hide
Free keywords:
-
Abstract:
Irony is one of the most frequently used form of figurative language and plays an important role in our daily communication. To comprehend ironic sentences additional, i.e. contextual and pragmatic, information is necessary. This suggests a more demanding processing than literal language understanding. Controversial findings from behavioral studies (Gibbs 1999, 2002, Giora & Fein 1999, Schwoebel et al. 2000) yield in three different views of irony processing differing in their time course of context influence. In the standard pragmatic model (Grice 1975) context effects are only assumed after literal (non-ironic) meaning activation of ironic sentences leading to a semantic integration problem into the previous context, that results in further inferential processes. In contrast, the direct access view (Gibbs 2002) assumes an initial context influence on lexical access, which allows a direct comprehension of the context relevant (i.e. ironic) meaning without an incompatibility phase. A third view, the graded salience hypothesis (Giora 2002), predicts a simultaneous computation of lexical and contextual information. During lexical access the most salient meaning is activated initially and since context is processed in parallel, it can immediately be effective on the availability of meanings. If the salient meaning is not in accordance with context information, this incompatibility would slow down comprehension processes.
In order to examine when and how ironic and non-ironic sentence processing differs, electrophysiological evidence is presented. According to the standard pragmatic model and the graded salience hypothesis, an integration problem of ironic sentences into a foregoing context should lead to an N400 at the critical word. If processing of both sentence types is due to initial contextual constraints (i.e. direct access view), no irony related ERP effect should be present.
In Experiment 1, event-related brain potentials (ERPs) were recorded from adults listening to ironic and non-ironic discourses. Target sentences as That’s really rich achieved an ironic meaning in negative contexts (i.e. receiving a very small dish in a restaurant), and a non-ironic meaning in positive contexts (i.e. receiving an opulent dish). ERPs on the sentence final word revealed no irony related N400 effect, but a long lasting anterior negativity that started around 100 ms and posterior positivity between 500-900 ms. The posterior positivity in the absence of an N400 effect was replicated in a second experiment, in which stimuli were presented visually. The absence of an N400 effect is inconsistent with the assumptions of the standard pragmatic model, as well as the graded salience hypothesis. The results do also not endorse the direct access model because differences in ERPs at the end of target sentences were found. The posterior positivity elicited in ironic discourses was interpreted as reflecting an evaluation process, e.g. estimating pragmatic information. Such evaluation may have caused more demanding inferential processes and suggests that additional processes seemed to be involved in a complete understanding of irony.
