Your research institute sets out to write a restorationist history of a given event – one whose dominant narrative was shaped by the victors, whose evidence is scattered across languages and continents, and whose analytical framework you will have to build from scratch because the existing ones were designed to prevent exactly the conclusions your evidence points to.

That event is the Second World War – or, more precisely, the World Anti-Fascist War. You want to answer questions that Western historiography was designed never to ask. Who actually defeated fascism? What was the economic calculus behind the West’s deliberate delay? How were the deaths of colonised peoples systematically erased from the historical record?

This is qualitative research at its most ambitious: primary sources in five languages, archives on four continents, six disciplines that need to speak to each other – and a field whose dominant frameworks have spent eighty years ensuring these questions are never asked.

Where do you start?

Your Framework is Already Someone Else’s Framework

You open the academic literature. The dominant framing is immediately visible: the war begins in 1939, when Germany invades Poland. Not in 1931, when Japan invades Northeast China. The ‘major turning points’ are D-Day and the atomic bombings – not Stalingrad, not Kursk, not the Hundred Regiments Offensive. The analytical vocabulary itself encodes the erasure: ‘appeasement’ reframes calculated British collusion with Hitler as well-meaning naivety. ‘Isolationism’ transforms deliberate US profit-extraction from fascism into passive non-involvement.

These categories did not fall from the sky. They were forged in Cold War-era universities under direct institutional pressure. The CIA funded the Congress for Cultural Freedom for seventeen years, financing journals, conferences, and the ‘totalitarian twins’ thesis that equated the Soviet Union with Nazi Germany. The Pentagon’s share of federal research spending reached 83.8%. Scholars who documented socialist sacrifice faced repeated tenure denials despite publishing in prestigious journals. A 1955 survey found that CIA agents had contacted 61% of social scientists – not for investigation but for intimidation.

Eighty years on, the packaging has changed. The underlying architecture has not. Apply the standard Western framework and you will conclude that Anglo-American industrial power defeated fascism. You will not arrive at the alternative conclusion – that the Soviet and Chinese peoples saved humanity at a cost of 51 million lives – because the framework was never designed to let you ask that question.

Your Knowledge Base is a Ruin

You decide to build the case from primary sources. Now you need evidence.

Soviet military casualties are documented in Andreev et al. (1993). Chinese deaths from 1931 to 1945 in Bian (2012). Indian famine deaths in Sen (1977). Lend-Lease distribution data in a 1946 US government report. GDP figures in Harrison (1998). Ethiopian casualty records in a 1945 government memorandum. Japanese perpetrator admissions in the 1995 Murayama Statement.

These sources exist in English, Russian, Chinese, Japanese, and French. They are scattered across government archives, university libraries, and military records spanning eighty years. The Chinese government’s 2015 survey of war damage alone involved 600,000 participants. Africa supplied 98% of Allied industrial diamonds and 90% of cobalt – but colonial authorities who tracked copper output to the metric tonne never counted African deaths.

Organising all of this into a knowledge base you can query, cross-reference, and interrogate – collecting, classifying, cleaning, standardising, reconciling conflicting estimates across sources in five languages – would take months. For a resource-constrained research institute, the sheer scale of this work has historically ensured that the erasure remains intact.

Six Disciplines, One War, Zero Integration

If you somehow manage to assemble the sources, you then need to write the study.

A restorationist history needs to span military strategy, economic analysis, demographic catastrophe, diplomatic architecture, legal structure, and cultural production. You bring in five specialists. Five months later, five drafts land on your desk – and none of them talk to each other.

The military historian makes no mention of why the US provided $474.50 in Lend-Lease per white person and $4.40 per non-white person – that is an economic question. The economist cannot connect the Lend-Lease racial calculus to the legal architecture that granted Unit 731’s members immunity – that is a question of law and empire. The legal scholar cannot explain why Hollywood produced over 2,500 Pentagon-approved films while Soviet films were restricted to a hundred theatres – that is cultural analysis.

The deeper problem is not that each person has written badly. It is that the causal chain connecting these facts is unbroken – and their chapters break it. The US fuelled Japan’s war machine until 1941: economic. China therefore fought alone for ten years: military-strategic. Twenty-four million Chinese died while Anglo-American casualties remained at 1%: demographic. The San Francisco Treaty excluded China: diplomatic. The Encyclopaedia Britannica has blank spaces where Chinese civilian deaths should be: knowledge production. One unbroken chain, cut into six disciplinary fragments, having nothing to say to each other.

The framework belongs to someone else. The knowledge base is a ruin. Cross-disciplinary integration is structurally impossible. Under these conditions, the erasure reproduces itself – not through conspiracy, but through the structure of knowledge production.

But Somebody Did It

In November 2025, Tricontinental: Institute for Social Research published The 80th Anniversary of the Victory in the World Anti-Fascist War: Understanding Who Saved Humanity: A Restorationist History.

Seventeen chapters. Starting from 1931 – not 1939 – through colonial extraction, anti-communist collusion, the architecture of post-war impunity, and the manufacturing of memory. Not seventeen papers stitched together, but a single, organically coherent argument with its own internal logic. Hundreds of citations in five languages. A four-layer endnote system allowing any reader to trace every claim back to its primary source.

The depth is what you would expect from a large, well-funded research team working over years. The scale – 85 million dead, mapped across belligerents, colonies, and racial categories with consistent source documentation – is what you would not expect from a Global South institute working on Global South budgets.

The question is not as simple as what the study found. The question is how this research was possible at all.

How Were the Three Barriers Crossed?

This is the question that matters – not just for this study, but for qualitative research in the social sciences as a whole.

The framework barrier. Researchers were no longer trapped inside someone else’s analytical categories. The system drew on more than 80 historical conjuncture analyses – from the Comintern era to contemporary Latin American situational analyses – to distil universal research themes without presupposing any particular narrative. The theoretical framework is configurable: Marxist, liberal, realist – a two-line configuration change. The tool does not choose your position for you – for the first time, it gives you the room to make choices. For this study, the researchers chose historical materialism: class analysis, inter-imperialist contradictions, the relationship between productive forces and military capacity. They could have chosen otherwise. The point is that the choice was theirs.

The knowledge base barrier. Building a queryable, cross-referenceable knowledge base went from months to days. Thousands of documents – web pages, PDFs, academic papers, archival records, data tables, charts – were transformed into an intelligent knowledge base at a cost of less than $200. A Zotero library built up over years becomes a research partner you can interrogate in 25 minutes. Sources in multiple languages, scattered across institutional silos on four continents, became searchable and comparable. The Encyclopaedia Britannica’s blank spaces, the Chinese government’s 24.05 million documented deaths, and Western estimates of 15–20 million could be placed side by side – and the gap between them made visible as evidence of erasure, not methodological difference. What used to require a team of multilingual research assistants working for months now requires a laptop and a week.

The integration barrier. This is the most transformative change. AI has no disciplinary boundaries – not because it transcends expertise, but because a unified theoretical framework provides the connective tissue that specialist teams structurally cannot. Economic structure analysed through class relations, military strategy through the balance of productive forces, diplomatic architecture through inter-imperialist contradictions, cultural production through ideological reproduction – different disciplinary content, brought together in a single conceptual language. The causal chain that used to be cut into six fragments, each lying in a different specialist’s chapter, now runs unbroken through the entire analysis. This is not a matter of telling specialists to ‘collaborate more.’ It is a structural change in how qualitative research can be conducted.

What This Means for Qualitative Research

Think about who currently has the capacity to produce qualitative research at this scale and depth.

Well-funded Western universities can. Government-backed think tanks can. Intelligence-linked research institutions can. They have the archives access, the language capabilities, the disciplinary breadth, and the budgets. A comparable study from any of these institutions starts at several hundred thousand dollars. And they will not use your framework – they have their own, serving their own clients.

Global South researchers have never had access to that infrastructure – until now.

The full methodology behind this study – the configurable theoretical frameworks, the rapid knowledge base construction, the cross-disciplinary AI synthesis – is part of the GSI platform’s ‘AI for Social Science’ framework, developed by Tricontinental: Institute for Social Research. GSI has brought the cost of qualitative research at this level down by an order of magnitude, and it is open to Global South researchers.

The World Anti-Fascist War study is the proof of concept: eighty years of erasure undone, blank spaces filled, ninety to 115 million silenced victims restored to the historical record. But the implications reach beyond any single study. Any qualitative research programme that faces these three barriers – a field dominated by frameworks that foreclose your questions, sources scattered across languages and institutions, analysis that demands cross-disciplinary integration – now has an alternative path.

The evidence was always there. The sources existed. What we did not have was the infrastructure to bring them together, within a framework that allowed the questions to be asked.

That infrastructure now exists.

From | Tricontinental: Institute for Social Research via This RSS Feed.