Giulia Bencini,1 Kathryn Bock2 & Adele Goldberg2
1 New York University & 2 University of Illinois at
Urbana-Champaign
Structural priming provides the strongest evidence for the existence of processes that build the forms of sentences independent of individual words and the meanings that the sentences convey (Bock, 1986; Bock & Loebell, 1990). The basic finding is that people tend to repeat the structure of a sentence they have previously used (Bock, 1986; Bock & Loebell, 1990, Branigan et al.,1995; Potter and Lombardi, 1998). One way to elicit structural priming experimentally is shown in the example below. Studies have shown that structural priming occurs even when primes and targets do not overlap in content words, function words (Bock, 1986, 1989), or semantic roles (Bock & Loebell, 1990), suggesting that the locus of the effect is at the level of abstract constituent structure. Another hypothesis is that the generalizations are less abstract, grounded in meaning (Hare & Goldberg, 1999) or specific function words associated with syntactic frames. To date no single study clearly distinguishes these hypotheses. The key demonstration of priming without overlap in semantics was Bock and Loebell (1990, Experiment 2). They found that by-passives and intransitive locatives were equally effective primes for passives, relative to active controls. This suggests that meaning overlap is not necessary for priming, because the semantic roles in locatives and by passives are uncontroversially distinct. In passives like "The 747 was alerted by the control tower", control tower is the agent of the action. In a locative with the same structure ("The 747 was landing by the control tower"), control tower plays the role of location. However, the materials did not rule out a contribution from the function words. The intransitive locative primes shared with the passives the auxiliary be and the preposition by. The present study was designed to determine whether the priming of by-passives from intransitive locatives is in part due to the preposition by, providing a more stringent test of the hypothesis that priming is abstract and operates at the level of constituent structure. We replicated the results of Bock and Loebell, finding priming of passives from both passives and locatives with the preposition by. However the results showed that by was necessary for priming to occur: locatives without by showed no priming. Thus these data indicate that the relevant generalizations for structural priming are at a level less abstract than pure constituent structure.
Example
[A priming trial in a structural priming experiment] Participants hear and repeat a priming sentence such as "The new graduate was hired by the software company." They then see and describe a pictured event, for example the picture of a dog chasing a mailman. They might say something like "The mailman is being chased by an angry poodle." On the same trial another group of participants hears and repeats the sentence "The new graduate left the software company." Structural priming is said to occur when participants match the structure of the sentence prime in their subsequent picture description.
References
Bock, J. K. (1986). Syntactic persistence in language production. Cognitive Psychology, 18, 355-387.
Bock, J. K. (1989). Closed-class immanence in sentence production. Cognition, 31, 163-186.
Bock, K., & Loebell, H. (1990). Framing sentences. Cognition, 35, 1-39.
Branigan et al. (1995). Syntactic priming: Investigating the mental representation of language. Journal of Psycholinguistic Research, 24, 489-506.
Hare, M., & Goldberg, A. (1999). Structural priming, purely syntactic? Cog. Sci. Proceedings, 208-211.
Potter, M., & Lombardi, L. (1998). Syntactic priming in immediate recall of sentences. Journal of Memory and Language, 38, 265:282