We investigated the idea that people use metaphor to help make sense of tex
t-based vignettes about everyday dilemmas. First, we found evidence for the
metaphor framing hypothesis across 6 experiments in which framing descript
ions of dilemmas within various metaphor fi-ames influenced reasoners' deci
sions about how to resolve the dilemmas. Second, we found evidence for the
metaphor processing termination hypothesis across 3 experiments: When a met
aphor was not necessary to increase understanding of a dilemma or when it a
ctually increased ambiguity, the metaphor framing effect was attenuated. In
particular, when the vignette contained or primed knowledge that could be
used to understand the vignette, reasoners were less likely to make decisio
ns consistent with the metaphor, were less likely to rate the metaphor as a
pt, and were less likely to choose metaphor-consistent responses on a subse
quent verbal analogy test. Results are consistent with a constructivist the
ory of text comprehension, which specifies the conditions under which learn
ers use metaphor to make sense of text vignettes.