Parsing requires the quantitative information of syntactic functions of part of speech.Based on the Probabilistic Valency Pattern Theory,this paper studied syntactic functions of Chinese nouns by using the Chinese Dependency Treebank.According to the frequency,we divided syntactic functions of nouns into typical functions and atypical functions,and proposed the model of Correlated Markedness and probabilistic valency pattern of grammatical functions of nouns.From this quantitative analysis,we validated and complemented the conclusions of previous studies,which helped us to have a clearer understanding of syntactic functions of Chinese nouns and provided a reference for teaching Chinese as a second language.