Category learning with(out) language: congenital deafness as a critical test for the role of language input in early category learning

当出生在听力正常的家庭时，重度失聪的婴儿几乎无法接触到他们家中使用的语言。人工耳蜗植入是目前最常见的干预途径，然而，尽管早在9个月大的时候就可以进行手术，植入物的声学效果也有所改善，但语言发育的迟缓一直持续到童年。重要的是，在植入之前接触手语对后来的口语和一般认知发展都有好处，但我们对为什么会这样还不太了解。我们建议研究一种新的假设，即语言输入（口头或手语），以照顾者给婴儿遇到的物体贴上标签的形式，帮助婴儿将世界解析为猫或汽车等类别。这种分类能力改变了学习的游戏规则，包括词汇量的增长——将一个新物体识别为一辆汽车，这是一个我们熟悉的类别，意味着我们可以立即推断出这个物体被称为什么，而且它移动得很快。我们提出了一个创新的研究计划，将提供早期语言和概念发展之间的相互作用的新认识。1)我们将打破传统的基于实验室的类别学习范式，即低估语言输入的重要性，在更自然的环境中研究学习。与实验室研究不同的是，在日常生活中，婴儿通过在整洁的背景下一只接一只地看到各种各样的猫来学习猫的类别，而在遇到不同的猫之间可能会经过几分钟、几小时或几天；记忆衰退和干扰将使婴儿很难看到这一类成员之间的共同之处。我们将在诺维奇城堡博物馆（Norwich Castle museum）带领婴儿走上一条真实的或虚拟的学习之路（Learning Trail），我们会问，被告知物体的名称是否能起到记忆线索的作用，通过减少记忆衰退来帮助分类学习。2)我们将开发新的方法来量化儿童在日常生活中获得的类别知识。虽然测量词汇增长的方法多种多样，但发展科学还没有办法测量儿童概念知识的增长。我们将使用脑电图和眼动追踪来绘制听力婴儿两岁前类别知识的增长情况，作为耳聋婴儿研究的对比。3)我们将首次对出生在听力健全家庭的失聪婴儿进行纵向队列研究，通过测量无语言情况下的类别知识，为语言输入在类别学习中可能发挥的作用提供关键测试。使用一套全面的早期概念和交际发展措施，我们将处于一个独特的位置，以了解是什么导致了语言和学习的延迟在这个人群。该提案汇集了一个独特的研究团队，他们在监督纵向队列研究、脑成像和听力和失聪儿童的早期认知发展方面具有互补的专业知识，并参与了临床护理和失聪婴儿教育的专业人员。与听力学家和聋人教师密切合作，将确保研究是由聋儿及其家庭的需求驱动的，更重要的是，研究结果反馈到临床实践中（例如，制定分类知识的措施，作为聋儿临床评估的一部分），以及早期教育中(例如，揭示替代，照顾者可以使用非语言策略来支撑婴儿的概念发展，作为低负担高影响的父母介导干预的一部分)。我们与诺里奇城堡博物馆早期教育团队的密切合作将改变包容性博物馆早期教育的交付方式——例如，通过开发方法让孩子们参与分类，而不是孤立的例子，以确保对历史文物的学习超越博物馆的大门。

When born into hearing families, profoundly deaf infants have little access to the language used in their homes. Cochlear implantation is now the most frequent intervention route, yet, despite surgery being offered as early as 9 months of age and improvements in the acoustics of implants, delays in language persist into childhood. Importantly, exposure to sign language prior to implantation benefits both later oral language and general cognitive development, but we have poor understanding of why this is the case. We propose to investigate a new hypothesis, that language input (verbal or signed), in the form of caregivers labelling the objects infants encounter, helps infants parse the world into categories such as cats or cars. This ability to categorize is a game changer for learning, including for vocabulary growth - recognising a new object as a car, a category we are familiar with, means that we can immediately infer what this object is called and that it moves very fast. We put forward an innovative research programme that will provide new understanding of the interaction between early language and conceptual development.1) We will break with classical lab-based category learning paradigms, which have underestimated the importance of language input, to investigate learning in more naturalistic environment. Unlike lab-based studies, where infants learn the category of cats, for example, by seeing various cats one after the other, on uncluttered backgrounds, in their daily lives, minutes, hours or days may pass between encountering different cats; memory decay and interference will make it difficult for infants to see what is common between the members of this category. Taking infants on a real or virtual Learning Trail in the Norwich Castle museum, we will ask whether being told what objects are called acts as a memory cue, helping category learning by reducing memory decay.2) We will develop new methodologies to quantify the category knowledge that children acquire in their daily lives. While a variety of methods exist to measure vocabulary growth, developmental sciences have as yet no means to measure the growth in children's conceptual knowledge. We will use electroencephalography and eye-tracking to chart the growth in category knowledge in the first two years of life, in hearing infants, which will serve as comparison for the study of deaf infants.3) We will provide, for the first time, the critical test for the role language input may have in category learning by measuring category knowledge in the absence of language, in a longitudinal cohort of deaf infants born into hearing families. Using a comprehensive set of measures of early conceptual and communicative development, we will be in a unique position to understand what causes the delays in language and learning in this population. This proposal brings together a unique team of researchers with complementary expertise in overseeing longitudinal cohort studies, brain imaging and early cognitive development in hearing and deaf children with professionals involved in clinical care and education of deaf infants. Working closely with audiologists and teachers for the deaf will ensure that research is driven by the needs of deaf infants and their families and, more importantly, that research findings feed-back into clinical practice (e.g. developing measures of category knowledge to be used as part of clinical assessments of deaf infants), as well as in early education (e.g. revealing alternative, non-linguistic strategies caregivers could use to scaffold infants' conceptual development as part of low burden high impact parent-mediated interventions). Our close collaboration with Norwich Castle Museum Early Years Team will transform the delivery of inclusive museum early education - e.g. by developing ways to engage children with categories rather than isolated exemplars to ensure that learning about historical artefacts generalises beyond the museum's gates.