History & Social Studies: Primary source interrogation

What you'll be able to do: Question a document about who made it and what it isn't saying, which is what turns a source into evidence rather than a quotation.

The mistake that produces quote-as-proof essays

Students read sources for content: what does it say?

That's the least interesting question available, and it produces the essay that quotes a document as though quoting settled something. A source saying X is not evidence that X was true. It's evidence that someone, in a particular position, writing for a particular audience, at a particular moment, said X, and every one of those clauses is doing work.

A source is not a window onto the past. It's an artefact, made by someone, for someone, for a purpose, under constraints. Reading through it to the situation that produced it is the historian's skill, and the questions that do it are standard and learnable.

The seven questions

1. Who made this, and what could they see from where they stood? Position determines access as much as bias does.
2. For whom? A private letter, an official report and a public speech about the same event are three different kinds of document, not three tones.
3. Why now? Documents are usually responses to something.
4. What does it take for granted? The assumed knowledge is the richest seam. What the author didn't think needed explaining reveals the world they were in.
5. What is it arguing against? Most documents are implicitly in dispute. Finding the absent opponent is frequently the key move.
6. What's absent? What would you expect to be here that isn't.
7. How did it survive? This one is rarely taught and reframes everything: survival isn't random. Whose documents get kept, and whose were never written, because they couldn't write, shapes what history can be written at all.

Question 7 turns the archive itself into a historical artefact, which is where the interesting methodological questions live.

How to run it

Read it yourself first, without commentary, and form a view. Then:

Here's a primary source [paste]. Don't summarise it. Ask me the seven interrogation questions one at a time, author's position, audience, occasion, assumed knowledge, what it argues against, what's absent, how it survived. Wait for my answer to each, then tell me what I missed.

And the verification clause, which is not optional in this subject:

Everything you've told me about this document, is it from the text I gave you, or from what you know about documents of this type? Mark which.

The failure mode here is a confident characterisation of a document the tool hasn't read. The source is the authority.

The richest question

What does this document assume its reader already knows? List everything it doesn't bother to explain, that's the world it was written into.

A parish register assumes you know what a parish is for. A 1930s memo assumes you know which department outranks which. A pamphlet assumes you know the argument it's answering. None of that is stated, all of it is recoverable, and it's where the texture of a period actually lives.

Across the subjects

History. Works on anything, a will, a ledger, a photograph, a menu, a timetable. Ledgers and inventories are particularly good precisely because they're dull: nobody was performing for posterity.

Political science. Official documents interrogated for audience and purpose. A white paper is an argument, not a description.

Sociology. Official statistics interrogated for who collected them, how, and to what end. Categories on a census form are political artefacts.

Anthropology. Field notes read for the observer's position, methodologically central to the discipline.

Geography. Maps as arguments. Projection, inclusion, naming, what's left blank. A map is the clearest case of a source that looks neutral and isn't.

International relations. Diplomatic documents, where the absent opponent question is almost always productive.

Pitfalls

1. Reading for content only. The failure this method exists to fix.
2. Summarising. Destroys every one of the seven questions.
3. Accepting characterisations you haven't verified. Check against the text.
4. Assuming bias means worthless. Every source has a position; that's what makes it evidence of something. A partisan pamphlet is excellent evidence about the partisan.
5. Over-reading absence. Something missing may be missing for dull reasons. Ask what else would explain it.
6. The tell: you can say what the document says and not what it's evidence for. Those are different, and only the second is history.

Try this today

Take any source from your course, or any document at all, including a modern one.

Ask only question four: what does this assume its reader already knows? List everything it doesn't explain.

That list is the world it was written into, and it's usually more interesting than the content.