Summary - Modules Reconsidered: Varieties of Modularity

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

Modules Reconsidered: Varieties of Modularity

In recent decades, neuroscience has challenged the orthodox account of the modular mind. As I have shown, one way of meeting this challenge has been to go for increasingly “soft” versions of modularity, and one version in particular, which I dub the “system” view, is so soft that it promises to meet practically any challenge neuroscience can throw at it. But an account of the mind that tells us that the mind can do different things, even interesting things, is not itself necessarily an interesting account. In this chapter, I have reconsidered afresh what we ought to regard as the sine qua non of modularity, and offered a few arguments against the view that an insipid “system” module could be the legitimate successor of the traditional notion. In part my arguments can be read as a plea for the precise use of language, but there is more than pettifogging behind this plea.

1 Defenders of massive modularity also part company with Fodor’s “central”/“peripheral” distinction. Fodor’s hypothesis is that only peripheral systems are likely to be modular “to some interesting extent” (1983, p. 37); i.e., sensory input and motor systems. Proponents of massive modularity think that the central systems will be modular, too; i.e., those involved in higher perceptual function, belief-fixation, and inferential reasoning (Sperber 1994, 2002; Carruthers 2006; see also Barrett & Kurzban 2006 and Prinz 2006 for reviews). See my § 7.2.2.

2 In § 5.2, I consider whether it is possible for systems consisting of shared domain-general parts to be functionally dissociable. This is the same as asking whether higher-level/gross cognitive functions could persist as functional modules. For now, we can assume the answer is no.

3 Its modification may of course indirectly impede a comparable “downstream” system; i.e., one at the receiving end of its efferent projections.

4 What I am calling the “primitives” {p1, p2, p3 . . . pn} need not be neurons: they could be specific subneural processes or ions.

5 Brain regions that are domain-general in the way envisaged by theories of neural reuse may of course ultimately prove not to sustain completely generalizable and projectable accounts of local function. The ability of a brain region to maintain a set of stable input-output relations, and hence to be truly dissociable, may be compromised by the effects of the neural network context. I pursue this topic in Chapter 5.

6 Decomposability and modularity do come apart. Boone and Piccinini (2016, p. 1524) outline “a mechanistic version of homuncular functionalism, whereby higher-level cognitive capacities are iteratively explained by lower-level capacities until we reach a level in which the lower-level capacities are no longer cognitive in the relevant sense.” While this might entail modularity for some lower-level elements (they do not say as much), it does not entail modularity for higher-level elements composed predominantly of shared parts (indeed, the word “modularity” or “module” appears nowhere in their paper): see McGeer’s (2007) helpful discussion of the cognitive neuropsychologist’s understanding of modularity. Prinz (2006) is actually explicit that, as long as the units of decomposition do not exhibit the properties associated with Fodorian modularity, we should proceed with decomposition but abandon the label of modularity. See my remarks, later in this chapter, for further clarification of this point.

7 Of whatever variety—strict or formal (see §§ 2.4.3 and 5.1).

8 Of course, what corresponds to a node in network neuroscience is somewhat arbitrary. For certain purposes, neuroscientists may use a neuroanatomically delimited region as a node, whereas for other purposes they may use another one. I certainly would not wish to say that a module is something that depends on the interest of the neuroscientist. (I thank an anonymous reviewer for bringing this point to my attention.) Additionally, the notion of a “module” in network neuroscience is, as we saw in Chapter 4, different from the notion of a “node.” A module in network theory refers to a community of nodes.