Specificity and flexibility in early word recognition

识别口语的能力在人类互动中起着至关重要的作用。然而，单词识别是一个非常复杂的过程：同一个单词在不同的语境和说话者之间听起来可能非常不同。例如，当来自多伦多的说话者和来自纽约市的说话者发出单词“dog”中的元音时，元音是完全不同的。 很少有工作已经解决了在何种程度上，年轻的单词学习者可以适应不同的语境和扬声器的发音变化，但它是一个重要的问题。如果一个来自多伦多的孩子听到“dog”的纽约发音，她应该把它理解为熟悉的单词“dog”的错误发音版本，还是一个新词？因为年轻的词汇学习者有一个有据可查的偏见，认为每个新单词都意味着一个新的含义，所以在不同的语境中识别同一个单词的失败可能会减缓词汇的习得。 这项拟议中的研究探讨了影响幼儿是否以及何时能够适应新的单词发音的因素，包括发音本身的细节（与典型发音的差异）和现实世界知识的类型（例如，视觉环境，关于说话者的知识）幼儿可以用来解释不寻常的发音。例如，当场景中有不熟悉的物体时，听到单词“狗”的不寻常发音可能会使幼儿比所有物体都具有已知标签时更有可能将其视为新词。这项工作还研究了学习新口音的抽象层次。学习是特定于特定个体所说的特定单词，还是发生在更抽象的层面上，概括到词典中的其他单词或其他说话者？ 这项研究将帮助我们了解儿童是否以与成人相同的方式识别单词。此外，了解成功的单词识别的基础能力可能会导致早期诊断和治疗儿童理解困难。

The ability to recognize spoken words plays a crucial role in human interactions. Yet, word recognition is an extraordinarily complicated process: the same word can sound very different across contexts and speakers. For example, the vowel in the word "dog" is quite different when produced by a speaker from Toronto and a speaker from New York City. Little work has addressed the extent to which young word learners can adapt to variability in pronunciations across contexts and speakers, but it is an important issue. If a child from Toronto hears the NYC pronunciation of "dog", should she interpret it as a mispronounced version of the familiar word "dog", or as a new word? Because young word learners have a well-documented bias to assume that each new word signals a new meaning, a failure to identify the same word in different contexts has the potential to slow vocabulary acquisition. The proposed research explores the factors that influence whether and when toddlers are able to adapt to new pronunciations of words, including particulars of the pronunciation itself (how it differs from the typical pronunciation) and the types of real-world knowledge (e.g., visual context, knowledge about the speaker) toddlers can use to interpret unusual pronunciations. For example, hearing an unusual pronunciation of the word "dog" when there is an unfamiliar object in the scene may make toddlers more likely to consider it a new word than when all of the objects have known labels. This work also examines the level of abstraction at which learning about a new accent occurs. Is learning specific to particular words spoken by particular individuals, or does it occur at a more abstract level, generalizing to other words in the lexicon, or to other speakers? This research will help us understand whether or not children recognize words in the same way as adults. Further, understanding the abilities that underlie successful word recognition may lead to earlier diagnosis and treatment of children struggling with comprehension difficulties.