This study examines the role of syntactic information in word recognition. Subjects made a word-nonword decision regarding a target string that was preceded by a syntactically appropriate word, a semantically related word, or an unrelated word. In Experiment 1, with syntactic and semantic trials assigned to separate blocks, syntactically and semantically appropriate context significantly reduced lexical decision for subsequent target words, compared with unrelated contexts. In Experiment 2, the syntactically and semantically primed trials were either blocked separately or mixed within the same block. Significant syntactic and semantic effects were both observed in the blocked condition, but only the semantic effect was obtained in the mixed condition.It is a well established fact that context facilitates word recognition (e.g., Tulving & Gold, 1963), and there is a considerable body of work suggesting that some of these effects are due to semantic priming (Fischler & Goodman, 1978;Meyer, Schvaneveldt, & Ruddy, 1975). However, there has been little work specifically designed to examine the effects of syntactic context. One relevant experiment was reported by Miller and Isard (1963), in which they showed that both semantic and syntactic information can improve the processing of sentences. In their first experiment, Miller and Isard's subjects listened to stimulus sentences and were told to repeat the sentences to the experimenter. The stimulus sentences were of three types: meaningful, grammatical sentences (e.g., "The academic lecture attracted a limited audience"), grammatically correct sentences that were semantically anomalous (e.g., "The odorless lecture became a filthy audience"), and ungrammatical sentences (e.g., "From hunters house motorists the carry"). Subjects were most accurate with the meaningful, grammatical sentences, but they were more accurate on the grammatical-anomalous sentences than on the ungrammatical sentences. Thus, appropriate syntactic context resulted in better performance even in the absence of meaningful semnatic information.Given Miller and Isard's (1969) findings, questions remain regarding the nature of the effect of semantic and syntactic information on linguistic processing. First, do semantic and syntactic information affect word recognition or some other aspect of processing required by the task? In the Miller and Isard experiment, conThis research was supported by NSF Grant BNS76-14830 to James L. McClelland. Thanks are due to Marguerite Moreno for her assistance in data collection and analysis and Jim Hollan for his helpful comments on an earlier version of this paper. Requests for reprints should be sent to James L. McClelland, Department of Psychology, C-009, University of California, San Diego, La Jolla, California 92093. textual information could have had the effect of improving subjects' performance by providing a structure to facilitate memory retention and/or retrieval, with linguistic structure having no effect on the recognition of the words.Some results relevant t...