Omar’s Reviews > The Mind Doesn't Work That Way: The Scope and Limits of Computational Psychology > Status Update

Appendix: Fodor illustrates how hasty can be some conclusions in the NS discussing the proposal we have a cheater detection mechanism, which could have evolved as a result of natural selection. Fodor shows the alleged contrast follows instead some general inferential rules and does not prove the existence of specific modules.

Ch.5:Darwin among Modules:NS is committed to the idea that cognitive architecture is an evolutionary adaptation. Fodor argues that this is also unsupported because we do not know enough about cognition and because it isn't clear why our minds are adaptations resolved by natural selection. He also points out the lack of connection between this thesis and modularity. It seems that alternative explanations are possible.

Ch.4:How many modules...? Abduction/global inferential processes wouldn't be so problematic for NS if massive modularity (MMT) is assumed. However, Fodor argues that there is no reason to think that MMT is true and that NS's claims are unsupported. The Input Problem is one of Fodor's main arguments, whereby it is shown that many aspects such as perception cannot be modular (or less modular, less domain-specific).

Ch.3:Two Ways that you can't probably explain abduction:CMT with a heuristic approach as well as connectionist models cannot deal with global inferences because computations are by definition local and dependent on their syntax. Since we do not know what type of architecture would explain abduction, we should remain skeptical and restrict to domains where CMT seems safer in scientific terms.

Ch.2:Syntax and Its Discontents: CMT finds problems with non-modular aspects of cognition (not everything seems to be computations). Discussed: generality of context invariance, simplicity and conservatism. Syntactic computations in thought processes do not seem local or insensitive to context in the general case expressing our limitations in AI and robotics. Much of the field would be in denial of this fact.

Introduction: Still snowing. Fodor introduces the main purpose of the book: a reply to what he calls New Synthesis approach represented by Pinker and Plotkin. He confesses not to be very optimistic about the picture of the Computational Theory of Mind depicted by them, in particular the idea that language is an instinct and is a product of evolution of the human brain. Fodor lays out the content five chapters.