Skilled readers use multiple heuristics to guide their eye movements during reading. One possible cue that readers may rely on is the way in which information about word identity is typically spread across words. In many (but not all) languages, words are, on average, more informative on the left, predicting that readers should have a preference for left-of-center fixation when targeting words. Any such effect will, however, be modulated by important perceptual constraints and may be masked by various confounding factors. In three experiments with artificially constructed lexicons, we provide causal evidence that the way in which a language distributes information affects how readers land on words. We further support our analyses with a Bayesian cognitive model of visual word recognition that predicts where readers ought to fixate in order to minimize uncertainty about word identity. Taken together, our findings suggest that global properties of the lexicon may play a role in isolated word targeting, and may therefore make a contribution to eye movement behavior in more natural reading settings.

中文翻译：

---

读者将目标词定位在他们希望最大限度地减少不确定性的地方

熟练的读者在阅读过程中会使用多种启发法来引导他们的眼球运动。读者可能依赖的一种可能线索是有关单词同一性的信息通常在单词之间传播的方式。在许多（但不是所有）语言中，平均而言，左侧的单词信息量更大，这表明读者在定位单词时应该更倾向于偏左的注视。然而，任何此类效应都将受到重要的感知约束的调节，并且可能被各种混杂因素所掩盖。在三个人工构建词典的实验中，我们提供了因果证据，证明语言分发信息的方式会影响读者对单词的理解。我们进一步使用视觉单词识别的贝叶斯认知模型来支持我们的分析，该模型可以预测读者应该关注的位置，以最大程度地减少单词身份的不确定性。总而言之，我们的研究结果表明，词典的全局属性可能在孤立词定位中发挥作用，因此可能对更自然的阅读环境中的眼动行为做出贡献。