Tomac, Goran.
(2018).
Linguistic Markers of Deception in Computer-Mediated Communication: An Analysis of Politicians' Tweets.
Diploma Thesis. Filozofski fakultet u Zagrebu, Department of English Language and Literature.
[mentor Grubišić, Marina].
Abstract
The aim of this master‘s thesis was to examine the lies of English-speaking politicians by determining whether relevant scientific data on deception applies to the statements they communicated on social media. Specifically, the goal was to analyse the studies on deception and see if one could make use of the data to detect deception in their messages. In addition to set format of this work, the reason for concentrating solely on messages transmitted via website such as Twitter is its popularity, availability and overall use of it among politicians. In order to analyse dishonesty, falsehood and disinformation in messages they communicate, the author first had to define deception, describe the characteristics of participants in a deceptive exchange and point out cues that signal deceptive behaviour. He compiled a summary of several studies which focused on describing the profile of deceptive behaviour and enumerated the linguistic features that characterize deceitful messages. Finally, given that the author looked into statements published on the Internet, it was also necessary to become acquainted with aspects of computer-mediated communication and the features of deception and its detection in this medium. In the following analysis the objective was to recognize those features in the selected false statements in order to discover if one can rely on language components when determining the truthfulness of a politician‘s proclamation, testimony or assurance. Therefore, the author presented examples of several American politicians‘ tweets containing different linguistic markers which, according to Interpersonal Deception Theory and several additional studies, point to deception. Namely, these are levellers, modifiers, negative emotion words, sensory words and qualifiers. Additionally, it was demonstrated that, when it comes to transmitting messages via Twitter, the rates of group references as opposed to self-references and the choice of verb tense are not reliable as indicators of deception. On the other hand, at the beginning of the section the author enumerated motion verbs as another marker which he attempted to identify in the false tweets; however, he was not able to come across any of them. Lastly, in addition to false tweets which contained no markers of deception, the author provided a handful of examples of truthful tweets, which suggest the markers can appear in truthful statements as well. Taking into account the characteristics of computer-mediated communication and a limited number of examined tweets, it can be argued that identifying the markers may be used as a method of detecting deception in statements published on Twitter. However, the method is far from being failsafe and these findings strengthen the importance of non-verbal cues, some of which, as we know, are necessarily omitted in text-based computer-mediated communication.
Actions (login required)
 |
View Item |
Statistics
Statistics