Sources, Evidence & Bias
CollegeArchival method, quantitative history and the silences of the archive
Theory Key ideas
University-level source criticism formalises what earlier chapters introduced, dividing evaluation into two stages:
- External criticism — establishing authenticity and provenance: is the document what it claims to be, who transmitted it, has the text been altered in copying or translation? This draws on diplomatics (the study of document form) and philology (the study of textual transmission).
- Internal criticism — once authenticity is established, assessing the content: the author's competence, proximity to events, purpose and likely bias, and the internal consistency of the claims.
Beyond individual documents, historians use serial sources — large runs of repetitive records (parish registers, tax rolls, ship manifests, price lists) — to reconstruct patterns statistically. This underpins cliometrics, the application of quantitative economic and statistical methods to history (pioneered by Robert Fogel and Douglass North), and techniques like family reconstitution in historical demography, which rebuilds birth, marriage and death patterns from parish records to estimate fertility, mortality and life expectancy in periods before census data existed.
Oral history raises its own methodological questions: memory is reconstructive and can be shaped by later events, group narratives and the interview situation itself — yet it remains often the only access to the perspectives of people who left few or no written records.
Finally, historians must reckon with the archive itself as a historical object. Archives are not neutral repositories; they were assembled by institutions (states, churches, companies) that decided what to keep and what to discard, whose voices to record and whose to ignore. Michel-Rolph Trouillot's Silencing the Past (1995) analyses this as occurring at four moments: the making of sources, the making of archives, the making of narratives, and the making of history itself.
Explanation Why it matters
Every historical claim ultimately rests on this evidentiary infrastructure, so its limits are not a technical footnote — they are the boundary of what can be known at all. Colonial and slaveholding archives, for instance, were built overwhelmingly from the records of literate, propertied, often European or colonial-administrative voices; the enslaved, the colonised and the illiterate appear mainly as objects of others' record-keeping (a bill of sale, a court case, a missionary's account) rather than as narrators of their own experience. This is what Trouillot means by an archival silence: not an absence of any evidence, but a structural imbalance in whose perspective the surviving evidence represents.
This has direct methodological consequences. Historians recovering marginalised voices (a project central to social history, subaltern studies and much recent scholarship) must often read colonial or elite-produced sources against the grain — asking what a court record, plantation ledger or missionary diary reveals inadvertently about the people it was not designed to represent, while being explicit about the interpretive leap this requires. Quantitative methods carry a parallel caution: a statistically rigorous dataset (say, ship manifests recording the transatlantic slave trade) can establish scale and pattern with real precision, but it cannot, by itself, recover subjective experience — for that, historians must triangulate with whatever qualitative and oral evidence survives.
Trouillot argued the Haitian Revolution (met in Chapter 12) was largely "unthinkable" to contemporary Western observers even as it happened, because it contradicted their basic assumptions about enslaved people's capacity for organised political action — producing a historiographical silence that took nearly two centuries to substantially correct.
Practical Triangulating quantitative and qualitative evidence
Suppose you are researching living standards in an early industrial city (recall Chapter 13's "standard of living debate"). A rigorous approach combines methods:
Serial/quantitative evidence: Real-wage series, height-and-health data (historical demographers use recorded heights as a proxy for childhood nutrition — a method called anthropometric history), infant mortality rates from parish or civil registers, and census occupational data. These establish scale and trend with statistical confidence, but flatten individual variation and say nothing about subjective experience.
Qualitative/testimonial evidence: Royal Commission testimony (like Sarah Gooder's, Chapter 13), diaries, union records, oral histories where they exist. These recover lived experience and meaning, but are unevenly distributed — likelier to survive for the literate, the organised, or the specifically selected by reform-minded investigators — and cannot establish scale on their own.
Method: Use the quantitative data to ask "how representative is this testimony?" and use the testimony to ask "what does this trend feel like to live through, and what does the average number conceal?" Neither source type alone supports a strong causal claim; convergence between independently-derived quantitative and qualitative evidence is what justifies confidence.
Q&A Test yourself
Distinguish external from internal source criticism, with an example of each.
External criticism asks whether a document is authentic — is this really a 15th-century manuscript, or a later forgery or corrupted copy? (Example: dating parchment and ink, tracing a manuscript's chain of custody.) Internal criticism, applied once authenticity is accepted, asks whether the document's content is reliable — was the author well-placed to know the facts, and what purpose or bias might shape their account? (Example: assessing whether a general's memoir exaggerates his own role.) Both stages are necessary; an authentic document can still be unreliable in content.
What is cliometrics, and what is one common critique of it?
Cliometrics is the application of quantitative, statistical and economic-modelling methods to historical questions — for example, using price and wage series to model economic growth, or regression analysis to test causal claims about historical events. A common critique is that it can impose modern economic categories and assumptions of rational-actor behaviour onto societies that did not think or organise their economies that way, and that statistical elegance can create a false sense of certainty about causally complex, data-poor historical situations.
What does Trouillot mean by "silences in the archive," and why does it matter for whose history gets written?
Trouillot argues that silence enters the historical record at multiple stages — which events get recorded as sources in the first place, which of those records are preserved in archives, which archived material historians choose to narrate, and which narratives become accepted as "what happened." Because archive formation has historically privileged literate, powerful and often colonial-administrative voices, whole perspectives (the enslaved, the colonised, women, the poor) are structurally underrepresented — not because nothing happened to or among them, but because the record-keeping apparatus was not built to capture their voices. Recovering their history requires deliberately reading existing sources against the grain and seeking out non-traditional evidence (material culture, oral tradition, indirect references).
Why is oral history methodologically both valuable and contested?
It is valuable because it can recover the perspectives of people who left few written records, giving access to lived experience, meaning and memory unavailable elsewhere. It is contested because human memory is reconstructive rather than a fixed recording — it can be reshaped by later events, by the narratives a community has since built around them, and by the dynamics of the interview itself (who is asking, why, and what they seem to want to hear). Rigorous oral historians treat testimony as evidence to be cross-examined and corroborated like any other source, not simply transcribed and trusted.
How the ideas connect
Every key idea in this chapter, branching from the core concept — use it to see the whole picture at a glance.
The key facts, visualised
Worked problems, step by step
Follow each solution line by line, then try to reproduce it on paper before moving on.
Example 1How do you triangulate quantitative and qualitative evidence on a social question?
- Gather statistics such as census or wage data
- Gather qualitative sources like letters or testimony
- Compare where they agree, differ and fill gaps
Example 2Explain why the silences in the archive matter to historians.
- Consider who created and kept records
- Consider whose voices were excluded
- Consider the effect on our picture of the past
Now you try
Work each one out first, then tap to reveal the worked answer.