The aim of this study is to test the graded salience hypothesis, according to which salient meanings should be processed initially before less salient meanings are activated (Giora, 1997). A meaning of a word or an expression is considered salient if it can be retrieved directly from the lexicon. According to the graded salience hypothesis, processing an utterance such as 'What a lovely day for a picnic' (which has one salient interpretation - the literal) should involve activation of its literal meaning in both types of contexts - in the context biased towards the literal meaning and in the context biased towards the ironic meaning. However, while the ironically biased context should activate the ironic meaning as well, the literally biased context should activate only the literal meaning. In our study, subjects were presented with 'target sentences' (ironies) at the end of either ironically or literally biasing contexts. They were asked to complete one of two fragmented words (such as t-b-e) with the first word that came to mind. The target words were related to either the ironic or the literal meaning of the target sentence, so that activation of the different meanings could be assessed. Findings support the graded salience hypothesis. They show that, contrary to current beliefs, irony interpretation involves processing the literal meaning. They further reveal that ironic and literal interpretations do not involve equivalent processes. (Rachel Giora and Ofer Fein)