School of Psychology and Counselling, Faculty of Health, Queensland University of Technology, Brisbane, QLD, Australia.
College of Nursing and Health Sciences, Flinders University, Adelaide, SA, Australia.
Q J Exp Psychol (Hove). 2024 May;77(5):943-963. doi: 10.1177/17470218231184940. Epub 2023 Jul 6.
A century of research has provided evidence of limited size sound symbolism in English, that is, certain vowels are non-arbitrarily associated with words denoting small versus large referents (e.g., /i/ as in and /ɑ/ as in ). In the present study, we investigated more extensive statistical regularities between surface form properties of English words and ratings of their semantic size, that is, , and its impact on language and memory processing. Our findings provide the first evidence of significant word form typicality for semantic size. In five empirical studies using behavioural megastudy data sets of performance on written and auditory lexical decision, reading aloud, semantic decision, and recognition memory tasks, we show that form typicality for size is a stronger and more consistent predictor of lexical access during word comprehension and production than semantic size, in addition to playing a significant role in verbal memory. The empirical results demonstrate that statistical information about non-arbitrary form-size mappings is accessed automatically during language and verbal memory processing, unlike semantic size that is largely dependent on task contexts that explicitly require participants to access size knowledge. We discuss how a priori knowledge about non-arbitrary form-meaning associations in the lexicon might be incorporated in models of language processing that implement Bayesian statistical inference.
一个世纪的研究为英语中有限的大小声音象征提供了证据，也就是说，某些元音与表示小与大的指称的单词有非任意的联系（例如，在 中是 /i/，在 中是 /ɑ/）。在本研究中，我们研究了英语单词的表面形式特征与语义大小的评级之间更广泛的统计规律，即 ，以及它对语言和记忆处理的影响。我们的发现为语义大小的显著单词形式典型性提供了第一个证据。在五个使用行为巨量研究数据集的实证研究中，我们在书面和听觉词汇判断、朗读、语义判断和识别记忆任务中的表现，我们表明，大小的形式典型性是词汇理解和产生过程中词汇访问的更强和更一致的预测指标，除了在口头记忆中发挥重要作用。实证结果表明，在语言和口头记忆处理过程中，会自动访问关于非任意形式-大小映射的统计信息，而语义大小则主要取决于明确要求参与者访问大小知识的任务上下文。我们讨论了如何在实施贝叶斯统计推断的语言处理模型中纳入词汇中关于非任意形式-意义关联的先验知识。
