| A polysemous morpheme has several sememes. It may have different characteristics when forming compound words or referring to an individual word with different sememes. In this thesis, we define a morpheme with a specific sememe as a"morpheme item", and the written form of a morpheme as a"morpheme form". We analyzed 581 morpheme items relevant with 50 morpheme forms, 12083 compound word items, and set up an attribute database of morpheme items. Based on the database, we did some research about the grammatical function, morphological productivity, distribution in words, frequency of usage as an individual word, degree of freedom and average meaning-showing degree of morpheme items.Different morpheme items have different morphological productivity. In respect of grammatical function, nominal, verbal and adjective morpheme items have the highest productivity. Prepositional, particle and conjunctive morpheme items have the lowest productivity. In respect of syntagmatic relations, semantic determinant attributes of sememes restrict the productivity of morpheme items. In respect of paradigmatic relations, there are competition and divisions among synonymous morpheme items. Morphological productivity of morpheme items is relevant to their distribution in words too. Delocalized morpheme items have the highest productivity, and the morpheme items which can only stand in the front of a word have the lowest productivity.We discuss degree of freedom based on morpheme items, not on morphemes as a whole. According to the morphological productivity and frequency of occurring as an individual word, we divide morpheme items into 5 groups, which are A (only occurring as a compounding element), B(most of the time occurring as a compound element), C(occurring as an individual word or a compounding element), D(most of the time occurring as an individual word) and E(only occurring as an individual word).It turns out that the morpheme items of group C always have the highest morphological productivity and highest frequency of occurring as individual words. And in term of"syntactic morpheme items", the lower freedom degree they have, the more dependent they are on lexical meaning on the other morpheme items; the higher freedom degree they have, the more dependent they are on grammatical meaning on the other morpheme items. In TCSL, different methods should be adopted when teaching morpheme items in different groups.Previous researches about the relationship between the meaning of words and the meaning of compounding elements always regarded"word"as the first step of research, which is aimed at formation or cognition of word meanings. In this thesis, we regard"morpheme"as the first step. By gathering all the compound word items formed by a certain morpheme item and analyzing its meaning-showing type in every word item, we calculate average meaning-showing degree. In TCSL, different methods should be adopted when teaching morpheme items of different average meaning-showing degree.In the end, proceeding from the reality of TCSL, based on qualitative study and quantitative study, we try to classify and grade the morpheme items, and scientifically arrange their studying order in certain grades. We also add teaching method tags to the morpheme items, aiming at finally setting up an attribute database of morpheme items for TCSL, so that we can apply our study results directly to teaching practice and textbook compiling. |