How History Works

Core Concepts

What "history" actually is

When we talk about history, we tend to mean a story reconstructed from sources — documents, records, archives. But that definition carries a hidden assumption: that the past left a textual trace, and that the trace survived, and that it was created by people with the means and motive to write.

None of those conditions holds for most of human history and most of human beings.

Historiography — the study of how history is written — has spent roughly a century interrogating what that gap means. What follows is the conceptual ground floor: the vocabulary and frameworks you need before examining what gets recovered when we look beyond the text.

---

Source types: what historians actually work with

Conventional historical practice has a default hierarchy. At the top: official documents, chronicles, state papers, legal records. These are treated as primary sources — the closest thing to unmediated access to the past.

But "document" is only one category in a much wider field. The Annales school — a French intellectual movement founded in 1929 — substantially expanded this evidentiary base by treating parish registers, notarial archives, climate records, manorial accounts, folk songs, judicial files, and material culture as primary sources equivalent to state chronicles and literary texts. This reconceptualization enabled the study of non-elite populations whose experience was not recorded in traditional narrative sources.

E. P. Thompson, working from a different tradition, drew on workshop customs and rituals, union records, threatening letters, popular songs, ballads, parish records, and other material traces of working-class life. He demonstrated that the agency and consciousness of ordinary people can be reconstructed from unexpected sources — documents created not to document working-class life, but that inadvertently register it through official anxiety, criminal charges, or administrative notation.

Expanding beyond the written document opens the following categories of evidence:

• Material culture: objects, tools, clothing, built environments
• Oral transmission: genealogies, songs, epic narratives, place names
• Landscape: field systems, territorial boundaries, topographic names
• Bodies: skeletal remains, burial practices, bodily adornment
• Climate and ecology: pollen records, ice cores, soil composition
• Judicial and administrative records: created by power, but legible against the grain

---

Elite-written bias: a structural problem, not an accident

The documents we have are not a random sample of what happened. They are a sample of what literate, powerful people chose to record — and what survived their choices.

Two forces converge to produce what historians call elite-written bias.

Production bias. Documents are produced by people with time, resources, and reasons to write. In most historical societies that means state institutions, religious organizations, merchant houses, and scholarly elites. The Confucian historiographic tradition — which produced China's Shiji and subsequent dynastic histories — was compiled by and for the scholar-official class, organized around the concerns of officialdom, and structured to privilege written records over oral testimony. Regions, merchants, women outside imperial courts, ethnic minorities, and popular religious movements appear in the record primarily as deviation, threat, or exotic background — not as agents with their own historical significance.

This pattern is not uniquely Confucian or Chinese. It is a structural feature of state-aligned historiographic traditions worldwide. The texts we inherit were shaped by Confucian assumptions about who counted as a historical subject: those whose deeds could be evaluated by moral standards appropriate to elite rulers.

Survivorship bias. Even what was written does not survive equally. Museum collections of clothing disproportionately represent formal, elite, or special-occasion garments — wedding gowns and court dress — rather than the everyday workwear that ordinary people wore to destruction. Most histories of costume in early modern Britain concentrated on elite clothing both because such clothing was documented in written sources and more likely to survive in collections. Small-sized children's garments survive because children outgrew them before wear destroyed them; adult work clothes did not survive because they were worn to fragments.

These two biases compound. The people least likely to produce documents are also the people whose material remains were least likely to be preserved or collected.

---

Documents as artifacts of power

Thompson's insight went one level deeper. It is not simply that elite documents are incomplete — it is that they are actively distorting. Legal documents, trial records, administrative papers, and official reports were created by states, courts, and employers to serve their purposes: to criminalize, regulate, or suppress subordinate groups.

The historian's craft is to read these documents "against the grain" — extracting evidence of the agency and perspectives of the people these documents were designed to marginalize or discredit. A trial record documents not objective crime but the legal system's encounter with working-class resistance. What the official record marks as "sedition" or "riot" may reveal a logic and morality that the court refused to recognize.

This epistemological stance — treating sources as products of power relations rather than transparent evidence — reframes the politics of historical evidence itself.

---

The evidential paradigm: reading traces

Carlo Ginzburg articulated what he called the evidential paradigm (also: the conjectural paradigm). The claim is that small-scale historical analysis requires not only different investigation techniques but different standards of evidence and proof than broader quantitative studies.

In this framework, the historian operates like a detective: using traces, clues, and marginal details to reconstruct hidden dimensions of past experience. Ginzburg traced this method to Giovanni Morelli's approach to art authentication (1874), which relied on close observation of apparently insignificant details — the shape of an ear, the rendering of a fingernail — to establish authenticity. The same logic appears in diagnostic reasoning in medicine and in detective fiction.

The implication: the expansion of valid sources is not a minor methodological preference. It is an epistemological claim about what counts as knowledge of the past.

---

The colonial distortion

The problem of source privilege is amplified by colonial historiography. Nineteenth- and twentieth-century colonial administrators explicitly dismissed itihasa-purana — the Indian epic-genealogical tradition — as "myth" incompatible with "real" history. Robert Orme and James Mill argued that rational, factual history emerged in India only with British rule, treating pre-colonial Indian historical consciousness as absent or deficient.

This was not a methodological assessment. It was an ideological one, encoding a preference for Western written documents over performative-genealogical narrative and oral transmission.

Contemporary scholarship has overturned this dismissal. Inscriptions across India frequently corroborate genealogical information preserved in Puranic literature and oral transmission systems, providing archaeological evidence that validates the accuracy of dynastic records. The corroboration contradicts colonial claims that oral tradition was inherently unreliable — demonstrating instead that systematic mnemonic practices in genealogical transmission produced historically verifiable information.

The Annales school's insistence on interdisciplinarity — on dialogue between history, sociology, and anthropology — opened one path out of these inherited biases. Subaltern Studies scholars opened another, arguing that the problem was not only who got included as a historical subject, but what counts as valid historical knowledge and evidence in the first place.

---

History without monuments

The debate about valid sources is not merely academic. It touches on which communities are considered to have histories at all.

Inuit place names operate as a landscape archive functionally equivalent to monumental infrastructure. Language draped over geography — place names embedded in speech and travel — constitutes a distributed, dynamic archive that is reactivated each time the name is spoken in context on the land or in story. This challenges Western archaeological and heritage frameworks that center monuments as the primary medium of historical memory.

Historiography and memory transmission do not require permanent ritual or architectural features. They require systems of transmission — which may be oral, spatial, material, or bodily.

---

Common Misconceptions

"Written sources are more reliable than oral ones." This conflates medium with accuracy. As the convergence of Indian inscriptions and Puranic genealogies demonstrates, systematic oral transmission practices can produce historically verifiable information. The reliability of a source depends on the transmission system, not the medium. Written sources produced under conditions of political pressure or ideological constraint can be actively misleading.

"Elite bias is a Western problem." It is a structural feature of state-aligned historiographic traditions globally. The Confucian historiographic tradition demonstrates the same dynamics: organized around official concerns, structured to privilege written records, and marginalizing non-elite populations as deviation or background. The form varies; the structure recurs.

"The archive is a neutral repository of the past." Archives are collections built by people with interests. What gets preserved, what gets catalogued, and what gets discarded reflects decisions made by institutions with power. The historian's task is to read not just what the archive contains but how and why it was constructed.

"History only exists where there are documents." This is precisely the assumption this curriculum interrogates. Landscape, body, language, and material culture all carry historical information. Communities without writing did not lack history — they preserved it through other systems. The question is whether the historian has the methods to read those systems.

---

Analogy Bridge

Think of the archive like a city library rebuilt after a fire. The books that survived are the ones that happened to be checked out, or stored in the basement, or duplicated in another branch. The collection you can browse is not a representative sample of what the library once held — it is a sample of what survived a specific catastrophe, with all the accidents that implies.

Now imagine that the library was also built by one particular neighborhood for its own use. The books that were most thoroughly collected in the first place reflected that neighborhood's concerns, languages, and interests. Other neighborhoods kept their knowledge in songs, in family stories, in the way they named the streets — none of which made it into the collection.

That is the archive. The question is: what are the equivalents of those songs, stories, and street names? And how do you read them?