Shinekhen Buryat, a Mongolic language spoken in Inner Mongolia, is one of the 'typical' agglutinative languages. In this language, nouns do not usually take any conjugational suffixes, nor do verbs take any case suffixes. It seemed to be easy to classify its words into parts of speech morphologically. However we can find some kinds of words difficult to classify. To classify these words 'objectively', I will point out that it is important to define 'STEM' with its inflectional/case paradigm.
　　It is clear that STEM definition for word classification is not so effective for all languages. For example, it is of no use for isolating languages like Chinese. But we can consider that Chinese has some parts of speech, so that Chinese speakers can distinguish Noun/Verb without any morphological clue. If the speakers distinguish, there must be some differences not morphologically, but syntactically or semantically or functionally. So, it is important to classify words into some parts of speech for grammatical description.