add "word" locator - any objections?

Any objections to adding a “word” locator, for reference works:

17.238 Reference works
Well-known reference books, such as major dictionaries and
encyclopedias, are normally cited in notes rather than in
bibliographies. The facts of publication are often omitted, but the
edition (if not the first) must be specified. References to an
alphabetically arranged work cite the item (not the volume or page
number) preceded by s.v. (sub verbo, “under the word”; pl. s.vv.)

Thanks,
Elena

I’m a little worried about this one.

As a preface, as I’ve said before, I don’t think it should be our goal
to support every little detail of every single style, particularly in
cases like this where the rules themselves are vague (words like
"often" and “normally”). So the criterion we base these decisions on
is not “because style X says so.” In general, if it’s needed, and
consistent with existing variables and such, then we just add them.

But I see this example a little differently.

First, it’s basically one of those dumb rules that’s legacy from
centuries of practice, but which makes little sense in the 21st
century. That kind of citation is hostile to readers (the
latin-derived acronyms). It also offers no advantages to authors with
automated tools like we’re discussing. If I cite a definition from the
OED, doesn’t a page number or URL (or even a prefix or suffix) seem
more than adequate, and clear?

I’d also think it would add implementation overhead, since I’m not
sure this is just a question of adding a new variable, It seems to
require special handling according to the bit you cite above?

Bruce

First, it’s basically one of those dumb rules that’s legacy from
centuries of practice, but which makes little sense in the 21st
century. That kind of citation is hostile to readers (the
latin-derived acronyms). It also offers no advantages to authors with
automated tools like we’re discussing. If I cite a definition from the
OED, doesn’t a page number or URL (or even a prefix or suffix) seem
more than adequate, and clear?

fine w/ me–i agree, this is a rare case that can be ignored.

I’d also think it would add implementation overhead, since I’m not
sure this is just a question of adding a new variable, It seems to
require special handling according to the bit you cite above?

i was going to add the locator but it would be impossible to add the
extra conditionals–how would zotero decide if the book is a “well-
known reference work” and should be cited without “facts of
publication” or a “little-known” one that should include them?

there are quite a few silly rules like in CMS that make little sense
and are impossibly hard to program.

elena> Bruce

Bruce–

The issue of encyclopedia articles came up again in a conversation
about Zotero types and it’s pretty clear that Zotero has to support
"s.v." citation form for reference works in Chicago, however archaic
(see the original description/discussion below). It will require some
extra conditionals, but that won’t be a problem–I just need a new
locator for “s.v.” and “s.vv.”

Given that this is necessary for Zotero, can I add a "sub verbo"
locator?

Thanks,
Elena

But can you respond to the concerns I outlined in my original
response? I don’t recall the details, but I was in general worried
that we’d be imposing an undue implementation burden on all CSL
implementations just to support this. The “well-known” exception seems
particularly problematic.

Bruce

Bruce–

Well, there isn’t any burden for CSL implementations, except to create
and map an encyclopedia/reference work type (which is already allowed
in CSL). If implementers don’t want to do that, they can just omit the
type and have users cite encyclopedias as chapters in books. Do you
see any other implementation problems?

After asking around it seems that while there are some leeway on
deciding whether a particular reference work is well-known, enough
scholars expect some frequently cited reference works (such as OED) to
be cited with “s.v.” in Chicago to make this rule necessary rather
than optional.

Thanks,
Elena

Mainly conditional logic specific to this example, and how tools are
supposed to distinguish different classes of reference works

Maybe this isn’t really that complicated though. If someone really
wants to use this method, then they can just add that point locator,
and the CSL logic to print it is easy (and so trivial to implement for
the code). If they don’t (I, for example, would just cite the OED
without it), then that’s fine too.

E.g. leave it to the user to determine it through their choice of
whether to indicate the poit locator (or not).

Does that make sense?

Bruce

Bruce–

Maybe this isn’t really that complicated though. If someone really
wants to use this method, then they can just add that point locator,
and the CSL logic to print it is easy (and so trivial to implement for
the code). If they don’t (I, for example, would just cite the OED
without it), then that’s fine too.

E.g. leave it to the user to determine it through their choice of
whether to indicate the poit locator (or not).

Does that make sense?

So what you’re saying, we’ll add the locator label to the locales
files and the implementor would decide whether to put it in the
pulldown menu for the point-locator? That would be great, thanks! If
that’s your thinking, I’ll go ahead and add the locator.

Best,
Elena

So what you’re saying, we’ll add the locator label to the locales
files …

And in this case that locator is “word”?

and the implementor would decide whether to put it in the
pulldown menu for the point-locator?

… and, in turn, the user.

That would be great, thanks! If
that’s your thinking, I’ll go ahead and add the locator.

OK.

Bruce