Abstract
This paper employs a usage-based approach to grammar, the theory of frame semantics, and the corpus-based method referred to as the attraction-reliance measure, for the purpose of examining the mutual association between a noun and the be of N-pattern: in other words, to determine nouns that are strongly associated with this construction. On the basis of the data gleaned from the Corpus of Contemporary American English (COCA), the paper aims to indicate that there are specific categories of nouns occurring in the construction under scrutiny, and that this construction performs specific functions in discourse.