Begin
Step-by-step instructions article — coming soon.
FANlines
Every experienced researcher knows the instruction: don’t just find your ancestor in the census — read the whole page, and the pages on either side of it. The enumerator walked a route, and the households recorded around your subject were, in most cases, the people who lived around your subject: the neighbors who witnessed their deeds, administered their estates, married their children, and moved west alongside them. Reading those surrounding households across a single census is straightforward. Tracking a dozen families across five or six censuses is a correlation problem that swallows research hours without adding judgment.
FANlines does that clerical work so you can spend your time on the judgment. You paste the table view of each census page from your subscription record site, decade by decade. FANlines parses each page, segments it into households, and follows your research subject’s FAN network — Family, Associates, and Neighbors — across the enumeration windows you supply. You confirm every cross-decade family link yourself, against the evidence, exactly as you would in a proof argument. The result is a standalone report of research findings & leads you can save or print. For a full walkthrough of the tool and how to read its report, see the FANlines Research Guide.
What FANlines delivers
- A cross-decade correlation grid — your subject family’s recorded attributes laid out decades-across and attributes-down, in the manner of a proof argument. Consistent progressions read clearly; expected changes such as a head succession or a migration are labeled; genuine conflicts, like a birthplace that disagrees with three other decades, are flagged for resolution against the images.
- Neighbor stability — the confirmed family groups near your subject, ranked by how persistently they appear and at what proximity in each decade, with flags for the strongest FAN signals: families present on both sides of your subject’s move, or families that reappear in a new location, consistent with chain migration.
- Household evolution — a decade-by-decade account of how the subject household changed: who departed, who arrived, where the head changed hands, and derived facts such as the child deaths implied by a mother’s children-born versus children-living figures.
- Surname recurrence — which surnames turn up across your pasted window and in how many decades, including the spelling variants an index scattered, with an option to catch first-letter transcription errors like Hanks and Lanks.
- Maiden-surname and daughter-marriage hypotheses — candidate origins for the wives in your family, drawn from nearby different-surname households in earlier decades, and candidate married surnames for daughters who vanish from the household, drawn from later ones. Each is ranked for consideration for further analysis.
- Recommended next steps and negative findings — the strongest patterns translated into record-specific research suggestions, plus a documented record of what was searched and not found, which is a step toward a reasonably exhaustive search.
Everything is delivered as a self-contained HTML report — data inventory, citations, the analyses you select, and a full methodology statement — that you can open, share, or print to PDF. FANlines is free, runs entirely in your browser, no accounts, and uploads nothing.
1 · Start
2 · Enter census pages
3 · Link households
4 · Analysis & report
Your data never leaves your browser. Pages you paste are processed entirely on your own computer; nothing is uploaded anywhere. Your research lives in this browser tab, an optional local autosave, and the JSON file you download — which is yours.
Page metadata
Only the census year is required — everything else improves your citations. Paste, per decade, at minimum the page before the subject’s page, the subject’s page, and the page after: the enumerator’s walking path means the adjacent pages hold the nearest neighbors. Pasting more pages widens the net.
Paste the census page
Copy the spreadsheet-style table view of the census page on your record site — start the selection at the very top, including the column titles — then click below and paste.
Click here, then paste (⌘V)…
Surnames-of-interest watchlist (optional)
Surnames from your prior research. The analysis will highlight these wherever they appear in surname recurrence, and document any that do not recur in the report’s negative findings — your record of a reasonably exhaustive search.
FANlines proposes household matches across decades from names, ages, and birthplaces. You decide every one. Nothing is linked automatically, and no match score is shown — you confirm against the evidence, the way you would in a proof argument. Adjacent-decade subject households are always offered first.
Candidate links to review
No undecided candidates. Designate subject households across at least two decades on the previous screen to generate candidates.
Confirmed family groups
No links confirmed yet.
Rejected links (0)
Surname-variant grouping:
Run the analysis to generate the panels below. Every result is a research lead, never a conclusion.
FAQ
What does FANlines do?
It structures the census pages you paste into households, follows your research subject across the decades you supply, and surfaces neighbor, surname, and marriage origin leads. FAN stands for Family, Associates, and Neighbors.
Where does it get its data?
Only from what you paste. FANlines does not search any database, fetch any image, or connect to any service. It analyzes the table view census pages you give it and nothing else.
Is my research uploaded anywhere?
No. Everything runs in your browser on your own computer. Nothing is transmitted or uploaded. Your data lives only in this browser session and the JSON file you download.
Does my work save automatically?
Yes. FANlines autosaves to your browser every time you commit a page, set a subject household, confirm or reject a link, resolve an age question, or edit a page. A long session survives a closed tab or a crash, losing at most the one paste in progress.
Will my work still be there after I close the browser?
Yes, in most cases. The autosave persists across closed tabs and restarts. When you reopen the tool you get a Resume previous session prompt. Clearing your browsing data, using the Clear saved session button, or working in private browsing will remove it. It is recommended to save your progress using the JSON file option.
Then why should I download the JSON file?
The autosave is one browser on one machine, and a routine clear browsing data click will erase it. The JSON file is your lasting copy. Save it often.
What is the JSON file for?
It is your complete research object: every page, every correction, every linkage decision, and every analysis run. Reopen it any time to continue, add more census pages, or to reanalyze the same data for a different subject.
Which censuses are supported?
Current scope is United States federal population schedules, 1830 through 1950.
How many pages should I paste per decade?
At minimum the page before your subject, the subject page, and the page after. The enumerator walked a route, so the adjacent pages hold the nearest neighbors. Pasting more widens the net.
Why was my paste rejected?
The only hard rejection is a paste with no detectable header row. Recopy from the very top of the table view, including the column titles. Every other issue is a warning you can read and move past.
Why does the header row matter so much?
FANlines identifies each census format by its column headers and maps fields by name. Without a header it would have to guess column positions.
Do I have to fill in every metadata field?
No. Only the census year is required. State, county, and division improve your citations, and the tool marks each citation as complete or partial so you always know what is missing.
Can I fix an index error before it enters the analysis?
Yes. The preview before you commit a page shows every parsed row and lets you correct it while the actual census page is loaded or in your hand. This is the best opportunity to catch and correct an indexing error.
Does correcting a value overwrite the record?
No. FANlines stores your corrected value alongside the recorded one and discloses the correction in the report, with an optional note for your reasoning. Values always display as recorded.
What is the subject household?
The one household per decade that you mark as your research subject. It is the anchor every proximity and FAN analysis is measured from. You can change it any time without losing other work.
Can I analyze the same pages from a different family’s view?
Yes. Redesignate the subject household and rerun. No repasting, no lost linkage work. The same data yields a new perspective.
Why does the tool ask me to confirm every household link?
Because an automated linkage error would corrupt downstream results. You confirm against the evidence, the way you would in a proof argument. Nothing links automatically at any strength.
Why is there no numerical match score or percentage on candidates?
By design. The internal ranking only orders the queue. What you see and decide on is the evidence list, which is the defensible artifact. A visible score would invite the shortcut the Genealogical Proof Standard exists to prevent.
Can it link a family across a missing decade?
Yes. Candidates are generated across all decade pairs, not just adjacent ones, so a family absent from one enumeration is still matchable directly across the gap.
What are the proximity tiers?
Same structure means a shared building. T1 is within 5 dwellings of your subject, T2 within 15, and T3 is in your pasted window but farther. These distinctions were selected as an aid but are arbitrary. A dash in the strip means the group was absent from a decade you entered.
What do the surname grouping toggles do?
Exact groups identical spellings. Soundex groups spelling variants and lists every member. Soundex first letter blind also catches first letter errors like Hanks and Lanks, at the cost of sometimes grouping unrelated surnames. Grouped rows always list their members.
Why did my analysis go out of date?
Any change after a run, such as adding a page, correcting a cell, or confirming a link, marks the results out of date. Rerun to update. The report cannot be generated from stale results, so it is never based on outdated analysis.
What are the maiden and daughter hypotheses?
The backward scan looks for each wife as a daughter in a nearby different surname household in an earlier decade. The forward scan looks for each departed daughter as a wife under a new surname in a later decade. Both are intended as research leads.
Why did it flag a birthplace conflict I know is a civil boundary change?
The tool contains a large set of known boundary changes; however, this message indicates the given pair is not in the tool’s equivalence table yet. The tool is simply flagging it. You can resolve it against the images and note it in your log.
My sheet numbers look gapped but I verified the pages are consecutive. What do I do?
Use the consecutive pages control in the page list. It asks you to assert that you checked the images, merges the runs, and discloses the assertion as a footnote in the report.
Does FANlines assert any probabilities?
No. Co-location counts and proximity tiers are descriptive lead ranking heuristics, not significance tests. No probability is claimed for any pattern. Every output is a research lead under the Genealogical Proof Standard.
What is the report I get at the end?
A single self-contained HTML file with your data inventory, source citations, the analyses you chose, next steps, negative findings, and a methodology statement. Save or print it to PDF.