Phonological computation and missing vowels: Mapping lexical involvement in reading.

Abstract

The role of assembled versus addressed phonology in reading was investigated by examining the size of the minimal phonological unit that is recovered in the reading process. Readers named words in unpointed Hebrew that had many or few missing vowels in their printed forms. Naming latencies were monotonically related to the number of missing vowels. Missing vowels had no effects on lexical decision latencies. These results support a strong phonological model of naming and suggest that even in deep orthographies, phonology is not retrieved from the mental lexicon as a holistic lexical unit but is initially computed by applying letter-to-phoneme computation rules. The partial phonological representation is shaped and completed through top-down activation.

Prelexical phonologic computation in a deep orthography: Evidence from backward masking in Hebrew

“…This process produces an impoverished phonological representation that includes mostly consonantal information. The vowel information is probably available in a second cycle from top-down lexical information (Berent & Perfetti, 1995;Frost, 1995). Our findings, therefore, provide additional support for a strong phonological model of reading that considers phonological computation a primary stage of visual word processing, regardless of the depth of the orthography.…”

“…Whereas some theorists suggest that phonological recoding is a late component ofprint processing, occurring mainly for low-frequency, regular, or consistent words (e.g., Coltheart, 1980;Paap & Noel, 1991;Seidenberg, Waters, Barnes, & Tannenhaus, 1984), other theorists contend that the recovery of phonological information from print is the primary cognitive operation in visual word perception (e.g., Frost, 1995;Katz & Frost, 1992;Van Orden, Johnston, & Halle, 1988).…”

Age of acquisition, word frequency, and the role of phonology in the lexical decision task

“…If one assumes that the phonological representations of earlyacquired words are easier to activate than those of lateacquired words, the matching process would take place more rapidly for the former than for the latter. There is also a "weaker" version ofthe phonological mediation hypothesis, which allows for parallel activation of prelexical and addressed phonology (Frost, 1995). Although the studies reported in this article were not designed to evaluate this hypothesis directly, we are inclined to think that our results are best explained by the automatic and compulsory activation ofphonology in word recognition (and certainly in the LDT).…”

SOA does not reveal the absolute time course of cognitive processing in fast priming experiments☆

“…It poses, however, interesting constraints on the ability of the model to account for lexical access occurring via fast prelexical phonological encoding (see Rastle & Brysbaert, 2006, for a detailed discussion). In contrast, to dual-route models, the strong phonological theory (e.g., Frost, 1995Frost, , 1998 Lukatela & Turvey, 1994a,b;Van Orden, Pennington & Stone, 1990) regards prelexical phonological computation as the initial cognitive operation launched in the reading process, and considers it to be very fast. Thus, findings regarding the speed of phonological computation are expected to provide distinguishing evidence between the two Correspondence should be addressed to Ram Frost,…”