New directions in discourse processing Edited by Roy O. Freedle (review)

450LANGUAGE, VOLUME 56, NUMBER 2 (1980) ------; D. M. Taylor; W. E. Lambert; and G. Albert. 1976. Dimensions of ethnic identity: An example from Northern Maine. Journal of Social Psychology 100.1119 . Lambert, Wallace E. 1974. Culture and language as factors in learning and education. Cultural factors in learning and education, ed. by Frances E. Aboud & R. D. Meade. Bellingham, WA: Western Washington State College. Rubin, Joan. 1968. National bilingualism in Paraguay. The Hague: Mouton. Scotton, Carol M. 1972. Choosing a lingua franca in an African capital. Edmonton: Linguistic Research. —¦—. 1976. Strategies ofneutrality : Language use in uncertain situations. Lg. 52. 919-41. ------. 1977. Linguistic performance as a socio-economic indicator. Journal of Social Psychology 102.35-45. Tajfel, Henri. 1974. Social identity and intergroup behaviour. Social Science Information 13.65-93. ------ (ed.) ms. Studies in intergroup behaviour. London : Academic Press. Taylor, Donald M., J. Bassili; and F. E. Aboud. 1973. Dimensions of ethnic identity: An example from Quebec. Journal of Social Psychology 89.185-92. [Received 12 June 1979.] New directions in discourse processing. Edited by Roy O. Freedle. (Advances in discourse processes, 2.) Norwood, NJ: Ablex, 1979. Pp. xv, 336. $23.95. Reviewed by Brian MacWhinney, University of Denver This collection of articles shows how frame theory can be used to describe a variety of issues in the area of discourse processes. From this evidence, it seems that such application of frame theory is here to stay. It has often been noted that modern frame theory is not unlike the older theories of schemata proposed by writers such as Kant, Herbart, James, Köhler, Bartlett— and even Aristotle. But, in its modern incarnation, this theory has also emerged as an interlingua facilitating communication between the various sub-specializations that constitute the field of cognitive science. In their introduction to the current volume, M. Adams and A. Collins (? schema-theoretic view of reading', 1-22) show how this interlingua can be used to account for letter recognition, word recognition, sentence parsing, and story-grammar processing. In a somewhat different vein, Deborah Tannen ('What's in a frame?', 137-81) shows how speakers can operate within a nested set of sociolinguistic role frames, including that of an 'experimental subject', a 'film viewer', or a 'film critic'. On yet another level, Tannen uses frame theory to handle grammatical phenomena such as ellipsis, false starts, and relativization. This surprisingly powerful approach is really quite simple. Frames (or schemata or scripts) are viewed as data structures composed of a set of 'slots', which can be filled by variables which may assume default (i.e. presupposed) values. Frames may embed themselves recursively or hierarchically. The activation of a given frame during processing can occur on the basis of data from its components (bottom-up), or its membership in a higher-order frame (top-down), or cross-level expectations. (For further discussion of the notion of a frame, see the introductory article by Adams & Collins, the first half of the article by Tannen, and articles published elsewhere—Minsky 1975, Rumelhart 1977, and Rumelhart & Ortony 1977). REVIEWS451 It is important to note that most current uses of frame theory involve postdiction rather than prediction. This is, given a set of behavioral data, writers often find it possible to propose a set of frames that might have produced those data. In only a few studies have frames been used in an attempt to predict actual behavioral data; and in these, there is often a variety of simple alternatives to the frametheoretic account. Thus, the study by N. Stein and C. Glenn ('An analysis of story comprehension in elementary school children', 53-120) attempts to demonstrate the reality of certain story frames by scrambling sentences within passages. However , as is argued in the adjacent article by Warren, Nicholas & Trabasso (p. 49), poor recall of scrambled passages can also be explained in terms of a disruption of normal inferential processes. In fact, experimental demonstrations of the psychological reality of frames generally turn up results that can be attributed to such alternative processes as analogy, response competition, or simple inferencing. Despite these problems, attempts to provide some experimental basis for frame theory will undoubtedly continue. Meanwhile, the theory can play an important role in...