Journal of Experimental Psychology: Human Perception and Performance 1981, Vol. 7, No. 1,216-230 Copyright 1981 by the American Psychological Association, Inc. 0096-1523/81/0701-0216$00.75 The Invariance of Sentence Performance Structures Across Language Modality Francois Grosjean, Harlan Lane, Robbin Battison, and Hartmut Teuber Northeastern University Native users of American Sign Language were asked to manipulate sentences in four different ways: sign them at slow rate, parse them, make relatedness judgments of pairs of signs taken from each sentence, and recall the sentences. The data obtained from these four tasks (pause durations, parsing values, indices of relatedness and probe latencies) were used to construct hierarchical perfor- mance structures for each of the sentences. The resulting structures were highly similar across tasks; that is, performance structures are not task specific. The four measures at each sign boundary in each sentence were well predicted by a performance model, elaborated by Grosjean, Grosjean, and Lane for speech, that combines a parsing measure with a symmetry measure. Thus performance structures appear to be founded in the processing of language, be it visual or oral, and not in the properties of any particular communication modality. Considerable research in the 1960s and early 70s was aimed at demonstrating the psychological reality of surface structures of sentences as they are characterized by mod- ern linguistic theory. Experiments in recall, perception, direct scaling of relatedness, and pausing, among others, were used to au- thenticate the role of syntactic units in pro- cessing spoken languages. Some studies were most interested in showing that clauses are functional units in speech processing (Fodor & Bever, 1965; Holmes & Forster, 1970; Jarvella, 1971). Other studies were more in- terested in showing that lower order surface constituents are also functional units in speech processing (N. Johnson, 1965, 1968; Suci, Ammon, & Gamlin, 1967). Levelt (Note 2), for example, used a method of direct scaling of relatedness. He presented subjects with all pairs of words from a sen- tence and asked them to rate each pair on a 7-point scale for relatedness of its mem- bers. A clustering algorithm (S. Johnson, This research was supported in part by grants from the National Science Foundation (768 253) and the U.S. Department of Health, Education, and Welfare (RR 07143 and NS14923). The authors would like to thank Joanne Miller and Steve Harkins for their useful comments and sugges- tions. Requests for reprints should be sent to Francois Gros- jean, Department of Psychology, Northeastern Univer- sity, Boston, Massachusetts 02115. 1967) was applied to the half-matrix of sim- ilarity scores', thus ascribing a hierarchical structure to the sentence, which proved to be much like its phrase structure. Levelt found that the degree of relatedness between words was inversely related to the number of phrase boundaries separating one word from the other. Studies of pausing in reading and spon- taneous speech have also shown that sen- tences are made up of functional units: to a hierarchy of pause frequency and duration corresponds a hierarchy of constituents (Brown & Miron, 1971; Goldman-Eisler, 1972; Grosjean & Collins, 1979; Grosjean & Deschamps, 1975). Whether the speech is spontaneous (conversation) or not (read- ing, recitation) there will be, at any given rate of speech, more and longer pauses at the ends of sentences than within them, and more and longer pauses at the breaks be- tween the noun phrase (NP) and verb phrase (VP) than within these constituents. In sum, all of these psycholinguistic stud- ies, involving such diverse tasks as recall, perception, direct scaling of relatedness, and pausing, have shown that a relation exists between the listener's (and speaker's) seg- mentation of the stream of speech and a structural description of the sentences com- prising that stream, and that it is possible, to some extent at least, to delineate pro- cessing units from the behavioral data that 216