The Effects Of Linear Distance And Structural Distance On The Processing Of Relative Clauses

The relative clause(RC)is a typical long distance dependency,also known as filler-gap dependency.The head noun modified by the relative clause is considered to be the filler moved out of the relative clause,which will leave a gap in its original position.Relative clauses can be divided into subject relative clauses and object relative clauses.The existing research on the relative clause mainly focuses on the processing asymmetry between the subject relative clause(SRC)and the object relative clause(ORC).In the study on the processing of the relative clause in Indo-European languages,it is unanimously found that the subject relative clause is easier to process.The distance between the filler and the gap is an important perspective in explaining the processing preference of the subject relative clause.Distance is defined in two ways: linear distance and structural distance.In the RC-head-initial Indo-European languages,the linear distance and the structural distance are synchronous,so both of them are decisive factors in the processing of relative clauses,that is to say,both of them can explain the processing preference of the subject relative clauses in these languages.However,in the RC-head-final Mandarin,which also has an SVO,the linear distance and structural distance in relative clauses are asynchronous,and their influence on the processing of Mandarin relative clause remains to be investigated.Which of these two factors plays a decisive role in Mandarin relative clauses? Or do they work together? If they do so,how do they interact with each other? These are the questions that researchers need to answer.Hsu &Chen(2013)investigated two hypothesis in filler-gap dependencies in Mandarin SRCs through eye movement experiments.It is found that the main effect of structural distance is significant,but the main effect of linear distance is not significant.So far,no research has ever studied the influence of these two distances in the processing of Mandarin object relative clauses.Therefore,this study intends to give us a more comprehensive understanding of the role of these two kinds of distance in the processing of relative clauses.This study aims to answer three questions:(1)Does the linear distance between the filler and the gap affect the processing of Chinese object relative clauses?(2)Does the structural distance between the filler and the gap affect the processing of Chinese object relative clauses?(3)Is there any interaction between linear distance and structural distance in the processing ofChinese object relative clauses? If so,how do they interact?Two experiments are designed in this study.Experiment 1 used the experimental sentences with background information(as shown in Appendix 1),and Experiment 2 used the experimental sentences without background information(as shown in Appendix 2).Both experiments were designed by two factors: linear distance(long / short)x structural distance(long / short).The main indexes were first fixation time(FF),first reading time(FP),gaze duration(GD)and total reading time(TT).Please see Appendix 1 and Appendix 2 for examples and divisions of experimental materials.The experimental results showed a main effect of linear distance,a main effect of structural distance and their interaction effect were found in the processing of Mandarin object relative clauses in both Experiment 1 and Experiment 2.In Experiment 1,in the processing of R6(DE),the total reading time showed a main effect of linear distance,a main effect of structural distance and an interaction effect between linear distance and structural distance.In the processing of R7(the head noun of RC),gaze duration showed a main effect of linear distance,and the total reading time showed a main effect of structural distance and an interaction effect.The interaction showed that when the linear distance is smaller,the structural distance effect is more significant.It is worth noting that the linear distance effect in R7 is completely opposite to the expectation: when the linear distance increases,the processing of R7 decreases.In the process of processing R4(the subject of RC),the total reading time showed a main effect of structural distance and the interaction between linear distance and structural distance.In the processing of R5(the verb of RC)and R11-13,the total reading time showed a main effect of structural distance.The results of all the regions except for R7 supported linear distance hypothesis;the results of all the regions supported structural distance hypothesis,and at R4,R6 and R7 the interaction effect was found.In this experiment,it was expected that the longer the linear distance is,the longer the processing time will be,while the data in R7 showed the opposite result.We speculate that the place adverbial in the background part of the sentence with shorter linear distance cause the participants to consume more working memory.In order to testify this assumption,this result,Experiment 2,where the background of the experimental sentence is deleted,was designed.In Experiment 2,in the processing of R6(DE),the total reading time showed a main effect of linear distance;the first fixation time,the first pass time,the gaze duration and thetotal reading time all showed a main effect of structural distance.The interaction between linear distance and structural distance was not significant.In the processing of R7(the head noun of RC),the gaze duration showed a main effect of linear distance,while the gaze duration and the total reading time showed a main effect of structural distance,and the interaction between linear and structural distance was not significant.In the processing of the R4(the subject of RC),the total reading time showed a main effect of linear distance and a main effect of structural distance at the same time.In the processing of R5(the verb of RC),the total reading time showed a main effect of linear distance,a main effect of structural distance and an interaction effect.In the processing of R11-13(adjective phrase),the total reading time showed a main effect of structural distance.The results of Experiment 2 showed that the linear distance effect in R7 was the same as that in Experiment 1: the longer the linear distance was,the shorter the processing time would be.The comparison between Experiment1 and Experiment 2 showed that the results of R7 had nothing to do with place adverbials.It is speculated that the hinting function of adjective phrases accelerates the processing of NP.In Experiment 2,the results of all the regions except for R7 supported the linear distance hypothesis;the results of all the regions supported the structural distance hypothesis.The interaction between linear distance and structural distance was found in R5.The background sentence before the object relative clause has an effect on the processing of some of the interest areas.