Some groups of people attract far more online hate and abuse than others. In this post and accompanying video, LSE’s Professor Shakuntala Banaji explains that her research shows how online hate is entangled with different kinds of oppression across the world, and how to challenge it.

Is it true that anyone who goes online a lot receives more hate than anyone who stays unconnected to social media? Are global tech platforms controlling hateful content adequately through algorithms and AI-based moderation? And is hateful misinformation largely a matter of media illiterate publics accessing mobile technologies? In 2022 Ram Bhat and I published a book called Social Media and Hate based on research we’d been doing in India, Brazil, Myanmar and the United Kingdom online and with people who make, circulate and receive hateful messages online.

Watch Shakuntala explain the hierarchy of hate here

Our analysis of thousands of distinct pieces of hateful and dehumanising audiovisual data – short films circulated online and through social media, memes, gifs, images, slogans, speeches, animations, quotes and other mediated content – led us to propose and test a typology of the most-circulated violent social media content that extends from the bizarre and extreme (massive traffic accidents, industrial disasters) to the dehumanising, the violent and hateful (random acts of violence, misogyny, xenophobia, racism, casteism, Islamophobia, fat-phobia, transphobia, homophobia and various iterations of such abuse and violence with actual images of violent attack or atrocity).

In some of these videos, groups of people were actively calling for pogroms or genocide against racialised or caste groups in national or local contexts. In others, they were making fun of them or perpetuating more subtle derogatory myths, stereotypes and lies to dehumanise minorities. In yet others, disinformation was being deployed about the minority to make a majority community seem to be endangered or victimised, and to perpetuate a sense of collective grievance against a minoritised group. Often this was the most-circulated and watched content, and the most liked or supported content, despite breaching the guidelines on hate speech of the various platforms.

Much of this content was deeply intertwined with the public political ideologies and positions of dominant groups, and political parties in the countries in question. Furthermore, we found clear patterns in terms of who was propagating and circulating such content and who was at the receiving end. And so, in our book, we set out what we call here a hierarchy of hate, linked to histories of colonial, precolonial and postcolonial inequality and oppression.

While they are often white in white-majority countries, the aggressor groups tend to be Muslim in many Muslim-majority countries (Indonesia, Pakistan, Saudi Arabia), Jewish in Jewish-majority countries (Israel) and Buddhist in some Buddhist majority countries (Myanmar, Sri Lanka). To be precise, it is not just outspokenness, and not just the intersection of race or caste with gender, sexuality, disability or age, that makes some individuals and groups more likely to be targets: even within racial and sexual groups there are subgroups who are most frequently and virulently targeted ‘just for existing’.

So Black women – especially those in the public eye, working class, Muslim, gay, and trans Black women – along with trans men, face some of the highest levels of violence and abuse on and offline, while Asian women/ Dalit women and visibly Muslim women (especially those who veil or those who openly reject the veil) are targets of multiple forms of discriminatory and abusive communicative behaviours. This is the case even when they do not speak out publicly about issues of rights and justice.

These discriminated groups are closely followed by Black men, Dalit men and Muslim men in all these intersecting categories in all our case study countries (and in particular any men in these categories who advocate for the poor and/or minoritised and oppressed groups and are broadly on the political left). Notably, hate speech is also aimed from within oppressed groups at other minoritised and oppressed groups, either intersecting with or outside of that group identity – so for example, from some diasporic migrants at refugees, or from some groups of women at other groups of women and at queer people, and so on.

In order to challenge and defeat this hierarchy of hate, we argue, it is important to pursue many different technological, legal, social and political means, including by forcing platforms to govern and work better, to recognise and take down such hate far more swiftly, and to support those at the receiving end. However, we are very clear: simple technocentric explanations and suggested solutions that do not include complex attempts to recognise and work against social injustices and discrimination and challenge the normalisation of violence and supremacy in the media and political speech, will end in failure and further bolster those who circulate hate on and off-line.

Learn more about this research in the book Social Media and Hate (Routledge, 2022) which is available free to download: https://www.routledge.com/Social-Media-and-Hate/Banaji-Bhat/p/book/9780367537272

This post represents the views of the author and not the position of the Media@LSE blog nor of the London School of Economics and Political Science.