Much work has been conducted that reports on the strong relationships between speech production and reading behaviour. However, disentangling the relative contribution of the various underlying speech mechanisms (i.e., articulation, somatosensory and/or auditory) that contribute to this relationship remains unknown. Oral stereognosis refers to the ability to discriminate shapes using our tongue/mouth, in the absence of seeing the shape beforehand, and is a measure of the somatosensory sub-system of the speech mechanism. In this work, we aimed to address the extent to which oral stereognosis predicted visual word recognition. Methods: Healthy participants ranging from 18 to 51 years old (N = 60; Mage = 22.7) completed the Florida Oral Recognition Measurement (FORM) task and two visual word recognition tasks. The FORM required participants to identify carrots that were cut into the following letters’ shapes: E, R, T, S, L, N, O, V. Carrot letters were randomly presented to participants. With respect to the visual word recognition tasks, participants were required to press a button: 1) if the letter string spelled a word or 2) if the letter string sounded like a word. Response time and accuracy were measured for all tasks. Results: Participants were divided into high (accuracy >50%) vs. low (accuracy ≦ 50%) oral stereognosis recognition groups. For the high FORM group, oral stereognosis recognition predicted visual word recognition in the spell task (Adj R² = .114). For the low FORM group, oral stereognosis recognition predicted visual word recognition in the sound task (Adj R² = .307). Together, these findings provide evidence that oral stereognosis is predictive of word recognition performance, and that performance on the FORM differentially predicts word recognition tasks (spell vs. sound). Overall, these findings inform current print-to-speech frameworks by identifying a relationship between sensorimotor information, specifically, oral stereognosis ability, and reading performance.

已经进行了许多工作，以报告言语产生与阅读行为之间的紧密关系。然而，弄清促成这种关系的各种潜在语音机制（即，发音，体感和/或听觉）的相对贡献仍然未知。口腔立体诊断是指在没有事先看到形状的情况下，使用我们的舌头/嘴部辨别形状的能力，是对言语机制体感子系统的一种度量。在这项工作中，我们旨在解决口腔立体诊断预测视觉单词识别的程度。方法：健康参与者的年龄范围为18至51岁（N  = 60；M年龄= 22.7）完成了佛罗里达口腔识别测量（FORM）任务和两项视觉单词识别任务。表格要求参与者识别切成以下字母形状的胡萝卜：E，R，T，S，L，N，O，V。将胡萝卜字母随机呈现给参与者。关于视觉单词识别任务，要求参与者按下按钮：1）如果字母字符串拼写了单词或2）如果字母字符串听起来像单词。测量了所有任务的响应时间和准确性。结果：参与者分为高（准确度> 50％）和低（准确度≤50％）口腔立体诊断识别组。对于高FORM组，口腔立体诊断识别可预测咒语中的视觉单词识别任务（Adj R ²  = .114）。对于低FORM组，口头立体诊断识别可预测声音任务中的视觉单词识别（Adj R ²  = .307）。在一起，这些发现提供了证据，即口腔立体诊断可预测单词识别性能，并且FORM上的性能可差异性地预测单词识别任务（拼写与声音）。总体而言，这些发现通过识别感觉运动信息，特别是口腔立体诊断能力和阅读性能之间的关系，为当前的“打印到语音转换”框架提供了信息。