Philosophy Concordance - online philosophical quotations

No one has ever taken seriously the idea that ‘cognitive architecture is Classical Turing architecture’, so the central premise of TMDWTW—that a Turing Machine is unsuited to solve a certain kind of problem that the human mind easily solves—is not relevant to anything.

Abduction really is a terrible problem for cognitive science, one that is unlikely to be solved by any kind of theory we have heard of so far’ (p. 41).

Scientific inference may really sometimes be abductive; but then, science is social, whereas quotidian cognition, of the kind psychologists care about, is carried out in single heads.

Quine?s metaphors of a fabric and a force field, with constraints along their edges that propagate across their surface, are reminiscent of the kind of computational system sometimes known as soap-film or soap-bubble models and which cognitive scientists call constraint satisfaction networks (Attneave, 1982; Marr and Poggio, 1976; Rumelhart et al. Waltz, 1975).

A and in tasks drawn from domain B. But though the creature reliably uses the information to perform one kind of task, it seems unable to do so when it is required to perform tasks of the other kind’ (p. 62).

Fodor acknowledges that a mind with some kind of modular design could, in principle, meet the abduction challenge.

He identifies the central thesis of HTMW as ‘massive modularity’, which ‘means that there is a more or less encapsulated processor for each kind of problem it can solve’ (p. 64).

Fodor claims that the problem of routing information to the appropriate reasoning system—for example, detecting that an event is an example of a social exchange in order to activate a cheater-detecting mechanism— requires nothing less than a full solution to the abduction problem, leaving us where we started: ‘Nobody has any idea what kind of cerebration is required for figuring out which distal stimulations are social exchanges’ (p. 76).

Faced with this puzzle, the only alternatives to natural selection are deliberate engineering by a deity or extraterrestrial; some kind of mysterious teleological force that allows future benefit to affect present design; and simply not caring.

For example, if the explanation of biological functionality in terms of natural selection is correct, we can rule out adaptations that work toward the greater good of the species, the harmony of the ecosystem, beauty for its own sake, benefits to entities other than the replicators that create the adaptations (such as horses which evolve saddles), functional complexity without reproductive benefit (e.g. an adaptation to compute the digits of pi), and anachronistic adaptations that benefit the organism in a kind of environment other than the one in which it evolved (e.g., an innate ability to read, or an innate concept of ‘carburetor’ or ‘trombone’).