Integrative Priming 2 Abstract Lexical priming, whereby a prime word facilitates recognition of a related target word (e.g., nurse doctor), is typically attributed to association strength, semantic similarity, or compound familiarity. Here we demonstrate a novel type of lexical priming that occurs among unassociated, dissimilar, and unfamiliar concepts (e.g., horse doctor). Specifically, integrative priming occurs when a prime word can be easily integrated with a target word to create a unitary representation. Across several manipulations of timing (stimulus onset asynchrony) and list context (relatedness proportion), lexical decisions for the target word were facilitated when it could be integrated with the prime word. Moreover, integrative priming was dissociated from both associative priming and semantic priming, but was comparable in terms of both prevalence (across participants) and magnitude (within-participants). This observation of integrative priming challenges current models of lexical priming, such as spreading activation, distributed representation, expectancy, episodic retrieval, and compound cue models. We suggest that integrative priming may be explained by a role activation model of relational integration. KEYWORDS: associative priming, integrative priming, lexical decision, relational integration, semantic priming, word recognition.
Integrative Priming 3One of the most robust phenomena of psychology is the lexical priming effect, whereby responding to a target word is facilitated by the prior presentation of a related prime word (for extensive review see Hutchison, 2003;Lucas, 2000;McNamara, 2005;Neely, 1991). In a seminal experiment, Meyer and Schvaneveldt (1971) simultaneously presented two letter strings in a lexical decision task (LDT; i.e., word/nonword judgments). They found that related words (e.g., BREAD -BUTTER) elicited faster lexical decisions than unrelated words (e.g., NURSE -BUTTER). Since that classic demonstration (see also Meyer, Schvaneveldt & Ruddy, 1975;Tweedy, Lapinski & Schvaneveldt, 1977), researchers have distinguished two broad classes of relation between primes and targets. Association refers to a relation in language use, typically defined as the proportion of participants who produce a given target word (e.g., PEPPER) in response to a cue word (e.g., SALT) in the free association task. Similarity refers to a relation in word meaning, typically defined as the degree of feature overlap between the concepts (e.g., DOG -WOLF). Thus one may distinguish associative priming, which is based on association strength, from semantic priming, which is based on featural similarity (see e.g., Fischler, 1977;Hutchison, 2003;Lucas, 2000;McNamara, 2005;McRae & Boisvert, 1998; Moss, Ostrin, Tyler & MarslenWilson, 1995;Shelton & Martin, 1992;Thompson-Schill, Kurtz & Gabrieli, 1998). The purpose of the present study is to demonstrate and characterize another source of lexical priming that is independent of both associations and similarity. Namely, we examine the influence of relational in...