University of Arizona
Psycholinguistics Laboratory

Research Papers


Tamar H. Gollan and Kenneth I. Forster
University of Arizona

Ram Frost
The Hebrew University


Translation priming between Hebrew and English is examined with a masked priming paradigm to determine whether the superiority of priming for cognates (words that share semantic and phonological properties) over noncognates is obtained when different scripts are used. Stronger translation priming effects were obtained for cognates, but only when the prime was in the dominant language (L1). Interestingly, strong translation priming was also obtained for noncognates, whereas previous studies have shown unstable priming effects for such stimuli. Overall, more facilitation was obtained when primes were in L1 and the targets were in L2. These results show that script differences play an important role in cross-language priming, and are interpreted in terms of a dual lexicon model, in which one lexicon must be accessed before the other. It is argued that the script provides a strong cue to the lexical processor, allowing the appropriate lexicon for the prime to be accessed in time to produce priming for noncognates. Without this cue access of the prime is too slow. For same-script cognates, however, no cue is necessary, since it is assumed that these words are jointly represented in both lexicons, and hence translation priming is obtained whether the languages differ in script or not.

Gollan, T. H., Forster, K. I., & Frost, R. (1997). Translation priming with different scripts: Masked priming with cognates and noncognates in Hebrew-English bilinguals. Journal of Experimental Psychology: Learning, Memory, & Cognition, 23(5), 1122-1139.


Kenneth I. FORSTER and Csaba VERES
University of Arizona


The hypothesis that the pattern of priming effects with orthographically related primes depends on awareness of the prime is investigated. The work of Humphreys, Besner and Quinlan (1988) indicates that form-primes (e.g., contrast-CONTRACT) are effective only when the prime is masked. However, Veres (1986) found that nonword primes (e.g., controct-CONTRACT) were effective whether the prime was masked or not. Using an improved design, the present study showed that in a lexical decision task with clearly identifiable primes (500 ms SOA), priming was obtained only with nonword primes. However, when the prime was masked (50 ms SOA), the same pattern of priming was observed, which fails to confirm earlier findings. Subsequently, it was shown that the prime lexicality effect with masked primes (weaker form-priming with word primes) depends on using nonword distractors that are very close to a particular word (e.g., UNIVORSE). When more distant nonword distractors are used (e.g. ANIVORSE), priming with word primes is restored, and there is no difference between word and nonword primes. This pattern is also observed with a semantic categorization task, and with lexical decision when a mask is interpolated between the prime and the target. These findings are interpreted in terms of the multiple readout model of Grainger and Jacobs (in press), and the entry-opening model of Forster and Davis (1984).

Forster, K. I., & Veres, C. (1998). The prime lexicality effect: Form-priming as a function of prime awareness, lexical status, and discrimination difficulty. Journal of Experimental Psychology: Learning, Memory, & Cognition, 24(2), 498-314.

Cascaded versus noncascaded models of lexical and semantic processing: The turple effect.

University of Arizona


The density of the orthographic neighborhood surrounding an item has been shown to have an inhibitory effect for nonwords in a lexical decision experiment. Four experiments are reported investigating whether a similar pattern holds for a semantic categorization task (animal vs. non-animal). 21-32 undergraduates participated in the 4 experiments. In the 1st experiment, no effects of neighborhood density were found for nonexemplars, whether they were words or nonwords. The absence of any inhibitory effect for nonwords implies that close orthographic neighbors are ignored. However, the 2nd experiment showed that if the nonword has a neighbor that is an animal name (e.g., turple), an interference effect is observed, implying that neighbors do exert an effect if they have the right semantic properties. The same items showed no additional interference in lexical decision. These results suggest the involvement of semantic properties very early in the processing cycle. A cascaded processing system monitoring activation in semantic features can explain these results, but cannot explain the frequency effect observed for nonexemplar words or the fact that variation in density is irrelevant when one of the neighbors is an exemplar.

Forster, K. I., & Hector, J. (2002). Cascaded versus noncascaded models of lexical and semantic processing the turple effect. Memory & Cognition, 30(7), 1106-1116.

Subject-verb agreement processes in comprehension.

Janet NICOL, Kenneth FORSTER, & Csaba VERES
University of Arizona

Examined the comprehension of sentences containing a subject noun phrase containing several daughter noun phrases which matched or mismatched in number with the head noun phrase, explored the effects of notional number on the comprehension of sentences in which the verb is inflected for number, and investigated the nature of the syntactic representation that readers compute during sentence comprehension. Five reading experiments with a total of 276 college students revealed the following: (1) patterns of reading times mirror the production error (subject-verb agreement error) asymmetry found in sentence production; (2) a phrase which is conceptually plural but grammatically singular produces no more reading difficulty than one which is conceptually and grammatically singular, a result which mimics J. K. Bock and C. A. Miller's (1991) production results; and (3) interference from an intervening plural depends on a close syntactic link to the head noun phrase. Results suggest that although the computation of agreement may be accomplished differently in the sentence production and comprehension systems, interference may arise whenever a structure containing a singular head and intervening plural is computed, whether during production or comprehension.

Nicol, J. L., Forster, K. I., & Veres, C. (1997). Subject-verb agreement processes in comprehension. Journal of Memory & Language, 36(4), 569-587.