The N400Effect Of Frequency And Font In Chinese Two-character Compound Words

Recently, there are a lot of cognitive theories to explain N400mechanism, such as lexical integration theory proposed that the difficulty caused by different semantic integration change the N400amplitude, namely the N400amplitude is small when semantic integration processing is easier. In the lexicion decision task, N400priming effect usually interpreted by pre-lexical processing theory or lexical processing theory. Lexical processing theory proposed the change in N400amplitude reflects the process of orthography or phonetic analysis which is not involved in semantic processing stage. Thus, the theories of N400effect have not reached a consensus. Now the problem is that, the effects of surface form still provide limited evidence for understanding reading processes. When a single factor is manipulated in experiments, the results can often be explained in a variety of different ways. In other words, many competing explanations-and several different (computational) models-may be equally accurate in explaining a main effect. However, when two factors are jointly manipulated, the data pattern is much more complex and it might help to falsify some of the various accounts. So some researchers using the effect of semantic priming and other variables to explore the cognitive processing mechanism of N400is worthy of our reference.In this study, we adopted the interaction between stimulus quality (such as font) and frequency to explore the processing mechanism of N400, the two factors were related to semantic.The activation model (IA model) can be better to explain these interactions. The IA model proposed the concept of a baseline level of activation, and assumed word frequency and context (including semantic priming) can improve the baseline word node (resting) activation level. It is due to the different levels of baseline activation in different word frequency and context, N400amplitude will be different, which is the reason of word frequency effect and context effect. In addition, the model predictes any factors impact on context, such as stimulus quality will also affect the word frequency, because the frequency and context affect the resting activation levels of the same processing stage. In the final analysis is that frequency and stimulus quality affect the same components of processing.On the contrary, multistage activation model, the processing of printed words begin with activation of that word’s representation in the orthographic input lexicion system (OIL). Then the OIL activates the semantic system. Stimulus Quality is assumed to influence only OIL and to have its effect by slowing the rate of activation. Word frequency influences processing in the pathways that map representations in the orthographic input lexicon onto representations in the semantic system and other subsystems, but it does not affect the word detector itself (not directly affect the semantic system itself). Therefore, multiple stages of activation model predicted that stimulate quality has an additive effect with frequency. However, which model can better explain the cognitive processing mechanism of N400? In this research, by simultaneously manipulating stimulus quality (such as font) and frequency to explore the N400cognitive processing mechanisms. The study was consisted of two experiments, one experiment used low-frequency words material, the other used high-frequency words as experimental material to investigate the differences effect of stimulus in the different frequency.In Experiment1, the experimental variables are word status (two-character compound words, two-character compound pseudo-words) and font (Song Ti, XingKai Ti). By making a lexicion decision task to record the event-related potentials of scalp. In the300-500ms time window, Results showed that whatever word status is, XKT induced more negative N400amplitude than ST. The study verified the unique N400effect of font exist in Chinese and enriched the theory of lexical analysis. The interaction between font and word status was found on N400amplitude. A further statistic analysis showed the font effect under words lever was larger than pseudo-words. The experimental results indicated font could improve the baseline word node (resting) activation level and provided evidence for the interactive activation modelExperiment1replicated the font N400effects of low frequency words, in order to further explore whether the font effects exist in high frequency words, and how word frequency influnce lexical processing of font. In the second experiment, we used the same lexical decision task, the results found in the300-500ms time window, pseudo-words elicited a more negative N400amplitude than words, XKT produced a more negative N400amplitude than ST. Combing the two experiments data between subjects, there is a three-way interaction among word status, font and frequency on N400amplitude. Further analysis found that, in the word condition, stimulus quality (such as font) and frequency existed an interaction. According to the prediction of post-lexical processing theory, stimulus quality, frequency and other factors which are not involved in semantic processing could not affect N400amplitude, thus denied post-lexical processing theory. In this paper, the results of interaction between stimulus quality (such as font) and frequency support the expectations of interactive activation model, which is not anticipated by multistage activation model.Through the research of the two experiments, we can draw conclusions as follows: First, this study explored the font effect of Chinese two-character compound words in visual word recognition and found different types of font influnce the amplitude of N400. So the inherent characteristic of word (such as orthography) plays an important role in semantic access. The results of this study enriched the lexical processing theory while it againstded simply regarded the change of N400amplitude as semantic integration processing.We put forward to use interactive activation theory to explain the amplitude of N400with the lexical features change, that is font affect the baseline word node (resting) activation level.Second, in lexicion decision task, throughing the processing mechanism of stimulus quality (such as font) and frequency, results showed that in the word condition, there is an interaction between stimulus quality and frequency. The results support interactive activation model and deny the multistage activation model. According to Sternberg (1969) additive factors logic, we believe that frequency and font affect at least onre commom stage of information processing in the Chinese language.