Anomaly in novel metaphor and experimental tests

Abstract

In this article, we assess some of the empirical work available in relation to anomaly in novel metaphor. This revision allows us to argue that the results of reaction time experiments do not provide, as many theorists have argued (Gibbs and Gerrig 1989; Keysar and Glucksberg 1992), evidence against any version of anomaly; at most they can be used against anomaly as categorial falsity. In addition, we assess the argument against anomaly based on the results of reaction time experiments to show that it is unsound. Thus, we show that contextual abnormality, a sub-propositional version of anomaly (Romero and Soria 1997/1998), cannot be rejected as one of the necessary identification conditions of novel metaphor. Furthermore, contextual abnormality is supported by the results of recent empirical studies on metaphor processing designed by neuropsychologists to test hypotheses related specifically with anomaly in novel metaphor (Tatter et al. 2002; Ahrens et al. 2007).