认知与发展心理学研究室知识库

词内字序编码是指人脑获取一个词内的字符的相对顺序信息的过程。一直以来，人们将字序编码与词内字符视觉位置表征等同起来。然而，字序编码并不完全等同于字符视觉位置的表征。有研究证据提示，至少对于中文文本，即使是从右向左逆向排版，或者从上到下纵向排版，读者仍然能较高效地识别词汇。这提示即使视觉信息与大脑中己有的对词汇字符视觉位置的表征不符，中文读者仍可以依据阅读方向线索完成词汇的字序编码。
既然中文读者可以仅依据阅读方向线索较高效地进行词汇字序编码，那么字符视觉位置表征对中文词汇字序编码是否还有独立贡献呢?为了检验这个问题，我们采用一个眼动实验(实验一)考察了逆向阅读(文本从右向左排版)中的两种情况:情况一中词序和字序都是从右向左排列;情况二中词序从右向左排列，但组成词内的字从左向右排列。结果发现，情况二比情况一的阅读时间显著更短，注视数量也显著更少。这初步提示了词内字符视觉位置表征在词内字序编码中有独立的贡献。
实验一的情况二下所有词内部字序都与文本阅读方向相违背，规律性较强，读者可能会在任务中形成特定策略。为了进一步检验实验一的结论，在实验二中，我们只对句子中的一个关键词进行相应的操纵，并加入了正向阅读情况作为对照。我们考察了两种阅读方向(从左向右和从右向左)下读者对句子中的一个目标词在两种呈现条件(正常字序和反转字序)下的加工，从而更严格地考察词内字符视觉位置表征和阅读方向对词内字序编码的贡献。我们发现，从右向左阅读的注视时间显著长于从左向右阅读;反转字序词的注视时间长于正常字序词;重要的是，两个因素有显著的交互作用:在从左向右阅读时，反转字序词的注视时间显著长于正常字序词，但在从右向左阅读时，两种呈现条件没有显著差异，并且反转字序(也即词内字序从左向右排列)的条件有阅读时间更短的趋势。这再次提示词内字符视觉位置表征在词内字序编码中有独立于阅读方向线索的贡献。
上述两个实验的结果一致提示，在中文阅读中，词内字符视觉位置表征会得到自动化加工，从而影响词内字序编码。

This research mainly examined the influence of visual position representation of characters within a word on the Chinese character order encoding. The character order encoding is defined as the process of the brain acquiring the relative order information of characters within a word. For a long time, character order encoding has always been equated with the visual position representation of characters within a word. However, character order encoding is not identical to the visual position representation of characters. There is some evidence that, at least for Chinese readers, when Chinese text  is presented in a non-conventional typography, for example, right-to-left format or upto-down typography, Chinese readers could still recognize the word. It suggests that Chinese readers could rely on the cue of reading direction to accomplish the encoding of Chinese character order, even when the visual information does not accord with the existing visual position representation in the brain.
Since Chinese readers could still encode character order when they only depend on reading direction, then it raise a question: whether character visual position representation has an independent contribution to character order encoding? To investigate this question, an eye-movement experiment (Experiment 1) was conducted. In Experiment 1, we manipulated two text presentation conditions where reading was from right to left. In one condition, both word order and character order were presented from right to left. In the other condition, the word order was presented from right to left, but the character order within a word was in a left-to-right format. The results showed that the latter condition had significantly shorter reading time, faster reading speed and less fixations. These results preliminarily tells us that character visual position representation within a word makes an independent contribution to character order encoding.
Given that all the character order within a word in the latter condition of Experiment 1 were always contrary to the text reading direction, which had high regularity, readers might develop some reading strategies during the experiment. Therefore, to further test the conclusion of Experiment1, only the target word in each sentence was manipulated in Experiment 2. In addition, the normal reading direction (left-to-right) were included as controls. We explored how reading directions, including LR (reading from left to right) and RL (reading from right to left), and the presentation type, which includes normal character order and reversed character order, influence reading times on the target words. The results showed that the RL condition had significantly longer fixation times than the LR condition, and the reversed character order presentation condition had significantly longer reading time than the normal character order presentation condition. What’s more, the interaction of the two factors was significant: the reversed character order presentation condition had significantly longer fixation times than the normal character order presentation condition in the direction of reading form left to right; however, when reading from right to left, the two presentation condition had no significant difference, and the reversed character order presentation condition had a tendency to read faster. Again, it suggests that character visual position representation within a word has an independent contribution to the processing of Chinese character order coding.
In conclusion, the two experiments suggest that character visual position representation within a word is automated processing during reading, thus affecting the Chinese character order encoding.