From amusia to dyslexia: the role of feedback

大多数成年人都知道自己唱走调了。有音乐的人不会；他们不知道怎么唱歌。如果音乐家不能分辨走调和走调的区别，他们怎么能学习呢？反馈在学习中的作用是本研究计划的重点。我的长期目标是开发一种认知学习障碍的神经生物学模型。我和我的学员将继续研究学习障碍，如失音症和阅读障碍，因为它们为复杂系统的学习提供了最好的证据来源之一。这种方法的逻辑是一种逆向工程，即行为分解模式与精确的神经成像技术相结合，可以揭示音乐和语言大脑的内部工作。根据我最近的研究，我提出听觉皮层和额叶皮层之间不可靠的反馈回路是先天性失音症学习缺陷的原因[Peretz， 2016，附件]，其中研究最多的形式是音高聋，它影响1%至4%的人口的细颗粒音高加工。节拍性耳聋是最近发现的一种形式，它会影响时间的微调。也有大量证据表明，阅读障碍患者的节拍处理受损，可能代表了从听觉到额叶的自下而上连接受损。我计划在这些学习障碍中结合脑电图（EEG）和在线反馈，以建立纠错的行为效应与障碍的大脑特征之间的因果关系。本研究提出了三个关键问题：1)在没有额叶反馈的情况下，在听觉皮层中捕获的音乐知识是什么？本研究将通过一项新颖的即兴歌唱任务来探讨音乐成人和3- 4岁正常幼儿的调性知识内容，这两组幼儿也是隐性地获得有限反馈的音乐知识。2)纠错能缓解音乐障碍吗？喜欢音乐的成年人将在检测旋律中异常音高的同时获得在线反馈，同时学习如何在键盘上演奏这些旋律。3)反馈微调时间是否适用于节拍性耳聋和阅读障碍？为了回答这个问题，我们首先需要获得新的计时机制的神经标记。为此，节拍失聪和诵读困难的大脑电反应将被记录下来，使用时钟错觉，其中每隔一个节拍就会有一个主观的口音。该研究项目主要基于音乐缺陷，为人类大脑如何有效学习提供了一个测试模型。在最佳时间提供足够的反馈可能是至关重要的，这对大多数（如果不是全部）认知领域的学习具有广泛的影响。

Most adults know when they sing out-of-tune. Amusic individuals do not; they have no idea how they sing. If amusics cannot tell the difference between in-tune and out-of-tune, how could they ever learn? The role of feedback in learning is the focus of this research proposal. My long-term objective is to develop a neurobiologically-informed model of cognitive learning disorders. My trainees and I will continue to study learning disorders, such as amusia and dyslexia, because they provide one of the best sources of evidence regarding the learning of complex systems. The logic of this approach is one of reverse-engineering, whereby the behavioural breakdown patterns coupled with precise neuroimaging techniques can reveal the inner working of the musical and language brain. Based on my recent research, I have proposed that the unreliable feedback loop between the auditory cortex and the frontal cortex is the cause of the learning deficits in congenital amusia [Peretz, 2016, attached], of which the most studied form is pitch deafness that affects fine-grained pitch processing in 1 to 4% of the population. Beat deafness is a more recently discovered form and affects fine-tuning of timing. There is also substantial evidence that beat processing is impaired in dyslexia and likely represents impaired bottom-up connections from auditory to frontal regions. I plan to combine electroencephalography (EEG) and online feedback in these learning disorders, in order to establish a causal relationship between the behavioural effects of error correction and the brain signatures of the disorders. The proposed research examines three key questions: 1)What is the musical knowledge captured in the auditory cortex without frontal feedback? The content of tonal knowledge will be probed through a novel sung improvisation task in amusic adults and in 3-to-4-year-old normal toddlers, who also have acquired musical knowledge implicitly with limited feedback. 2)Can error correction alleviate musical disorders? Amusic adults will receive online feedback while detecting anomalous pitches in melodies and while learning how to play them on a keyboard over several sessions. 3)Can feedback fine-tune timing in beat deafness and in dyslexia? In order to answer this question, we first need to obtain novel neural markers of timekeeping mechanisms. To this aim, beat-deaf and dyslexic electric brain responses will be recorded using the clock illusion, in which a subjective accent is ascribed to every other beat. The research programme, based primarily on musical deficits, provides a test model of how the human brain learns effectively. The provision of adequate feedback at the optimal time may be vital, with wide-ranging implications for learning across most, if not all, cognitive domains.