Oral histories citations include incorrect n.d. reference when date is present

[yonyitz]

N.d. is erroneously present after Science History Institute in all oral history records.

See example here: "Fred Basolo, interviewed by James J. Bohning in Northwestern University (Evanston, Ill.) on March 1, 1991. Philadelphia: Science History Institute, n.d. Oral History Transcript 0091."

Please remove n.d. in citation unless there is a reason why it is present.

[eddierubeiz]

@sanfordd notes: "What's occurring is that the date of interview is being used in the title section when they say "on March 1, 1991" and then it's been set to a nil value which the CSL (Citation Style Language) tool we're using reads as no date and inserts as n.d.in the date area."

[jrochkind]

Hmm, I wonder if a new release of the ruby CSL gem changed behavior there.

We could fairly easily make the date appear twice.

"Fred Basolo, interviewed by James J. Bohning in Northwestern University (Evanston, Ill.) on March 1, 1991. Philadelphia: Science History Institute, March 1, 1991. Oral History Transcript 0091."

Or something like that. But that's not what you're asking for.

We may not very easily be able to make the date not appear at all in that second slot. The CSL style for "chicago style" that we are using believes there must be a date at that position in a chicago-style citation. Either an actual date, or "n.d." for no-date.

We could look into defining our own custom CSL style instead of using the built-in "chicago", that has different behavior.

It might also be worth looking at old chf_sufia app to see if it behaves identically. If it behaves differently and does not put the "n.d." there, then looking into how the code may differ may give us a clue as to a solution. I feel like it might have been behaving differently at one point, as I feel like we would have noticed oral histories doing this when we worked out citation output for oral histories.

[eddierubeiz]

For published oral histories, there's a distinction to be made (if we want to) between the interview date and the publication date. Ideally, the interview date would go in the title, and the publication date would go where it currently says n.d..

[jrochkind]

I don't believe we have the publication date recorded in our metadata, or ever have.

Looks like we did do some custom work on "cite as" for Oral Histories, at https://github.com/sciencehistory/chf-sufia/issues/1169

Still curious what it was looking like in chf_sufia; not sure if it's been this way since we "improved" it, and we didn't notice the problem (or thought it was okay at the time); or if it somehow got broken in between chf_sufia days and this app. The main reason to be curious is if it was working better in chf_sufia, that will give us a path to make it work better again. If it never was, there may no good way to get it to.

[yonyitz]

Really great point about the difference between the publisher date and interview date. Also some records have multiple interview dates. Not sure how to handle that...

[eddierubeiz]

Ashley she says the library needs a "publication date" to manufacture the call numbers affixed to the bound copies of the oral history transcripts. They just use the interview date for that purpose.

[yonyitz]

Cool. What if there are multiple interview dates?

[eddierubeiz]

They use the first.

[yonyitz]

Ok. That makes sense! Let me know if I need to do anything on my end!

[HKativa]

I'm on board with following the library's example and using the first interview date as the "publication" date for the purposes of citation. Also, having the date appear twice in the citation, per Jonathan's example at the top of the thread, works for me. As it stands, the issue isn't with the inclusion of "n.d." or another date in the citation, but rather that there exists a date we COULD input there rather than using "n.d."

[jrochkind]

As I discovered finding the issue on chf_sufia, what happened here is we were trying to make the citation look more like it looks in the "cite as" section of the OPAC for these records. (Of course, that wasn't necessarily "Chicago" style, but it was close).

Part of that was putting "on DATE" at the end of the title in the citation, even though it wasn't in our title metadata.

We thought we could have it only there, but at least in the current app with present versions of CSL libraries, that results in the n.d..

So, we aren't accomplishing what we meant to.

We could have the date "in two places". We could instead just stop trying to put it in the title as "on DATE" to match the OPAC-suggested citation, and have it only at the end in the proper Chicago slot. Is the latter preferable @HKativa ?

[HKativa]

@jrochkind thanks for the extra context. Are the options outlined above roughly equivalent in terms of developer time?

[jrochkind]

Yep, now that we've figured out what was going on (to some extent), it ought to be relatively easy to make either of those changes.

(I'm still a bit curious if we had this n.d. there from the start, or at one point had it working without a date there at all and no n.d.; cause I'm surprised we merged/deployed with the n.d., but maybe we just weren't reviewing carefully enough at that time).

[jrochkind]

Just for my own curiosity... yes, the "n.d." has been there since we did the work to improve OH citations ~1 year ago, which can be verified cause a test was included at the time for what we expected citation to look like, and "n.d." was there, ok! So OH citations have looked this way for a year, the app migration did not change it.

https://github.com/sciencehistory/chf-sufia/pull/1214/files#diff-e401084378064cd527614519b0825b75R393

[apinkney0696]

Not a priority.