Premodification in evaluative of-binominal noun phrases

Premodification patterns play a central role in the analysis of the evaluative binominal noun phrase (EBNP; a beast of a man). This is, on the one hand, because the EBNP sometimes demonstrates non-canonical premodification: modifiers in front of the first noun can be selected by the second noun (e.g. a bitchy iceberg of a woman), and, on the other hand, because some of the EBNPs have evolved into evaluative modifiers (EMs; a beastuva day), with [N₁ of a] integrating itself into the pre-existing premodification patterns. In the context of the premodification distinctions relevant to the evaluative of-binominal family, this study will juxtapose a hierarchical, Functional Discourse Grammar (FDG; Hengeveld & Mackenzie 2008; Keizer 2015) analysis of the premodification patterns against those of a linear zone-based, Construction Grammar (Ghesquière 2014) approach to premodification. In particular, this chapter, using corpus data from the Corpus of Contemporary American English and the Corpus of Historical American English, compares the categories proposed by each theory and discusses to what extent each model is able to account for the shifting premodification patterns found in the evaluative of-binominal family. This chapter concludes that although a zone-based account can, to a great extent, model these differences, this form of model does not provide an explanation as to why these changes have taken place. On the other hand, FDG can model and provide an explanation for the irregular premodification patterns, and captures the critical distinction between pragmatic and semantic modifiers, which also plays a role in this explanation. This study also provides evidence that the [N₁ of a] chunk has integrated into pre-existing premodification patterns, and that, therefore, an integration of these two approaches to modeling premodification may be possible.