Single-word naming in a transparent alphabetic orthography
The cognitive processes involved in single-word naming of the transparent Turkish orthography were examined in a series of nine naming experiments on adult native readers. In Experiment 1, a significant word frequency effect was observed when matched (i.e. on initial phoneme, letter length and number of syllables) high- and low-frequency words were presented for naming. However, no frequency effect was found in Experiment 2, when an equal number of matched (i.e. on initial phoneme, letter length and number of syllables) nonword fillers were mixed with the target words. A null frequency effect was also found in Experiment 3 when conditions were mixed-blocks, i.e. high- and low frequency were words presented in separate blocks mixed with an equal number of matched nonword fillers. Experiment 4 served the purpose of creating and validating nonwords (to be used in Experiments 5 and 6) that could be named as fast as high- and low-frequency words by manipulating the letter length of nonwords. A significant word frequency effect emerged with both the mixed-block design (Experiment 5) and mixed design (Experiment 6) when the nonword fillers matched the target words in speed of naming. Experiment 7, however, found no frequency effect when high- and low-frequency words were mixed with word fillers that were slower to be named (longer in length) than the target words. In Experiment 8, frequency was factorially manipulated with imageability (high vs. low) and level of skill (very skilled vs. skilled) which found significant main effects for word frequency and level of skill, and a significant 2-way interaction of skill by imageability and a significant 3-way interaction of skill by imageability by frequency. In Experiment 9, however, there was only a main effect for frequency when previously skilled readers performed on the same words used in Experiment 8. These findings suggest that whilst a lexical route dominates in naming the transparent Turkish orthography, an explanation that the readers shut down the operation of this route in the presence of nonword fillers is not entertained. Instead, the results suggest that both routes operate in naming, with the inclusion of filler stimuli and their “perceived difficulty” having an impact in the time criterion for articulation. Moreover, there are indications that a semantic route is involved in naming Turkish only when level of skill is taken into account. Implications of these findings for models of single-word naming are discussed.