It literally says in your excerpt "if there are more than six, list only the first three, followed by “et al.”" in the bibliography and says that notes without bibliography should follow the bibliography rules for et al
Et Al after three authors for works with more than 6 is implied by "as in the bibliography" in 13.23 -- you're welcome to ask CMOS about this, they do answer emails, but I promise you adunning's interpretation is what the Manual intends. It would make no sense to list more authors in notes than in a bibliography, not would the Manual establish a third different et al rule.
Huh. We read that very differently! I just queried a scholarly editors group about how to read that. I'll let you know. (I may end up writing CMoS for clarification)
Adam, the scholarly editors group is equally divided on how to read that. I've submitted the question to the Chicago Q&A. I'll report back when and if they respond.
@rudyleon Thanks for following up on that. The Chicago editors typically respond to queries within a month. The styles follow official guidance as closely as the software allows; I’m happy to make adjustments if my interpretation is inaccurate.
I've recently been using CMOS styles to write a paper, for which I use Zettlr for the Markdown and Zotero (via BetterBibTeX) to handle the referencing. One can then export the markdown file via Zettlr's internal pandoc to generate a new format; I generally send it to Word, since that is what is requested. Zettlr allows you to use the Zotero style files to do the rendering.
I recently swapped from 17 to 18 and noticed that with the change to 18 (and the reverse-engineering of 17 mentioned in these fora), that some of the forms of referencing have stopped working, whereas they worked perfectly in the previous version (of 17).
Specifically, the text used to indicate a chapter in a ref (i.e., "chap.") no longer appears to work, whereas "part", "figure", "Table", etc all seem to.
I find this to be so in most of the styles I've tried: notes; notes-bibliography; shortened-notes-bibliography; etc.
Example: the reference written in markdown in Zettlr as
[@kocs19io, chap. 7]
would when exported render, correctly, under the older CMOS 17 style as
(Kocs 2019, chap. 7)
However, under CMOS 18 and the revised 17, it now renders as
So, I am wondering: is the new CMOS deprecating the "chap." designator, or is there possibly a typo/bug in the style files that handle the "chapter" referencing?
I ask because using, e.g., Harvard Cite Them Right style, does insert the "chap." text in the Matlock citation above.
The idea here is to meet the requirement in Chicago (CMOS 14.143–54), and many other styles, for complex locators that do not require a label. I wonder however whether the numeric detection is too likely to create undesirable results.
I have a lot of references for the paper, and I have noticed that while most are rendered in title capitalisation, as expected, a handful are done in sentence caps.
This is odd, because I store all refs in sentence caps, with the APA caveat of a capital after the colon in a compound title (i.e., main title and a subtitle; a throw back to olden times... I think CMOS does do this now). I've double-checked this, and when I change the citation of one ref which only has a title to another with a subtitle (i.e., a later paperback edition), the title caps work as expected.
However, the sentence caps also occurs for a ref which has a compound title, so is there perhaps another pandoc obscurity that causes this?
@jvoros -- could you post the metadata for one or two of the items in question either here (in <code> tags) or on a github gist or so and link to them? Easier to just look at them then to go through all the possibilities.
@jvoros I suspect that you're encountering the difference in pandoc versus citeproc-js (the citation processor in Zotero) in the interpretation of languages. Pandoc follows the CSL specification strictly, which states that the value of language should be an ISO 639-1 (= IETF) language tag, i.e. en for English. Zotero interprets this more loosely.
Thus, if you have an item with English or eng as the Language, it will be rendered in title case in Zotero but sentence case in pandoc. Changing the language to en will fix the problem.
@dunning Yep, that fixed all of them! Amazing. I wish this forum supported Markdown, then I could do a :tada: party popper emoji to say thanks!
I also had one item for which the language was listed as "multiple" (it is a letter to the SecGen of the UN). It also rendered in sentence instead of title. Removing the text from the language entry fixed that too. All the anomalies I mentioned above are now gone... BUT, there are some new ones (of course ;-). Next post...
Many thanks for this; I have a query about citing manuscript letters and the like located in archival collections. The csl for CMOS 17 (full note, short title subsequent) rendered letters in archival collections as
Writer to Recipient, Date, Loc. in Archive, Archive.
Which is what I see recommended in CMOS 18 too, more or less; see for example item two in 14.120.
Unless I'm doing something wrong, it looks to me as if the csl for Chicago 18 (notes and bibliography) renders the same item as
Writer to Recipient, Date, Archive (Loc. in Archive).
Is the rendering of location in archive in parentheses actually a new recommendation in CMOS 18, and I'm failing to find it? I admit this is entirely possible and would be grateful for guidance. Otherwise, is there some way to revert to the way CMOS 17 rendered it?
@erauchway I fixed this bug a couple of weeks ago, but it is awaiting review. In the meantime, you can also fix this by adding the archive-place and/or archive_collection fields to Extra, as described here:
OK, so there seems to be a difference between how Zotero (using, I now understand, citeproc-js) renders podcast episodes and how Zettlr does it (using an 'internal' pandoc).
CMOS18 14.168 suggests the following template for podcasts (using notes style):
Name of Podcaster, host, Title of Podcast in Title Caps, podcast, season num, episode num, "Title of Episode in Title Caps", Publisher/Source, Date, URL.
No mention is made of what to do when there is a guest on the podcast: i.e., should the guest be the cited name or should it still be the podcaster?
Using the CMOS18 notes-with-bibliography style in the Item Pane header, the following is produced (it is the Bibliography form of course since the Item Pane header does not allow the choice of a Note rendering -- at least, not yet):
So, is this an example of the difference between citeproc-js and pandoc? At any rate, I like the use of the guest's name as the citation anchor (when there is one), since that seems to be of more interest (and it is what I ended up doing for the paper I wrote). But CMOS does not seem to have a stance on this, except insofar as they say to cite podcasts in a similar way to TV eps and series.
I am not criticising here; I'm just wondering where the disconnect might be, in the spirit of trying to work out what is going on. And for how to get these two excellent programs to work together even better (as I said over on the Zettlr forums).
BTW, the metadata for the above citation is:
Item Type: Podcast Title: Is there such a thing as a rules-based international order? Podcaster: Gideon Rachman Guest: G. John Ikenberry Date: 2023-04-20 URL: https://www.ft.com/content/664d7fa5-d575-45da-8129-095647c8abe7 Series Title: The Rachman Review podcast Running Time: 27m
Hmm. I knew about podcaster not being part of CSL (from reading through the comments in @dunning's CSL files).
I found that by deleting Podcaster in the above data, and putting
Director: Rachman || Gideon
into Extra, the second of the above citations is produced in the Item Pane header.
But it's still a bit strange -- I looked through the JSON file that BetterBibTeX produces, and anyone listed as podcaster is output as "director" in the Better CSL JSON file, so it really shouldn't have made any difference.
However, if I export the library using CSL JSON (not Better CSL JSON), then indeed the podcaster is rendered as an author.
Zotero incorrectly maps its 'podcaster' field to CSL author rather than host;
BetterBibTeX is overriding an author on any film/broadcast items (this isn't specific to podcasts) as director to fix Zotero's formerly incorrect mapping of directors as authors;
Zotero provides a Series Title field for podcasts rather than the Program Title used on radio/TV broadcasts, which maps to the CSL collection-title rather than the container-title variable that the specification requires for broadcasts.
I made the same assumption as you that if a radio or podcast episode has both a guest and a host, it should be cited under the name of the guest, on the model of an interview. For examples of this (again requiring the Extra field), I have an episode of BBC In Our Time in the Test Items Library and an episode of Just One Thing.
@dunning This is fantastic to know. I've updated the podcasts mentioned in my paper and the formatting works as needed except if there is no episode number (I have a couple of these, since not all podcasts number their episodes). I've tested this on author-date and notes.
For example, the Rachman-Ikenberry item shown above renders with the container-title and title reversed (i.e., the "title" first, then the italic container-title second).
It's too late to change them in the paper, since that has been submitted and is now on its way to the interweb. Nonetheless, I am redoing all the podcasts in my library to use this format for future use. Thanks!
@jvoros There are no examples of broadcast episodes in CMOS without an episode number, so I chose to use the reversed format only if there is both a title and episode number. I have absolutely no idea which approach is correct, but I'll write CMOS to find out. I've added your example to the Test Items Library for future testing:
@dunning It strikes me that, for consistency of form, the same format would make sense, with or without a season/episode number to necessarily distinguish the specific episode. If I try to think like a CMOS editor might, I would be led down that path, since the specific episode comes after the collection title, whether or not it is distinguished by a season/number (which I would consider like a locator in the collection). This is the same way of thinking that led me to use the guest name as the citation anchor above.
That said, I can also see how it would work the other way, like an analytic work title inside a monographic work title. :shrug:
I've now signed up to CMOS and, dammit, will have to learn CSL too, at this rate :-)
@jvoros They seem to be treating a broadcast as something in between a serial publication and a multivolume monographic work. Compare MHRA 7.8 for a similar hybrid. I've written CMOS and hope that we'll find out the answer in a month or two.
@jvoros See the Buolamwini example in the Quick Guide for placing the episode name of a broadcast before the programme title without an episode number:
@dunning Yes, I see that it "works" as a citation. Like I said above: it works both ways, and I can see a certain logic in having the non-numbered title before the programme title, modelled on interviews (and the analytic-monographic nesting. Did I say that I used to use ProCite back in the day, and BibTeX before that?).
I want to reiterate @erauchway 's grateful thanks. Twenty-five years ago I used to geek out with the library staff around citation styles, and was involved in creating one for the Masters program I was teaching into. The students were sometimes a bit rebellious about it but, in a lovely twist of fate, one of them ("for my sins", she said) later became the manager of the Australian Govt Style Manual renewal process. Ha! The Universe sometimes does have a sense of humour.
I've peered into the innards of the style files you wrote (with much trepidation) and now really want to learn to code in CSL. Back as a doctoral student (*ahem* quite a while ago now) I wrote my own LaTeX format for my very-mathematical thesis. Then LaTeX2e came out and made it unnecessary, but I do understand the attraction of bending the code to one's will :-)
Of course this raises another edge case (naturally!):
an un-numbered podcast episode without a guest. Analytic-monographic, or monographic-analytic?
Upgrade Storage
pandocto generate a new format; I generally send it to Word, since that is what is requested. Zettlr allows you to use the Zotero style files to do the rendering.I recently swapped from 17 to 18 and noticed that with the change to 18 (and the reverse-engineering of 17 mentioned in these fora), that some of the forms of referencing have stopped working, whereas they worked perfectly in the previous version (of 17).
Specifically, the text used to indicate a chapter in a ref (i.e., "chap.") no longer appears to work, whereas "part", "figure", "Table", etc all seem to.
I find this to be so in most of the styles I've tried: notes; notes-bibliography; shortened-notes-bibliography; etc.
Example: the reference written in markdown in Zettlr as
[@kocs19io, chap. 7]
would when exported render, correctly, under the older CMOS 17 style as
(Kocs 2019, chap. 7)
However, under CMOS 18 and the revised 17, it now renders as
(Kocs 2019, 7).
Whereas, something like
[@matlock10si, chap. 1; @matlock24lbla, part III]
renders as
(Matlock 2010, 1; 2024, pt. III).
So, I am wondering: is the new CMOS deprecating the "chap." designator, or is there possibly a typo/bug in the style files that handle the "chapter" referencing?
I ask because using, e.g., Harvard Cite Them Right style, does insert the "chap." text in the Matlock citation above.
Anyone else notice this behaviour?
<else-if locator="chapter line verse" match="any"/>from the CSL file?Tried it in
shortened-notes-bibliographyand it worked!I just commented out the line.
https://github.com/jgm/citeproc/issues/166
The idea here is to meet the requirement in Chicago (CMOS 14.143–54), and many other styles, for complex locators that do not require a label. I wonder however whether the numeric detection is too likely to create undesirable results.
This is odd, because I store all refs in sentence caps, with the APA caveat of a capital after the colon in a compound title (i.e., main title and a subtitle; a throw back to olden times... I think CMOS does do this now). I've double-checked this, and when I change the citation of one ref which only has a title to another with a subtitle (i.e., a later paperback edition), the title caps work as expected.
However, the sentence caps also occurs for a ref which has a compound title, so is there perhaps another
pandocobscurity that causes this?enfor English. Zotero interprets this more loosely.Thus, if you have an item with
Englishorengas the Language, it will be rendered in title case in Zotero but sentence case in pandoc. Changing the language toenwill fix the problem.Yep, that fixed all of them! Amazing. I wish this forum supported Markdown, then I could do a
:tada:party popper emoji to say thanks!I also had one item for which the language was listed as "multiple" (it is a letter to the SecGen of the UN). It also rendered in sentence instead of title. Removing the text from the language entry fixed that too. All the anomalies I mentioned above are now gone... BUT, there are some new ones (of course ;-). Next post...
Unless I'm doing something wrong, it looks to me as if the csl for Chicago 18 (notes and bibliography) renders the same item as Is the rendering of location in archive in parentheses actually a new recommendation in CMOS 18, and I'm failing to find it? I admit this is entirely possible and would be grateful for guidance. Otherwise, is there some way to revert to the way CMOS 17 rendered it?
https://forums.zotero.org/discussion/comment/496135/#Comment_496135
citeproc-js) renders podcast episodes and how Zettlr does it (using an 'internal'pandoc).CMOS18 14.168 suggests the following template for podcasts (using
notesstyle): No mention is made of what to do when there is a guest on the podcast: i.e., should the guest be the cited name or should it still be the podcaster?Using the CMOS18 notes-with-bibliography style in the Item Pane header, the following is produced (it is the Bibliography form of course since the Item Pane header does not allow the choice of a Note rendering -- at least, not yet): Interestingly, when this reference is exported from Zettlr, using the same style as used in the Item Pane header, this is the output: So, is this an example of the difference between citeproc-js and pandoc? At any rate, I like the use of the guest's name as the citation anchor (when there is one), since that seems to be of more interest (and it is what I ended up doing for the paper I wrote). But CMOS does not seem to have a stance on this, except insofar as they say to cite podcasts in a similar way to TV eps and series.
I am not criticising here; I'm just wondering where the disconnect might be, in the spirit of trying to work out what is going on. And for how to get these two excellent programs to work together even better (as I said over on the Zettlr forums).
BTW, the metadata for the above citation is:
podcasternot being part of CSL (from reading through the comments in @dunning's CSL files).I found that by deleting Podcaster in the above data, and putting
Director: Rachman || Gideon
into Extra, the second of the above citations is produced in the Item Pane header.
But it's still a bit strange -- I looked through the JSON file that BetterBibTeX produces, and anyone listed as podcaster is output as "director" in the Better CSL JSON file, so it really shouldn't have made any difference.
However, if I export the library using CSL JSON (not Better CSL JSON), then indeed the podcaster is rendered as an author.
Interesting...
- Zotero incorrectly maps its 'podcaster' field to CSL
- BetterBibTeX is overriding an author on any film/broadcast items (this isn't specific to podcasts) as
- Zotero provides a Series Title field for podcasts rather than the Program Title used on radio/TV broadcasts, which maps to the CSL
It's possible to fix these problems using the Extra field: for a model, see the example podcast used in the CMOS 14.168 and the quick guide.authorrather thanhost;directorto fix Zotero's formerly incorrect mapping of directors as authors;collection-titlerather than thecontainer-titlevariable that the specification requires for broadcasts.I made the same assumption as you that if a radio or podcast episode has both a guest and a host, it should be cited under the name of the guest, on the model of an interview. For examples of this (again requiring the Extra field), I have an episode of BBC In Our Time in the Test Items Library and an episode of Just One Thing.
This is fantastic to know. I've updated the podcasts mentioned in my paper and the formatting works as needed except if there is no episode number (I have a couple of these, since not all podcasts number their episodes). I've tested this on author-date and notes.
For example, the Rachman-Ikenberry item shown above renders with the container-title and title reversed (i.e., the "title" first, then the italic container-title second).
It's too late to change them in the paper, since that has been submitted and is now on its way to the interweb. Nonetheless, I am redoing all the podcasts in my library to use this format for future use. Thanks!
https://www.zotero.org/groups/2205533/items/BS7KWXQI
It strikes me that, for consistency of form, the same format would make sense, with or without a season/episode number to necessarily distinguish the specific episode. If I try to think like a CMOS editor might, I would be led down that path, since the specific episode comes after the collection title, whether or not it is distinguished by a season/number (which I would consider like a locator in the collection). This is the same way of thinking that led me to use the guest name as the citation anchor above.
That said, I can also see how it would work the other way, like an analytic work title inside a monographic work title. :shrug:
I've now signed up to CMOS and, dammit, will have to learn CSL too, at this rate :-)
https://www.chicagomanualofstyle.org/tools_citationguide/citation-guide-1.html#cg-interview
Yes, I see that it "works" as a citation. Like I said above: it works both ways, and I can see a certain logic in having the non-numbered title before the programme title, modelled on interviews (and the analytic-monographic nesting. Did I say that I used to use ProCite back in the day, and BibTeX before that?).
I want to reiterate @erauchway 's grateful thanks. Twenty-five years ago I used to geek out with the library staff around citation styles, and was involved in creating one for the Masters program I was teaching into. The students were sometimes a bit rebellious about it but, in a lovely twist of fate, one of them ("for my sins", she said) later became the manager of the Australian Govt Style Manual renewal process. Ha! The Universe sometimes does have a sense of humour.
I've peered into the innards of the style files you wrote (with much trepidation) and now really want to learn to code in CSL. Back as a doctoral student (*ahem* quite a while ago now) I wrote my own LaTeX format for my very-mathematical thesis. Then LaTeX2e came out and made it unnecessary, but I do understand the attraction of bending the code to one's will :-)
Of course this raises another edge case (naturally!):