关键词: 语义联系, 关联再认, 熟悉, FN400, LPC

Abstract: Previous research has revealed two different old/new effects, the early mid-frontal old/new effect (a.k.a. FN400 component) and the late parietal old/new effect (a.k.a. LPC component), which reflect dual-process model’s two processes of familiarity and recollection respectively. Though associative recognition is thought to be more based on recollection rather than familiarity compared to item recognition, recent studies have confirmed that familiarity makes greater contribution to associative memory when to-be-remembered items of a pair have a stronger semantic relation. However, it remains unclear whether the modulations of different subtypes of semantic relations on the old/new effect of associative recognition are different. The roles of the processes of familiarity and recollection of a single item that belongs to a pair also require further investigations. Using event-related potential technique (Neuroscan), the current study aimed to explore the modulations of different subtypes of semantic relations on the waveforms of associative recognition. The current experiment was a within-subject design, which had semantic relation and paired type of pictures as variables, the levels of semantic relation were thematic and taxonomic relations, and the levels of paired type of pictures were new pair, old pair, rearranged pair, and “old + new” pair. A total of 22 right-handed adults, aged 23.83 ± 5.21 years, participated in the current experiment. In the experiment, 672 named pictures (336 pairs) which were all from Hemera Color Gallery served as stimuli, and the procedure was conducted via the continuous recognition paradigm. During the whole experiment, participants were instructed to study paired pictures with two different semantic relations (thematic and taxonomic relations) and to make discriminations among new, old, rearranged, and “old+new” pairs. The results showed that, under both thematic and taxonomic relation conditions, the FN400 component (during 350–500 ms) that related to familiarity process was confirmed for old, rearranged, and “old+new” pairs. In other words, compared to new pairs, more positive-going waveforms were confirmed over frontal region for old, rearranged, and “old+new” pairs. Under the thematic relation condition, reliable LPC component (during 500–650 ms) that related to recollection process was observed for old, rearranged, and “old+new” pairs. That is to say, more positive-going waveforms over parietal region were elicited by old, rearranged, and “old+new” pairs than by new pairs. For the taxonomic relation, the LPC component was only elicited by old pairs, showing more positive-going waveforms over parietal region for old pairs versus new pairs. To conclude, the old/new effect differences between the studies of current pictures and previous words indicate that the old/new effect of associative recognition is not only semantic-relation-sensitive but also stimulus-specific. The old/new effect differences between the pairs of thematic relation and of taxonomic relation indicate that the old/new effect of associative recognition is sensitive to semantic relation. The old/new effect elicited by “old+new” pairs suggests that its effect can be affected by the familiarity of a single item of this type of pairs. These results reinforce the dual-process model.

Key words: semantic relation, associative recognition, familiarity, FN400, LPC