What Primary Source Analysis Actually Looks Like When You're Doing It
I spent three weeks last semester trying to get my students to stop treating primary sources like confirmation machines. They'd pull a quote from a diary, a government report, or a newspaper editorial, then rearrange the rest of the document to fit the quote. It's a habit that's hard to break because it's efficient in the wrong way. You get an answer quickly, but the answer is usually shallow or wrong. The method I settled on was to make them interrogate the source before they used it at all. That means working through a set of Primary Source Analysis Questions that forces them to treat the document as an artifact first, evidence second. The sequence matters. Too many people start with content and skip the frame.
How to Build Your Primary Source Analysis Questions
Start with the basics that take thirty seconds to answer but save you two hours of rewriting. Who created this? What was their role, formal and informal? When was it produced relative to the event being described, and in what physical form does it survive today? Why was it made, and who was the intended reader or audience? From there move into the harder questions. What assumptions is the author making about the reader's knowledge? What choices were made about what to include and what to leave out, and who benefits from those choices? How does the medium shape what can be said—diplomatic correspondence versus a personal letter versus a public pamphlet will say completely different things about the same moment. What genre conventions is the document following or breaking? Then push into corroboration and context. Which other documents support this account and which ones actively contradict it? What broader political, economic, or social conditions would make this document plausible to its original audience? Are there gaps in the archive that might explain why certain voices are missing here?
That sequence isn't sacred. Sometimes you start with genre and hit the author question as you go. I usually have people run through the whole thing once on a quick pass, then circle back to the corroboration questions after they've drafted an interpretation. The second pass catches things the first pass missed because the student now knows what claim they're building toward. I ran into a specific problem last fall with a collection of mid-nineteenth century factory inspection reports. The documents looked straightforward on the surface. Government inspectors writing about working conditions. But the provenance was messy. Several of the reports had been excerpted and reprinted in labor reform pamphlets twenty years after they were written, and the excerpts were cut to serve an argument the reformers were already making. When my students treated the pamphlet versions as the primary source, they missed entire paragraphs where the inspectors admitted uncertainty or noted contradictory testimony from mill owners. The workaround was simple but tedious. I made them trace every quote back to the original published volume in the state archives, note every passage that had been dropped, and record how the omission changed the meaning. It took longer. It also meant they couldn't use the pamphlet as evidence at all for their final paper, which was the right call. One thing beginners consistently get wrong is treating silence as absence. Just because a document doesn't mention something doesn't mean it didn't happen or wasn't known. A colonial administrator's dispatch might not mention food shortages because reporting priorities in the bureaucratic chain demanded attention to trade volume and tax revenue, not peasant subsistence. The absence is itself data. You have to ask what the genre and the reporting structure would have filtered out before you decide the source is incomplete.
Get the Full Details
Another counter-intuitive point is that contemporary sources are not automatically more useful than later ones. Memoirs written decades after an event often have better contextual awareness, access to archives the original author couldn't reach, and the benefit of seeing which details actually mattered. The trade-off is memory decay and retrospective framing. The trick is to use later sources for context and earlier sources for contemporaneous reaction, then check where they diverge. That divergence is usually where the interesting history lives. Here are the questions in a form you can copy into a worksheet or adapt for your own materials. Creator and perspective: Who produced this document? What was their position, authority, and possible bias? Did they have access to the information they present?
Origin and timing: When and where was it created? How close in time and space is it to the event it describes? Purpose and audience: Why was it created? Who was it meant for? How would the intended audience have received it? Content and argument: What does the document explicitly claim? What is it trying to accomplish—inform, persuade, record, defend, advertise?
Omissions and framing: What is left out? What choices shaped what appears here? Whose voice is centered and whose is absent? Medium and genre: What form is this in? How does the medium constrain or enable what can be said? Context: What historical conditions made this document possible? What larger events or structures is it responding to?
Corroboration: What other sources confirm or challenge this account? Where do they agree and where do they conflict? Significance: What does this document reveal that other kinds of sources might not? What questions does it raise that remain unanswered? A few practical notes about using these in a classroom or independent research setting. Keep the initial read short. Let students skim the document first and answer the creator, origin, and purpose questions before they dive into close reading. That first layer prevents them from getting lost in details that turn out to be marginal. Also keep a running log of which sources corroborate and which contradict your interpretation. The log becomes your citation map when you're writing the actual analysis, and it saves you from having to reconstruct your reasoning from scratch.
The main limitation of this approach is time. A thorough pass through all of these questions on a single dense document can take forty to sixty minutes for someone who is still learning to read archaic handwriting or unfamiliar administrative language. If you have a large document set, you'll need to be selective about which questions you apply deeply and which you treat as a quick check. I usually reserve the full set for the documents that will carry the heaviest weight in a paper and compress the rest to the first five questions. Another limitation is that the framework assumes the document survives intact enough to interrogate. Fragmentary sources, damaged manuscripts, or records destroyed in archival loss events force you to work with less information than the questions expect. In those cases you shift from asking what the document proves to asking what the gaps themselves indicate about power, censorship, or disaster. That's a different analytical move and it requires you to be explicit about the uncertainty rather than filling it in. If you need a ready-to-use template, most university library research guides host downloadable versions of these question sets. The Library of Congress has a primary source analysis worksheet that aligns closely with this sequence, and several state archives publish similar forms tailored to their collections. I use a modified version I built from the LOC template, stripped down to the questions that matter for my course level and padded with a corroboration section that the original left thin.
The bottom line is that these questions work because they slow you down at the point where you're most tempted to rush. They catch the confirmation bias before it takes root. They also expose the places where your source simply cannot answer the question you think it can, which is arguably more valuable than any answer it provides.
