The Tamil Tigers were one of the most organized, focused, and influential organizations in Asian politics during the last three decades. Throughout the existence of the organization, the goal of the LTTE was to establish a separate Tamil state – Tamil Eelam - comprising the Northern and Eastern parts of Sri Lanka where Tamil-speaking populations were dense. The Sri Lankan civil war, which ended in May of 2009, claimed tens of thousands of lives, displaced hundreds of thousands of residents, and impeded the economic well-being of Sri Lanka. This thesis employs the operational code analysis as a way to better understand the character, behaviour, and norms of the Liberation Tigers of Tamil Eelam. An additional purpose of this thesis is to test the utility of this model in the case of a terrorist organization such as the LTTE. The Operational Code Analysis framework was premiered by Nathan Leites in his two-part study of the Bolshevik Party, and was subsequently revived and re-imagined by a number of academics. The iteration of the Operational Code framework used in this thesis is a qualitative analysis which is comprised of ten questions: five ‘philosophical’ and five ‘instrumental’. This examination reveals that while this model has much to offer in analyzing this terrorist organization, it is in some ways a flawed method of exposition. However, the model produces unexpected insights about the stated beliefs of the LTTE.

abstract: Operational code analysis (OCA) is a common method of content analysis within the foreign policy analysis (FPA) literature used to determine the “operational code” of state leaders and, by extension, the foreign policy behaviors of their respective state. It has been tried and tested many times before, on many different world leaders from many different time periods, to predict what the foreign policy behavior of a state/organization might be based on the philosophical and instrumental beliefs of their leader about the political universe. This paper, however, questions if there might be types of politicians that OCA, conducted using the automated Verbs In Context System (VICS), has problems delivering accurate results for. More specifically, I have theoretical reasons for thinking that populist leaders, who engage in a populist style of communication, confound VICS’ analysis primarily because the simplistic speaking style of populists obscures an underlying context (and by extension meaning) to that leader’s words. Because the computer cannot understand this underlying context and takes the meaning of the words said at face value, it fails to code the speeches of populists accurately and thus makes inaccurate predictions about that leader’s foreign policy. To test this theory, I conduct the content analysis on speeches made by three individuals: Donald Trump, Boris Johnson, and Narendra Modi, before and after they became the executives of their respective countries, and compared them to a “norming “ group representing the average world leader. The results generally support my hypotheses but with a few caveats. For the cases of Trump and Johnson, VICS found them to be a lot more cooperative than what I would expect, but it was also able to track changes in their operational code - as they transition into the role of chief executive – in the expected direction. The opposite was the case for Modi’s operational code. All-in-all, I provide suggestive evidence that OCA using VICS has trouble providing valid results for populist leaders. / Dissertation/Thesis / Summary Statistics and Calculations of Indices / Masters Thesis Political Science 2020