Mongolian (the Khalkha dialect) has three past tense suffixes which are added to the finite verb at the end of a sentence. The traditional works, such as Poppe (1951, 1970), Luvsanjav et al. (1976), Vietze (1978), Ozawa (1978) and Kuribayashi (1992), have explained the meanings of the suffixes., in terms of the assumption that there are striking differences among them. All these explanations certainly account for some aspects of these suffixes but fail to offer a comprehensive view on the whole matter. The purpose of this paper is to find out where the meaning differences come from, by reexamining available data as strictly as possible.
Section 1 introduces the main points at issue in the past studies. In section 2 we observe the environments in which the three suffixes occur, and determine that their basic meaning is past time reference. Section 3 deals with the idiosyncratic meaning of each suffix. It turns out that the differences between the suffixes in meaning are derived from the present time orientation of the reference domain. Section 4 concludes that the three past tense suffixes, operating interactively, construct the past time reference system as a continuum extending from the past time to the future time.
