Accessible Requires Authentication Published by De Gruyter June 21, 2018

The Global Language of Human Rights: A Computational Linguistic Analysis

David S. Law


Human rights discourse has been likened to a global lingua franca, and in more ways than one, the analogy seems apt. Human rights discourse is a language that is used by all yet belongs uniquely to no particular place. It crosses not only the borders between nation-states, but also the divide between national law and international law: it appears in national constitutions and international treaties alike. But is it possible to conceive of human rights as a global language or lingua franca not just in a figurative or metaphorical sense, but in a literal or linguistic sense as a legal dialect defined by distinctive patterns of word choice and usage? Does there exist a global language of human rights that transcends not only national borders, but also the divide between domestic and international law?

Empirical analysis suggests that the answer is yes, but this global language comes in at least two variants or dialects. New techniques for performing automated content analysis enable us to analyze the bulk of all national constitutions over the last two centuries, together with the world’s leading regional and international human rights instruments, for patterns of linguistic similarity and to evaluate how much language, if any, they share in common. Specifically, we employ a technique known as topic modeling that disassembles texts into recurring verbal patterns.

The results highlight the existence of two species or dialects of rights talk—the universalist dialect and the positive-rights dialect—both of which are global in reach and rising in popularity. The universalist dialect is generic in content and draws heavily on the type of language found in international and regional human rights instruments. It appears in particularly large doses in the constitutions of transitional states, developing states, and states that have been heavily exposed to the influence of the international community.

The positive-rights dialect, by contrast, is characterized by its substantive emphasis on positive rights of a social or economic variety, and by its prevalence in lengthier constitutions and constitutions from outside the common law world, especially those of the Spanish-speaking world. Both dialects of rights talk are truly transnational, in the sense that they appear simultaneously in national, regional, and international legal instruments and transcend the distinction between domestic and international law. Their existence attests to the blurring of the boundary between constitutional law and international law.


This paper was initially prepared for presentation at the Boston College Conference on Global Constitutionalism and Human Rights.I am grateful to all participants at the conference. I am especially indebted to Tom Ginsburg, for sharing his corpus of full-text constitutions and collaborating on the analysis of the topic model described here; Ran Hirschl, for his insightful and trenchant feedback as my discussant; and J.P. Kuhn, Leah Robis, and Rosana Tse for their diligent research assistance, especially in preparing the corpus of constitutions for automated analysis.

Appendix A: Effect of selected covariates on topic prevalence

CovariateTopicStatistical significanceEffect of covariate on topic prevalence
age3: municipalityp<0.01negative
age6: officeholderp<0.10positive
age8: governor-generalp<0.10positive
age10: social, economic & cultural rightsp<0.01negative
age12: Latin Americap<0.01negative
age13: legal systemp<0.01positive
age14: socialismp<0.05negative
age15: judiciaryp<0.05negative
age16: legislative chambersp<0.01positive
age17: public orderp<0.01positive
age18: government powersp<0.01negative
age21: republicp<0.01negative
age23: monarchyp<0.10positive
age24: generic rightsp<0.01negative
age26: foreign affairsp<0.05positive
age30: Francophoniep<0.01negative
length3: municipalityp<0.10positive
length6: officeholderp<0.01positive
length7: Commonwealth courtsp<0.01positive
length9: commissions & tribunalsp<0.01positive
length10: social, economic & cultural rightsp<0.10positive
length12: Latin Americap<0.10negative
length14: socialismp<0.10negative
length17: public orderp<0.05negative
length18: government powersp<0.05negative
length20: federalismp<0.05positive
length26: foreign affairsp<0.10negative
length27: parliamentarismp<0.01positive
length30: Francophoniep<0.01negative
legal family: British2: legislative powerp<0.10negative
legal family: British6: officeholderp<0.01positive
legal family: British7: Commonwealth courtsp<0.01positive
legal family: British9: commissions & tribunalsp<0.01positive
legal family: British20: federalismp<0.10negative
legal family: British21: republicp<0.10negative
legal family: British27: parliamentarismp<0.01positive
legal family: Spanish2: legislative powerp<0.10negative
legal family: Spanish6: officeholderp<0.10negative
legal family: Spanish7: Commonwealth courtsp<0.10negative
legal family: Spanish9: commissions & tribunalsp<0.05negative
legal family: Spanish10: social, economic & cultural rightsp<0.05positive
legal family: Spanish13: legal systemp<0.05positive
legal family: Spanish18: government powersp<0.05positive
legal family: Spanish26: foreign affairsp<0.10positive
legal family: Spanish27: parliamentarismp<0.05negative
legal family: French10: social, economic & cultural rightsp<0.10positive
legal family: French20: federalismp<0.10negative
legal family: French21: republicp<0.10negative
region: Sub-Saharan Africa4: civil servicep<0.10positive
region: Sub-Saharan Africa6: officeholderp<0.10positive
region: Sub-Saharan Africa9: commissions & tribunalsp<0.05positive
region: Sub-Saharan Africa20: federalismp<0.10positive
region: Sub-Saharan Africa21: republicp<0.01negative
region: Sub-Saharan Africa25: electionsp<0.10positive
region: Sub-Saharan Africa29: communismp<0.01negative
region: Sub-Saharan Africa30: Francophoniep<0.01positive
region: Middle East/N. Africa6: officeholderp<0.10negative
region: Middle East/N. Africa17: public orderp<0.05positive
region: Middle East/N. Africa21: republicp<0.05negative
region: Middle East/N. Africa23: monarchyp<0.05positive
region: Middle East/N. Africa29: communismp<0.01negative
region: Latin America3: municipalityp<0.05positive
region: Latin America6: officeholderp<0.05positive
region: Latin America7: Commonwealth courtsp<0.05positive
region: Latin America9: commissions & tribunalsp<0.10positive
region: Latin America12: Latin Americap<0.01positive
region: Latin America14: socialismp<0.05negative
region: Latin America15: judiciaryp<0.05positive
region: Latin America17: public orderp<0.05positive
region: Latin America18: government powersp<0.10negative
region: Latin America21: republicp<0.01negative
region: Latin America26: foreign affairsp<0.01positive
region: Latin America27: parliamentarismp<0.05positive
region: Latin America29: communismp<0.01negative
region: East Asia4: civil servicep<0.10positive
region: East Asia9: commissions & tribunalsp<0.05positive
region: East Asia10: social, economic & cultural rightsp<0.01negative
region: East Asia16: legislative chambersp<0.10positive
region: East Asia18: government powersp<0.10negative
region: East Asia21: republicp<0.05negative
region: East Asia23: monarchyp<0.05positive
region: East Asia29: communismp<0.05negative
region: East Asia30: Francophoniep<0.05negative
Published Online: 2018-6-21
Published in Print: 2018-6-26

© 2018 Walter de Gruyter GmbH, Berlin/Boston