claim. Every model is necessarily an imperfect representation of reality.