Derek: The Direct Encoding Routine for Evoking Knowledge
2012
Taylor (1982), who reported a related, very similar set of experiments. Their Experiment III used both nonword phonological primes and word graphemic primes in a tachistoscopic recognition task. Inspection of the list of stimuli (Humphreys et al., Appendix C) reveals orthographic target-prime pairs such as CIRCLE - PIRAIL, CLAW - BLET, KALE - DARL, QUEEN - FRENE, and CALM - RAIM. But the list also contains pairs such as DRAW - DREP, DREW - DROF, FALL - FIRL, FORM - FEAM, HAWK - HEEK, and MAUL - MOOL. I would argue that it makes a mockery of the idea of orthographic overlap to suggest that the primes in the first set could facilitate recognition of their targets (it must be noted that words in this condition were selected as orthographic controls for the phonological primes, not for their orthographic overlap with targets. Nonetheless, the point that priming in this condition could be due to phonological similarity between prime and target remains). In the case of the second set, flexible rule theory would be quite comfortable with a priming effect of primes on targets, based on the consonant frames of the stimuli. Thus, Humphreys et al. &s;s conclusion (and Evett and Humphreys&s; before them) that there is no phonological priming because there is no advantage of phonological over "graphemic" conditions is suspect.