The interplay of bottom-up and top-down processes in the segmentation of speech and action

自下而上和自上而下过程在言语和动作分割中的相互作用

• 批准号: 277140225
• 负责人: Professorin Dr. Birgit Elsner
• 金额: --
• 依托单位: Abteilung Entwicklungspsychologie
• 依托单位国家: 德国
• 项目类别: Research Units
• 财政年份: 2015
• 资助国家: 德国
• 起止时间: 2014-12-31 至 2023-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/277140225?language=en
• 关键词: interplay; bottom; up; top; down

Both heard speech and observed action-sequences consist of continuous streams of to-be- processed information. Boundaries in the signal help the listener/observer to segment or chunk the incoming information into meaningful units. In speech, intonation-phrase boundaries, which usually coincide with syntactic boundaries, are characterized by prosodic boundary cues, such as pre-boundary pitch changes, pre-boundary lengthening, and pause. The processing of those cues is an important prerequisite of language acquisition, and is accordingly found already in the first year of life. In action sequences, pauses are also relevant to identify action units, and respective processing has been shown for 11-month-old infants. Action segmentation can also rely on other kinematic cues, such as time intervals with high limb acceleration/deceleration and high motion velocity, but the processing of such has only scarcely been studied in infants. Besides these signal-driven bottom-up cues, adults also apply top-down knowledge to segment continuous streams of information (in the linguistic domain: lexical and syntactic knowledge; in the action domain: conceptual knowledge about the structure and functionality of actions). In the action domain, such knowledge-based segmentation has been proposed already for 11-month-olds. It is so far unknown how bottom-up and top-down mechanisms interact in infants’ and young children’ segmentation of speech or action, and to what extend their operation is domain-general. Using series of EEG- and eye-tracking investigations in young infants and children (from 6 to 42 months of age), our interdisciplinary project aims to characterize commonalities and/or differences in segmentation or chunking processes in speech perception and action observation. We will mainly focus on the roles of signal-based bottom-up processes and of knowledge-based top-down processes in the early development of segmentation abilities in the two domains. In addition, we will characterize the neurocognitive basis of the development of speech and action segmentation, which will further inform about the domain-generality or -specificity of the involved cognitive processes. Furthermore, we will gain a better knowledge on how the interplay of bottom-up and top-down processes in segmentation is related to the early development of general cognitive competences. In sum, the planned research will contribute to a more detailed understanding of the interplay of language, cognition, and the brain in young children’s segmentation of continuous streams of linguistic and non-linguistic information.

听到的语音和观察到的动作序列都是由连续的待处理信息流组成的。信号中的边界帮助听者/观察者将传入的信息分割或分块成有意义的单元。在语音中，语调-短语边界通常与句法边界一致，其特征在于韵律边界线索，如边界前音高变化、边界前延长和停顿。对这些线索的处理是语言习得的重要先决条件，因此在生命的第一年就已经发现了。在动作序列中，停顿也与识别动作单元有关，并且已经为11个月大的婴儿显示了相应的处理。动作分割还可以依赖于其他运动学线索，例如具有高肢体加速度/减速度和高运动速度的时间间隔，但是这些处理在婴儿中几乎没有被研究。除了这些信号驱动的自下而上的线索，成年人还应用自上而下的知识来分割连续的信息流（在语言领域：词汇和句法知识;在动作领域：关于动作结构和功能的概念知识）。在行动领域，这种基于知识的分割已经被提出用于11个月大的婴儿。到目前为止，还不清楚自下而上和自上而下的机制如何在婴幼儿的言语或动作分割中相互作用，以及它们的运作在多大程度上是领域通用的。使用一系列的脑电图和眼动跟踪调查的婴幼儿和儿童（从6至42个月的年龄），我们的跨学科项目的目的是表征的共性和/或差异分割或组块过程中的语音感知和动作观察。我们将主要集中在基于信号的自下而上的过程和基于知识的自上而下的过程在这两个领域的分割能力的早期发展中的作用。此外，我们将描述语音和动作分割的发展的神经认知基础，这将进一步告知所涉及的认知过程的域的一般性或特异性。此外，我们将更好地了解自下而上和自上而下的分割过程的相互作用如何与一般认知能力的早期发展相关。总之，计划中的研究将有助于更详细地了解语言，认知和大脑在幼儿分割连续的语言和非语言信息流中的相互作用。