Abstract: A particular function of Chinese characters is that their shapes signify their meaning directly; semantic radicals are the main carriers of this function. Many researches have examined the role of components in the recognition of Chinese characters; however, the primary findings pertaining to the cognition of Chinese characters are still controversial. While some researchers have assumed that the characters as whole are the primary units of visual recognition and mental representation, others have assumed that the components of the characters are the prime units. Then, there are some that have assumed that the accessing units of the characters, depending on the frequencies of whole words, position of the components, and the processing tasks, are the prime units. Based on previous researches, this study investigated the role of familiarity of semantic radicals in the recognition of highly familiar Chinese characters. In addition, the effects of various attributes (i.e., orthographic and semantic) and the time period for the relevance activation of semantic radicals in highly familiar characters were explored.
Sixty university students participated in this experiment. In a character decision task, pictophonetic compound targets (composed of a semantic radical and a phonetic component) followed primes that (1) shared the target’s radicals and were semantically related (R+S+), (2) shared the target’s radicals and were not semantically related (R+S–), (3) had no related radicals but were semantically related (R–S+), and (4) had no related radicals and were not semantically related (R–S–). Further, the target radicals varied depending on the frequency of the semantic radicals which valued from 1–7 assessment of college students. The targets followed the primes immediately after the three types of stimulus onset asynchrony (SOA) (43 ms, 72ms, and 243 ms). In addition, the reaction times and error rates of the recognition of targets were measured.
The results indicated that when SOA = 43 ms, the target latencies following R+S– primes were facilitated relative to R–S– controls under the highly familiar semantic radical condition; however, this was not the case under the less familiar condition. When SOA = 72 ms, the target latencies following R+S– primes were slowed relative to R–S– controls under both highly familiar and less familiar semantic radical conditions. Further, when SOA = 243 ms, neither facilitation nor inhibition effects were observed under both conditions. Facilitation effects of R+S+ and R–S+ primes were observed under all three SOA. Target latencies following highly familiar semantic radicals were slower than those following less familiar ones.
Moreover, the results indicated a dynamic interaction between the processing of semantic radicals and the processing of whole characters. The semantic of Chinese characters as a whole primed very early. The priming of the orthographic attributes of the highly familiar semantic radicals began earlier than those of the less familiar radicals. When the SOA was extended, priming occurred on both the semantic attributes of the highly familiar and less semantic radicals. On this SOA, while the priming of whole characters still existed, the semantic priming of the radicals disappeared. This was an indication of the completion of the integration process. All these results indicate that one type of component—the semantic radical—is processed in the course of Chinese character recognition, during which word frequency and radical familiarity are in a dynamic competition with one another.
These results suggest that: (1) There was a dynamic interaction between the activation of semantic radicals and the activation of whole characters in the cognition of Chinese pictophonetic compound characters. (2) The familiarity of the semantic radicals affected the semantic access of Chinese pictophonetic compound characters

Key words: semantic radical, semantic access, orthographic access