What matters most in anniversaries and archives

An anniversary is a date doing work. It turns a loss that happened once into something that recurs, and it gives families, institutions and publications a reason to return to a record they had otherwise set aside. An archive is where that record waited in the meantime. The two depend on each other more than is usually noticed, and this page is about both: what gets marked and why, what an archive actually holds, and how to find and cite an old record without misreading it.

No person is named on this page and no death is reported here.

What to take away

• Anniversaries are marked because round numbers and first years give people permission to remember publicly. There is no rule about which dates matter; the family's and the institution's choices are the only ones that count.
• Archives hold several different records of a death, produced by different bodies for different purposes. Each has its own access rules, and the body that holds it publishes them.
• An old notice found in an archive should be cited as what it is: a dated publication, by a named source, saying what it said. Not as a fact about the person.

What gets marked, and by whom

Three different parties mark anniversaries, and they mark different ones.

Families mark the first anniversary above all, and then the ones that fall on personal calendars: a birthday, a wedding date, the anniversary itself each year. Some place a short notice in a newspaper's in-memoriam column annually, a practice that is older than most people realize and that persists in many local papers. It is a private act carried out in public, and it should be read as nothing more than that.

Institutions mark anniversaries that align with their own calendars: a founder at a centenary, a member at an annual gathering, a colleague when a prize named for them is awarded. The date chosen is whatever the institution's own rhythm makes convenient, and it is not a claim about the significance of the date itself.

Publications mark round-number anniversaries of deaths that were widely reported at the time, because a round number is an occasion and an occasion is a reason to publish. Ten, twenty-five, fifty. The choice reflects what the publication thinks its readers will recognize, and it reflects the availability of archival material to draw on, which is why anniversaries of well-documented lives are marked far more than those of significant but poorly recorded ones.

None of this is a ranking. A date that is marked is a date somebody chose to mark. A date that passes unremarked has been forgotten by nobody who was there.

What an anniversary piece is

A publication returning to a death on its anniversary is doing something specific, and it is worth naming so that the piece can be read properly.

It is not news. Nothing has happened. The piece is an act of retrieval: the original coverage is reread, the record is reassembled, and the person is presented again to readers who may be encountering them for the first time. Its value lies in the reassembly, and its risk lies in the same place.

The risk is that errors in the original coverage are repeated as settled fact, with the added authority of a decade. A cause that was speculated in the first week becomes the cause. A date that was wrong in the first notice becomes the date. A survivor who was omitted stays omitted. An anniversary piece written from the archive alone inherits everything the archive got wrong.

The better practice is to treat the anniversary as a chance to check. Go back to the sources behind the original coverage, see what has since been corrected, and where the family or the institution has said something in the intervening years, prefer it to what was printed at the time.

What an archive actually holds

People say "the archive" as though it were one place. For a death, at least four separate records exist, held by four separate bodies, under four separate sets of rules.

Record	Held by	What it is
The civil register entry	The registration office for the place of death	The formal record. Access rules vary widely by jurisdiction and by how long ago the death was
The published notice	The newspaper's archive, and any library or service that holds that newspaper	What the family placed, as printed, on the date it ran
The reported obituary	The same newspaper archive, or the publication's own	What a journalist wrote, with the sourcing of the time
Burial or cremation records	The cemetery, crematorium, or the authority responsible for it	Where and when a committal took place, and sometimes who arranged it

Beyond these, there may be probate or estate records where such things are public, parish or congregational registers, institutional files, and family papers. Each was made for a purpose that was not answering a researcher's question, and each answers only what it was made to answer.

The single most useful habit in archival work is to find out, from the holding body itself, what it has and who may see it. Every registration office, library and cemetery authority publishes its own rules on access, on what periods are open, and on what has been digitized. Those rules are the facts; nothing on a page like this one can substitute for them.

Finding an old notice

The practical search is usually for a notice in a newspaper, and a few things make it tractable.

You need a date range, and you should expect the notice to run after the death, sometimes well after. You need to know which papers served the place where the person lived or died, which is a matter of local knowledge and of a library's holdings guide; the Chronicling America researcher guide is a good model of what such a guide tells you and how candid a good one is about the limits of searching scanned text. You need to allow for names spelled as they were heard, for initials in place of given names, and for a woman recorded under her husband's name. And you need to be prepared for the digitized text search to miss what is plainly there on the page, because scanned newspapers are read by software that fails on small print and poor microfilm.

If a search returns nothing, the honest conclusions are limited: the notice may be in another paper, in another range, under another spelling, or it may never have been placed. Many families placed none.

Citing an archived record

Once found, a record should be cited as what it is. A notice in a newspaper is a publication, on a date, by a named source, saying certain things. It is evidence that those things were said. It is not, on its own, evidence that they were true, though for the practical matters a notice exists to convey it is generally reliable.

A workable citation gives the publication, the date, the page where known, the archive or service through which it was consulted, and the date it was consulted. Where a fact in a notice conflicts with a civil record, the civil record generally governs the date and the identity, and the notice governs what the family chose to say. The civil side of that pairing has its own finding aids: the National Archives guidance on vital records sets out what exists and who holds it for one country. Both facts can stand; they are different kinds of fact.

The same discipline applies to reported obituaries, with the added caution that a journalist's account carries the sourcing of its time, and that later corrections may exist that the archived copy does not show.

Resurfacing, and its costs

Anniversary pieces, digitized archives and search engines all put old records back in front of people. For a researcher this is a gift. For a family it can be something else.

A notice placed in a local paper decades ago was placed for that paper's readers on that week. A family that did not expect it to be findable by anyone, anywhere, forever, may find its reappearance unwelcome, and a family whose loss was hard may find an anniversary retrospective painful. Publications that mark anniversaries tend to know this and to contact families beforehand; archives tend to publish their own policies on requests to limit findability. Neither practice is universal, and neither is a legal requirement in most places, but both are the considerate default.

For a reader, the corresponding practice is restraint. An old record found in an archive can be used for what it says. It is not an invitation to contact the people named in it.

What this site does

We keep no calendar of anniversaries and mark none. Where an old record is referred to on this site, it is cited as a dated publication, with its source, and nothing is added to it. Nothing here should be read as saying that any person died on any date.

For why the first days after a death produce no publishable record, see deaths today. For how a weekly window is defined, see deaths this week. For the units an old entry is filed under, see deaths by month and deaths by year. For why an old record's cause line should stay blank, see cause of death guides.

Common questions

Is there a standard set of anniversaries that get marked?

No. Families mark what matters to them, institutions mark what fits their calendars, and publications mark round numbers of deaths that were widely covered. None of it is a rule.

Where do I find out whether a civil death record is available?

From the registration office for the jurisdiction where the death occurred. It publishes its own access rules, including which periods are open and what has been digitized.

A notice I found says something different from a civil record. Which is right?

They record different things. The civil record generally governs the date and identity; the notice records what the family chose to say. A conflict on a fact of identity is worth resolving in favor of the register unless there is reason to think the register is wrong.

Can I republish an old notice I found in an archive?

The notice was a publication; quoting it briefly with attribution is ordinary practice. Reproducing it wholesale, or attaching it to the family's current whereabouts, is a different matter and the considerate answer is usually no.

Should an anniversary piece mention how the person died?

Only where the family or an official finding made it public at the time or since, and then attributed. A cause that was speculated in the original coverage should not be promoted to fact by repetition.