Emergent Metaphor Theory: Frequency, Schematic Strength, and the Processing of Metaphorical Utterances

Emergent Metaphor Theory, as described in Sanford 2012, asserts that metaphors are schemata that link cognitive domains, that differences in the frequencies of metaphors are a core aspect of metaphorical systems, and that metaphor is operated upon by frequency effects at the level of both overall mappings and individual utterances. This paper presents a corpus study and a series of experiments that support key predictions of Emergent Metaphor Theory. The corpus method makes use of a preliminary survey to elicit core terminology from speakers of English on ten metaphorical source domains; these are used as search terms in identifying similes that instantiate each of ten metaphorical mappings. The survey method supports the assertion that metaphors differ in their frequency, as well as that some terms from a given source domain are used more frequently than others to invoke the mapping. A series of experiments uses the corpus frequencies of metaphors as established in the corpus study to test whether more frequent mappings are more productive, more accessible, and more acceptable to speakers than less frequent metaphors. Both the results of the corpus study and experimental approaches support the view that metaphor is a usage-based phenomenon, and that many of the properties of metaphorical utterances are best accounted for as arising from the interaction of the conceptual schemata that license cross-domain mappings, and syntactic schemata that link meanings to syntactic templates.