Suppr超能文献

形态屈折变化学习与泛化的神经基础。

The neural bases of the learning and generalization of morphological inflection.

作者信息

Nevat Michael, Ullman Michael T, Eviatar Zohar, Bitan Tali

机构信息

Language and Brain Plasticity lab, IIPDM, University of Haifa, Israel.

Department of Neuroscience, Georgetown University, United States.

出版信息

Neuropsychologia. 2017 Apr;98:139-155. doi: 10.1016/j.neuropsychologia.2016.08.026. Epub 2016 Aug 26.

Abstract

Affixal inflectional morphology has been intensively examined as a model of productive aspects of language. Nevertheless, little is known about the neurocognition of the learning and generalization of affixal inflection, or the influence of certain factors that may affect these processes. In an event-related fMRI study, we examined the neurocognition of the learning and generalization of plural inflections in an artificial language, as well as the influence of both affix type frequency (the proportion of words receiving a given affix) and affix predictability (based on phonological cues in the stem). Adult participants were trained in three sessions, and were scanned after the first and last sessions while inflecting trained and untrained words. Untrained words yielded more activation than trained words in medial frontal (including pre-SMA) and left inferior frontal cortices, which have previously shown activation in compositional grammatical processing. A reliance on phonological cues for untrained word inflection correlated positively with pre-SMA activation, but negatively with activation in the pars triangularis. Thus, pre-SMA may be involved in phonological cue-based composition, while the pars triangularis underlies alternative processes. Inflecting trained items yielded activation in the caudate head bilaterally, only in the first session, consistent with a role for procedural memory in learning grammatical regularities. The medial frontal and left inferior regions activated by untrained items were also activated by trained items, but more weakly than untrained items, with weakest activation for trained-items taking the high-frequency affix. This suggests less involvement of compositional processes for inflecting trained than untrained items, and least of all for trained inflected forms with high-frequency affixes, consistent with the storage of such forms (e.g., in declarative memory). Overall, the findings further elucidate the neural bases of the learning and generalization of affixal morphology, and the roles of affix type frequency and affix phonological predictability in these processes. Moreover, the results support and further specify the declarative/procedural model, in particular in adult language learning.

摘要

词缀屈折形态学已被作为语言生成方面的一个模型进行了深入研究。然而,对于词缀屈折的学习和泛化的神经认知,或者可能影响这些过程的某些因素的影响,我们知之甚少。在一项事件相关功能磁共振成像(fMRI)研究中,我们考察了人工语言中复数屈折的学习和泛化的神经认知,以及词缀类型频率(接受给定词缀的单词比例)和词缀可预测性(基于词干中的语音线索)的影响。成年参与者接受了三个阶段的训练,并在第一阶段和最后阶段之后进行扫描,同时对训练过和未训练过的单词进行屈折变化。在额内侧(包括前辅助运动区)和左下额叶皮质,未训练的单词比训练过的单词产生更多激活,这些区域先前在成分性语法加工中已显示出激活。对未训练单词屈折变化时对语音线索的依赖与前辅助运动区的激活呈正相关,但与三角部的激活呈负相关。因此,前辅助运动区可能参与基于语音线索的合成,而三角部是替代过程的基础。对训练过的项目进行屈折变化仅在第一阶段双侧尾状核头部产生激活,这与程序记忆在学习语法规则中的作用一致。未训练项目激活的额内侧和左侧下部区域在训练项目中也被激活,但比未训练项目弱,对于采用高频词缀的训练项目激活最弱。这表明与未训练项目相比,对训练项目进行屈折变化时合成过程的参与较少,对于带有高频词缀的训练屈折形式参与最少,这与这些形式(例如,在陈述性记忆中)的存储一致。总体而言,这些发现进一步阐明了词缀形态学学习和泛化的神经基础,以及词缀类型频率和词缀语音可预测性在这些过程中的作用。此外,结果支持并进一步明确了陈述性/程序性模型,特别是在成人语言学习中。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验