Think Like A Historian

Historical Thinking Is Mostly About Not Trusting Anything

The Stanford History Education Group published a framework called "Thinking Like a Historian" around 2016, and since then it has become the standard way history departments and AP courses teach students to process primary and secondary sources. The core idea is straightforward: historians don't just absorb information. They interrogate it through three lenses — sourcing, contextualization, and corroboration. Close reading gets folded in as a supporting skill. That's basically the entire model. I spent several years building curriculum around this framework for high school and undergraduate courses, and the thing nobody tells you is that most students can name the three pillars but they can't actually apply them under time pressure. They'll tell you "you need to consider the source" but then proceed to cite a tweet from 2023 as if it carries the same weight as a diplomatic memo from 1947. The framework doesn't fix that by itself. You have to drill the habit.

Think Like A Historian in Practice

Here's what the actual workflow looks like when you sit down with a document. Start with sourcing before you read a single line of the text. Who wrote it? What was their position, access, and possible motivation? When was it created relative to the events described? This takes roughly thirty seconds per document if you're disciplined, and it prevents about sixty percent of the errors students make when they treat every source as equally reliable. Contextualization comes next. You have to place the document inside its broader historical moment — not just the year it was written, but the political climate, the intellectual traditions, the material conditions, the audience it was meant for. A letter from a Union soldier in 1863 means something different depending on whether he was writing home to his wife or drafting a petition to a congressman. The words might be nearly identical. The meaning shifts entirely. Corroboration is where most people shortcut. You compare multiple sources on the same event and look for convergence and divergence. Where they agree, you can be more confident. Where they disagree, that's usually where the interesting historical question lives. I once had a student trying to write a paper on the 1918 influenza pandemic who only used medical journal articles and completely missed municipal health department records, newspapers, and personal diaries. The medical journals described the virus. The diaries described the panic. Neither told the whole story. She got a C+ and didn't understand why until I made her read two weeks of Kansas City newspaper archives from November 1918 side by side with the CDC's retrospective analysis.

Close reading ties everything together. You annotate. You look for word choice, rhetorical structure, what's left unsaid. Historians call this lateral reading sometimes, though that term has gotten muddied by digital literacy researchers. The point is you're not passively consuming text. You're reverse-engineering it. The framework has real limitations that people don't talk about enough. It works best with written documents. When you're working with material culture, oral histories, or archaeological evidence, the sourcing and contextualization steps still apply but the mechanics change significantly. A pot shard doesn't have an author in the same way a letter does. You have to infer provenance through stratigraphy and typology, which is a completely different skill set. The model doesn't account for that well. There's also the problem of source availability bias. The documents that survive tend to come from wealthy, literate, powerful people. Applying the framework mechanically to those sources can reinforce the very silences you're supposed to be aware of. I've seen teachers use this model and end up teaching students to "think like historians" while only ever looking at government records and newspapers. The framework gives you tools for critical reading, but it doesn't generate sources. You need archival work or deliberate source hunting to fill the gaps.

If you want to download or reference the original Stanford framework materials, they're available through the Stanford History Education Group's website at shg.stanford.edu. The Civic Reasoning section has the full lesson sets with source packets and rubrics. There are also versions adapted for different grade levels. The core PDFs are free. One thing that trips people up consistently: sourcing and contextualization are not the same thing. Sourcing asks who and when. Contextualization asks what was happening around that who and when. Students merge them constantly, which means they think they're doing rigorous work when they're really just restating the document's metadata. I had a graduate student once who wrote a three-page sourcing analysis of a Federalist Paper that was entirely accurate and entirely useless because she never pushed past the surface details into the actual political tensions of 1787-88. The framework requires you to go deeper. The rubric won't catch that. You have to. The model also assumes a level of archival access that most people don't have. In a classroom, you get curated document sets. In real historical research, you're often working with fragmented, damaged, or incomplete records where sourcing is genuinely uncertain and corroboration is impossible because the surviving evidence is thin. The framework was designed for pedagogy, not professional practice. It's a scaffold, not a complete methodology. Professional historians layer in additional approaches — cliometrics, microhistory, historiographical analysis, discourse analysis — depending on their subfield. Think Like a Historian is the foundation. It's not the building.