Gabriel,

I would like to get into a dialogue with you about this. I could give a talk (Zoom) that poses the question of what role evidence from human natural language plays in computational neuroscience. I think, for example, you either need to accept that nouns and verbs in natural (spoken) language are observable facts or not. No human can observe every natural language but if anyone who is fluent in a natural language can prove this is not the case for that language, that would be interesting. We know for example for hundreds if not thousands of years, people picking up a new language relied on the assumptions of grammar in order to quickly pick up those languages. Those experiments were done, even if you do not recognize them.

The theory of theory questions come spilling quite rapidly after that with, I believe, a lot of positive consequences for a motivated and productive form of computational neuroscience. And, there are plenty of quantitative results.

For example, in my book I point out an observation anyone can make about verbs. First, virtually all verbs in any language are action verbs that allow an agentive role. In any such, there is always a dual sense possible. One is that the agent intended to do it, the other not and it was perhaps forced by a different agent. Such dual senses are at many levels of natural language. A more obvious one is that all languages have agreed on opposites. There are literatures, some quite old, where these kind of observations have stood the test of time and experience. Do you relegate them to no importance, or perhaps, of great importance? I believe any acceptable theory of computational cognitive neuroscience must account for these observations which we can find self-evident, or axiomatic.

Other classes include modelling and simulation results. I have long argued that the particular details of the discrimination functions in neural computations are probably unimportant for a more general theory of computational neuroscience.

Here are some more current medium papers on these subjects.

https://medium.com/liecatcher/how-your-brain-computes-41ebe7428ff9

Here are some testable conclusions about the evolution of natural language that come from a study of the computational basis for human lies.

https://medium.com/liecatcher/the-natural-evolution-of-human-lies-655e983ee6c6