Skip to content

Political Indoctrination

Authoritarian regimes have always existed, even against the will of their citizens. Sometimes, especially against their will. Social science literature has pointed to three mechanisms that were employed by authoritarian regime leaders in order to maintain their control over these citizens and society overall. These mechanisms are repression, cooptation and indoctrination. More directly, people accept the legitimacy of the authoritarian leaders because otherwise they might face violence, because they are rationally co-interested in the upkeep of the regime, which may bring them direct advantages as well, or they are manipulated into believing in the legitimacy – or even perfection – of the regime. 

While repression and cooptation have been relatively amply studied in the political science literature, indoctrination has not been studied so much. And here is where my PhD research comes in. I am PhD researcher in the Social and Political Sciences Department of the European University Institute in Florence, Italy. I am working on my doctoral thesis under the supervision of professor Elias Dinas, previously at Oxford University. 

My thesis aims to offer a broad perspective on the indoctrinating intentions of communist governments in Eastern Europe, how they compare across countries and time, but especially how they compare in indoctrinating intentions to the textbooks of capitalist/liberal democratic regimes to the West. I plan to devise a research framework and a means to study political indoctrination in such regimes. Indoctrination itself is a highly nuanced phenomena that is encapsulated mostly in text that is written or disseminated by the authority of the political regime. In my research precisely, I focus on the written material with the highest likelihood to influence the belief system of young citizens of a country: their education. I will focus mostly on communist authoritarian regimes, as these offer the cases with the highest probability to find a wide, systematic effort to develop and inject these politically indoctrinating beliefs. 

Many researchers believe that textbooks used in communist countries were extremely loaded with ideas, biased readings of historic events or models of behavior that were meant to make children increase their acceptance of the authority of the communist party and state. Research in history has insisted on these effects in time, but the empirical tests for these ideas have been limited, mostly because of the monumental size of such a project. The belief was that indoctrination is a subtle phenomena, therefore it would have to be studied by the human eye, and the manual study of hundreds of textbooks would have been impractical or difficult to do outside of a subjective qualitative schema. 

Nonetheless, recent discoveries in the field of automated text analysis and machine learning have opened up an avenue for such texts to be studied in a more systematic framework, with quantitative tools. Dictionaries models, unsupervised learning, embeddings, these are just some of the tools that have been developed in recent years that could be used to study such subtle phenomena as political indoctrination. For this purpose, though, a large weight of the research itself is set on acquiring the texts that were injected with indoctrination in a clean and sleek machine-readable format. Automated text analysis tools will only be as good as is the text they are applied to. 

For that, the main thing this research needs is a reliable and faithful transfer of text from photocopies of textbooks that are sometimes up to 70 years old into a good machine-readable format. At this point, I have acquired close to 150 photocopied textbooks from multiple countries, in multiple languages. In order to transfer them into a functional machine-readable format, I have decided the best thing to do is use Abbyy Cloud OCR SDK – https://www.ocrsdk.com/ – which is the state of the art in OCR in the world. 

Ultimately, the quality of the texts I get to parse into R’s quantitative text analysis tools will be a large and important driver of the quality of results that I achieve. For that purpose, I am as interested as can be to make sure that these textbooks are OCR-ed to the best quality possible. 

Published inArticle