Accommodating ‘unaccustomed pragmatic spaces’ in Arbib's model Comment on “Towards a Computational Comparative Neuroprimatology: Framing the language-ready brain” by Michael A. Arbib

Abstract

Arbib's article [1] offers a sophisticated and convincing account of the evolution of human language that does not shy away from nailing together neurophysiology and the forms and functions of language. The core recognition of what language does, rather than just what language looks like or how its forms are generated, gives the model a high level of explanatory significance. This commentary explores interaction in the context of his account of comprehension, section 4.3.2, offering some observations that potentially test and support the model. They expand Arbib's existing consideration of agrammatism, where he suggests that difficulties with managing the linguistic system (Grammatical Route, G, including Light Semantics, LS) are compensated for using world knowledge (Heavy Semantics, HS). That is, if, in comprehension, one is not sure how the words go together, one increases attention to the pragmatics generated by the observable context and general memory to fill in the gaps. I introduce the opposite situation, where, in dementia of the Alzheimer's type, compromises to world knowledge and pragmatics create unstable HS, and may generate ambiguity in the linguistic input. I consider how Arbib's model sheds light on the typical fixes for this eventuality, and also venture a link back to the protolanguage account.