Sources and Evidence: How Historians Evaluate the Past

History is not simply a list of facts recovered from old documents. Historians build accounts of the past from surviving evidence, and every surviving item was created by a person or institution with a particular purpose, audience, and point of view. Learning to read that evidence carefully is one of the central skills of historical research.

Primary, secondary, and tertiary sources

A primary source was created during the period being studied or by someone directly connected to it. Letters, laws, diaries, photographs, maps, newspapers, oral histories, buildings, artifacts, financial records, and government files can all serve as primary sources. The label describes the source’s relationship to a research question, not its reliability. A memoir written decades after an event may be primary evidence for how the author remembered the event, even when it is unreliable about exact details.

A secondary source analyzes or interprets the past after examining primary sources and other scholarship. Scholarly books and articles are common examples. Secondary sources help readers understand the existing debate, learn how evidence has been interpreted, and identify important primary collections.

A tertiary source summarizes established knowledge from primary and secondary work. Encyclopedias, timelines, textbooks, and reference databases are often tertiary sources. They are useful starting points, but a serious investigation normally follows their citations to the underlying evidence and scholarship.

Source type	Typical examples	What it can help answer
Primary	Letters, artifacts, photographs, court records, oral testimony	What did people record, experience, create, or claim at the time?
Secondary	Scholarly books, journal articles, historical essays	How have specialists interpreted the evidence?
Tertiary	Encyclopedias, handbooks, general timelines	What is the broad outline, and where should research begin?

Start with provenance and authenticity

Before interpreting a source, historians ask where it came from. Provenance is the documented history of an object or record: who created it, how it was preserved, and how it reached its present location. A source without reliable provenance may still contain useful information, but uncertainty about its origin limits what it can prove.

Authenticity is not the same as accuracy. A document can be genuinely old and still contain lies, rumors, propaganda, mistakes, or selective memories. Conversely, a later copy of a lost original may preserve valuable text even though the physical copy was not created at the time described.

Useful opening questions include:

• Who created the source, and when?
• Is this the original, a copy, a translation, an excerpt, or a later edition?
• Has any part been altered, removed, or taken out of sequence?
• Where has the source been held, and is that chain of custody documented?
• Do its materials, language, and format fit the claimed date and origin?

Read for purpose, audience, and context

Sources were rarely created to answer a future historian’s questions. A royal decree asserted authority. A merchant’s ledger tracked transactions. A newspaper sought readers and operated within political and commercial constraints. A private letter might be candid, guarded, performative, or written with the expectation that others would eventually read it.

That is why historians place evidence in context. They investigate the circumstances of creation, the creator’s position, the intended audience, and the conventions governing that kind of source. Calling a source “biased” is only the beginning. The more productive question is how its perspective shaped what it recorded, emphasized, omitted, or assumed.

Corroborate instead of trusting a single account

One source can establish that a claim was made. It often cannot establish that the claim was true. Historians compare independent evidence to find agreement, contradiction, and meaningful silence. This process is called corroboration.

The strongest comparison does not merely count sources. Ten reports copied from the same original rumor are not ten independent confirmations. Researchers trace relationships among accounts, look for evidence produced for different purposes, and consider whether the sources had access to the information they reported.

A reconstruction of a battle, for example, might combine official reports, soldiers’ letters, maps, archaeological evidence, casualty records, weather observations, and accounts from opposing forces. These sources answer different questions and carry different limitations. Their points of overlap can support a conclusion; their conflicts can reveal uncertainty or competing interests.

Notice what the archive does not preserve

Archives are shaped by power, institutions, accidents, collecting practices, and survival. Governments and wealthy organizations often leave extensive written records. People excluded from formal institutions may appear mainly through records created about them by officials, courts, employers, missionaries, or colonizers. Other experiences survive through oral tradition, material culture, archaeology, music, or community memory.

This problem is sometimes described as archival silence. Silence does not prove that something did or did not happen. It tells historians to examine why evidence is absent, whose voices were preserved, and whether other kinds of sources can address the gap.

Red flags that require more investigation

• The item has no identifiable creator, date, repository, or chain of custody.
• A dramatic claim appears only in a much later account.
• Several accounts use nearly identical language and may share one source.
• A quotation circulates without a reference to the complete document.
• A translation removes disputed wording or does not identify its edition.
• A source is presented without its intended audience or political setting.
• The conclusion goes beyond what the evidence can actually establish.
• Evidence that complicates the preferred interpretation is ignored.

None of these automatically makes a source worthless. They change the questions that can responsibly be asked of it and the confidence attached to the answer.

From source to historical claim

Historical arguments should distinguish among observation, inference, and interpretation. “This letter contains these words” is an observation. “The writer privately believed them” is an inference that depends on authorship, audience, and context. “The letter represents public opinion” is a much broader interpretation requiring additional evidence.

Good historical writing makes that chain visible. It cites evidence, acknowledges uncertainty, explains why particular sources are persuasive, and avoids presenting a disputed inference as an uncontested fact.

How this applies to Daily History AI

Daily History AI event records include source metadata so readers can inspect the basis for a summary and continue their own research. The presence of a citation is a starting point rather than a guarantee. Sources still differ in proximity, expertise, independence, and evidentiary value.

The site’s confidence and review fields help identify claims needing extra attention, while summaries distinguish the event description from the explanation of why it matters. For the full publication workflow, see Sources and Fact Checking and the Editorial Policy.

A practical source-evaluation checklist

When reading a historical claim, ask:

1. What exactly is being claimed?
2. What source supports it?
3. Who created that source, when, and for whom?
4. Is the source authentic, complete, and presented in context?
5. Was the creator in a position to know?
6. What interests or conventions shaped the account?
7. Do independent sources support or contradict it?
8. What evidence may be missing?
9. Does the conclusion match the strength of the evidence?
10. What new evidence would change the conclusion?

Thinking through these questions does not eliminate disagreement. It makes disagreement more precise, transparent, and productive.