Three experiments were conducted to examine the nature of event-related brain protentials (ERPs) in a priming situation. In Exp. 1 and 2, we used the SOV paradigm, where simple Japanese sentences were presented visually in the order of subject (S), object (O), and verb (V), of which the subjects judged the veracity. S was a concrete noun, V was affirmative (is) or negative (isn't).
The effect of probability of the REP was examined in Exp. 1. In L coneition, the REP was represented by 25% and in M conditin, 50%. DCN elicited the N400 and REp elicited a positive wave. The positivity elicited by REP fdid not differin either condition, suggesting that the positivity did not reflect the rare occurrence of the event, but waas pure reflection of repetition.
In Exp. 22, the effect of SOA was examined. In S condition, the SOA from S to O was 200 ms, and in M condition, 800 ms. In both conditions, SOA from S to V was 1600 ms. Again, DCN elicited the N400 and REP elicited positivity. When the SOA w
as short, the N400 was delayd, but there was no effects on the REP positivity.
Exp. 3 was conducted using a traditional priming paradigm with a lexical decision task. In C condition, 43% of the target fell into the "prime" category. In A condition, about 43% of the target was an antonym of the prime word. In both conditions, about 43% of the target was nonword. facilitation and inhibition in reaction time were greater in the A than in the C condition. Although there was no condition effect on N400, greater facilitation was associated with earlier positivity, suggesting that the N400 did not reflect the inhibition observed in a behavioral measure.