- Shopping Bag ( 0 items )
In The Algebraic Mind, Gary Marcus attempts to integrate two theories about how the mind works, one that says that the mind is a computer-like manipulator of symbols, and another that says that the mind is a large network of neurons working together in parallel. Resisting the conventional wisdom that says that if the mind is a large neural network it cannot simultaneously be a manipulator of symbols, Marcus outlines a variety of ways in which neural systems could be organized so as to manipulate symbols, and he shows why such systems are more likely to provide an adequate substrate for language and cognition than neural systems that are inconsistent with the manipulation of symbols. Concluding with a discussion of how a neurally realized system of symbol-manipulation could have evolved and how such a system could unfold developmentally within the womb, Marcus helps to set the future agenda of cognitive neuroscience.
The MIT Press
In The Algebraic Mind, Marcus dives into the difficult waters of the connectionist-symbolic debate.
|Ch. 1||Cognitive Architecture||1|
|Ch. 2||Multilayer Perceptrons||7|
|Ch. 3||Relations between Variables||35|
|Ch. 4||Structural Representations||85|
|Ch. 6||Where Does the Machinery of Symbol Manipulation Come From?||143|
Posted June 3, 2013
Posted April 6, 2004
Much of what is stated and described by Gary Marcus has already been widely discussed in several prior publications in the area of neuro-computation by artifical neural networks. In 'Introduction to the Theory of Neural Computation',1990,John Hertz et al cover all computational and algebraic aspects of artificial networks and there is no sense whatsoever that artifical networks can not serve as symbol manipulators, depending upon the task. Hopfield nets are precisely storage units of symbols and serve an essential part of the connectionist paradigm. Moreover,it does not take much imagination to construct artificial neuronal logic gate systems based upon boolean logic that provide prescribed operations on inputs. Thus, everything that is algebraic in nature can be easily built into an artificial neuronal network system. This fact is so obvious that there was never any need to extensively discuss it until Gary Marcus's book laid claim to being the originator of the concept. The fact that classical connectionist theory has concerned itself with neuronal computations that algebraic based systems could not accomplish alone should not be taken to imply that the algebraic based hardwiring and Hopfield symbol storage units were excluded from the complete model. Gary Marcus attempts to oversimplify a complex topic. The Book has no bearing on the professional literature and was written for an immature, naive audience. The author addresses topics that are so obvious that no one else has ever felt a need to discuss them. Through hardwiring and coding, neural networks can, clearly, serve as manipulators of symbols. It is obvious and understood that connectionist theory assumes an adjunct logic gate hardwiring that can serve as a processor or encoding of neuronal activations relying on algebraic symbol manipulation suitable to the task. Gary Marcus has told us nothing that wasn't already known or inconsistent with traditional connectionist theory. What Gary Marcus has done, however, is opportunized the ignorance of the reading public. (This review refers to both hardcover and softcover editions.)Was this review helpful? Yes NoThank you for your feedback. Report this reviewThank you, this review has been flagged.