This study deals with the discourse functions of the response particles in a split interrogative (hereinafter, SI), thereby contributing to effectively identifying the characteristics of the tag part in SI. I provide a piece of evidence, through the corpus data, that the tag part must be a non-wh-question and essentially functions as a biased polar question. The data tell us that the interlocutors can choose whether to use the SI or a regular question. Through the data that the speakers select a regular question, this paper tries to provide the evidence. In the process of analyzing their functions in SI, this study demonstrates that these particles must have a referential scope in picking up their antecedents in the previous question: a wide or narrow scope in referring to their antecedents. This characteristic leads us to identify the biased property of the tag question, therefore arguing that this tag must be a biased polar question. In the course of these works, we figured out that the addressees apparently take three strategies to respond to the split interrogative: to respond only the wh-question, the tag part, or both.

1. Introduction
2. Previous Analyses
    2.1. Split Interrogatives
    2.2. Response Particles
3. Corpus Studies
    3.1. Result 1
    3.2. Result 2
    3.3. Result 3
4. Discussion
    4.1. How to Interpret the Tag
    4.2. Explicit Antecedents
    4.3. Reporting Verb
    4.4. Grammatical Functions of Response Particles
    4.5. Biased Polar Question
5. Conclusion
References

• Mija Kim(Kyung Hee University.)