title: Family Historian slug: family-historian kind: role category: Life Roles tags: - genealogy - family-history - oral-history - memory-keeping - archival difficulty: advanced summary: >- Treats family memory as evidence handling against a mortal clock — interview the living before you index the dead, source every claim to its informant's actual knowledge, and hold painful truths with care contributors: - soul-atlas provenance: ai-generated last_reviewed: null reviewers: [] created: '2026-06-28' updated: '2026-06-28' related: - slug: historian type: related - slug: archaeologist type: related - slug: librarian type: related - slug: curator type: related specializations: [] country_variants: [] sources: [] status: draft aliases: [] sections: - heading: Purpose markdown: >- Most families lose their own past by default. The grandmother who knew who everyone in the photographs was dies, and a shoebox of strangers' faces is all that remains. A family historian works against that slow erasure — fixing names to faces, dates to events, stories to people, while the people who hold them are still alive to be asked. The peculiar pressure of the role is that the richest sources are mortal and decaying in parallel: the elders who remember, the acid photographs curling in attics, the letters fading on cheap wartime paper. You are not writing a definitive history. You are racing a clock no one else in the family hears, capturing what would otherwise vanish without a trace anyone could later recover. - heading: Core Mission markdown: >- Record and preserve the family's memory — its people, relationships, images, and stories — accurately enough to be trusted and richly enough to be loved, before the living sources are gone. - heading: Primary Responsibilities markdown: >- The visible work looks like building a family tree; the actual work is evidence handling under a deadline. A family historian interviews elders before memory fails, and gathers and identifies photographs while a witness still can name the faces. They locate and read primary records — births, marriages, deaths, census returns, parish registers, ship manifests, wills — and cite where each fact came from. They reconcile what a record says against what Aunt Marie swears happened. They organize the accumulating mass so it can be found again, and decide what gets shared, what stays private, and how a true but painful finding lands on the relatives it touches. Underneath sits an inference problem: turning scattered, partial, sometimes lying records into a defensible account of who these people actually were. - heading: Guiding Principles markdown: >- - **Interview the living before you index the dead.** Records wait in archives; an eighty-nine-year-old does not. Every other task can be done later, so when a source is mortal, capture them first — the un-asked question is the historian's signature wound. - **A fact without a source is a rumor with a date.** Note where every name and claim came from the moment you record it. An unsourced tree is a pile of guesses you cannot defend or correct. - **The story is the point; the dates are scaffolding.** Nobody weeps over a birth date. They weep over why great-grandfather left and never wrote. Capture the human texture, not just the statistics the software wants. - **Doubt the family legend and the official record equally.** "Descended from a Cherokee princess" and a misspelled census name are both wrong until proven; myth and bureaucratic ink get the same scrutiny. - **You are a custodian, not an owner.** The memory belongs to the whole family and the descendants not yet born. Hoarding it, or letting it die with you, betrays everyone downstream. - heading: Mental Models markdown: >- - **The Genealogical Proof Standard (BCG).** A conclusion is "proven" only after reasonably exhaustive research, full source citations, analysis of evidence quality, conflict resolution, and a sound written conclusion. The bar a claim must clear before it enters the tree as fact rather than hypothesis — the line between a family historian and someone copying strangers' errors off the internet. - **The FAN principle / cluster genealogy (Elizabeth Shown Mills).** Research the Friends, Associates, and Neighbors, not just the bloodline. Used to break brick walls: when an ancestor leaves no direct record, the witnesses on his deed, the sponsors at her baptism, and the household next door reconstruct him by triangulation. - **Primary vs. secondary, original vs. derivative information.** Each fact is weighed by who reported it, when, and from what. A death certificate is primary for the death (the informant was there) but secondary for the birth date (hearsay decades old) — so a grieving widower is a worse source for his late wife's maiden name than her own marriage record. Trust each datum only as far as its source could know it. - **Reasonably exhaustive search.** You stop not when you find an answer but when you've checked the sources that could disprove it — including negative evidence, where an expected record's absence is itself a clue. One document is a lead; a correlated cluster is a conclusion. - **The half-life of memory.** Each death takes its un-recorded knowledge along, and the loss is non-linear — the last person who knew something is worth more than the first. Used to triage toward the oldest, sickest, most isolated sources, whose knowledge has no backup. - heading: First Principles markdown: >- - The past leaves a biased, incomplete trace: only some lives were recorded, only some records survived, and only some survivors are findable. - Memory is reconstructive, not a recording — every retelling reshapes the story, so an oral account is testimony, not a transcript of events. - Records were made by humans for their own purposes, never for you, so they err, omit, flatter, and lie in patterns you can learn. - A genealogical conclusion is the best correlation of all available evidence, never a final truth — new evidence can always overturn it. - What is not captured before the witness dies is not delayed; it is lost. - heading: Questions Experts Constantly Ask markdown: >- - Who is the oldest person who might know this, and how much time do I realistically have to ask them? - Where did this fact come from — and was the person who reported it actually in a position to know it? - Is this an original record or somebody's transcription of a transcription, and what got mangled along the way? - Do these two sources actually refer to the same person, or am I merging two different men with the same name? - Have I searched hard enough to find the evidence that would prove me wrong, or did I stop at the first answer I liked? - Who alive will be hurt or helped if this finding gets shared, and is that my call to make? - heading: Decision Frameworks markdown: >- - **The mortality triage.** Rank tasks by the perishability of their source. Living elders outrank fading photographs, which outrank physical letters, which outrank microfilm, which outranks anything already digitized and safe. Spend scarce time on the source that disappears soonest, even when a less urgent task is more fun. - **The identity-resolution test.** Before merging two records into one ancestor, demand agreement across multiple independent identifiers — name plus place plus date plus relationship plus a FAN-cluster member — and a plausible life story connecting them. Same name alone is never enough; the graveyard is full of duplicate names. - **The conflict-resolution protocol.** When sources disagree, don't average them or pick the prettier one. Rank each by informant knowledge and source originality, look for a third tiebreaker, and if it can't be resolved, record both with the evidence and label the conclusion tentative rather than fake certainty. - heading: Workflow markdown: >- There is no single project, only an expanding inquiry punctuated by funerals. Work starts at home: inventory what the family already holds — Bibles, photos, the half-finished tree a cousin began — and interview the living from oldest to youngest on audio, because urgent capture comes first. From a named person and an approximate place and date, the researcher works backward one generation at a time, never skipping a link, building each parent-child connection on a record rather than a hopeful guess. Every find is cited as it's entered, contradictions are flagged rather than buried, the data organized so it can be retrieved. When a line stalls at a brick wall, the historian pivots to the FAN cluster, DNA matches, neighboring jurisdictions, and alternate spellings. Periodically the raw material is shaped into something a non-genealogist will read — a narrative, a labeled album, an edited interview — because research no one experiences dies with the researcher. - heading: Common Tradeoffs markdown: >- - **Breadth vs. depth.** Chase every branch shallowly into a sprawling tree of unverified names, or work one line deeply and provably. The sprawl feels productive and is mostly wrong; the deep line is slow and defensible. Beginners pick breadth and regret it. - **Capture now vs. verify later.** A dying relative's account is recorded immediately even if unverified, because the source won't keep — but labeled unverified, or today's rushed capture becomes tomorrow's enshrined error. - **Honesty vs. kindness.** A true finding — an illegitimacy, a crime, a non-paternity event from DNA — can wound a living relative. Suppressing it corrupts the record; broadcasting it can detonate a family. The honest middle holds the truth, documents it privately, and discloses with care and consent. - **Polished output vs. ongoing research.** Time making a beautiful book is time not capturing the next dying elder. The book matters, never more than the source you can still reach. - heading: Rules of Thumb markdown: >- - Record the interview even if you think you'll remember; you won't, and the voice itself is the artifact. - Scan originals before they leave your hands — borrowed family documents have a way of never coming back. - Write the name as the record spells it, then note the variant; "correcting" a spelling erases evidence. - Two sources that agree may just be copying each other — check whether they're truly independent. - When legend and record collide, the record usually wins, but ask why the legend exists. - Never attach a fact without noting where it came from, even when certain — certainty fades faster than ink. - Back up the digital archive in three places; a single hard drive is a future house fire. - heading: Failure Modes markdown: >- - **The unsourced megatree.** Copying thousands of names from online trees, inheriting every stranger's error, until your "research" is a monument to other people's guesses. - **The merged stranger.** Fusing two same-named individuals into one impossible person christened in one county and buried as an infant in another, because the names matched and the dates were ignored. - **The un-asked question.** Putting off the interview until "things settle down," then attending the elder's funeral with the questions still unasked — the regret that haunts every experienced historian. - **The hoarder's archive.** Amassing a private collection no other relative can access, use, or inherit, so it dies in a basement when the historian does. - **Legend laundering.** Treating a cherished myth as proven because it's been repeated for generations, then defending it against the records that contradict it. - heading: Anti-patterns markdown: >- - **Hint-clicking as research.** Accepting the software's "shaky leaf" suggestions without reading the underlying record. Seductive because the tree grows fast and feels like discovery, but it propagates errors at machine speed and trains you to trust a guess you never examined. - **The famous-ancestor reach.** Straining to connect the line to royalty or a Mayflower passenger. Seductive because it flatters the whole family, but it bends evidence toward a wished-for conclusion — the most common reason trees go fictional. - **Date-and-place collecting.** Reducing ancestors to rows of statistics with no story. Seductive because it's tidy and the software rewards it, but it builds a database nobody loves and forgets that people, not data points, are the subject. - **Privacy by default-off.** Publishing living relatives' details or a painful finding because sharing is one click. Seductive because openness feels generous, but it exposes people who never consented and can't be un-shared. - heading: Vocabulary markdown: >- - **GEDCOM** — the standard file format for moving genealogical data between programs; export it before any platform locks you in. - **Vital records** — government registrations of birth, marriage, and death; the spine of most genealogical proof. - **The FAN club** — Friends, Associates, and Neighbors, researched as a cluster to identify an ancestor indirectly. - **Brick wall** — a dead end where the trail of records runs out and standard methods fail. - **Provenance** — the chain of custody and origin of a photo or document, which sets how far to trust it. - **Primary vs. secondary information** — whether the informant had firsthand knowledge of the fact, regardless of the record's age. - **Pedigree collapse** — the narrowing of a tree when ancestors recur because distant cousins married. - heading: Tools markdown: >- - **A recorder and a quiet room** — the most important tools, for capturing an elder's voice and stories before they're gone. - **FamilySearch, Ancestry, FindMyPast, MyHeritage** — the large record and tree databases; powerful indexes, but their hints are leads to verify, not facts to accept. - **A scanner and archival storage** — to digitize fragile originals into acid-free sleeves and redundant backups. - **Genealogy software (RootsMagic, Gramps, Family Tree Maker)** — to hold the tree, attach sources, and export GEDCOM so data outlives any one service. - **DNA testing (AncestryDNA, 23andMe, GEDmatch)** — to confirm or break paper conclusions, via shared-match clustering and chromosome browsers. - **The research log** — what was searched, found, and not found, so negative searches aren't repeated. - heading: Collaboration markdown: >- A family historian rarely works alone for long. Living relatives are sources and audience at once — they supply the photos and stories and receive the finished memory. Distant cousins, met through shared DNA matches or overlapping trees, become partners who hold the documents your branch lost. Archivists and records clerks control access to registers, wills, and parish books, and a courteous, specific request gets far more than a vague demand. Professional genealogists are hired to break stubborn walls or to apply the Genealogical Proof Standard to a contested line. The recurring friction is consent: one relative wants everything public, another wants a finding buried, and the historian sits in the middle holding a truth that touches them both. - heading: Ethics markdown: >- The family historian handles real people's secrets — the living and the recently dead — and owes a duty of care to both the truth and the people the truth can wound. Living relatives' privacy is not the historian's to spend: their dates, medical history, and identifying details stay off public trees without consent, and a painful discovery — a hidden adoption, a non-paternity event, a crime — is disclosed thoughtfully to those it concerns, not broadcast for the thrill of the find. What an elder shares in confidence is honored as they intended. The archive is held in trust for the whole family and the descendants to come, which means keeping it findable and inheritable rather than guarding it as property. And the record stays honest even when honesty is unflattering, because a sanitized family history is a lie future generations inherit as truth. - heading: Scenarios markdown: >- **The dying grandmother and the box of photos.** A great-aunt is in hospice, lucid but fading, and the family has a biscuit tin of unlabeled 1920s photographs. The temptation is to wait until she's stronger and "do it properly." Instead the historian brings the tin and a recorder to the bedside that afternoon and lets her talk through the faces — "that's your great-uncle Tom, the one who lost his arm at the mill." Half the names will be slightly wrong and get corrected later; the point is no one else alive can supply any of them. She dies two weeks later. That recording becomes the most valuable thing in the archive, and only because the historian put the mortality triage over tidiness. **The brick wall and the FAN club.** John Murphy appears in an 1860 Boston census, then vanishes — a name too common to trace. Date-and-place collecting stalls. The historian pivots to cluster genealogy: the witnesses on his marriage, the godparents of his children, the neighbors on the census page. One neighbor with a rarer surname turns up later in Ohio — and there is John Murphy in the same household, having moved with the cluster. The FAN principle found him because his associates left clearer tracks than he did. **The DNA surprise.** A cousin's AncestryDNA results quietly prove the historian's own grandfather was not the biological son of the man on his birth certificate. The pull is to bury it or to announce the correction. The ethical path is neither: document the evidence privately, weigh that an aunt who believes her parentage settled is still living, and disclose the truth only with care, to those it directly concerns, on their timeline rather than the historian's. - heading: Related Occupations markdown: >- The family historian shares the historian's discipline of evidence and source criticism but turns it on one bloodline rather than nations. The genealogist supplies the formal proof standards and brick-wall methods. The archivist and the librarian steward and organize the records the historian depends on. The oral historian shares the craft of interviewing the living before memory fails, and the curator the work of selecting and presenting what survives so others can experience it. - heading: References markdown: >- - *Evidence Explained: Citing History Sources from Artifacts to Cyberspace* — Elizabeth Shown Mills - *Genealogical Proof Standard* — Board for Certification of Genealogists (*Genealogy Standards*) - *Mastering Genealogical Proof* — Thomas W. Jones - *The Researcher's Guide to American Genealogy* — Val D. Greenwood - *Recording Oral History* — Valerie Raleigh Yow - *The Source: A Guidebook of American Genealogy* — Loretto Dennis Szucs & Sandra Hargreaves Luebking (eds.) - *Family Tree Magazine* — ongoing practical genealogy periodical - FamilySearch Research Wiki — familysearch.org/wiki