An ERP investigation of orthographic priming with superset primes.

Prime stimuli formed by inserting unrelated letters in a given target word (called "superset" primes) provide a means to modify the relative positions of the letters shared by prime and target. Here we examined the time-course of superset priming effects in an ERP study using the sandwich-priming paradigm. We compared the effects of superset primes formed by the insertion of unrelated letters (e.g., maurkdet-MARKET), or by the insertion of hyphens (e.g., ma-rk-et-MARKET), with identity priming (e.g., market-MARKET), all measured relative to unrelated control primes. Behavioral data revealed significantly greater priming in the hyphen-insert condition compared with the letter-insert condition. In the ERP signal, letter-insert priming emerged later than hyphen-insert priming and produced a reversed priming effect in the N400 time-window compared with the more typical N400 priming effects seen for both hyphen-insert priming and identity priming. The different pattern of priming effects seen for letter-insert primes and hyphen-insert primes suggests that compared with identity priming, letter superset priming reflects the joint influence of: (1) a disruption in letter position information, and (2) an inhibitory influence of mismatching letters.