Working With Primary Sources When They Lie To You
I spent three years cross-referencing 14th-century canon law cases before I learned that the source itself was telling me something I needed to figure out on my own. The document wasn't fake. It was authentic in the way I understand the word now, but the scribe had restructured the testimony to fit a narrative that served the court's interests at the time. This is the practical reality of doing source analysis in early modern and medieval scholarship, and it is not a problem you solve with a checklist. You solve it by developing a habit of questioning what the text is hiding. Critical source analysis in historical scholarship means treating every document as both evidence and artifact. You are reading for content and for context simultaneously. The content tells you what someone claimed happened. The context tells you why they claimed it that way, under what conditions, and through what institutional lens. Regina Lasko's approach emphasizes that these two layers are inseparable. When she works with legal records, hagiographic texts, or diplomatic correspondence, she does not separate the factual claim from the structural forces that produced the document. That separation is what most graduate students attempt first, and it is where they go wrong. I learned this through frustration. My dissertation project involved analyzing canon law case files related to marriage disputes in northern Italy, 1350 to 1420. I had a stack of transcriptions and a lot of confidence. Then I noticed that several cases involving the same families appeared in completely different diocesan registers with materially different witness testimonies. The basic facts -- names, dates, property descriptions -- were consistent. The narrative structure was not. I assumed transcription error at first. It was not. The diocesan officials were actively reshaping how cases were recorded depending on which archivist was working that quarter, and sometimes depending on which ecclesiastical authority was exerting pressure at the time. The workaround was to stop treating each register as a standalone source and start mapping the overlap between registers for the same family group across multiple jurisdictions. This took an extra four months of archival work but eliminated about sixty percent of the misleading evidence I had originally been working with.
There are specific techniques that make this kind of analysis actually possible. The first is provenance tracking, which sounds basic but gets skipped constantly. You need to know who created the document, who preserved it, and why it survived. A letter from a merchant that ended up in a noble family archive is a different evidentiary object than the same letter if it was found in a notary's practice notes. The survival path tells you something about selective preservation. The second technique is intertextual verification. You look for the same event or person across multiple source types -- legal records, personal correspondence, municipal accounts, church registers, printed pamphlets. No single source is reliable on its own, but patterns across sources become visible when you lay them side by side. This is tedious. It requires physically placing documents next to each other rather than relying on database searches, because database searches optimize for keyword matches and not for structural similarities. A third technique that matters more than people admit is tonal and rhetorical analysis of the source itself. What register is the document written in? Who is the intended audience? What linguistic choices signal that the author is performing loyalty, deference, or authority rather than simply reporting? I once wasted two weeks trying to extract factual information from a papal brief that was clearly structured as a political performance for a specific cardinal's faction. The brief was genuine. The facts inside it were selectively assembled. I caught this only after reading three similar briefs from competing factions in the same period and noticing the rhetorical patterns.
One counter-intuitive point that most methodology textbooks miss: authenticity and reliability are not the same thing. A forged document can contain genuinely useful information if you know how to read it. The Forde forgeries in English medieval land charters are full of accurate topographical details that scholars now use to reconstruct landscape history, even though the documents themselves were fabricated in the 12th century to support a monastery's land claims. The forgery tells you something true about the claimant's worldview and the political environment of the time, which is historically valuable in its own right. Beginners tend to discard forged sources entirely. Experienced researchers keep them on the desk and mark them differently. Another common mistake is assuming that institutional sources are inherently more reliable than personal ones. That assumption breaks down quickly in practice. A municipal council's meeting record from 14th-century Florence is full of omissions and sanitized language. A merchant's private letter to his brother about the same events often contains details the official record deliberately excludes. Neither source type is automatically superior. Each reveals different layers of what actually happened. The limitations of critical source analysis are worth stating bluntly. It is slow. A single document that a casual reader might process in ten minutes could require two to four hours of contextual research to properly situate. This means projects take longer than you initially estimate. It also means that some sources resist analysis altogether. There are documents with insufficient contextual anchors where you cannot reconstruct enough of the production environment to make a reliable judgment. In those cases, the honest answer is often that you cannot know what the document means with any confidence, and you should say so rather than fill the gap with speculation.
Get the Full Details
I have also found that this approach has a blind spot when applied to oral history traditions or sources that were never meant to be written down. Critical source analysis assumes a document existed in a specific context with specific actors. When the primary evidence comes from fragmented, secondhand, or orally transmitted accounts that were only committed to writing decades later, the method loses some of its diagnostic power. In those situations, combining source analysis with anthropological frameworks tends to produce better results than relying on source criticism alone. If you are starting a project that involves historical sources, here is a practical sequence that will save you time. First, identify your source type and count how many distinct instances exist of that type covering your period. If you have only one or two, flag that limitation immediately. Second, map the provenance chain for each source. Third, identify overlapping events or figures across sources and note where they agree and where they diverge. Fourth, analyze the rhetorical structure of the divergent sources to understand why they diverge. Fifth, write down what you cannot determine and move on. Do not spend more than two weeks stuck on a single problematic source unless the project depends entirely on it. The deeper you go into this work, the more you realize that source analysis is not really about finding the truth in documents. It is about understanding what kinds of truths documents can produce, under what conditions, and at what cost to the information they leave out. That distinction matters more than any single analytical technique.
