关键词: 联结记忆, 语义可整合性, 新旧效应, 项目提取, 关系提取

Abstract:

According to dual-process theory, recognition memory is supported by two distinct retrieval processes known as familiarity and recollection. Evidence coming from studies using event-related brain potentials (ERPs ) showed that FN400 old/new effect (300~500 ms, anterior) is related to familiarity and parietal old/new effect (400~800 ms, posterior) is related to recollection. Opinions about how familiarity and recollection support item recognition and associative recognition are changed with time. At beginning, it was found that both familiarity and recollection have contribution to item recognition, but associative recognition is mainly supported by recollection. However, results coming from source memory and associative memory indicated that when items, which are shown simultaneously, can be integrated or engage integrated processing, familiarity also affects associative recognition. The current experiment aims to give a further discussion about how familiarity and recollection affect associative recognition by manipulating the semantic integration between items. The present study used study-recognition paradigm. During study phase, subjects were required to learn two kinds of materials with different integration level. One kind of picture-pairs can be linked by four character Chinese idioms, which have high integration level. We called those picture-pairs as idiom condition. The other kind of pictures can’t be linked by four character Chinese idioms, which have low integration level. We called those picture-pairs as non-idiom condition. In idiom condition, subjects were required to report the idioms linked with picture-pairs. In non-idiom condition, subjects were required to report the names of pictures presented on the screen. In both conditions, subjects were required to remember the picture-pairs and the relationship between the pictures composing the picture-pairs. During test phase, subjects were required to judge if picture-pairs were presented in the same pairing as at study, were rearranged from at study, or were entirely new. ERPs results at test phase showed that old/new effects were obtained in both conditions at 200~800 ms period. Idiom condition has larger distribution than non-idiom condition. But only idiom condition showed old/rearrangement effect at 400~800ms period. The ERPs evoked by old judgments and rearrangement judgments in idiom condition were more positive than that in non idiom condition. However, there was no significant difference between the two conditions for new judgment. Topographic maps analysis showed that there were no significant distribution differences between old/new effects and old/rearrangement effects in both conditions. All results above indicate that item integration affects the efficiency of memory in associative recognition. Highly integrated items result in better recognition performance. Increasing the semantic integration of items can enhance the effects of familiarity and recollection in associative recognition, and can also enhance the effects of conceptual priming in associative recognition.

Key words: associative memory, semantic integration, old/new effect, item retrieval, relationship retrieval