Making Zotero work for historians doing archival research

Roy Rozensweig wrote an article in the current issue of the Organization of American Historians' Newsletter suggesting that Zotero will replace the historian's traditional reliance on 3x5" index cards. Having read his article and explored the Zotero site, it looks amazing--except that it cannot process unpublished archival sources. For instance, while the "Info" tab for the letter and manuscript formats allows you to enter some archival information, these fields are not exported into the Chicago Manual note. Also, the formats that accept archival information do not include some essential citation fields such as the box and folder (or for microfilm, reel) in which the document was found. The letter format uses the "author," "title" fields of a published source rather than the "From" and "To" fields necessary for letter citations. Lastly, since there seems to be much interest in making Zotero compatible with legal research and writing, it would be wonderful if there were formats for common legal documents such as memoranda, briefs, and petitions.

Unfortunately, this is a big "except" for the many historians who use unpublished archival sources. Thanks to Rosenzweig's article, the demand for these functions is likely to increase dramatically. I noticed in another discussion that there is some plan to make Zotero work for archival sources. It would be great to get a more concrete sense if and when Zotero will add these capabilities and be ready to replace the index card (or other non-open-source reference programs) for historians working in archives.
  • We've been talking about this for the development of the RDF model I hope to see Zotero use for import/export. I'll let the Zotero people comment on whether or not this fits their thinking.

    In the RDF model (and indeed the citation style model) there's a generic property called "locator." So a volume, issue, page number and so forth are all examples of locators.

    While it's certainly possible, I don't think you can realistically expect to be getting dedicated fields for boxes, folders and reels. OTOH, a generic field for this information seems reasonable? In that case, the field content would be "Box 12, Folder 23". Does that work for you?

    Letters? Certainly recipient is important, but is not a letter "from" the author? E.g. the roles are not "from" and "to" but "author" and "recipient" respectively.

    We'll cover legal stuff too, though we could use some help there. Let's take your example of the brief. Would it be fair to say that it is a document that is related to a legal case? In other words, these are two distinct data objects: brief and case?

    I ask because in most existing programs and data formats, there's just a generic "case" type, which strikes me as less-than-useful. I myself have cited both briefs and decisions, and it seems to me they are separate objects.
  • Yes, I think a generic field where "Box #, Folder #" could be entered would work. Would you be able to sort items by whatever terms were entered in this generic field?

    As for letters, "Author" and "Recipient" would definitely work. What I was really getting at is that most citation styles require that you cite a letter "X to Y." As a result, an author field alone doesn't suffice.

    I agree that briefs are different from opinions, particularly because the latter is often a published source while the former often is not. Under the Bluebook rules, brief citations require a case or docket number, in addition to the data included in a typical case citation. Chicago accepts Bluebook conventions, so having some way to input a docket number would be important.

    And just to add one more issue to the pot: if you are citing an archival document other than a letter, Chicago requires that you specify the document type (e.g., memo, telegram). Perhaps there could be two "Generic Archival" options:
    --one with the "X to Y" fields that also had a generic field where the type of document could be specified and
    --one with author / title fields that also had a generic field where the type of document could be specified (for instance, I often cite to press releases)
    This would be a lot simpler than developing formats for every type of document one can encounter in an archive.
  • I agree that briefs are different from opinions, particularly because the latter is often a published source while the former often is not. Under the Bluebook rules, brief citations require a case or docket number, in addition to the data included in a typical case citation. Chicago accepts Bluebook conventions, so having some way to input a docket number would be important.
    As I've been thinking of it, from the data standpoint, a docket number is just a generic document number.
    And just to add one more issue to the pot: if you are citing an archival document other than a letter, Chicago requires that you specify the document type (e.g., memo, telegram). Perhaps there could be two "Generic Archival" options:
    --one with the "X to Y" fields that also had a generic field where the type of document could be specified and
    --one with author / title fields that also had a generic field where the type of document could be specified (for instance, I often cite to press releases)
    This would be a lot simpler than developing formats for every type of document one can encounter in an archive.
    That's our intention. We just haven't decided exactly how to do it (whether a simple dumb string, a URI and so standardized, or both).

    There will be a generic Document type, and a subclass of Manuscript. In addition, there will be a separate type of PersonalCommunication and a subclass of it and Document called PersonalCommunucationDocument.

    Sorry to get technical, but I figure you get the general idea. Im short, we'll have you covered on the input/output end (with usual caveats that I can't make promises for whether Zotero will implement it and so forth).
  • edited August 14, 2007
    zoila--

    While CSL revisions that Bruce is talking about above may take a while to implement, the Zotero chicago note citation will be fixed in a few days -- there is a ticket for this.

    Most features you request for entering archival sources already exist in Zotero:
    The letter format uses the "author," "title" fields of a published source rather than the "From" and "To" fields necessary for letter citations.
    When you enter "creators" in the "info" tab, there is a pulldown menu next to the name to enter creator's role. Use "author" for "from" and "recipient" for "to." (see FAQ for exact steps). The title field is there because it's used to select records in the middle pane (although sometimes memos do have titles).
    the formats that accept archival information do not include some essential citation fields such as the box and folder (or for microfilm, reel)
    Various archives use many different ways to indicate location of documents (besides the ones you mention, citations for documents in the National Archives in the US use RG for "record group," for example). For that reason, Zotero uses "Loc. in Archive" where users can enter box, folder, record group, or any other location. (see also FAQ). It doesn't really make sense to create separate fields for each of these. This field is not in the middle pane yet but can be easily added for sorting in future versions.
    If you are citing an archival document other than a letter, Chicago requires that you specify the document type (e.g., memo, telegram). Perhaps there could be two "Generic Archival" options:
    --one with the "X to Y" fields that also had a generic field where the type of document could be specified and
    --one with author / title fields that also had a generic field where the type of document could be specified (for instance, I often cite to press releases)
    Use "type" field in "letter" item type to enter "telegram," etc, and "type" in "manuscript" item type to enter "press release."

    All these fields will be incorporated into the Chicago note/bibliographic citation shortly.

    Elena
  • Thanks Elena--this is very helpful! I had gone through the FAQs but must have missed the one about the creator drop down menus. I'm sold.
  • edited August 27, 2007
    This might be a cavil, but isn't there the possibility that "Loc. in Archive" conflates arrangement and physical location ("Records Group 75, Box 43, Folder 5" vs. "Johnson City Public Library, Floor 3")?

    Are the CSL revisions going to make it possible to deal with issues such as the fact that an item may be a part of (say) "The Michael Hartley Collection" which is part of the Springfield Natural History Museum's Special Collections? In such a case "Special Collections" may refer primarily to location (Floor 5, room 512) or it may refer primarily to arrangement (with the items making up Special Collections located on different floors and/or buildings of different cities, as in NARA Record Groups).

    I know that most people aren't as OCD about this as I am, but it would be nice to have the option to create different sorts of records to deal with these different kinds of information. For example, could there be a class of Zotero records like "Level of arrangement" where I could enter information about NARA Record Group 75? This class of records could be hierarchical, so that one could describe the series within RG 75 (like "Enumeration and Enrollment Censuses, 1893 - 1913") as well as the boxes and folders within it if I wished. At the same time, there could be a class of Zotero records like "Institution/Repository". That would allow me to enter the address, coordinates, contact information, and history of the NARA repository in Morrow, Georgia.

    I'm offering this as food for thought, I know not everyone needs or want to go to this level of detail. But there must be a way to develop Zotero so that its structure is general enough to allow this sort of information (as opposed to creating a lot of specific types and categories).
  • edited August 29, 2007
    This doesn't address all your concerns, but the upcoming hierarchical item structure should include an "archive" or "archival collection" item type. It may not have as many fields as you request here--probably just name of collection, name of repository, city/country where it is located, and dates for a particular collection. But you could then attach a note to an "archive" item and enter contact information, etc there.
  • Can I ask you to elaborate a little on how one might achieve this at the moment?

    I have an Access database with about two thousand records. It's in a relational arrangement but I can easily produce the flat file format necessary to import to Zotero using Bibtex format. I need Author, Recipient, Title, DocLocation, Repository, DocDate, Keyword, VerbatimNote, ResearchNotes -- Bibtex seems to do this best.

    To import my Endnote database, including PDF attachments and tags, I used the Change fields tool in Endnote to move the Custom data into Fields recognized by RIS and then changed the field descriptors with Notepad++ until I had just about everything I wanted. I'm very happy with the result.

    I think that I can do the same with the Access data, but I'll end up with two authors instead of Author-Recipient (which for historians is critical). So here's the question: if I capture the flat file data in to Zotero, and then export the whole recordset with Zotero RDF is there a relatively simple way to change the second author to recipient in an editor and then suck the whole thing back in to the Zotero? (I've looked at the Zotero RDF for a couple of hours, experimenting with changing Author to recipient, but I can't make head or tail of it.)

    My thanks,
    Keith Breckenridge
  • To answer my own question (after five days of battling):

    The best option is to use the Mods xml syntax, wrapping it around the Access fields.

    If you have a large Access database with large Memo fields you cannot export the contents as a report because that injects infuriating errors that screw up the effort to Find & Replace the field labels with the xml tags from the mods template. You need to save the Access Query as an Excel sheet.

    You can then import that Excel sheet in to an XML sheet using the excellent Oxygen editor, and do the necessary replacements. You'll need to match up the sheets exactly and check that there are no errors in your syntax. Oxygen is great at this -- although it seems to run out of memory after four large find & replace attempts.

    You then need to break the large XML file in to smaller files or the Zotero import will time-out.

    Finally, there is a great SQLite Client (SQL data browser) which allows the kind of programmatic replacement of creator types I was talking about.

    I'm very happy with the result.
Sign In or Register to comment.