Maryland responds on linguistics and LLMs

Critiques of "optimism" in target article.

Last summer in Behavioral and Brain Sciences, Richard Futrell (Irvine) and Kyle Mahowald (UT Austin) "offer[ed] an optimistic take on the relationship between language models and linguistics." This summer the journal has a special issue of responses, including several from past and present Terrapins: Sathvik Nair, Katherine Howitt, Philip Resnik, Jeff Lidz, Colin Phillips and Alayo Tripp *19. 

• Sathvik Nair and Colin Phillips / Across the levels of analysis: Explaining predictive processing in humans requires more than machine estimated probabilities

  Under the lens of Marr’s levels of analysis, we critique and extend the authors’ two points about language models (LMs) and language processing: first, predicting upcoming linguistic information based on context is key to language processing, and second, that many advances in psycholinguistics would be impossible without LLMs. We also outline directions combining LLMs’ strengths with psycholinguistic models.

• Philip Resnik / Are language models models?

  Futrell and Mahowald claim language models (LMs) “serve as model systems,” but an assessment at each of Marr’s three levels suggests the claim is clearly not true at the implementation level, poorly motivated at the algorithmic-representational level, and problematic at the computational theory level. LMs are good candidates as tools; calling them cognitive models overstates the case and unnecessarily feeds large language model hype.

• Jeff Lidz and Katherine Howitt / LLMs are not children: They have to earn our love

  Futrell and Mahowald argue that the success of large language models should move the field away from the formal structures of generative linguistic theory. The limited success of these models falls short of formal linguistic theory in explaining both the character of human languages and understanding the trajectory of child language acquisition.

• Alayo Tripp / Sociopolitical ramifications of language models make them worth worrying about

  Regarding the utility of language models for linguistic research, Futrell and Mahowald advance a crackpot realism, wherein the concerns of a powerful elite are portrayed as “realistic” in a sense which is technocratic and detached from broader human consequences.