The “language of thought hypothesis” (LoTH), famously associated with the late Jerry Fodor and not particularly popular in recent years, appears to be recovering some ground, according to Mandelbaum and colleagues, whose “Problems and Mysteries of the Many Languages of Thought” I cited in Chapter 1.
Mandelbaum develops this further with colleagues Quilty-Dunn and Porot in their “The Best Game in Town: The Re-Emergence of the Language of Thought Hypothesis Across the Cognitive Sciences” by proposing the “six core properties of LoTs: (i) discrete constituents; (ii) role-filler independence; (iii) predicate-argument structure; (iv) logical operators; (v) inferential promiscuity; and (vi) abstract content. These properties cluster together throughout cognitive science. Bayesian computational modeling, compositional features of object perception, complex infant and animal reasoning, and automatic, intuitive cognition in adults all implicate LoT-like structures. Instead of regarding LoT as a relic of the previous century, researchers in cognitive science and philosophy of mind must take seriously the explanatory breadth of LoT-based architectures.”
What I find fascinating about this is that it might also describe the way that Google’s PALM and other LLM entities “think” (my scare quotes intentional here), especially since it seems that these may soon be able to outdo humans in both “inferential promiscuity” and “abstract content.” Or, if it does not, it may be even more fascinating (and would have been to Fodor himself, I suspect) as AI researchers start to look inside the black box of LLM models.
Keywords In and Out of Context 