Prior research has suggested that the identification and encoding of letter positions within letter strings might be influenced by orthography. Letters in transparent languages (e.g., Greek) with regular grapheme-to-phoneme correspondences are processed sequentially, whereas letters in deep languages (e.g., English) are processed in parallel. In three experiments, we used a visual search paradigm to test this hypothesis on Russian—a relatively transparent language. In Experiment 1, we measured the identification speed of Cyrillic letters at each position in the five-element real words or pronounceable pseudowords. In Experiment 2, the performance was compared to random letter strings, and in Experiment 3, to non-linguistic symbol strings. Our results reveal a search pattern similar to English, excluding strictly serial letter computation, which is inconsistent with the orthography hypothesis. Moreover, we showed that the lexical status and the nature of the string (linguistic/non-linguistic) affect response times for Russian and therefore must be accounted for in models of visual word recognition.

先前的研究表明，字母串中字母位置的识别和编码可能会受到拼写法的影响。具有规则字素到音素对应关系的透明语言（例如希腊语）中的字母按顺序处理，而深层语言（例如英语）中的字母则并行处理。在三个实验中，我们使用视觉搜索范式在俄语（一种相对透明的语言）上测试了这一假设。在实验1中，我们测量了五元素真词或可发音伪词中每个位置的西里尔字母的识别速度。在实验 2 中，将性能与随机字母字符串进行了比较，在实验 3 中，将性能与非语言符号字符串进行了比较。我们的结果揭示了一种类似于英语的搜索模式，不包括严格的序列字母计算，这与拼写假设不一致。此外，我们表明词汇状态和字符串的性质（语言/非语言）会影响俄语的响应时间，因此必须在视觉单词识别模型中加以考虑。