On the need for separate holdings and item records in the Codex

codex

#1

@andrealoigman and I have been working through some of the consequences of the combined items/holding record in the latest version of the FOLIO Codex model, and there may be a need to have them separated. The “holding record” concept is useful in cases where many items have nearly the same attributes and are only distinguished by a few properties. For example, long runs of materials are housed together and share the same place; the ‘location’ property is in common among the physical pieces and specific pieces particular volume numbers and barcodes. Another example is the situation where there are multiple copies of multi-volume sets. The diagrams below describe what is needed when a library holds three copies of a four-volume set.

Holding-and-Item, Base Model

Holdings-Item model, base
This is the original model minus “location” and “coverage”. (They will be added back later.) The “item/holding” record has been split into a “holding” record and an “item” record.

Holding-and-Item, Physical Materials

Holdings-Item model, physical
When modeling physical items – in particular multiple copies of a multi-volume set – “holding” and “item” records are used. Locations are a property of both holdings and items; if the item location is not populated, the item inherits the holding location; if the holding and item location are both populated the item location takes precedence.

All told there are three “holding” records (perhaps representing three separate purchases of the set) and twelve “item” records (four item records for each of the three sets).

In this version of the model, it might make sense for acquisition records to be attached to holdings records in the case when multi-volume sets are purchased at different times or from different suppliers.

Holding-and-Item, Electronic Materials

Holdings-Item model, electronic
In the case of electronic resources, the “holding” record contains the supplier of the content. Content can come from many suppliers, and each “holding” record has its own coverage statement. This starts to look a little bit like the Title-Package-Platform construct that has served KB+ well. (@ianibbo and @marcjohnson – am I off base here with the electronic materials?)

Andrea has other concerns about combining holding and item together, some of which she described here earlier: Location/Classification question (also for RM/MM SIGs)