Faísca Luís, Reis Alexandra, Araújo Susana
Universidade do Algarve, Faro, Portugal.
Universidade de Lisboa, Lisbon, Portugal.
J Cogn Neurosci. 2025 Jul 1;37(7):1202-1215. doi: 10.1162/jocn_a_02293.
Behavioral research has shown that inconsistency in spelling-to-sound mappings slows visual word recognition and word naming. However, the time course of this effect remains underexplored. To address this, we asked skilled adult readers to perform a 1-back repetition detection task that did not explicitly involve phonological coding, in which we manipulated lexicality (high-frequency words vs. pseudowords) and sublexical spelling-to-sound consistency (treated as a dichotomous-consistent vs. inconsistent-and continuous dimension), while recording their brain electrical activity. The ERP results showed that the adult brain distinguishes between real and nonexistent words within 119-172 msec after stimulus onset (early N170), likely reflecting initial, rapid access to a primitive visuo-orthographic representation. The consistency of spelling-to-sound mappings exerted an effect shortly after the lexicality effect (172-270 msec; late N170), which percolated to the 353- to 475-msec range but only for real words. This suggests that, in expert readers, orthographic and phonological codes become available automatically and nearly simultaneously within the first 200 msec of the recognition process. We conclude that the early coupling of orthographic and phonological information plays a core role in visual word recognition by mature readers. Our findings support "quasiparallel" processing rather than strict cognitive seriality in early visual word recognition.
行为研究表明，拼写与发音映射的不一致会减缓视觉单词识别和单词命名的速度。然而，这种影响的时间进程仍未得到充分探索。为了解决这个问题，我们要求熟练的成年读者执行一项1-back重复检测任务，该任务没有明确涉及语音编码，在这个任务中，我们操纵了词汇性（高频词与假词）和次词汇拼写与发音的一致性（视为二分法——一致与不一致——以及连续维度），同时记录他们的脑电活动。事件相关电位（ERP）结果表明，成年大脑在刺激开始后119 - 172毫秒内就能区分真实单词和不存在的单词（早期N170），这可能反映了对原始视觉正字法表征的初始快速访问。拼写与发音映射的一致性在词汇性效应出现后不久（172 - 270毫秒；晚期N170）产生影响，这种影响渗透到353 - 475毫秒的范围，但仅针对真实单词。这表明，在熟练读者中，正字法和语音编码在识别过程的前200毫秒内自动且几乎同时可用。我们得出结论，正字法和语音信息的早期耦合在成熟读者的视觉单词识别中起核心作用。我们的研究结果支持早期视觉单词识别中的“准并行”处理而非严格的认知序列性。
