Why Neurons Don't Have Conventional Content

Finally, Adams and Aizawa reject the notion that cognitive content could be conventionally determined. They do this because agreement on what an artifact means is dependent on the artifact being publically accessible; for example, we can make "bad" or "cool" into positive adjectives by agreement. However, we cannot do this with neuronal states; we cannot agree that a group of neurons will mean something by agreement. This, Adams and Aizawa claim, gives us reason "to believe that cognitive content is not normally derived via any sort of social convention" (ibid., p. 73).

I'm not sure that our argument here comes through in Menary's explication.  Take things like white flags, idiot lights on automobile gas tanks, and the bells on timers.  One thing that seems to be requisite to establishing their conventional content is that they be publicly accessible.  One can easily and directly tell when they are tokened.  By contrast, we cannot easily and directly tell when particular brain states, such as when certain neurons are active, are tokened.  We might infer that some brain-internal meaningful state occurred, but that would not be like the detection of a token of surrender flag.

Thinking about this now, I can see someone going after this argument.