An Experimental Study Of The Cognitive Processing Model Of Chinese Scalar Implicature

This study is intended to find a most rational account for one of the three debated cognitive processing models in scalar implicature (SI):the context driven account, the default account and the standardization account. Three experiments were conducted and their statistic results were applied to figure out the psychological reality of these three processing models.I first reduplicated Experiment 1 by Katsos (2011) to assess whether the processing of classic scale was prior to the poset scale among both children and adult groups.Ⅰ presumed that in the context driven account, the processing rate of the classic scale and the poset scale would have no difference in the child and adult group. In the default account, the processing rate of the classic scale would be higher in the child and adult groups than the poset scale. In the standardization account, the processing rate of both the classic scale and the poset scale would be higher the adult group than in the child group. The participants were asked to judge the answer of a character according to the scenario shown on the computer.Ⅰrandomly chose 16 children under 7 years old as the participants of the child group, who were pupils from the Little Swan Kindergarten in Lanzhou, China. For the group of adult participants, Ⅰ randomly chose 16 students who are 20 to 25 years old at the Lanzhou University. All the participants had no knowledge of linguistics and related fields. The adult group finished the experiment in the Experimental Pragmatic Lab at Lanzhou University, while the child group finished the experiment in a quiet room at the kindergarten. (The experimental sites were the same in the Experiment 2 and the Experiment 3.) The results were as follows:(1) The processing rate of the classic scale was lower than that of the poset scale in the child group, while in the adult group, the processing rate of the classic scale was much higher than that of the poset scale.(2) The processing rate of both the classic scale and the poset scale was much higher in the adult group than that in the child group. (3) The adults used more direct response strategies to the classic scale, while the children used all direct response to the classic scale and the poset scale. The analysis was as follows:(1) The child group did not process scalar implicature under the influence of the underinformativeness caused by the poset scale. (2) The high processing rate in the adult group implied the possibility of standardization. (3) The response strategies in both the child and the adult group further testified the first two analysis above. No result showed a clear evidence to the three theories. Further evidence were still needed to testify the three processing models.Experiment 2 focused on contrasting the processing differences of the classic scale without context between child and adult groups. I presumed that in the context driven account, the scalar implicature would not be processed in the situation without any context in both child and adult groups. In the default account, the processing rate and reaction time of child group would have no difference with the adult group. In the standardization account, the adult would have a higher processing rate and shorter reaction time than the child group. The materials, picture with different objects and a basket on it, were shown by the software of DMDX on the computer screen. The participants were asked to speak out the number of objects they decide to put in the basket. There were 18 pupils aged between 5 to 7 and 20 adult participants aged between 22 to 25 years old. All of them were from the same participant groups mentioned in the Experiment 1. The results were as follows:(1) There was no distinctive difference in the SI processing rate of the child and the adult group. (2)The reaction time of the adult group was shorter than the child group. The analysis was as follows. (1) The retained process of SI in the situation without context strongly denied the context driven account. (2) The shorter reaction time in the adult group might support the standardization account. However, some adults tended to be confused when facing testing items totally without context constraints, which might influence their reaction in the experiment. Thus it would still be early to reach the final result.Experiment 3 examined the processing rate of SI when the upper bound context (UBC) and the lower bound context (LBC) were added. This experiment especially focused on distinguishing between the default account and the standardization account, for a solid evidence to deny the context driven account has been revealed in the Experiment 2. In the default account, the processing of scalar implicature would be abstained in LBC with a longer reaction time due to the cancellation of SI. In the standardization account, SI would still be processed even in LBC with a similar reaction time to that in UBC. The materials, presented through the software of DMDX on the screen, showed a scenario in which the participant went to shopping and decided the number of objects put in the basket according to the specific vocal demand in each testing item.20 adult participants aged between 22 to 25 years old took part in the experiment. All of them were from the same participant groups mentioned in the Experiment 1 and Experiment 2. The results were as follows. (1) The processing rate of SI in UBC was higher than that in LBC. (2) There was no significant difference in the aspect of reaction time in both UBC and LBC. The analysis was as follows. (1)The still processing of scalar implicature in LBC strongly denied the default account and was in accordance with the standardization account. (2) The similar reaction time to the scalar implicature processing in both UBC and LBC further showed a closer relation to the standardization account. Thus, the final result testified the possibility of the standardization account in the Chinese SI processing.