Key takeaways
- A primary source is evidence created at the time by someone with direct involvement; a secondary source interprets primary evidence afterwards.
- The label is relative to the question: a 1950s textbook is a secondary source about the 1800s and a primary source about 1950s education.
- Primary does not mean reliable — participants have interests, misremember and lie, and being present is not the same as being right.
- Tertiary sources such as encyclopaedias summarise secondary work; they are good starting points and poor citations.
- The practical skill is asking who made this, when, why, and what they were in a position to know.
A primary source is evidence created at the time by someone with direct involvement. A secondary source interprets that evidence afterwards. Simple enough — and the useful part is what the definition does not say.
The categories
| Type | What it is | Examples |
|---|---|---|
| Primary | Created at the time, by a participant or observer | Letters, diaries, official records, photographs, artefacts, treaties, contemporaneous reporting |
| Secondary | Interprets primary evidence afterwards | History books, journal articles, documentaries, biographies |
| Tertiary | Summarises secondary work | Encyclopaedias, general reference works, overview textbooks |
Tertiary sources are for orientation. They are a good place to find out what the questions are and a poor place to end up citing, because they compress arguments into statements and drop the disagreement.
The label is relative to the question
This is the point that turns the distinction from a definition into a usable tool.
A source is not primary or secondary in itself. It is primary or secondary for a particular question.
- A 1950s history textbook is a secondary source about the nineteenth century.
- The same textbook is a primary source about how history was taught in the 1950s.
- A newspaper's account of a battle is a primary source about contemporary reporting and a secondary source about the battle.
So the first question is never "is this primary?" It is "primary about what?"
Primary does not mean reliable
The most common misuse of the distinction is treating "primary" as a synonym for "trustworthy". It is not. Proximity and accuracy are different properties.
Primary sources fail in specific, predictable ways:
Interest. A participant writing about their own decisions has reasons to present them favourably.
Memory. A memoir written decades later is a primary source about the writer's recollection, which is not the same as a primary source about the events.
Position. Being present does not mean being able to see everything. A soldier's account of a battle is authoritative about their part of it and not about the whole.
Institutional framing. Official records record what an institution chose to record, in the terms it chose. Absence from the record is not absence from history.
Survival. What survives is not a random sample. Durable materials, valued documents and the records of powerful institutions survive at higher rates.
That last one is structural and easy to forget: the evidence base is shaped by what lasted, not by what mattered.
A practical evaluation procedure
Four questions, applicable to any source:
1. Who made this? Their position, their access, their stake in the outcome.
2. When? At the time, or reconstructed later? Both are useful; they are useful for different things.
3. Why? A private letter, a public report and a legal deposition are produced for different audiences under different pressures.
4. What were they in a position to know? Present at the event, told about it, or inferring it?
The U.S. National Archives' document analysis worksheets formalise essentially this procedure for teaching, and they are a good template even for casual reading.
Where to find primary sources
More is freely available than most people assume. Two substantial, freely readable examples:
- The Foreign Relations of the United States series, the State Department's official documentary record of foreign policy.
- The National Archives' America's Founding Documents.
National archives, university digital collections and government publishing offices in many countries make comparable material available.
The experience of reading a document rather than a summary of it is genuinely different, and it is the fastest way to understand why historians disagree — the raw material is more ambiguous than any account of it.
Why this matters for a timeline
A timeline presents dates as points. Behind each point is a body of evidence of varying quality, and the presentation flattens that.
"1492" and "c. 3000 BCE" look equally definite on a screen and are not remotely equivalent claims. How historians date events covers where dates come from and how confidence varies, and BCE and CE explained covers the notation itself.
Where our app fits
World History Timeline Sim is an overview tool — closer to a tertiary source in the taxonomy above. That is a legitimate role: orientation, seeing how things relate in time, and finding out what questions exist.
It is not a substitute for reading evidence, and we would not claim it is. If a period interests you, the archives linked above are where the actual material lives.
The app is free to download on Google Play, supported by ads, with optional in-app purchases. The major eras of world history covers how periods get defined, and the rest of what we make in this area is under education and brain games.
Good to know
Frequently asked questions
What is a primary source?
What is a secondary source?
Can something be both?
Does primary mean more reliable?
What is a tertiary source?
Sources
- Document Analysis Worksheets (opens in a new tab)U.S. National Archives and Records Administration — accessed
- Foreign Relations of the United States (Historical Documents) (opens in a new tab)Office of the Historian, U.S. Department of State — accessed
- America's Founding Documents (opens in a new tab)U.S. National Archives and Records Administration — accessed