Implicit versus explicit mechanisms of vocabulary learning and consolidation

Abstract

Abstract Previous research has suggested that integration of novel words into lexical competition benefits from a consolidation delay containing a period of sleep (Dumay & Gaskell, 2007). However, a recent study argued that learning novel words via a relatively implicit Hebb repetition task leads to later lexical integration independently of sleep (Szmalec, Page, & Duyck, 2012). It is not clear whether this different time course of lexical integration is a consequence of the learning method chosen, as opposed to other between study differences. Four experiments directly compared the learning of novel words using explicit and implicit methods, namely phoneme monitoring on isolated tokens vs. Hebb repetition of syllable sequences. The impact of the learning was tested at a range of later time-points using two tests of explicit knowledge (recognition and recall) and a test of lexical integration (pause detection on related existing words). Between experiments, we also manipulated exposure frequency, the impact of syllable grouping cues in Hebb repetition and the level of mismatch between novel and real words. The results suggested that learning novel words via Hebb sequence repetition does not confer a benefit on lexical integration prior to or after sleep. We observed an engagement in lexical competition only in the case where a good level of explicit training was followed by a consolidation delay. Recognition and recall performance was generally poorer for Hebb learning. We conclude that Hebb-style implicit learning of words does not allow consolidation processes to be bypassed in lexical integration.