关键词: 阅读, 快速自动命名, 视觉注意广度, “级联”加工

Abstract:

Rapid Automatized Naming (RAN) is an important predictor of reading. However, why and how RAN predicts reading remains controversial. Some researchers believe that the essence of reading is to recognize individual words in text in rapid succession, Individual word recognition rate can predict the fluency of sequential text reading through phonological awareness, orthography and processing speed. Therefore, individual word processing efficiency dominated the effect of RAN on reading. Other studies have found that Single word reading efficiency cannot fully explain individual differences in the process of serial reading. readers process multiple words during fluent reading in a "cascade" of processing, which is why RAN predicts reading. Therefore, There is no consensus on the mechanism by which RAN predicts reading. This paper analyzes existing studies and finds that they have not focused on the moderating effect of Visual Attention Span (VAS) on lexical processing, which may lead to inconsistent conclusions on the underlying mechanism of RAN in predicting reading. Readers with low VAS level recognized a limited number of orthographic units and read individual words one by one in discrete form during reading, so the relationship between RAN and reading was dominated by individual word processing efficiency; Readers with high VAS regard words as a whole unit and carry out parallel processing among multiple words in sequence. This "cascade" processing dominates the relationship between RAN and reading. However, due to the stable correlation between RAN and reading, many researchers extend the efficiency of single word processing to continuous text reading, believing that reading is the rapid and automatic reading of single words in succession. The researchers did not notice that in early reading, the reader's discrete RAN was closely correlated with the serial RAN, and that in serial reading, the reader was also engaged in rapid single-word reading. In adulthood, the reader's discrete RAN is independent of the serial RAN, and discrete and serial reading are not the same. Although some researchers noticed this later, but they still did not further pay attention to the fact that the lexical processing modes of readers at different stages correspond to different levels of VAS. With the potential changes of readers' VAS, their continuous reading modes are also changing. In addition, the selected subjects in previous studies were screened for basic reading ability, They all have basic reading skills. RAN materials were all high-frequency words or words with short characters, The VAS of the subjects was sufficient to support the “sight word reading” of these high-frequency short words. Therefore, although the subjects had different levels of VAS, they were similar in reading performance. By analyzing existing studies, we found that the level of readers' VAS seemed to explain the discrepancy in the conclusions on the relationship between RAN and reading. Future studies should focus on individual differences in VAS during reading, organically integrate different theories, and further explore the relationship between RAN and reading as well as the influence of VAS at different levels on it. To improve the reasons for RAN prediction of reading and related theories, Enriching the theory of nonverbal interventions for developmental dyslexia.

Key words: reading, rapid automatized naming, visual attention span, cascade processing