A reliable history blog post depends on more than finding an interesting document. This source-checking checklist gives you a repeatable workflow for verifying primary and secondary sources, recording provenance, testing transcriptions, checking permissions, and producing citations that readers can inspect. Use it before publication, whenever you revise an older article, and again when a source link, archive record, or image-rights status changes.
Overview
Source checking is a quality-control process, not a final formatting task. It asks whether a source is what you say it is, whether your interpretation matches the evidence, and whether another reader can locate the same material.
Begin by separating the two jobs involved. Verification tests the source and the claim. Citation documents the path a reader can follow to the source. A polished citation cannot compensate for an uncertain provenance, a misleading quotation, or an unmarked editorial alteration.
For primary sources, check who created the item, when and where it was created, what form it takes, and how it reached the repository or publication in which you found it. For secondary sources, identify the author, publication context, edition or version, argument, and evidence base. A secondary source can be useful even when you disagree with its interpretation, but your article should make clear whether you are using it for a fact, a scholarly argument, a historiographical context, or a lead to further research.
Keep a source log while researching rather than reconstructing one after drafting. A simple spreadsheet or notes table is enough. Give each source an identifier, such as PS-01 or SS-03, and connect every important claim or quotation to that identifier. This creates an audit trail and makes later updates much faster.
What to track
1. Provenance and descriptive metadata
Record the information that establishes what the item is. For an archival document, this may include the creator, title or supplied description, date, collection, series, box or folder information where available, repository, and item identifier. For a digitized item, also record the catalog or finding-aid page rather than relying only on a viewer link.
For a book or article, capture the author, full title, publication, publisher or journal, date, edition, volume and issue where relevant, page range, and persistent identifier if one is provided. Do not silently fill gaps. Mark missing information as “not supplied,” “not located,” or “to verify.”
2. The source’s relationship to your claim
Write a one-sentence evidence note for each major source: “This item supports the date of the meeting,” or “This article provides a later interpretation of the policy.” This prevents a common error in history blogging—using a source that discusses a subject generally as if it directly proves a specific event.
Distinguish direct evidence from reported evidence. A letter may show what its writer claimed or believed; it does not automatically prove that every statement in the letter was accurate. A later historian may offer a carefully researched explanation; that explanation remains an interpretation rather than a firsthand record.
3. Exact quotations, transcriptions, and OCR
Compare every quotation against the page image or authoritative text. Preserve original spelling when it matters to the meaning or historical context, and use brackets or an editorial note when you expand, omit, or clarify text. If you modernize spelling, say so.
Optical character recognition can make a digitized document searchable, but it can also misread names, dates, punctuation, and columns. Treat OCR as a finding aid until you have checked the relevant passage against the scan. In your source log, note whether the quotation was checked against the image, a verified transcription, or OCR only. If a better scan becomes available, recheck quotations and update that note.
4. Stable access and retrieval details
Save the repository’s stable item page, catalog record, or persistent identifier. A temporary viewer URL may stop working when a session ends or an interface changes. Record the date you accessed the item, especially when the resource is likely to change or is presented through a dynamic database.
Download or save available metadata according to the repository’s terms. A local research copy helps you identify what you used, but it does not replace linking to the authoritative record. Keep a short note explaining any access limitation, such as an image that requires a reader account or a collection that is available only through a catalog description.
5. Rights and reuse information
Separate the right to read a source from the right to reproduce it. Before adding a scan, photograph, map, or excerpt to a post, check the repository’s rights statement and the item-level information. Record the stated license, rights holder, restrictions, attribution wording, and the date you checked. A public-domain statement, Creative Commons designation, or “rights unknown” notice should be quoted or linked accurately, not paraphrased into a broader permission claim.
When permission is unclear, use a link and descriptive caption rather than reproducing the image, or seek clarification from the rights holder or repository. This is a practical risk-control step, not a substitute for legal advice.
Cadence and checkpoints
Use the following workflow for a new post and as a maintenance routine for an existing one.
- At discovery: Create a source-log entry immediately. Record the repository or publication, item title, creator, date, identifier, stable URL, and why the source may matter.
- Before drafting: Classify the item as primary, secondary, or another type of evidence for your specific question. Write the claim it can support and note any limitations.
- During drafting: Attach source identifiers to quotations, dates, statistics, disputed interpretations, and claims that a reader may reasonably want to verify. Avoid relying on memory or an untraceable search-result snippet.
- Before publication: Open every important link, compare quotations with the original, inspect names and dates, check citations for completeness, and confirm that image-use information matches the item actually displayed.
- Monthly or quarterly: Review links and high-value sources in an update log. A monthly check suits frequently edited posts; a quarterly check may be sufficient for stable evergreen articles. Replace broken links with the repository’s current stable record, then document the replacement.
- Every six to twelve months: Review archive access notes, rights and licensing information, citation-style guidance, and the availability of better scans or transcriptions. The appropriate interval depends on how often the repository and your own site change.
Use this compact checklist at the end of your editorial notes:
- Provenance and creator verified.
- Date, title, collection, and item identifier recorded.
- Claim-to-source relationship written in plain language.
- Quotation checked against the original image or authoritative text.
- OCR or transcription limitations recorded.
- Stable URL or persistent identifier captured.
- Access date and relevant metadata saved.
- Rights statement and attribution requirements checked.
- Citation format applied consistently.
- Link, citation, caption, and image checked before publication.
How to interpret changes
A changed archive page does not necessarily mean the underlying evidence has changed. First determine whether the problem is only a new interface, a replaced viewer, or a corrected catalog record. Keep the original item identifier and update the access path. If the repository changes the description, compare the old and new metadata before revising your prose.
When a transcription changes, compare the revised text with the scan. A correction may improve accuracy, or it may reflect a different editorial convention. For a quotation that affects your argument, retain a note about which version you used and explain any material difference.
When a new secondary source appears, do not automatically rewrite the article around it. Ask what has changed: the factual record, the interpretation, the available bibliography, or simply the way the topic is presented. Update the post when the new work alters a central claim, supplies a stronger source, corrects an error, or gives readers important context. Otherwise, add it to a further-reading section or research note.
Broken links and rights changes deserve separate treatment. A broken link is an access problem; it is not proof that the source has disappeared. A changed rights statement affects reuse, not necessarily the historical content. Record each type of change separately so that editorial decisions remain clear.
When to revisit
Run the full checklist when you publish a new history article, add a newly discovered primary source, replace an image, or prepare a post for peer review or academic citation. Recheck the affected entries rather than assuming an old approval covers new material.
For older posts, schedule a monthly or quarterly link review and a broader source audit every six to twelve months. Prioritize articles that receive sustained traffic, contain many digitized images, depend on OCR, cover contested subjects, or link to repositories with changing interfaces. Also revisit an article when a reader identifies a citation problem, a repository corrects its metadata, or a style guide or site citation policy changes.
Finish each update by recording the date, editor, sources checked, links changed, rights information reviewed, and substantive revisions made. If nothing required changing, record that result too. This small maintenance note prevents repeated work and shows future editors what was actually verified.
For a practical publishing sequence, pair this checklist with the history article editing checklist, then use the guide to cite historical sources in blog posts for format examples. When an article is revised, the historical content maintenance workflow can help you document the wider update.
