Abstract:

Recent research has indicated that humans exhibit better item recognition when working with their second language (L2) than in their first (L1). Associative and item recognition are based on different retrieval information and retrieval processes, even though they share certain characteristics. In the present study, we investigated whether bilingual associative recognition performance was better in L2 than in L1. We asked participants to complete two study-test tasks that were presented in Chinese or Uygur, as appropriate. During the study phase, participants were instructed to remember either compound or unrelated word pairs. Participants were then asked to indicate whether word pairs were intact, rearranged, or new.
According to the dual-process model of recognition memory, recognition can be mediated by two functionally distinct processes known as familiarity and recollection. Familiarity is a subjective feeling of prior encounter associated with an early (300~500 ms) frontal old/new effect (FN400). Recollection provides access to detailed information about the prior occurrence of an item and its associated episodic context, which is reflected by a later (500~800 ms) left parietal old/new effect (LPC). Traditionally, most researchers have assumed that associative recognition depended only on recollection, but more and more researchers have suggested that familiarity could also support associative recognition under unitized encoding conditions.
In the present study, we manipulated levels of unitization (LOU) through semantic relations of word pairs. In the unitization condition (compound word pairs), two words can be processed as a single coherent entity or an object. In contrast, in the non-unitization condition (unrelated word pairs), two items can only be treated as two separate objects. The current experiment found (1) associative recognition was more rapid in L2 than in L1 for both compound and unrelated word pairs, and the accuracy of associative recognition was higher in L2 for compound word pairs but equal for unrelated word pairs; and (2) associative recognition was better for compound than for unrelated word pairs both in L2 and in L1. The event-related potentials (ERPs) showed that in the unitization condition, recognition in L1 elicited both FN400 and LPC effects, indicating the unitization effect kept consistency in different language. However, recognition in L2 only elicited the FN400 effect. In addition, participants accomplished associative recognition at a time of 650 ms in L2. However, associative recognition was not completed until 900 ms in L1. This result pattern indicated that associative recognition in L2 can rely solely on familiarity. In the non-unitization condition, there was no FN400 effect, but the LPC effect occurred in both L2 and L1.
Together, these results indicate similar to item recognition, bilingual associative recognition is better in L2 than in L1 in the unitization condition. In addition, unitization increases the relative contribution of familiarity to subsequent associative retrieval. The practical significance of this study is that it provides a cognitive neuroscientific basis for promotion of the national common language in minority regions of China.

Key words: associative recognition, bilingualism, unitization, familiarity, recollection