Abstract: Chinese script is normally printed as a continuous string of characters, where there are only small spaces between characters, and these are the same size regardless of whether they appear between or within words. Other than punctuation marks between characters, there is no other visual word boundary information to help Chinese readers segment and identify the words within sentence contexts. The lack of visual word segmentation cues, such as word spacing, leads to an intriguing but fundamental theoretical question in relation to Chinese reading: How do Chinese readers segment character strings to identify words? There has been an amount of research to examine the role of word spacing on Chinese word identification during reading. There has, however, been less research focusing on the issue that how word spacing affected the cognitive processes of word acquisition when novel words are encountered during reading. In the present study, we endeavor to explore the role of word spacing on the acquisition of Chinese novel words, especially to examine whether the word spacing effect was modulated by the characteristics of each morpheme which was comprised of the novel word. Two types of two-character pseudowords were constructed as novel vocabularies: the first type of novel words was made up of both familiar initial and final constituents, which was considered as easy novel word; whilst the second type was made up with unfamiliar initial and final constituents, which was considered as difficult novel word. Each pseudoword was embedded into two sets of sentences for the learning and the test phase separately in order to describe each novel word into a special semantic category, such as fruit, animal and so on. Each sentence frame provided a chance for readers to understand the novel word’s meaning. In the learning phase, half participants read a set of sentences in word-spaced format, and half in unspaced format. In the test phase, all participants read another set of sentences in unspaced format. In order to check that adult readers were able to learn the pseudowords from the sentences that we created, we included a semantic category judgment for each target word in the test phase of the experiment. The results showed that, readers made shorter reading times on pseudowords when inserting spaces into words than that in traditional, unspaced format in the learning phase, indicating that the facilitatory role of word spacing was observed in the leaning phase. However, the benefit of word spacing in the learning phase did not maintain to the subsequent test phase, here, all readers read in the unspaced format. Furthermore, greater benefit of word spacing was found for pseudowords which were comprised of two familiar constituents than when pseudowords were made up of two unfamiliar constituents, indicating that the benefit of word spacing was modulated by the different familiarity of constituents composed of pseudowords when acquiring them within sentential contexts. On the bases of the main findings in the present experiment, we argue that it would deepen our knowledge of word segmentation mechanism in Chinese reading and would provide important teaching suggestions for Chinese children and readers of Chinese as second language.

Key words: word learning, morpheme familiarity, word spacing

摘要： 为考察词素熟悉性是否会影响视觉词切分线索在新词学习中的作用，本研究中新词由两类假词构成：第一类假词由两个高频字（高熟悉性词素）组成，第二类假词由两个低频字（低熟悉性词素）组成。实验采用学习-测试范式，将新词镶嵌在句子中供大学生被试阅读。结果发现，相比由低熟悉性词素构成的新词，词间空格在由高熟悉词素构成的新词中起到的促进作用更大。表明在汉语阅读过程中，词素熟悉性可能作为一种线索参与词切分。

关键词: 新词学习, 词素熟悉性, 词间空格