This paper discusses the evolution of the parallel prepositional connectives YU and JI((because YU and JI had many forms in the history of Chinese language, we hereby use theire Pinyin to represent them)in Archaic Chinese. Our viewpoint is based on an assumption that there is always differece between the so-called synonyms. That means theoretically, not a single word in Chinese has an identical copy. The differences between synonyms could be semantical, functional or stylistic. For instance, in Chinese, there are many words expressing the meaning of \"OFTEN\", such as: 經常, 常常 ; 總是, 老是 ; 通常, 往往 which mean \"often\", \"always\" and \"usually\" respectively. And also, among each group, of the two words sharing the same meaning, the one is more formal and of more literary style than the other. Another example is \"他好些了\", \"他好多了\". These expressions differ in the degree of recovery. And both of them are different in meaning from \"他更好了\". As a reflect on this phenomenon, Li(1980) has proposed a theory of \"refinement\" for diachronic linguistics claiming that languages are becoming subtler and subtler. Why there existed a greater amount of pronouns and parallel connetives in Archaic Chinese compared to modern Chinese? Can we differentiate them according to their functions as many linguists intend to do? The answer to the first question can be generalized as thatthe initial phase of a development is always in a chaotic status. The answer to the second question is \"No\". \"Refinement\" implies somthing need to be refined, i.e.to be clarified. At the first stage of development of a language, everything needs refining.