Support Structures for Linguistic Creativity: A Computational Analysis of Creative Irony in Similes

Abstract

Language provides a wealth of stock images and formulaic expressions to speakers who wish to describe a topic in figurative terms.Speakers can also craft their own similes and metaphors, and frequently do, but their most creative efforts run the risk of being misunderstood.Ironic comparisons, in particular, can mistakenly impart the very opposite of what a speaker intends, by virtue of superficially saying the opposite of what is actually meant.Creative similes are often accompanied by explanations that explicitly draw out their meaning, but such explanations run counter to the playful nature of irony and humour.We hypothesize that speakers employ more subtle support structures when minting humorous (and risky) comparisons on the fly, and empirically explore the role of the marker "about" as a channel for telegraphing humorous and creative intent.Our analysis is founded on two large corpora of similes that are harvested from the web, and reveals that the presence of "about" is a very reliable predictor of a simile's true affective meaning.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 44 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.