Musicians with february birthdays: what to know and why

Libraries have been describing performers for well over a century, and they solved the identity problem long before anyone tried to build a birthday page. What they produced is worth studying, because it shows how much can be recorded about a person without a public date attached.

What to take away

• A catalog record describes a work and points at a separate record that describes the person, and keeping those apart is what makes both correctable.
• The person record carries whatever is needed to tell two people apart, which is usually far less than a full date.
• A field that exists in a schema is not a field that has to be filled or published.

Two records, not one

A bibliographic record describes an item: a recording, a score, a published edition. It says what the thing is, who is responsible for it and how to find it. It does not try to describe the people involved beyond naming them in a controlled form.

The description of the person lives elsewhere, in a name record whose job is to establish one preferred form of the name and link every variant to it. When a cataloger needs to distinguish two performers with the same name, they add a qualifier to the name record, not to every item. Correct the person record once and every work attached to it inherits the correction.

That separation is the whole trick, and it is the opposite of how a scraped celebrity database is usually built, where the person's details are copied into every row that mentions them and each copy then ages independently.

What the format actually asks for

The MARC standards that libraries have used for machine-readable cataloging are old, verbose and unglamorous, and they encode a set of decisions that are still right.

The decision	How it shows up	Why it holds up
Names get a preferred form and variants	One heading, many cross references	The reader can arrive by any name and land in one place
Dates are a disambiguator, not a biography	Added when needed to separate two people	Nothing is recorded because it is interesting
Every element is tagged	Fields and subfields rather than a sentence	A machine can act on a part without parsing prose
Provenance is part of the record	Where a heading came from and who set it	A later editor can see the basis for a decision
Records are meant to be revised	Headings change as knowledge changes	Nothing is presented as permanent

The second row is the one worth borrowing. In a catalog a date exists to answer a question that arose. It is not collected on the chance it might be useful, and where the two people can be separated another way, it is not there at all.

What a music entry can carry instead

A performer can be described very completely without a date. The works they contributed to, the roles they took on those works, the period their output spans, the ensembles they appeared with and the identifiers other systems use for them are all descriptive, all checkable and all derived from published material rather than from personal records.

That set of fields also degrades better. If one work is misattributed, the entry is wrong about one work. If a date is wrong, the entry is wrong about the person, and the error travels with the name. The musicians on the January page covers why an identifier rather than a date is the field to chase; the argument here is the same one from the cataloging side.

Applying it to a month page

• The month is a heading, and the entries under it are pointers to descriptions. Nothing about the heading is a claim about anyone.
• An entry should be improvable by adding a work rather than by adding a personal detail.
• Where two performers share a name, separate them by their output, which is public, rather than by a date, which is not. The March musicians page shows how release chronology gets confused with personal chronology when this is not done.
• A correction to a person should propagate everywhere they appear. If it does not, the data model has copies where it should have references, as the February index describes.
• The inclusion rule is editorial and belongs in public. The January index states this site's.

Common questions

Do library records contain birth dates?

Sometimes, where one was needed to separate two people with the same name and a published source supported it. That is a narrower purpose than a birthday page, and the presence of a field in one system is not a reason to copy it into another.

Is a catalog record a reliable source for a date?

It is a reliable record of a cataloging decision, which is not the same thing. It says a qualified professional judged this value good enough to distinguish two headings, and it usually points at what they relied on. Follow that pointer rather than treating the record as the origin.

Why does any of this matter for a page about a month?

Because a month page is a small catalog with the same problems and a fraction of the discipline. Everything it gets wrong, catalogers got wrong first and then fixed, and the fixes are documented.

What is the one habit worth copying?

Describe the person in one place and refer to it everywhere else. Almost every stale detail on the internet is a copy that stopped being updated when the original changed.