The Modularity of Mind (A Bradford Book) Kindle Edition

This book introduced a new concept to the philosophy of mind--"module" and revived another -- "faculty". Modularity of Mind is a work of historical importance, despite a confusing style. The author attempts to construct by a priori reasoning a general picture of the human brain.

The term "faculty" was used in psychology since ancient times to mean any mental capability. There was a faculty of memory, a faculty of judgment, and so on. By the 19th century the faculties were known to be located in the brain. Observing that the human skull was rather bumpy in ways that varied from one person to another, Franz-Joseph Gall proposed that the brain was divided into areas representing various faculties and instincts and the size of a bump represented the size of the faculty beneath. There was in fact no relation between bumps and faculties and this so-called science of "phrenology" was discredited. For the next 150 years little was said about faculties. The reigning doctrine came to be "associationism", according to which brain material was generally homogenous. Mental structure was built up from perceptions by a process of association. The proposed regularities in the way mental structure could be built up according to associationism became more complex as psychology advanced and appeared similar to rules for the finite-state machines and Turing machines of the dawning science of cybernetics.

The author begins his cybernetic model by reconceiving the mind as a complex Turing machine. He admits at once this analogy by itself is inadequate. It takes no account of the environment with which the mind interacts. Therefore he envisions the Turing central processor as embedded in a matrix of "input systems" which provide the central processor with information about. the outside world. In a human they represent the senses. An input system gets its information from the outside world via what he calls a "transducer" and converts it to something accessible to the central processor's "thought".

Input systems have certain common characteristics that warrant giving them a specific classification. There is at least one for each sense of hearing, sight, etc. First, they are "domain specific", i.e., each one deals with one source of data from the outside world and data from other sources is excluded. So in this way they are encapsulated; another frequent descriptive term is "modular". Associated with each are sub-processors that hypothesize about what in the real world is being seen or heard. Second, these entities are encapsulated, excluding all other data sources. Input systems and their sub-processors are always in operation and operate rapidly, which is necessary to warn of danger. (The author gives an example of a ferocious animal nearby; best play safe even if this means jumping to a conclusion.) Their outputs are "shallow", meaning that the input system does not do a great deal to interpret or compute with the data it receives from the transducer. (But the system does interpret to some degree, as the Muller-Lyer optical illusion shows.) There must be a region where the pathways from input systems terminate. This is part of what the author calls the central system. In contrast to the vertical input systems, each covering a single cognitive domain, the central system is horizontal, sharing cognitive data from more than one input system.For example, language processing uses both auditory and visual data. The author devotes much space to the question of the relative extent of processing of input within the input system versus within the central processor. The central system is not modular. In its decision-making process, the central system employs information computed from the output of its vertical input systems to act on its environment and thereby change it. Therefore there will be new representations of the environment which the input systems reflect. There is circumstantial evidence for this process even though we actually see little of it going on.

The relative lack of modularity in the central processor provides it greater access to the information from its input systems but even so it may not be able to verify new information about the external world. The author considers this question of obtaing information in terms of isotropy and Quineianism, i.e., how valid an estimate of scientific truth may become when scientific knowledge in other, seemingly unrelated fields changes. He suggests that isotropy and Quineianism are more likely in large, complexly interrelated bodies of knowledge. Considering Artificial Intelligence systems, he infers the problem of framing in AI is very difficult when a system needs to modify its frame (interrelational database) in the light of new information and may not know what needs to be modified. He provides an amusing example: a system looking up a phone number that turns out to be inaccessible. But just as the scope of an AI system may be limited by the total capacities of its input systems, the human mind is to some degree encapsulated by its inputs and likewise limited. This is referred to as 'epistemic boundedness', and is the final topic of the book. The answer is not encouraging, but not wholly negative either.

This book is historically very significant and the argument is cohesive. I wish thr writing were more lucid.

------------------------------------------------------------------------------------------------

This is a phenomenal book - probably the finest extended work of empirically informed philosophy of mind of the last fifty years. I enjoy Fodor's other work, but this is his magnum opus. His rich empirical insights are as relevant as ever, and none of the major ideas in the book have been settled one way or another by more recent data (though the debates continue to rage as fiercely as ever). Fodor writes with clarity and style, and presents complex experiments and ideas very appealingly. There's no-one else writing today (except perhaps Ned Block) who manages to pull this off so well. While this book might be a bit of a challenge to a non-philosopher, it's certainly readable, and full of insights. For an undergraduate philosopher of mind, it's positively essential.

Fodor lays it out clearly and precisely in this discussion of the physiological connections between neurology, conceptualization and behavior.

Fodor, stop fodoring!

Joke aside, a good book to get acquainted with if you're interested in different theories in cognitive science.

Thank you

Without a doubt the most difficult vocabulary, and consistently difficult vocabulary, in any book I have every read completely:

It’s only 145 pages but it’s unlikely anyone not an academic knowledgeable in the field (psychology and neurobiology of behavior) will every read it completely.

There were two reasons for my completion of this book — it was quite good and for the shear challenge of doing so.

Despite having an extremely large vocabulary, having study psychology for my degree (later in psychology), and a long time interest in the subject it was necessary to look up words in almost every sentence.

Since I read it before the Internet was available and personal computing devices were ubiquitous this required a tremendous effort and extreme perseverance.

I would also agree with **Rolf Dobelli’s 3-star review:**

This classic in cognitive science has a great deal to say, but an awkward way of saying it. Author Jerry A. Fodor's style is academic and dense, a potential barrier to all but the most determined, well-prepared reader. Arcane and brilliant, Fodor intersperses colloquial jests with jargon-burdened exposition, leading one to believe that he could have written a book more accessible to the lay reader had he wished to do so. We find, however, that the book repays the persistent, dedicated reader. The reward is a fascinating exploration of the mind, drawing on the literature of epistemology and psychology, with occasional detours down the rarely explored byways of phrenology.

Fodor usually writes the most arcane books in the cog sci set. The Language of Thought was so opaque that even Hilary Putnam couldn't understand it (circa Language and Learning by Piatelli-Palmarini, ed). This book is written in such a way that even lunkheads such as myself can get it. What he's saying in Psychosemantics I don't know either. But it's nice that he's written one popular philosophy book. When you take this book, together with Stephen Wolfram, you probably get the Language of Thought. When you add Richard Dawkins you get Steven Pinker. Not a bad piece of work.