Default sarcastic interpretations of attenuated and intensified similes

This study tests the Defaultness Hypothesis (Giora et al., 2015b, 2018), aiming to show that, “fully abstract phrasal patterns” (Goldberg, 2003, 2006), strongly attenuating (i.e., weakening) highly positive concepts, generates sarcastic interpretations by default – i.e., automatically, without relying on supportive context or internal incongruity. According to the Defaultness Hypothesis, for an interpretation, activated automatically, to be considered as default, it has to be novel, free of internal incongruity, and free of contextual support. The results of our studies support the prediction of the Defaultness Hypothesis, showing that strongly attenuated highly positive concepts (of novel, isolated similes, involving no internal incongruity) generates sarcastic interpretations by default. Additionally, our experimental results diverge from previous corpus-based findings, in which attenuated similes, containing negative internal incongruity, were found to be more sarcastic than their non-attenuated counterparts (Veale, 2013). Given that Veale's corpus-based studies relied heavily on items' internal incongruity, his findings cannot be considered default interpretations. However, our experiments, involving items free of internal incongruity do reveal a significant difference between the attenuated and non-attenuated similes. We conclude that, as predicted by the Defaultness Hypothesis, it is attenuating highly positive concepts that affects sarcasm by default, i.e., even without relying on internal incongruity.