Copyright © 1999 Massachusetts Institute of Technology
Schemata are the psychological constructs that are postulated to account for the molar forms of human generic knowledge. The term frames, as introduced by Marvin Minsky (1975), is essentially synonymous, except that Minsky used frame as both a psychological construct and as a construct in artificial intelligence. Scripts are the subclass of schemata that are used to account for generic (stereotyped) sequences of actions (Schank and Abelson 1977).
Although the term schema was used by the philosopher Immanuel KANT and the developmental psychologist Jean PIAGET, the direct line of intellectual descent for this construct in cognitive science is through the work of British psychologist Sir Frederic BARTLETT. Bartlett (1932) was investigating the recall of folktales and noticed that many of the errors that occurred in the recall protocols tended to make the recalls more conventional than the original text. In order to account for this class of memory errors, Bartlett proposed that human beings have substantial amounts of generic knowledge in the form of unconscious mental structures (schemata) and that these structures interact with incoming information to produce schematized (conventionalized) errors in recall. Bartlett's schema construct was not compatible with the world view that was dominant in psychology at the time (see BEHAVIORISM), and therefore the schema concept was not incorporated into mainstream MEMORY research in psychology (cf. Brewer and Nakamura 1984).
The schema construct was reintroduced into modern cognitive science through the work of the computer scientist Marvin Minsky (1975). Minsky was attempting to develop machines that would show humanlike intelligence. Minsky read Bartlett's book on memory (Bartlett 1932) and concluded that much of human intelligent behavior derived from the use of generic knowledge. This led Minsky to argue that in order to make machines intelligent it would be necessary for them to be provided with large amounts of knowledge. This proposal had an enormous influence on the development of the field of artificial intelligence (cf. Dyer, Cullingford, and Alvarado 1990; Maida 1990).
Minsky's more specific proposal was to introduce the construct of frames to represent knowledge of ordinary aspects of the world (e.g., rooms). Frames are knowledge structures that contain fixed structural information. They have slots that accept a range of values; each slot has a default value that is used if no value has been provided from the external world. For example, if a person or a machine is trying to represent a particular college classroom the generic classroom frame will contain the fixed information that the room will have walls, a ceiling, and a door. The frame will contain a slot for type of lighting. If no information is provided about this aspect of the world (e.g., if an individual has just glanced at a room without looking up to see the lights) then the frame provides a default value (in this case, that the lights are fluorescent). Thus the frame construct can be used to give an account for why someone walking into a room without a ceiling will be surprised and why an individual might recall that a particular classroom had fluorescent lights when it actually did not. Note that in this example there is a generic frame for classrooms in long-term memory, and to represent a specific classroom the generic frame is instantiated by a specific episodic representation. In general, Minsky's frames provided a much more structured account of KNOWLEDGE REPRESENTATION than previous proposals.
Work on frames in artificial intelligence has had a strong impact on psychological investigations of knowledge representation. Rumelhart (1980) developed a psychologically based theory of schemata derived from Minsky's work. (Psychologists working on these topics usually use the term schema, plural schemata, for these forms of knowledge representation.) In both psychology and artificial intelligence there has been much controversy over the relationship of schemata to other "simpler" forms of representation such as propositions, semantic nets, and CONCEPTS. In general, schema theorists (Minsky 1975; Rumelhart 1980) have made an ontological argument that there are molar phenomena (e.g., events, spatial scenes, discourse structure) in the (psychological?) world, and that schemata are the appropriate forms of knowledge representation for these molar phenomena (cf. Brewer and Nakamura 1984; Davis, Shrobe, and Szolovits 1993).
Schema theories have had a wide impact on empirical work in cognitive psychology. They provided an account for earlier work such as the finding of Bransford and Johnson (1972) that recall for an opaquely written passage is much improved when it is given with a schema-relevant title. They also generated much new research, such as the finding by Bower, Black, and Turner (1979) showing high rates of script-based intrusions in recall of script narratives, and the work of Brewer and Treyens (1981) showing schema-based intrusions in the recall of visual scenes (see Brewer and Nakamura 1984 for an extensive review of these empirical findings).
How are schema theories to be viewed within the overall study of knowledge representation in cognitive science? The original papers on frames and schemata invite the inference that schemalike representations will account for most of human and machine knowledge. Subsequent work has shown that although schema representations provide a powerful account for generic forms of knowledge, this type of representation is only a part of the total repertoire of human knowledge. It now seems clear (cf. Brewer 1987) that MENTAL MODELS (Johnson-Laird 1983), naive theories (Gopnik and Wellman 1992) and many other forms of representation (e.g., Schank 1982) are needed to give a comprehensive account of human and machine knowledge.
Bartlett, F. C. (1932). Remembering. Cambridge: Cambridge University Press.
Bower, G. H., J. B. Black, and T. J. Turner. (1979). Scripts in memory for text. Cognitive Psychology 11:177-220.
Bransford, J. D., and M. K. Johnson. (1972). Contextual prerequisites for understanding: Some investigations of comprehension and recall. Journal of Verbal Learning and Verbal Behavior 11:717-726.
Brewer, W. F. (1987). Schemas versus mental models in human memory. In P. Morris, Ed., Modelling Cognition. Chichester, UK: Wiley, pp. 187-197.
Brewer, W. F., and G. V. Nakamura. (1984). The nature and functions of schemas. In R. S. Wyer, Jr. and T. K. Srull, Eds., Handbook of Social Cognition, vol. 1. Hillsdale, NJ: Erlbaum, pp. 119-160.
Brewer, W. F., and J. C. Treyens. (1981). Role of schemata in memory for places. Cognitive Psychology 13:207-230.
Davis, R., H. Shrobe, and P. Szolovits. (1993). What is a knowledge representation? AI Magazine 14:17-33.
Dyer, M., R. Cullingford, and S. Alvarado. (1990). Scripts. In S. C. Shapiro, Ed., Encyclopedia of Artificial Intelligence, vol. 2. New York: Wiley, pp. 980-994.
Gopnik, A., and H. M. Wellman. (1992). Why the child's theory of mind really is a theory. Mind and Language 7:145-171.
Johnson-Laird, P. N. (1983). Mental Models. Cambridge, MA: Harvard University Press.
Maida, A. S. (1990). Frame theory. In S. C. Shapiro, Ed., Encyclopedia of Artificial Intelligence, vol. 1. New York: Wiley, pp. 302-312.
Minsky, M. (1975). A framework for representing knowledge. In P. H. Winston, Ed., The Psychology of Computer Vision. New York: McGraw-Hill, pp. 211-277.
Rumelhart, D. E. (1980). Schemata: The building blocks of cognition. In R. J. Spiro, B. C. Bruce, and W. F. Brewer, Eds., Theoretical Issues in Reading Comprehension. Hillsdale, NJ: Erlbaum, pp. 33-58.
Schank, R. C. (1982). Dynamic Memory: A Theory of Reminding and Learning in Computers and People. Cambridge: Cambridge University Press.
Schank, R. C., and R. P. Abelson. (1977). Scripts, Plans, Goals and Understanding. Hillsdale, NJ: Erlbaum.
Graesser, A. C., and G. V. Nakamura. (1982). The impact of a schema on comprehension and memory. In G. H. Bower, Ed., The Psychology of Learning and Motivation, vol. 16. New York: Academic Press, pp. 59-109.
Johnson-Laird, P. N., D. J. Herrmann, and R. Chaffin. (1984). Only connections: A critique of semantic networks. Psychological Bulletin 96:292-315.
Mandler, J. M. (1984). Stories, Scripts, and Scenes: Aspects of Schema Theory. Hillsdale, NJ: Erlbaum.
Rumelhart, D. E., and D. A. Norman. (1988). Representation in memory. In R. C. Atkinson, R. J. Herrnstein, G. Lindzey, and R. D. Luce, Eds., Stevens' Handbook of Experimental Psychology. Vol. 2, Learning and Cognition. 2nd ed. New York: Wiley, pp. 511-587.
Rumelhart, D. E., and A. Ortony. (1977). The representation of knowledge in memory. In R. C. Anderson and R. J. Spiro, Eds., Schooling and the Acquisition of Knowledge. Hillsdale, NJ: Erlbaum, pp. 99-135 .
Copyright © 1999 Massachusetts Institute of Technology