Non-LIFO Execution of Cognitive Procedures
Abstract
Many current theories of human problem solving and skill acquisition assume that people work only on the unsatisfied goal that was created most recently. That is, the architecture obeys a last-in-first-out (LIFO) constraint on the selection of goals. We argue that this restriction seems to be violated by some subjects on some tasks. In particular, we show that non-verbal protocols of 8 subjects in a sample of 26 can be precisely simulated by dropping the LIFO restriction and assuming instead that subjects' knowledge includes explicit goal selection preferences. Although there is a great deal of between- and within- subject strategy variation in the data, it is nearly completely accounted for by a few specific preferences that seem to be over-generalized, conditionalized or missing from the subjects' knowledge. On the other hand, LIFO-based models can not account for the strategy variations in any simple way. Thus, it seems that part of the flexibility in human problem solving comes from having a choice in which goal to work on next. We conclude by reviewing the theoretical problems that led to the adoption of the LIFO restriction, and showing how these can be simply solved inside a non-LIFO architecture without introducing any new architectural mechanisms. (sdw)
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 13, 1989
- Accession Number
- ADA222326
Entities
People
- Bernadette Kowalski
- Kurt VanLehn
- William Ball
Organizations
- Carnegie Mellon University