The Adaptable Mind: What Neuroplasticity and Neural Reuse Tell Us about Language and Cognition - John Zerilli 2021

A Pivot to the Neurosciences
Modules Reconsidered: Varieties of Modularity

Evidence of neural reuse points to an overall picture of the brain that has disruptive implications for the modularity of mind, particularly for classical varieties of the theory such as Fodor’s, massive modularity, and Adaptive Control of Thought—Rational (ACT-R), which all posit modules for high-level cognitive functions or proprietary domains. No doubt many will resist this assessment. I certainly have sympathy for the tradition of functional decomposition and shall not in any case be recommending that we dispense with modules here. Nonetheless, such evidence of reuse as we have clearly does point to “the need for a supplement to business as usual” (Anderson 2010, p. 249).

The central problem for modularity, at least as it has traditionally been understood, is that modules talk lends itself most naturally to the analogy of bricks and mortar, or the assembly of component parts. As an intricately dense network of synaptic connections, electrical signals, and neuromodulatory dynamics, however, the brain is nowhere obviously organized in this bricks and mortar sort of way, even where it sometimes proves fruitful to account for neurobiological function in mechanistic compositional terms (Craver 2007; Bechtel 2008b). The question is whether the bricks and mortar analogy is so far superseded by the network analogy that there is no longer any residual value in speaking of modules at all. If the brain is not obviously or even predominantly an assembly of functional components, surely it would not be unreasonable to hope that any theory having as its target the mind’s functional organization would adequately accommodate itself to this fact.

The brain’s network structure notwithstanding, metaphors, it seems, die hard, especially ones freighted with as much philosophical baggage as modularity. It may be that metaphors are all we have, but if so, we are going to need the right ones. To those reluctant to give up on the modular perspective, I hope my own recommendation of a substantial yet cautious reform may offer some consolation. My proposal is simple—that we recalibrate our notion of modules in deference to what currently passes for a module in contemporary mainstream neuroscience. Cognitive scientists and philosophers whose work is attentive to the neurosciences already think in these terms, and it is not hard to appreciate why: when it comes to modularity, which concerns the functional organization of the mind, psychological theorizing is even more constrained by issues of implementation than might generally be the case. For some reason, however, many philosophers continue to talk about modules in a manner conveying either ignorance of what neuroscience has to say about the structure of the brain, or else a breezy indifference. A reorientation towards neuroscience entails a shift of focus away from understanding modules as unimodal higher level cognitive mechanisms and towards a conception of modules as metamodal (i.e., reusable) nodes subserving exiguous low-level subfunctions. I argue that a module built on this pattern, sometimes called a “brain module,” can serve as an appropriate revisionary benchmark for cognitive scientists and philosophers of psychology still wedded to the idea of classical modules. Notice, however, that this proposal also entails a certain agnosticism regarding the prospects of modularity in the long run. Being sensitive to developments in the neurosciences means being willing to part with long-cherished notions if need be. It so happens that further evidence of neural reuse presented in the next chapter may necessitate a profounder shift away from traditional modules than the one I am currently recommending. Hence I am urging a recalibration in the face of developments that will either, if all goes well, allow us to safeguard a respectable (though revised) notion of modularity, or, should things not turn out so well, undermine its rationale comprehensively—this is where the real battle lines are being drawn. Later on I shall suggest one way that we might usefully conceptualize the issues presented by these developments. Still, the broader point remains: if modules exist at all—a question on which it pays not to be dogmatic one way or the other—they will not resemble the modules of classical cognitive science.

In the next section, I provide a rough sketch of the varieties of modularity one might expect to come across in the cognitive sciences. In this section I also defend what I take to be the sine qua non of modularity; namely, functional dissociability. This will be important in heading off an obvious objection to the argument I am making here: that modules can always survive qua abstract, higher level/gross functional “systems.” I follow this section with a basic account of the brain module. The next chapter pursues at greater length the central question of this book—Whither modularity?—in the hope of demonstrating why neural reuse points us in the direction of something like the brain module.