[ CogSci Summaries home | UP | email ]
http://www.jimdavies.org/summaries/

Holyoak, K. J. & Thagard, P. (1997). The analogical mind. American Psychologist.52(1)35--44.

@Article{,
  author = 	 { K. J. Holyoak and P. Thagard},
  title = 	 {The analogical mind},
  journal = 	 {American Psychologist},
  year = 	 {1997},
  OPTvolume = 	 {52},
  OPTnumber = 	 {1},
  OPTpages = 	 {35--44},
}

Author of the summary: Jim Davies, 2002, jim@jimdavies.org

Cite this paper for:

Kids at age 2 can think analogically. (35)

multiconstraint theory: use of analogy is guided by a number of general constraints that jointly encourage coherence in analogical thinking.

There are three general constraints on analogy: (36)

  1. similarity: direct similarity of elements
  2. structure: consistent structural parallels (Gentner 1983). There is evidence of this in Spellman and Holyoak 1992.
  3. purpose: it is guided by the reasoner's goals-- what the analogy is intended to achieve. Evidence in Spellman and Holyoak 1996.
They are not rigid rules.

The models ARCS and ACME make analogies by applying the multiconstraint theory.(41)

The structural contstraint in ACME allows objects to map to objects only and predicates to map to predicates only.

Structural consistency encourages that if Iraq = Germany, then Iraq != Hitler.
Also if Iraq = Germany then invade = occupy is encouraged. (42)

The purpose affects what initial analogies are hypothesized-- so if you want to make an analogy between Hitler and Husein, then other mappings are encouraged.

In the computational model, elements are units, pos and neg constraints are excitatory and inhibitory links, external constraints are represented by links to special units, and parallel constraint satisfaction is achieved by updating activations of the units.

Summary author's notes:


Back to the Cognitive Science Summaries homepage
Cognitive Science Summaries Webmaster:
JimDavies (jim@jimdavies.org)