How records arrived

Source of cataloging asks who wrote a record. This page asks how it travelled which is answered by a different field.

MARC 035 holds system control numbers, written (ORGCODE)number: the number, and the system the number belongs to. Records accumulate these as they move between union catalogues, knowledge bases and vendor platforms, so the set of namespaces on a record is a rough itinerary. It is also much better populated than the cataloging source: nearly every record carries an 035.

Two things to keep in mind:

• These channels overlap. A record routinely carries several system numbers, so the shares below are coverage and do not sum to 100%. That overlap is often the history: RLG merged into OCLC in 2006, so a great many records legitimately carry both an RLIN and an OCLC number.
• A system number means the record passed through, not that it originated there. An OCLC number says the record exists in WorldCat, not that OCLC created it — which is exactly why the 040 page shows almost no OCLC authorship while nearly every record here has an OCLC number.

The two charts below are the same field seen at two resolutions. The first sorts namespaces into a handful of categories, so one question has this record been through OCLC? can be put to every library and answered on comparable terms. The second drops the categories and shows the namespace strings as they were written. Neither is a summary of the other, and their numbers will not always agree; the list under the second chart says exactly where they diverge and why.

Which systems each library's records have been through

Share of each institution's records carrying a control number from each system. "Any system number" is the baseline, essentially every record has one.

The namespaces themselves

The raw 035 namespaces, uncategorized: the union of each institution's own twelve most common — a namespace must also reach a minimum share of that institution's records to earn a row — so a system that matters to one library isn't ranked away by the others. Cells are a share of all that institution's records, the same denominator as the chart above, which makes the two directly comparable.

Three ways they differ, each of them a reason both charts exist:

• One category can be several namespaces. OCLC / WorldCat above matches (OCoLC) together with its variants — (OCoLC-M), (OCoLC-I) and so on. A library that writes only the bare form shows the same figure in both charts; one that also writes the variants shows a lower figure here, split across several rows, because this chart does not add them up. That is the most likely reason two numbers you expect to match do not.
• One category can be a different namespace at each library. A local library system resolves to whatever each institution actually writes — usually its own code, sometimes a generic ILS product name shared by every library running that system. The category is what makes those one comparable row; only the raw view shows what any given row is made of.
• The categories are not exhaustive. Nothing above collects vendor platforms, special-collections databases, or other libraries' union catalogues, so those appear only here.

Unrecognized codes pass through raw rather than being dropped or guessed at.

Two rows of the first chart — A local library system and Another POD member's system — rest on a hand-curated map of namespace codes to institutions that POD has not ratified. See the institution code mapping for what it contains and what adding a member would change.