<> "The repository administrator has not yet configured an RDF license."^^ . <> . . "Category Induction and Representation"^^ . "A provisional model is presented in which categorical perception\n (CP) provides our basic or elementary categories. In acquiring a category we\n learn to label or identify positive and negative instances from a sample of\n confusable alternatives. Two kinds of internal representation are built up in this\n learning by \"acquaintance\": (1) an iconic representation that subserves our\n similarity judgments and (2) an analog/digital feature-filter that picks out the\n invariant information allowing us to categorize the instances correctly. This\n second, categorical representation is associated with the category name.\n Category names then serve as the atomic symbols for a third representational\n system, the (3) symbolic representations that underlie language and that\n make it possible for us to learn by \"description.\" Connectionism is one\n possible mechainsm for learning the sensory invariants underlying\n categorization and naming. Among the implications of the model are (a) the\n \"cognitive identity of (current) indiscriminables\": Categories and their\n representations can only be provisional and approximate, relative to the\n alternatives encountered to date, rather than \"exact.\" There is also (b) no\n such thing as an absolute \"feature,\" only those features that are invariant\n within a particular context of confusable alternatives. Contrary to prevailing\n \"prototype\" views, however, (c) such provisionally invariant features must\n underlie successful categorization, and must be \"sufficient\" (at least in the\n \"satisficing\" sense) to subserve reliable performance with all-or-none,\n bounded categories, as in CP. Finally, the model brings out some basic\n limitations of the \"symbol-manipulative\" approach to modeling cognition,\n showing how (d) symbol meanings must be functionally grounded in\n nonsymbolic, \"shape-preserving\" representations -- iconic and categorical\n ones. Otherwise, all symbol interpretations are ungrounded and\n indeterminate. This amounts to a principled call for a psychophysical (rather\n than a neural) \"bottom-up\" approach to cognition."^^ . "1987" . . . "Cambridge University Press"^^ . . . "Categorical Perception: The Groundwork of Cognition"^^ . . . . . . . . . . . "Stevan"^^ . "Harnad"^^ . "Stevan Harnad"^^ . . . . . . "Category Induction and Representation (HTML)"^^ . . . "harnad87.categorization.html"^^ . . . "Category Induction and Representation (Indexer Terms)"^^ . . . . . . "indexcodes.txt"^^ . . "HTML Summary of #1572 \n\nCategory Induction and Representation\n\n" . "text/html" . . . "Cognitive Psychology" . . . "Neural Nets" . . . "Philosophy of Language" . .