Cognitive models of sociolinguistics must support a wide range of goal‐oriented behavior (e.g. Eckert, 2000a) without suggesting unrealistic levels of deliberative control on the part of speakers. The current study investigates the limits of deliberative control in audiovisual face‐voice perception. Perceivers evaluated co‐present recorded speech and static face pictures, rating the stimuli on the scales ‘accented’ and ‘good‐looking’ in one of three conditions: as a combined voice and face; evaluating the face while ignoring the voice; and evaluating the voice while ignoring the face. Perceivers' ability to ignore social information from a face or voice upon instruction are taken as indicative of deliberative control in social evaluation. The results suggest that deliberative control and evaluative relevance both play a role in perception, but that available social information is difficult to ignore completely. They further suggest an asymmetry making voices more difficult to ignore than static faces and support a model of sociolinguistic perception and evaluation as a function of multiple competing processes under varying degrees of deliberative control.

中文翻译：

---

视听社会语言感知中的协商控制*

社会语言学的认知模型必须支持各种各样的面向目标的行为（例如Eckert，2000a），而又不能暗示说话者的协商控制水平不切实际。当前的研究调查了视听面部声音感知中的协商控制的局限性。感知者评估共同呈现的语音和静态面部图像，并在以下三种情况之一中以“重音”和“好看”的等级对刺激进行评分：在忽略声音的同时评估面部；并在忽略脸部的同时评估声音。感知者在接受指导时从脸部或声音中忽略社交信息的能力被视为社交评估中的协商控制的指示。结果表明，协商控制和评估相关性均在感知中发挥作用，但是现有的社交信息很难被完全忽略。他们进一步提出一种不对称性，使声音比静态面孔更难以忽略，并支持社会语言感知和评估模型，该模型是在不同程度的协商控制下多个竞争过程的函数。