Can masked priming effects be obtained with words?

Abstract

One important issue often neglected in the literature on visual-wordrecognition,despiteitsobviousecologicalvalidity,ishowthe cognitive system processes handwritten words. Althoughhandscript wastheinitial and onlyway of writing/reading untilthe 15th century, the vast majority of psycholinguistic experi-ments use spotless printed words in which, unlike inhandwriting, letters are physically separated within each wordand the instances of each letter are identical (e.g., comparedenied vs. ). Not surprisingly, there is some costassociated with the processing of handwritten words (Barnhart& Goldinger, 2010; Manso de Zuniga, Evett, & Humphreys,1991). Given the inherent noise in the bottom-up informationfrom handwritten words, the cognitive system has to relymore on more effortful, top-down processes (Manso deZuniga et al., 1991). Consistent with this interpretation,lexical effects are magnified with handwritten words—including the effects of word frequency, regularity, bidi-rectional consistency, and imageability (see Barnhart G see alsoGrainger, 2008, for a recent review), in which a brieflypresented prime word precedes the presentation of a targetword. One highly robust and replicated phenomenon in thiscontext is the masked repetition priming effect: Identifica-tion times to a target word are consistently faster when it hasbeen briefly preceded by the same word than when it hasbeen preceded by anunrelated word (Forster & Davis,1984;see Dehaene et al., 2001, for fMRI evidence; see Carreiras,Dunabeitia, & Molinaro, 2009, for ERP evidence).Can words produce masked priming effects?Clearly, if masked priming occurs with handwritten primes,it will be possible to examine the impact of handwrittenwords on orthographic, phonological, or morphologicalprocesses at the earliest stages of processing usingbehavioral and/or neurophysiological techniques. There-fore, we believe that it is critical to demonstrate, in the firstplace, the existence of masked repetition priming withhandwritten primes (i.e., the most robust form of maskedpriming). Previous research has found that masked primingeffects are quite abstract in nature; indeed, they can beobtained between upper- and lowercase typed words thatare visually dissimilar (e.g., edge–EDGE; see Bowers,Vigliocco, & Haan, 1998). This would suggest that maskedpriming should extend between written and typed primesand targets—at least for easily readable prime words(e.g., –CABLE). However, the only published experi-ment on the issue, conducted by Qiao et al. (2010), failed tofind a masked repetition priming effect with handwrittenprimes using a semantic categorization task (i.e., wordsreferring to man-made objects vs. words referring to natural