Orthographic and Semantic Representations: Consolidation and Role in Visual Word Recognition

这项研究的目的是加深我们对阅读认知基础的理解；这项研究代表着对单词同一性的特定组成部分在阅读中的作用的研究的继续。虽然单词同一性或心理表征的各方面通常被认为是大量阅读理论中的主要因果作用，但对这些原则的直接实证测试令人惊讶地有限。 当一个单词存储在内存中时，编码的信息涉及它的发音方式(音韵学)、拼写方式(正字法)以及所表示的含义/概念(语义)。大量阅读理论将成功的单词阅读解释为依赖于这种完整和完整的表征(例如，词汇质量假说：Perfetti，2007；三角形模型：Plaut et al.，1996)。此外，熟练的阅读并不总是与同样熟练的拼写联系在一起；许多阅读能力强的人在拼写方面遇到了困难。阅读和拼写之间的分离已经参考单词如何存储在存储器中来解释；当表征不完整时，它们的指定程度可能足以阅读(在视觉单词存在的情况下)，但不足以拼写(在单词必须生成的情况下)。这一论点的一个有趣的延伸是，在个体内部，一些单词的表征将准确地存储在内存中(高质量)，而其他单词的表征将是不完整或不准确的。从词汇通达的角度来看，由于内部表征和视觉刺激之间的不匹配，不完全表征可能需要更长的时间才能激活。因此，尽管不完整的陈述对一些阅读任务来说可能足够了，但对个人来说仍然应该是有代价的。我们最近通过NSERC资助的一系列研究直接测试了这一论点，并发现了令人信服的证据，尤其是参考拼写表示法支持这一解释(Martin-Chang等人，2014；Ouellette等人，2017；Martin-Chang等人，正在审查中)。 然而，到目前为止，实证研究的重点主要局限于正字法表征及其在高技能成人读者中的作用。这里提出的研究计划扩展了这一研究路线，包括对拼写和语义表征的评估和直接操作，同时也考虑了不同熟练程度和年龄的读者。在精心控制的实验设计中，通过训练范例操纵单词同一性的语义和正字法方面，并直接测量对阅读速度的任何结果影响。先进的混合效果建模计划将训练效果与参与者和项目(即单词)效果分开。这项研究的结果将对当前认知和阅读的理论模型做出重大贡献。

The goal of this research program is to add to our understanding of the cognitive basis of reading; this research represents a continuation of study into the role of specific components of word identity in reading. While facets of word identity, or mental representations, are commonly attributed a leading causal role in much reading theory, direct empirical tests of such tenets are surprisingly limited. When a word is stored in memory, information is encoded pertaining to how it sounds (phonology), how it is spelled (orthography), and the meaning/ concepts represented (semantics). Much reading theory has explained successful word reading as relying upon such intact and integrated representations (e.g., lexical quality hypothesis: Perfetti, 2007; triangle model: Plaut et al., 1996). Further, proficient reading is not always associated with equally proficient spelling; many strong readers struggle with spelling. The dissociation between reading and spelling has been explained with reference to how words are stored in memory; when the representations are incomplete, their degree of specification may suffice for reading (where the visual word is present) but not spelling (where the word must be generated). An intriguing extension of this argument, is that within the individual, some words will have representations that are accurately stored in memory (high quality) while other words will have representations that are incomplete or inaccurate. Viewed through the lens of lexical access, incomplete representations should take longer to activate potentially because of the mismatch between the internal representation and the visual stimuli. Thus, while incomplete representations may suffice for some reading tasks, there should still be a cost to the individual. We recently directly tested this contention through a series of NSERC funded studies, and have found compelling evidence in favor of this explanation with reference to orthographic representations in particular (Martin-Chang et al., 2014; Ouellette et al., 2017; Martin-Chang et al., under review). To date, however, the focus of empirical study has been primarily restricted to orthographic representations and their role in highly skilled adult readers. The program of research proposed here extends this line of inquiry to include evaluation and direct manipulation of both orthographic and semantic representation, while also considering readers of varying proficiency and age. Semantic and orthographic facets of word identity are manipulated through training paradigms within a carefully controlled experimental design, and any consequential effects on reading speed directly measured. Advanced mixed-effect modeling is planned to isolate training effects separate from participant and item (i.e., word) effects. The results of such inquiry stand to contribute significantly to current theoretical models of cognition and reading.