Features
Historical figures with February birthdays: the records
Historical figures with February birthdays: six dates, the records behind them, and how to label a back-calculated or self-chosen date on a calendar.
Historical figures with February birthdays sit on a birthday calendar beside living stars, but their dates rest on very different records: a state birth certificate for one, a parish register or a family bible for another, a self-chosen day for a third. A calendar entry for them should say which. Many of these dates reach calendars through compiled family trees, and a compiled family tree is not a source. It is somebody's conclusion, built from other people's conclusions, and the further back it runs the more of it rests on nothing anyone can now inspect.
What to take away
- Six February birthdays are listed below with the record type behind each date.
- A date computed from an age at death is a calculation, not a record.
- Where two sources disagree, the disagreement is the finding; do not resolve it silently.
- The February birthdays list linked above holds the day-by-day entries; this page explains the records under the historical ones.
February birthdays, and what holds each date
February birthdays
| Person | Date | What the date rests on |
|---|---|---|
| Rosa Parks | February 4, 1913 | Alabama state vital records; her own account in Rosa Parks: My Story |
| Ronald Reagan | February 6, 1911 | Illinois birth certificate, filed with the state |
| Charles Darwin | February 12, 1809 | Shrewsbury parish register, kept by the church |
| Abraham Lincoln | February 12, 1809 | Family bible and later affidavits, not a state record |
| Frederick Douglass | February 14, 1818 | His own choice of day; he never knew the date |
| George Washington | February 22, 1732 | A register entry in the Old Style calendar |
Douglass is the instructive one. He wrote that he had never met a slave who knew a birthday, and adopted February 14 himself. That is a self-declaration, and it should be labelled as one wherever it appears.
Washington's date was recorded as February 11, 1732, under the Julian calendar then in use in Virginia. The Gregorian shift moved it eleven days. Both figures are correct, in their own systems.
How a calendar should label each one
Put the record type next to the date, because the six entries above are not equal:
- Record made at the time:Rosa Parks, Ronald Reagan and Charles Darwin. Label the record, such as the state birth record or the parish register.
- Family record and later statements:Abraham Lincoln. Label it as a family record, which is weaker than a civil registration.
- Self-chosen date:Frederick Douglass. Label it as the date he adopted.
- Converted date:George Washington. Show the New Style date and note the Old Style original.
February adds one more labelling choice that no other month has. A person born on February 29 has a calendar date only one year in four, so a calendar must decide whether to show that birthday on February 28 or March 1 in common years, and say which rule it uses.
Where family-tree dates usually break
Many early dates on calendars come from compiled trees rather than from records like these. This is where those trees most often break.
What normally supports it
- Two people in one place with one name
- A record connecting them
- A gap between two documented generations
- An intervening record
- A descent from a well-known figure
- A documented line at every step
- A date given to the day for a distant ancestor
- A register entry
- Agreement across many published trees
- Independent research
What often actually supports it
- Two people in one place with one name
- The assumption that one name is one person
- A gap between two documented generations
- A plausible bridge someone proposed
- A descent from a well-known figure
- One weak link, inherited downstream
- A date given to the day for a distant ancestor
- A calculation from an age at death
- Agreement across many published trees
- One tree copied many times
The last row looks like the opposite of a problem. Ten trees agreeing is one claim with nine copies unless each cites its own document, and the copying is invisible because copies do not announce themselves.
What really supports each link
The link
- Same name, same place
- connecting record
- Gap between generations
- intervening record
- Descent from famous figure
- documented every step
- Exact date, distant ancestor
- register entry
- Many trees agree
- independent research
Real support
- Same name, same place
- one name = one person
- Gap between generations
- proposed bridge
- Descent from famous figure
- one weak link
- Exact date, distant ancestor
- age-at-death math
- Many trees agree
- one tree copied
Usual support
- Same name, same place
- Gap between generations
- Descent from famous figure
- Exact date, distant ancestor
- Many trees agree
Back-calculated dates and thin links
A date derived from an age recorded at another event is a computation. It inherits every rounding in the original. The March historical page works through that arithmetic and what it can support.
How one bad link spreads
- One mistaken link early in a line
- Many later claims inherit the error
- Claims appear to corroborate each other
- They share the same error, not evidence
The second failure is structural. Lines of descent converge going back, so the same ancestors reappear inside one tree. That is pedigree collapse. One mistaken link early in a line spreads into many later claims that appear to corroborate each other because they share the error.
Genealogy as a discipline separates a conclusion from the evidence for it. A sound tree states, for each link, what document supports it, where it is held and what it says. A primary source is a record made at or near the time by someone in a position to know.
How to read a claim about an early figure
How to read a claim
- Look for the document, not the tree. A tree that names none is unsupported however detailed it reads.
- Check what the document says, not what the tree says it says. Transcriptions drift.
- Treat unusual precision for a distant period as a signal to look harder, not as good research.
- Expect calendar problems on top. The January historical page covers what a converted date should carry.
- Record uncertainty in the entry rather than resolving it silently, the practice the January index builds on.
Common questions
Is an online tree useless?
No. It is a finding aid, a way to discover which documents might be relevant. Used as a pointer it is valuable; used as a citation it is a copy of an opinion.
Why do so many trees reach famous ancestors?
A line that reaches one is worth publishing and gets copied. The weak link in the middle is not visible in the copy, and selection does the rest.
What counts as enough evidence for a link?
A record that connects the two people specifically, rather than one showing both existed in the same place. Same name, same parish, same period is a hypothesis, and many published lines rest on exactly that.
Should a page like this repeat a historical date at all?
It can describe what a record says and what kind of record it is, as the table above does for six February figures. The record type is the useful contribution; a bare date is the kind of entry that spreads errors from one calendar to the next.
