Effects of association, frequency, and stimulus quality on naming words in the presence or absence of pseudowords

Abstract

According to classical dual-route theory, effects of associative priming and frequency on the naming of printed words arise from lexical access and should be weak or absent when word names are assembled prelexically. Assembled naming would be more likely in a shallow orthography, especially in the presence of nonwords. This hypothesis was examined with the shallow Serbo-Croatian orthography. Interactions between association, frequency, and stimulus quality were also examined in both Serbo-Croatian and English. Contrary to classical dual-route theory, both lexical effects were found for naming words in Serbo-Croatian, with or without nonwords. Neither interaction was significant in Serbo-Croatian and only association X quality was significant in English. Discussion focused on (a) the claim that lexical effects on naming in a shallow orthography constitute prima facie evidence against either prelexical phonology or the orthographic depth hypothesis, and (b) the possible factorization of frequency and active associative knowledge in naming words.The rapid naming task places few explicit demands on the reader. It does not, in principle, require that the reader know the meaning of the letter string or what words it is related to or how common it is or even if it is, in fact, a word. Nonetheless, the latency to initiate the pronunciation of a letter string can be affected by all of these factors. When a target is preceded by an associative context, it is named faster than when it is preceded by an unrelated context (e.g., Meyer, Schvaneveldt, & Ruddy, 1975;Seidenberg, Waters, Sanders, & Langer, 1984). High-frequency words tend to be named faster than lowfrequency words (e.g., Forster & Chambers, 1973; MeCann, Besner, & Davelaar, 1988), and words tend to be named faster than nonwords (e.g., Forster & Chambers, 1973;Frederiksen & Kroll, 1976). Association, frequency, and lexicality are understood to exert their influences only after the lexicon has been accessed. Granting this understanding, what do such effects suggest about how rapid naming is accomplished?Discussions of naming often are framed in terms of accessed versus assembled phonology. In the former, visual characteristics of the letter string are mapped in whole word fashion onto lexical representations where information about phonology is accessed. In the latter, the visual form of the letter string is recoded into corresponding phonological constituents, and this assembled

Visual and Associative Factors in Processing Serbo-Croatian Words

“…The upshot is that, for word naming, there ought to be di erent dependencies of stimulus quality on frequency and context. This prediction has been veri® ed for English (Borowsky & Besner, 1991;Carello et al, 1995) but has not been veri® ed for Serbo-Croatian (Carello et al, 1995).…”

“…In two experiments, the latency di erence between naming following an associated prime and naming following a non-associated prime was not magni® ed by degrading the target (Carello et al, 1995). The main e ects on naming latency of associative priming and stimulus degradation were reliable, but their relation was additive not interactive.…”

“…In Carello and co-workers' (1995) experiments, the SOA was 600 msec. A reasonable assumption is that the temporal lag of target onset relative to prime onset is an important determinant of priming and its interactions (Carello et al, 1995;Lukatela, Lukatela, Carello, & Turvey, 1993). Experiments that have manipulated the SOA and the strength of the prime± target association have found that the priming e ects on naming are apparent by 150 msec (the briefest SOA evaluated), increase to an asymptotic level that depends on the association strength, and do so at a rate that is independent of association strength (Lorch, 1982).…”

“…Clearly, in view of the ambiguity in Bajo and co-workers' experiments and the clear negative outcome in Carello and co-workers' experiments, the question of a stimulus quality´associative priming interaction in shallow orthographies remains open. In the present paper, we continue the investigations of Carello et al (1995) regarding the interaction between stimulus quality and associative priming in the naming of SerboCroatian words.…”

“…This is true for both lexical decision and naming (Becker & Killion, 1977;Borowsky & Besner, 1991;Carello, Lukatela, Peter, & Turvey, 1995;Massaro, Jones, Lipscomb, & Scholz, 1978;Meyer et al, 1975); semantic (Massaro et al, 1978;Sanford, Garrod, & Boyle, 1977) and associative priming (Becker & Killion, 1977;Meyer et al, 1975); degradation accomplished with a superposed regular dot pattern (Meyer et al, 1975), checkerboard pattern (Sanford et al, 1977) , 180°rotation of the target (Massaro et al, 1978), or reduced target intensity (Becker & Killion, 1977); and spoken (Massaro et al, 1978) or silently read contexts (Becker & Killion, 1977;Meyer et al, 1975). Behind the original interpretation of this interaction were the assumptions that a word's name and identity reside in its lexical entry, degradation retards the activation of a word's lexical entry by its sublexical components, and a preceding associate brings the lexical entry closer to threshold (Meyer et al, 1975).…”