A cognitive modeling approach on ironical phraseology in Twitter

Abstract

The development and growth of social networks evidence human creativity via the use of figurative language including irony. Recent studies on modeling irony and irony detection in social media have looked at it from a traditional perspective and have focused primarily on developing natural language processing systems, thus ignoring the mental processes the participants experience during ironic speech acts. As a result, irony has been misinterpreted and mixed by the experimental literature with other disparate phenomena, such as jokes, understatements or banter. On the other hand, scholars from the field of Cognitive Linguistics have studied the cognitive processes operating in the creation of ironic remarks. With regard to this, Ruiz de Mendoza’s [9] development of the echoic account focuses on ironic discourse and categorizes verbal irony. Yet, no study to date has explored ironical phraseology in terms of cognitive modeling based on bigdata. This study, therefore, aims to examine how Spanish-speakers conceptualize and express irony in Twitter. Results revealed that irony was frequently misconceived and, as a consequence, additional cues such as explicit ironic hashtags prevented readers from interpreting the message literally, especially in explicit-echoic ironic cases. A more frequent interaction between text-hashtag as compared to text-emoji was also evinced for all potentially ironic linguistic signs. It is expected that our findings contribute to research on Spanish as a foreign language (ELE in the native-language acronym) teaching by enhancing the intercultural sensitivity in the learner, as well as to the field of computational linguistics in adding feature types.