I think we'll need Archive and Loc. in Archive, just like other items types. That makes the logic of how this relates to hierarchical items a little difficult, simply because Archival Collection captures only part of the meaning of Loc. in Archive.

I think we'll want some sort of medium (electronic, microfilm, etc.) and extent (pages, number of items) variables/fields here too.

I'm still wary of how this is related to Data Set -- I thought Data Set was being considered for things like published quantitative data, genetic data, text corpora, etc. I believe that the citation requirements are different enough to make a shared treatment in CSL problematic.

If I'm not mistaken, this type is intended to represent Finding Aids, like http://dla.library.upenn.edu/dla/ead/ead.html?id=EAD_upenn_museum_PUMu1022

FWIW, when I designed CSL originally, I was writing a manuscript that included reference to many archival documents. But I never considered the possibility to cite the collections themselves. I was thinking of collections much like a periodical; a collection of documents you don't cite on your own.

which I think was a reasonable assumption, but turned out, as it has for periodicals (#23), that they are, in fact, cited as a whole, and at least for collections, not infrequently so--this comes up a ton when talk to historians.

I think it'd be worth clarifying here those use cases. Putting lists of
collections at the top of a bibliography, for example (which I've seen
historians do), is not citing as we tend to mean it here. So when do people
cite collections directly?

From Chicago manual 14.233

In a bibliography, the main element is usually either the collection in which the specific item may be found, the author(s) of the items in the collection, or the depository for the collection. (Entries beginning with the name of the collection or the last name of the author—which sometimes overlap—tend to be easiest to locate in a bibliography.)

That's different from footnotes, that do refer, normally, to specific documents, but, also you'll note, those collections are definitely listed in the bibliography--they make a specific note to how they should be sorted in there--not at the top/end etc.

I recently moved, and can't find my copy of the CMS (along with most of my
books!). That excerpt, though, sounds like it's describing how to reference
a document within a collection (with its mention of the "the specific
item")?

On Tue, Oct 14, 2014 at 2:51 PM, Sebastian Karcher <notifications@github.com

wrote:

From Chicago manual 14.233

In a bibliography, the main element is usually either the collection in
which the specific item may be found, the author(s) of the items in the
collection, or the depository for the collection. (Entries beginning with
the name of the collection or the last name of the author—which sometimes
overlap—tend to be easiest to locate in a bibliography.)

That's different from footnotes, that do refer, normally, to specific
documents, but, also you'll note, those collections are definitely listed
in the bibliography--they make a specific note to how they should be
sorted in there--not at the top/end etc.

—
Reply to this email directly or view it on GitHub
#27 (comment).

Adding support for this issue/feature, following discussions with fellow archivists. My experience from trying to implement Zotero support for http://archives.nypl.org/ has been that the current data models don't really support archival metadata/collections, so this is an exciting feature to see implemented.

There's a draft extension to Schema.org for representing archival repositories and collections (https://archival.github.io/schema-org/about/), which may be a good place to start regarding modeling/mapping archival collections. From the perspective of implementation, it would allow for Zotero to only need a single translator, which would work on any site supporting the extended standard -- as well as providing a strong use case for implementing Schema.org for archival repositories.

Regarding modeling Archival Collections - would these be modeled as Items or Collections? The advantage of using a Collection model would be the built-in support for hierarchical data, which would allow for importing full finding aids into Zotero. This would enable things like folder/item-level citation generation and the ability to attach digital images to granular archival components (ref. https://guides.library.harvard.edu/zotero_archival_research for a sample use case).

APA style also requires citing the geographic location of the the archive, so archive-place also seems like a necessary additional field.

Yes, of course -- archive-place is already in CSL and should be part of however Zotero decides to model archives.

Ah, forgot archive-place was already in CSL