Macken Bill, Taylor John C, Jones Dylan M
School of Psychology, Cardiff University.
J Exp Psychol Learn Mem Cogn. 2014 Sep;40(5):1257-70. doi: 10.1037/a0036845. Epub 2014 May 5.
The advantage for real words over nonwords in serial recall--the lexicality effect--is typically attributed to support for item-level phonology, either via redintegration, whereby partially degraded short-term traces are "cleaned up" via support from long-term representations of the phonological material or via the more robust temporary activation of long-term lexical phonological knowledge that derives from its combination with established lexical and semantic levels of representation. The much smaller effect of lexicality in serial recognition, where the items are re-presented in the recognition cue, is attributed either to the minimal role for redintegration from long-term memory or to the minimal role for item memory itself in such retrieval conditions. We show that the reduced lexicality effect in serial recognition is not a function of the retrieval conditions, but rather because previous demonstrations have used auditory presentation, and we demonstrate a robust lexicality effect for visual serial recognition in a setting where auditory presentation produces no such effect. Furthermore, this effect is abolished under conditions of articulatory suppression. We argue that linguistic knowledge affects the readiness with which verbal material is segmentally recoded via speech motor processes that support rehearsal and therefore affects tasks that involve recoding. On the other hand, auditory perceptual organization affords sequence matching in the absence of such a requirement for segmental recoding and therefore does not show such effects of linguistic knowledge.
在系列回忆中,真词相对于非词的优势——词汇性效应——通常归因于对项目层面语音的支持,这要么是通过重新整合实现的,即部分退化的短期痕迹通过语音材料的长期表征的支持得以“清理”,要么是通过长期词汇语音知识更强大的临时激活实现的,这种激活源于其与已建立的词汇和语义表征层面的结合。在系列识别中,词汇性效应要小得多,因为项目会在识别线索中重新呈现,这要么归因于长期记忆重新整合的作用极小,要么归因于在这种检索条件下项目记忆本身的作用极小。我们表明,系列识别中词汇性效应的降低不是检索条件的作用,而是因为之前的研究使用了听觉呈现,而我们在听觉呈现没有这种效应的情况下,证明了视觉系列识别中存在强大的词汇性效应。此外,在发音抑制的条件下,这种效应会消失。我们认为,语言知识会影响言语材料通过支持复述的言语运动过程进行分段重新编码的准备程度,因此会影响涉及重新编码的任务。另一方面,听觉感知组织在没有这种分段重新编码要求的情况下提供序列匹配,因此不会表现出这种语言知识的效应。