CMOS-18 misc issues

[usethe4ce]

Trying out the relatively new Chicago Manual of Style 18th edition (#7424, thanks for the huge feat!), and ran into a series of small issues (I'm using Juris-M 6.0.30m3), which I fixed for myself on my fork here. Making a PR looks extra complicated in this case, but here are the issues:

1. Title case (8.160) I was able to configure <style-options skip-words="..." /> to filter down the built-in list of stop words to exclude prepositions longer than four letters. Still not perfect, but strictly better and often good enough.
2. Books published before 1900 (14.31) I've worked around this by defining a custom field old-publisher:true in all my items published before 1900 and modifying the style to respect that field.
3. Translators ("trans.") Before: Frank Williams, tran., The Panarion of Epiphanius of Salamis, Books II and III. De fide, 2nd ed., Nag Hammadi and Manichaean Studies 79 (Brill, 2013). After: Frank Williams, trans., The Panarion of Epiphanius of Salamis, Books II and III. De fide, 2nd ed., Nag Hammadi and Manichaean Studies 79 (Brill, 2013).
4. Reviews ("review of") Before: Henri-Charles Puech, Les Lettres grecques en Occident. De Macrobe à Cassiodore, by Pierre Courcelle, Revue de l’histoire des religions 131 (1946): 187–97. After: Henri-Charles Puech, review of Les Lettres grecques en Occident. De Macrobe à Cassiodore, by Pierre Courcelle, Revue de l’histoire des religions 131 (1946): 187–97.
5. Number of volumes ("vols.") Before: René Henry, ed., Photius: Bibliothèque, 9 (Les Belles Lettres, 1959–77). After: René Henry, ed., Photius: Bibliothèque, 9 vols. (Les Belles Lettres, 1959–77).
6. Abbreviations (using Juris-M's abbreviation support) I added form="short" where missing for container-title and collection-title Before: F. Stanley Jones, “Marcionism in the Pseudo-Clementines,” in Pseudoclementina Elchasaiticaque inter Judaeochristiana: Collected Studies, Orientalia Lovaniensia Analecta 203 (Peeters, 2012). After: F. Stanley Jones, “Marcionism in the Pseudo-Clementines,” in Pseudoclementina Elchasaiticaque inter Judaeochristiana: Collected Studies, OLA 203 (Peeters, 2012).
7. For the "without URLs" variants, shouldn't web sites still include the URL? (Just ran into this, haven't fixed it yet.) Before: Evina Steinová, “The Innovating Knowledge Database,” accessed August 1, 2025. After: Evina Steinová, “The Innovating Knowledge Database,” accessed August 1, 2025, https://innovatingknowledge.nl/.

(In case not all of these changes are adopted in the style here, at least anyone trying to accomplish the same thing can follow my path.)

[bwiernik]

@dunning could you take a look?

[adunning]

Thanks for these ideas!

1–2. These are not part of the CSL specification so unfortunately they can't be included. 3–5. It looks as if you're using a very old locale file. This is probably due to Juris-M not having been updated in two years; I don't know whether it's possible to replace the files manually. You'll see other problems as well if you're using outdated files. 6. I don't believe that CMOS specifies this? I included something similar to this in the original version but Zotero users were confused by the abbreviations checkbox that appeared as a result, so it was subsequently removed. 7. The relevant section is CMOS 13.6:

For sources consulted online, authors should record a URL (e.g., https://www.chicagomanualofstyle.org/) as the final element in a citation that includes all the components described throughout this chapter and in chapter 14. Whether the URL is retained for publication may depend on whether readers will need the URL to locate or assess the source. Journal publishers will often create links to sources cited in their articles automatically, whether a URL is provided or not (a process that is facilitated by accurate citation data). Book and journal publishers may retain URLs in citations of sources that would be difficult to locate without one but not in citations of journal articles, books, and other formally published sources that would be easy to find online from a title and other basic details alone. For many readers, a URL, which is designed to be read by a computer, will be less helpful—even as a hyperlink—than a link to the cited source from its title or other reader-friendly (and preferably unique) component of the citation (using the same URL as the link target).

The only way I have seen to program this is to treat the display of URLs as an all-or-nothing affair, but there is obviously a point of ambiguity in 'sources that would be difficult to locate'. Web pages are not necessarily more difficult to locate than a journal article using a search engine, so I don't think that changing the display based on the type of item would make sense. I'm certainly open to other suggestions for ways to implement this section.

Your idea for dealing with the number of volumes https://github.com/citation-style-language/styles/commit/7a7dc7a2536074a6b816987f0927c279e9415e15 is very helpful and I've written a variant on this.

[usethe4ce]

Thanks for the detailed response! I guess then there's no action to be taken here.

[adunning]

I wouldn't say there is no action to be taken: see above for my pending adjustment to the display of the number of volumes in response to your suggestion, for which I'm grateful.

[usethe4ce]

Here's one more (for which I've also pushed an attempted fix on my fork):

8. For shortened notes with a volume title and only editors, the editors should be shown in place of authors Unshortened note: The Ante-Nicene Fathers, vol. 1, The Apostolic Fathers, Justin Martyr, Irenaeus, ed. Alexander Roberts and James Donaldson (Christian Literature, 1885), 77. Before: Ante-Nicene Fathers, 1:77. After: Roberts and Donaldson, Ante-Nicene Fathers, 1:77.

[adunning]

See CMOS 14.20 for the model that I followed. The Henry James example is the crux: I have interpreted this as meaning that the editor should be omitted in shortened notes in the same circumstances as the full note, although it's possible that they omitted the editor's name in this example because the are considering Henry James to be the author, whose name is part of the title (this is a situation that CSL cannot handle).

[usethe4ce]

I see. That sure is puzzling for a normative example. If your (first) interpretation is right, the author (Henry James, as in 14.19) has even been omitted in the unshortened citation simply because of the presence of a volume title, which doesn't make much sense to me. And overall that's what seems so strange to me here, that specifying a volume title would cause both author and editor to be suppressed (which, for the case of editor without author, seems contrary to the spirit of 14.5). So, your alternative idea seems more likely, that the author has been suppressed due to redundancy in the title, even though I don't see that principle expressed elsewhere and this would be an especially confusing occasion to implicitly introduce it. If that's correct, of course I wouldn't expect CSL to proactively eliminate any such redundancy (which presumably is optional) but would expect an author/editor to still appear in the shortened note.

[adunning]

I looked again and found that CMOS 13.80 specifies the case of omitting an author's name from the note if it appears as part of the title. That must be the explanation for sect. 14.20 as well as the John Donne and Euripides examples in sect. 14.24. It's doubly confusing because styles such as MHRA specify that editors should be always omitted from abbreviated notes.

I've added a commit to #7811 that includes your suggestion: https://github.com/citation-style-language/styles/pull/7811/changes/a20f259e8d8328e20b31252081b113ea29e7b951. For consistency, this also shifts to using the first listed forms of the two versions in CMOS of the order of titles in bibliographies. Thanks for catching this!