Positive Emotions in Resistance: A Sentiment Analysis of Speeches by Hezbollah

• In Event: Responding to and Coping with Violence

Abstract

The conflict literature typically associates resistance behavior with negative emotions, such as anger (Van Zomeren, Postmes and Spears 2008) or frustration (Gurr 1970). Recent findings from political psychology suggest an important additional role of positive emotions (Dornschneider 2021; Sabucedo and Vilas 2014), showing that they contribute to individuals’ motivation to participate in political dissent in a substantial way. This paper adds to this new literature by applying methods from quantitative text analysis to analyze discrete emotions and affect contained by Arabic speeches of a well-known resistance actor – Hezbollah.

The paper applies sentiment analysis, which is an automated method to identify affect-laden words from text. Although rarely applied to studies of political revolt, it is a well-known method which has been applied widely to examine text as varied as social media (e.g. Bakshi et al. 2016), news reports (e.g. Bautin, Vijayarenu and Skiena 2008), speeches (e.g. Proksch et al. 2019), or interviews (e.g. Dornschneider and Todd 2021). Sentiment analysis is notoriously difficult to conduct in Arabic, which is why the development of related lexica has been slow, and discrete emotions are not included by Arabic sentiment analysis tools to date (al-Ayoob et al 2019). This paper applies the NRC lexicon (Saif and Turner 2011), a validated tool that includes a translated dictionary in Arabic. To the best of our knowledge, this lexicon has not yet been applied to study resistance in the Arab World.

The analysis is performed on 228 speeches, obtained from scraping Hezbollah’s Arabic website. Overall, the corpus includes speeches between September 2001 and April 2022, which contain a total of 1.3 million words. The analysis focuses on overall positivity versus negativity as well as discrete emotions of anger (negative) versus trust (positive). The findings show that the speeches contain surprisingly high levels of overall positivity, underlining the importance of positive emotions and affect. On average, words associated with positive affect occurred 1.72 times more often than words related to negative affect. Consistent with this finding, words associated with the positive emotion of trust occurred 2.33 as often as words related to anger, on average. This underlines the importance of exploring positive emotions in addition to well-known negative emotions, such as anger.

Consistent with the recent literature on positive emotions, the findings offer a new and important contrast to common views associating resistance in general, and especially violent resistance in the Arab World, with negative emotions. Instead, they suggest a more nuanced view, according to which negative and positive emotions and affect co-exist, with positive emotions and affect being more frequently expressed. This offers an important new perspective of our understanding of resistance behavior, and especially of its violent forms in the Arab World.

Van Zomeren, Martijn, Tom Postmes, and Russell Spears. 2008. "Toward an integrative social identity model of collective action: a quantitative research synthesis of three socio-psychological perspectives." Psychological bulletin 134(4), 504-535.

Gurr, Ted R. Why Men Rebel. Princeton. Princeton University Press, 1970.

Dornschneider, Stephanie. 2021. Hot contention, cool abstention. Positive emotions and protest behavior during the Arab Spring. New York: Oxford University Press, Series in Political Psychology.

Sabucedo, José Manuel, and Xiana Vilas. 2014. "Anger and positive emotions in political protest." Universitas Psychologica 13(3), 829-838.

Bakshi, Rushlene Kaur, Navneet Kaur, Ravneet Kaur, and Gurpreet Kaur. 2016. "Opinion mining and sentiment analysis." In 2016 3rd International Conference on Computing for Sustainable Global Development (INDIACom), 452-455. IEEE.

Bautin, Mikhail, Lohit Vijayarenu, and Steven Skiena. 2008. "International sentiment analysis for news and blogs." In Proceedings of the International AAAI Conference on Web and Social Media 2(1), 19-26.

Proksch, Sven‐Oliver, Will Lowe, Jens Wäckerle, and Stuart Soroka. 2019. "Multilingual sentiment analysis: A new approach to measuring conflict in legislative speeches." Legislative Studies Quarterly 44(1), 97-131.

Dornschneider, Stephanie and Jennifer Todd. 2021. Everyday sentiment among unionists and nationalists in a Northern Irish town. Irish Political Studies 36(2), 185-213.

Al-Ayyoub, Mahmoud, Abed Allah Khamaiseh, Yaser Jararweh, and Mohammed N. Al-Kabi. 2019. "A comprehensive survey of arabic sentiment analysis." Information processing & management 56(2), 320-342.

Saif, Mohammed, and Fred Turner. 2011. “NRC Word-Emotion Association Lexicon.” Retrieved August 25, 2020 (http://saifmohammad.com/WebPages/NRC-Emotion-Lexicon.htm)

Authors

• Stephanie Dornschneider-Elkink, University College Dublin

• Luke Alan Horton, University College Dublin

• Sofia Quaglia, University College Dublin