Die Planung von Äußerungen aus sprachübergreifender Sicht

思维是整体的--对一头紫色奶牛的思考并不是不确定的、紫色的东西，而是顺序地对着一头奶牛思考--但语言的使用要求这些想法在时间上被组织起来(这样说话者就会一个接一个地说“一头紫色奶牛”)。说话，就像任何有运动输出的认知任务一样，需要提前计划。计划策略，例如，在开始发音之前提前计划了多少单词，或者单词在计划单元中编码的顺序，人们很难理解。多项研究结果不一，导致几位作者提出，规划策略是灵活的。本项目的目的是审查支配和限制这种灵活性的原则。它利用了这样一个事实，即在某些话语/语言中，规划策略受到语法和语音依赖的限制。例如，在法语中，限定词或形容词的发音通常取决于名词的性别(语法限制，例如le beau chat‘the good cat’与la belle plie‘the good rain’)，以及下一个单词的音系学(例如，le beau chat与le beláne‘the good donkey’)。为了无差错地产生这些话语，说话者别无选择，只能(1)在发音开始之前将这些话语计划到最后，(2)首先对后续单词进行编码。换句话说，对单词发音的语音和语法约束定义了这些发音的最小规划单位和单词的编码顺序。相比之下，在德语中，只有语法信息可以限制限定词和形容词的发音(例如，die Katze、der Tisch、groçe Katze、groüer Tisch)。该项目考察了跨语言和跨话语在语法和语音约束方面的差异对规划策略的影响，目的有几个。第一个目标是检验这样一个假设，即如果存在这些限制，规划的灵活性就会受到限制。第二个目标是表明，在这些限制之上，灵活性被用来满足交流需求。最后一个目标是确定规划策略是否遵守语言内部和跨语言的最大一致性(某种类型的所有话语都是相似的规划)或最优化原则(规划是根据每种话语的当地要求进行调整的)。三个系列的行为实验将被要求参与者产生不同结构的话语。将收集德语、法语和英语的数据，因为这些语言的话语在语法和语音方面对单词发音的限制有所不同。此外，为了对语言产生系统的认知结构提供新的见解，本项目将有助于长期存在的关于语言使用中的语言特殊性与语言共性的辩论。

Thoughts are holistic - thinking about a purple cow does not consist in thinking at something indefinite, something purple, and at a cow, sequentially- but the use of language requires that these thoughts be organized in time (such that the speaker says “a purple cow”, one word after the other). Speaking, like any cognitive task with a motor output, requires advanced planning. Planning strategies, e.g., how many words are planned ahead before the onset of articulation or in which order words are encoded within planning units, are poorly understood. Mixed findings across studies led several authors to suggest that planning strategies are flexible. The aim of the present project is to examine the principles that govern and limit this flexibility. It exploits the fact that in some utterances/languages, planning strategies are constraints by grammatical and phonological dependencies. In French for instance, the pronunciation of determiners or adjectives often depends on the gender of the noun (grammatical constraint, e.g., le beau chat ‘the nice cat’ vs. la belle pluie ‘the nice rain’), and on the phonology of the next word (phonological constraint, e.g., le beau chat vs. le bel âne ‘the nice donkey’). In order to produce these utterances without errors, speakers have no other choice but (1) to plan these utterances up to the end before the onset of articulation, and (2) to encode subsequent words first. In other words, phonological and grammatical constraints on word pronunciation define the minimal planning unit for these utterances and the order in which words are encoded. By contrast, in German, only grammatical information can constrain the pronunciation of determiners and adjectives (e.g., die Katze, der Tisch, große Katze, großer Tisch). The project examines the impact of cross-linguistic and cross-utterance differences in grammatical and phonological constraints on planning strategies, with several aims. The first aim is to test the hypothesis that flexibility in planning is limited by these constraints, when present. The second aim is to show that above these constraints, flexibility is used to accommodate communicative needs. The last aim is to determine whether planning strategies obey maximum consistency (all utterances of a certain type are planned similarly) or optimization principles (planning is adapted to the local requirements of each utterance) within and across languages. Three series of behavioural experiments will be conducted in which participants are asked to produce utterances of difference structures. Data will be collected in German, French, and English as utterances in these languages differ in the sets of grammatical and phonological constraints on word pronunciation. In addition, to providing novel insights into the cognitive architecture of the language production system, the present project will contribute to the long-standing debate about language-specificity versus language universals in language use.