Predictive Timing in Speech Perception and Production

这一扩展阶段的总体目标将是进一步加深我们对语音中适应行为和在线处理时间信息的理解。第一个项目阶段(阶段1)使用了听觉反馈的时间特性的实时操作。研究发现，受试者对音节密码子的干扰反应比音节密码子更强；这被解释为对密码子的运动计划有更大的延展性的证据。在特定对象的薪酬模式中发现了与此相对应的情况。所有受试者都接受了广泛的感觉运动技能测试。在同步敲击任务中，表现出更大补偿的受试者表现出更大的运动变异性(当运动计划不那么牢固地根深蒂固时，补偿更多)。我们现在的目标是扩大针对扰动的刺激操作的范围，并通过考虑特殊人群，即口吃者和音乐家来扩大参与者的感觉运动图谱。4个工作包的具体目标是：(1)在第一阶段，逐步和系统地引入扰动，使参与者变得可预测。分析的重点是调整存储的电机表示以抵消扰动。为了补充这种方法，我们现在将使用意外引入的扰动。随着话语的展开，这利用了在线监控过程(而不是适应性学习过程)。我们假设，干扰的知觉突显现在比运动表征的根深蒂固具有更大的权重。(2)阶段1的范式将被用来比较口吃和非口吃的参与者，再次侧重于对音节开头和尾部的干扰反应的比较。研究的目的将是确定口吃组是否只是对听觉反馈操作表现出整体较弱的反应(正如之前的一些研究所表明的那样)，或者这一组是否表现出与音节开头有关的特定加工缺陷的证据。(3)在第一阶段，我们收集了从儿童到年轻人的感觉运动技能(从时长辨别到预测性定时敲击任务)的广泛数据。此外，所有的非言语任务都有言语任务作为对应物。现在的目的是从发展的角度分析这些测试，以确定不同技能并行发展的程度。(4)在扩展阶段，我们的目标是比较音乐家和非音乐家，以提高对适应行为的程度和性质的理解。这从上面关于运动表征延展性的结果中浮现出来。到目前为止，这方面的讨论还很少，但在基于音乐的刺激的微扰实验中得到了一些关注。因此，为了实现我们的总体目标，我们将音乐家和非音乐家在语言和音乐刺激方面的明确比较视为一种特别有希望的途径。

The overall objective in this extension phase will be to further deepen our understanding of adaptive behavior and online processing of temporal information in speech. The first project phase (Phase 1) used real-time manipulation of the temporal properties of auditory feedback. It found stronger subject responses for perturbation of syllable codas compared to onsets; this was interpreted as evidence for greater malleability of motor plans for codas. A counterpart to this was found in subject-specific compensation patterns. All subjects were tested for an extensive battery of sensorimotor skills. Subjects showing greater compensation showed greater motor variability in synchronized tapping tasks (more compensation when motor plans are less strongly entrenched).We now aim to expand the range of stimulus manipulations for perturbation, and also to widen the sensorimotor profiles of participants by taking special populations into account, i.e. persons who stutter and musicians. Specific aims in 4 work-packages are:(1) In Phase 1 the perturbation was introduced gradually and systematically, becoming predictable for participants. Analysis focuses on adjustments to stored motor representations to counteract the perturbation. To complement this approach we will now use perturbations that are introduced unexpectedly. This taps into online monitoring processes as an utterance unfolds (rather than adaptive learning processes). We hypothesize that the perceptual salience of the perturbation now has more weight than the entrenchment of motor representations.(2) The paradigm of Phase 1 will be used to compare stuttering and non-stuttering participants, again focusing on a comparison of responses to perturbations in syllable onset vs. coda. The aim will be to determine whether the stuttering group simply shows an overall weaker response to auditory feedback manipulation (as some previous studies have indicated) or whether this group shows evidence of a specific processing deficit related to syllable onsets.(3) In Phase 1 we collected extensive data from children to young adults on sensorimotor skills (from duration discrimination to predictive timing tapping tasks). Moreover, all non-speech tasks had speech tasks as counterparts. The aim is now to analyse these tests from the developmental point of view to determine the extent to which different skills develop in parallel.(4) In the extension phase we aim to compare musicians and non-musicians to improve understanding of the extent and nature of adaptive behaviour. This emerges from the results above on the malleability of motor representations. This aspect has been little discussed hitherto for speech, but has received some attention in perturbation experiments with music-based stimuli. To achieve our overall aims we thus see explicit comparison of musicians and non-musicians for both speech and music-based stimuli as a particularly promising avenue.