Abstract
This study focuses on an unresolved issue -- the role of salience-based yet incompatible interpretations in shaping contextually compatible ironic interpretations in contexts strongly benefiting such interpretations by inducing an expectation for ironic utterances. According to the direct access view (Gibbs, 1986, 1994, 2002), if context is highly predictive of an oncoming ironic utterance it will facilitate that utterance relative to an incompatible alternative. According to the graded salience hypothesis (Giora, 1997, 2003), even if irony is facilitated, this will not block salience-based interpretations -- interpretations based on the salient meanings of the utterance's components (whether literal or non-literal) -- even if contextually incompatible. Review of the findings accumulated in the literature so far show that ironies took longer to make sense of than salience-based interpretations and involved salience-based incompatible interpretations even in the presence of a strong context inducing an expectation for an ironic utterance. This was true even when contextual information was heavily biased in favor of an ironic interpretation, whether observably promoting such an expectation (Giora et al., 2007) or involving more than one contextual factor supportive of that interpretation (Pexman, Ferretti & Katz, 2000).
