Primary and secondary source analysis for historical reasoning. Covers source classification, sourcing (author/context/purpose), corroboration across multiple sources, contextualization within time and place, and bias detection. Use when evaluating historical evidence, assessing source reliability, or constructing evidence-based historical arguments.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
Primary and secondary source analysis for historical reasoning. Covers source classification, sourcing (author/context/purpose), corroboration across multiple sources, contextualization within time and place, and bias detection. Use when evaluating historical evidence, assessing source reliability, or constructing evidence-based historical arguments.
type
skill
category
history
status
stable
origin
tibsfox
modified
false
first_seen
"2026-04-12T00:00:00.000Z"
first_path
examples/skills/history/source-analysis/SKILL.md
superseded_by
null
Source Analysis
History is not the past itself but the disciplined reconstruction of the past from surviving evidence. Source analysis is the foundational practice through which historians transform raw documents, artifacts, images, and oral testimonies into usable evidence. Without rigorous source analysis, historical claims rest on authority rather than evidence, and narrative replaces inquiry.
This skill covers four interconnected practices: classifying sources as primary or secondary, sourcing (interrogating authorship and context), corroborating claims across independent sources, and contextualizing evidence within its historical moment. Together these practices form the evidential backbone of all historical work.
Agent affinity: herodotus (chair, source criticism and historiographical method)
Convergence of unrelated witnesses strengthens claims
4
Contextualization
What was happening when and where this was created?
Political, social, economic, cultural backdrop
5
Bias detection
What perspectives are present or absent?
Silences, framing choices, selective emphasis
Practice 1 โ Primary vs. Secondary Sources
Definition. A primary source is a document, artifact, or record created during or close to the period under study, by a participant or direct witness. A secondary source interprets, analyzes, or synthesizes primary sources from a later vantage point.
The classification is relational, not absolute. The same document can be primary for one inquiry and secondary for another. A 1920s history textbook is secondary for the events it describes but primary for studying how history was taught in the 1920s. Classification depends on the research question being asked.
Categories of Primary Sources
Type
Examples
Strengths
Limitations
Official records
Treaties, laws, census data, court records
Institutional authority, systematic
Reflect state perspective, may omit dissent
Personal accounts
Diaries, letters, memoirs, oral histories
Individual perspective, emotional texture
Memory distortion, self-justification
Media
Newspapers, pamphlets, broadcasts, photographs
Contemporary framing, wide circulation
Editorial bias, commercial pressures
Material culture
Tools, clothing, architecture, art
Non-textual evidence, reveals daily life
Requires archaeological interpretation
Statistical records
Trade ledgers, tax rolls, parish registers
Quantifiable patterns, demographic data
Recording biases, incomplete survival
Categories of Secondary Sources
Type
Examples
Strengths
Limitations
Monographs
Academic books on specific topics
Deep analysis, peer-reviewed
Author's interpretive framework
Journal articles
Published in historical journals
Narrow focus, current scholarship
May assume specialized knowledge
Textbooks
Survey and introductory texts
Broad synthesis, accessible
Oversimplification, lag behind research
Documentaries
Film and video treatments
Visual engagement, wide audience
Narrative compression, dramatic license
The Spectrum Problem
The primary/secondary distinction is not a clean binary. Some sources occupy intermediate positions:
A memoir written decades after events is primary (authored by participant) but carries secondary-source problems (retrospective interpretation, hindsight bias).
A newspaper editorial from 1863 is primary for Civil War-era opinion but secondary for the battlefield events it describes from reports.
An archaeological survey report is secondary (interpretation of material evidence) but becomes primary for studying archaeological methods of its era.
Practical rule. Always state explicitly what question you are using the source to answer. The classification follows from the question, not from the source alone.
Practice 2 โ Sourcing
Sourcing is the systematic interrogation of a document's origin. Before using any source as evidence, a historian must establish who created it, when, where, for what purpose, and for what audience. Sourcing is not optional โ it is the prerequisite for all subsequent analysis.
The Five Sourcing Questions
Who created this source? What is the author's identity, social position, institutional role, and relationship to the events described? An enslaved person's account of plantation life carries different evidentiary weight than a slaveholder's inventory.
When was it created? Proximity to the event matters. A letter written the day of a battle differs from a memoir written thirty years later. Both are useful, but for different purposes.
Where was it created? Geographic proximity shapes what the author could have known. A London newspaper reporting on colonial events relied on delayed dispatches and intermediary accounts.
Why was it created? Every source has a purpose. A diplomatic dispatch aims to inform a government. A propaganda poster aims to mobilize a population. A diary entry may aim at self-reflection. Purpose shapes content.
Who was the intended audience? A private letter to a confidant may be more candid than a public speech. A petition addressed to a monarch follows formal conventions that constrain what can be said.
Worked Example โ Sourcing a Document
Source: Letter from Abigail Adams to John Adams, March 31, 1776.
Who: Abigail Adams, wife of a Continental Congress delegate, educated, politically engaged, managing family farm during his absence.
When: March 1776, months before the Declaration of Independence, during active debate about independence and new government.
Where: Braintree, Massachusetts โ away from Philadelphia where Congress met, relying on correspondence for political news.
Why: Personal correspondence maintaining a marital and intellectual partnership; also a deliberate intervention in political thought ("Remember the Ladies").
Audience: John Adams personally, though Abigail likely knew letters might circulate among trusted correspondents.
Sourcing conclusion. The letter is valuable as evidence of women's political consciousness during the Revolution and of the domestic-political nexus in elite colonial families. Its private nature suggests candor, but its rhetorical sophistication suggests awareness of broader stakes.
Practice 3 โ Corroboration
Corroboration is the practice of checking claims across multiple independent sources. No single source, however compelling, establishes historical fact on its own. Convergence of independent accounts strengthens a claim; divergence demands investigation.
Principles of Corroboration
Independence matters more than quantity. Three accounts derived from the same original report provide the strength of one source, not three. A newspaper article, a private diary, and an official report that each independently describe the same event provide genuine triangulation.
Disagreement is information, not failure. When sources conflict, the disagreement itself is evidence. It may reveal different vantage points, different stakes, or different biases โ all of which illuminate the event.
Silence is evidence. When a source that should mention an event does not, the silence requires explanation. The absence of women's voices in most political archives before the 19th century does not mean women were politically inactive โ it means the archival system was not designed to capture their activity.
Corroboration Matrix
When working with multiple sources on a single event, construct a corroboration matrix:
Claim
Source A
Source B
Source C
Assessment
Event occurred on date X
Confirms
Confirms
Silent
Strong (2 independent confirmations)
Crowd numbered 10,000
States 10,000
States 3,000
States "large"
Disputed โ investigate author positions
Leader gave specific speech
Quotes speech
Mentions speech
No mention
Moderate โ one direct, one indirect
Worked Example โ Corroborating Across Sources
Question: Did the Boston Massacre of 1770 involve unprovoked firing on civilians?
Source A (Paul Revere's engraving): Depicts disciplined soldiers firing in formation on a passive crowd. Created for propaganda purposes.
Source B (Captain Preston's deposition): Claims soldiers acted in self-defense after being attacked by the mob. Created for legal defense.
Source C (Anonymous account in the Boston Gazette): Describes soldiers as aggressors. Published in a patriot-sympathizing newspaper.
Source D (Trial testimony from witnesses): Mixed accounts; some witnesses describe a chaotic scene with provocation from both sides.
Corroboration assessment. No single account is reliable alone. Revere's image is explicitly propagandistic. Preston's deposition is self-serving. The Gazette had editorial bias. Trial testimony, being sworn and cross-examined, provides the most constrained accounts, and these suggest a more complex event than any partisan source admits. The historian's reconstruction must account for all sources while weighting them by independence and purpose.
Practice 4 โ Contextualization
Contextualization is the practice of situating a source and its claims within the broader historical conditions of its creation. A document cannot be understood apart from the world that produced it.
Dimensions of Context
Political context. What government structures, power relations, conflicts, or reforms were active? A petition to Parliament in 1832 means something different than the same petition in 1830 โ the Reform Act changed what was politically possible.
Social context. What social hierarchies, class structures, gender norms, or racial systems shaped the author's world? An enslaved person's narrative published by an abolitionist society was shaped by both the author's experience and the publisher's political goals.
Economic context. What economic conditions, trade patterns, labor systems, or resource pressures were relevant? The Irish Famine dispatches cannot be understood without knowing British free-trade ideology and landlord-tenant relations.
Cultural context. What religious beliefs, intellectual movements, artistic traditions, or shared assumptions were current? A medieval chronicle attributing a plague to divine punishment reflects sincere belief, not ignorance โ within its cultural context, supernatural causation was the dominant explanatory framework.
Technological context. What communication, transportation, production, or military technologies were available? The speed of information transmission fundamentally shapes what actors could know and when.
Worked Example โ Contextualizing a Source
Source: Woodblock print depicting Commodore Perry's arrival in Japan, 1853. Japanese artist, published in Edo.
Without context: An image of foreign ships arriving.
With context: Japan had maintained sakoku (closed country) policy for over 200 years. The Tokugawa shogunate controlled all foreign contact. Perry's "Black Ships" were steam-powered warships โ a technology Japan had never encountered. The print circulated in a society where information about the outside world was tightly controlled. The artistic conventions (ukiyo-e style) domesticated the foreign into familiar visual language. The print is evidence not just of the event but of how Japanese visual culture processed an unprecedented intrusion.
Practice 5 โ Bias Detection
Bias is not a flaw to be eliminated but a feature to be understood. Every source reflects a perspective, and every perspective includes and excludes. The historian's task is not to find "unbiased" sources โ they do not exist โ but to identify, characterize, and account for the biases present.
Types of Bias
Selection bias. What the author chose to include and exclude. A war correspondent who only interviews officers produces a systematically different account than one who interviews enlisted soldiers.
Confirmation bias. The tendency to interpret evidence in ways that confirm existing beliefs. A colonial administrator who believed in the "civilizing mission" would interpret resistance as proof of "savagery" rather than as rational political action.
Survivor bias. Sources that survive are not representative of all sources that existed. The literary records of the ancient world disproportionately represent elite male perspectives because those were the texts copied and preserved by institutional libraries.
Hindsight bias. Memoirs and retrospective accounts reshape the past in light of known outcomes. A general who lost a battle may, in his memoirs, emphasize the factors beyond his control.
Structural bias. Archives themselves are not neutral. Colonial archives preserve the colonizer's records. State archives preserve the state's perspective. The historian must actively seek sources from outside these institutional frames โ oral histories, material culture, community archives โ to reconstruct silenced perspectives.
The Bias Analysis Protocol
For any source under analysis:
Identify the author's position relative to the events and to power structures.
Identify the purpose โ what the source was meant to accomplish.
Identify what is absent โ whose voices, what perspectives, which facts are missing.
Compare the source's framing with other sources on the same events.
Assess whether the bias strengthens or weakens the source for your specific question. A propagandist's claims about events are unreliable, but the same document is excellent evidence of what the propagandist wanted people to believe.
Source Analysis Decision Tree
When encountering a new source:
Classify it. Primary or secondary for your research question?
Source it. Who, when, where, why, for whom?
Contextualize it. What was the world like when and where this was created?
Detect bias. What perspective does this represent? What is missing?
Corroborate it. Do independent sources support, contradict, or complicate the claims?
Assess its evidentiary value. What can this source reliably tell you? What can it not?
Common Mistakes in Source Analysis
Mistake
Why it fails
Fix
Treating primary sources as automatically trustworthy
Primary means close to events, not accurate
Always source and corroborate
Treating secondary sources as automatically inferior
Good secondary work synthesizes vast primary evidence
Evaluate the scholarship's methods and evidence base
Ignoring the author's purpose
Purpose shapes content systematically
Always ask "why was this created?"
Assuming bias invalidates a source
Every source is biased; bias is information
Characterize the bias and use the source accordingly
Reading a source in isolation
Single sources cannot establish historical claims
Always corroborate across independent sources
Imposing modern categories on past sources
Presentism distorts historical meaning
Contextualize within the source's own world
Treating absence of evidence as evidence of absence
Archives are incomplete by nature
Consider what sources may not have survived
Cross-References
herodotus agent: Department chair and primary source criticism specialist. Guides all source analysis workflows.
ibn-khaldun agent: Social and economic history analysis. Uses source analysis to reconstruct material conditions.
arendt agent: Political and modern history. Applies source criticism to 20th-century political documents.
causation-consequence skill: Source analysis feeds causal reasoning โ reliable evidence is prerequisite to causal claims.
historical-perspectives skill: Bias detection connects directly to understanding multiple perspectives.
historiography skill: Different historiographical schools prioritize different source types and analytical methods.
References
Bloch, M. (1953). The Historian's Craft. Translated by Peter Putnam. Vintage Books.
Collingwood, R. G. (1946). The Idea of History. Oxford University Press.
Tosh, J. (2015). The Pursuit of History. 6th edition. Routledge.
Wineburg, S. (2001). Historical Thinking and Other Unnatural Acts. Temple University Press.
Trouillot, M.-R. (1995). Silencing the Past: Power and the Production of History. Beacon Press.
Ginzburg, C. (1980). The Cheese and the Worms. Johns Hopkins University Press.
Herodotus. (c. 440 BCE). The Histories. Translated by Tom Holland (2013). Penguin Classics.
