Neighing dogs and ringing cars - studies on lexical processing during language production with a novel picture-sound interference paradigm

在日常生活中，我们经常被自然产生的声音所包围。这些声音中的许多都很容易识别，并且与特定来源密切相关。例如，当我们听到叫声时，我们就会自动地想起发出这种声音的动物，我们就会毫不犹豫地把它命名为“狗”。换句话说，许多自然发音允许快速且明确地激活相应的语义和词汇表示。在本研究项目中，自然语音和语义表征之间的这种联系被用来研究语言产生中的词汇过程。特别是，该项目旨在帮助回答一个备受争议的问题：词汇选择，即我们从一系列选择中快速选择合适单词的能力，是否是一个竞争性的过程。目前关于词汇选择本质的争论主要集中在图片-词汇干扰范式的证据上。在这项任务中，参与者说出了图片的名字，而忽略了分散注意力的单词。一个关键发现是，与不相关的干扰词(例如，词：汽车)相比，如果干扰词在语义上与图片相关(例如，图片：狗，词：马)，则图片命名速度较慢。这种语义干扰效应通常被认为是竞争性词汇选择的证据。然而，近些年来，这种解释一直备受争议，并提出了一种在非竞争性词汇选择框架内对这种影响的另一种解释。本研究的主要目的是通过研究自然语音和图片加工中的跨通道语义交互作用来检验对词汇选择的竞争性和非竞争性解释的预测。在该项目的第一阶段，自然声音被用作分心刺激(在图片命名任务中)和目标刺激(在声音命名和声音分类任务中)。该项目第一阶段的主要结果是观察了图片命名中的语义干扰效应(使用分心声音命名)和声音命名使用分心图片。这些发现表明，语义干扰并不局限于干扰词。这支持了语义干扰效应反映了言语产生中的竞争性词汇选择机制的假说。在项目的第二阶段，我们旨在通过直接比较单通道和跨通道干扰任务来阐明目标通道和分心通道(分别是它们的组合)是否调节语义干扰效应。此外，我们将研究声音和图像之间的跨通道语义干扰是否会概括为顺序命名任务。这将为在概念和词汇表征的界面上形成语言产生的增量学习机制提供新的见解。

In everyday life we are constantly surrounded by naturally occurring sounds. Many of these sounds can be easily identified and are strongly associated with specific sources. For instance, when we hear a bark, we are automatically reminded of the animal producing that sound and we would have no problem naming it as “dog”. In other words, many natural sounds allow for a fast and unambiguous activation of corresponding semantic and lexical representations. In this research project this association between natural sounds and semantic representations is used to investigate lexical processes in language production. In particular the project aims to help answer the much debated question of whether lexical selection, that is, our ability to swiftly select the right word among a number of alternatives, is a competitive process or not. The ongoing debate about the nature of lexical selection is highly focused on evidence from the picture-word interference paradigm. In this task participants name pictures while ignoring distractor words. A key finding is that picture naming is slower if the distractor word is semantically related to the picture (e.g., picture: dog, word: horse) compared to an unrelated distractor word (e.g., word: car). This semantic interference effect is often regarded as evidence for competitive lexical selection. However, this interpretation has been strongly disputed in recent years and an alternative explanation of the effect within a framework of non-competitive lexical selection has been proposed. The main goal of this research project is to test predictions of competitive and non-competitive accounts of lexical selection by investigating cross-modal semantic interaction in processing natural sounds and pictures. In the first phase of the project naturals sounds were used as distractor stimuli (in picture naming tasks) and as target stimuli (in sound naming and sound categorization tasks). Main result of the first phase of the project are observations of semantic interference effects in picture naming with distractor sound) and sound naming with distractor pictures. These findings show that semantic interference is not restricted to distractor words. This supports the hypothesis that semantic interference effects reflect a competitive lexical selection mechanism in speech production. In the second phase of the project we aim to elucidate whether target and distractor modality (respectively their combination) modulate semantic interference effects by directly comparing uni-modal and cross-modal interference tasks. Furthermore, we will investigate whether cross-modal semantic interference between sounds and pictures generalizes to sequential naming tasks. This will provide novel insights into the incremental learning mechanisms shaping language production at the interface of conceptual and lexical representations.