Greg Detre
Tuesday, April 29, 2003
�
Title :� Cognitive psychology.�
Published :� v. 1- Jan. 1970-�
Published :� San Diego [etc.] Academic Press.�
INTERNET LINK :� http://nrs.harvard.edu/urn-3:hul.eresource:cogpsych��
Locations/Orders :� Availability�
Location :� Countway Medicine Serial� Holdings� Availability�
Location :� Gutman Education Per. BF309 .C62� Holdings� Availability�
Location :� Networked Resource BF309 .C62 [Provides access to tables of contents and abstracts from 1993 and full text of articles from 1997.]� Holdings� Availability�
Location :� Psychology Research SERIAL� Holdings� Availability�
Location :� Widener Phil 12.130 [Current Issues: Periodicals Reading Room Stacks]� Holdings� Availability�
Cognitive psychology.
San Diego [etc.] Academic Press.
Location :� Psychology Research SERIAL�
Acquisition Data :� Status: Currently received�
Holdings :� v.1(1970)-v.44(2002)�
Title :� Cognition.�
Published :� v. 1-�
Published :� Lausanne [Switzerland, etc.] Elsevier Sequoia [etc.] 1972-�
INTERNET LINK :� http://nrs.harvard.edu/urn-3:hul.eresource:cognitio��
Locations/Orders :� Availability�
Location :� Gutman Education Per. BF309 .C6� Holdings� Availability�
Location :� Networked Resource [Table of contents and abstracts of articles only for 1994-1998; beginning with 1999- includes also brief articles]� Holdings� Availability�
Location :� Psychology Research SERIAL� Holdings� Availability�
Location :� Widener Phil 12.155 [Current Issues: Periodicals Reading Room Stacks]� Holdings� Availability�
Location :� Widener Phil 12.155 Current Issues: Periodicals Reading Room Stacks�
Acquisition Data :� Status: Currently received�
Holdings :� v.1(1972)-v.85(2002)�
Current Receipts :� v.86:no.1(2002:Nov.)�
�v.86:no.2(2002:Dec.)�
�v.86:no.3(2003:Jan.)�
�v.87:no.1(2003:Feb.)�
Index :� v.1/10(1972/1981) in v. 10.�
�v.1/20(1972/1985) in v. 20.�
�v.1/40(1972/1991) in July 1992.�
�v.1/60(1972/1996) in July 1996.�
�
To do this, artifical systems need to be able to form concepts in the same way that curious natural systems do
incompressible is incomprehensible
compressible is comprehensible
compression is comprehension
minimisation and inference are deeply intertwined
maximal descriptive compactness also corresponds to maximal inferential power
the debate over rules vs exceptions can be subsumed
Hofstadter on the representational flexibility of NNs:
�members of a category are not all represented by identical symbolic structures; rather, individual objects will be represented in subtly different ways depending upon the context in which they are represented�
�the great adaptability of high-level perception suggestst hat no module that produced a single context-independent representation could ever model the complexity of the process�
�Generalisation outwards from a conceptual centre is an automatic, unconscious process that pervades thought - indeed, it defines thought. It�s not as if there is just one rigid propositional or logical structure that captures what we understand when we read in the newspaper about a kidnapping or hear a throwaway remark about dieting. That is as far from the proper image of what thought is as one can get! Rather, all sorts of analogous events and related images from our own lives are activated to different degrees, and commingle and blur with aspects of the event itself to form a very complex, active, fluid structure, whose rules bear very little connection to those of any kind of formalisable logic�
Minsky
�the really useful �meanings� are not the flimsy logic chains of definitions, but the much harder-to-express networks of ways to remember, compare and change things�
Hofstadter: �Generalisation involves the ability to internally reconfigure an idea, by:
moving internal boundaries back and forth
swapping components or shifting substructures from one level to another
merging two substructures into one or breaking one substructure into two
lengthening or shortening a given component
adding new components or new levels of structure
replacing one concept by a closely related one
trying out the effect of reversals on various conceptual levels�
�the familiar and stable-seeming fluidlike properties of thought emerge as a statistical consequence of a myriad tiny, invisible, independent, subcognitive acts taking place in parallel�
game: an actiivty that is engaging and diverting, deliberately detached from real life
minsky som index
concept
105
179
211
game
29
57
130-31
143
semantic nets??? semantic content??? halos??? vagueness???
too arbitrary - no useful/meaningful features to anchor
chunking???
perhaps we consciously learn concepts differently from unconsciously learning (when presented with instances)???
Feldman_Miller
Harnad�s debunking of Rosch
natural kinds
johnson-laird
separated vs integral
hofstadter + minsky quotes
jorge luis borges� classification of animals
basic level categories
R. M. PIRSIG The motorcycle..a system of concepts worked out in steel.
prototypes vs exemplars???