Cross-national analysis of global security discourse using word embeddings

26 May 2020, Version 1
This content is an early or alternative research output and has not been peer-reviewed at the time of posting.


This study conducts a systematic investigation of discursive dynamics in the context of policy deliberations on international peace and security, by performing cutting-edge text analysis on the entire body of meeting records of the Security Council of the United Nations for the past quarter of a century (1994-2019). Focusing on one of the most consequential notions for the council’s policy-making, "threat to the peace," it employs an unsupervised machine learning model, broadly termed "word embedding," to analyze how this notion has been discussed by relevant members of the council, especially its five permanent members, during the period under investigation. The study reveals persistent patterns of cross-national convergence and divergence in security discourse, including, most notably, a considerable degree of correlation in how to conceive international threats found between the close-knit Western allies in the council and their Russian counterpart.


United Nations
international security
text analysis
word embedding
threat conception

Supplementary materials

This appendix provides further descriptions of the study's methodological aspects. It also reports on additional measurements and analyses that were conducted to examine the robustness of the results discussed in the main text.


Comments are not moderated before they are posted, but they can be removed by the site moderators if they are found to be in contravention of our Commenting Policy [opens in a new tab] - please read this policy before you post. Comments should be used for scholarly discussion of the content in question. You can find more information about how to use the commenting feature here [opens in a new tab] .
This site is protected by reCAPTCHA and the Google Privacy Policy [opens in a new tab] and Terms of Service [opens in a new tab] apply.