Chicago/MLA ellipsis versus the Unicode defined AP ellipsis

Tue Apr 18 16:22:00 CDT 2023

On 4/18/2023 11:43 AM, Mark E. Shoulson via Unicode wrote:
> I'm not entirely sure what the fuss is about.  As Ken said, "an 
> ellipsis is an ellipsis is an ellipsis," and I have to agree with 
> that.  U+2026 HORIZONTAL ELLIPSIS is only U+002E U+002E U+002E in 
> *compatibility* decomposition, as far as I can see, and compatibility 
> was never meant to cover fine points of typography. If you need an 
> ellipsis that "looks" more bunched-up or more stretched-out, those are 
> glyph variants, not even variation sequences, right?  And besides, 
> when actually typeset well, I think there isn't much difference as to 
> what a well-typeset ellipsis should look like for a given font (that 
> is, when looking at a book, on paper, do CMoS fans expect ellipses to 
> be that much more spaced out that AP fans?  Do they really look wrong 
> to each other?)  Maybe they do, as you speak about "submitting 
> manuscripts."

If you drill down, you find that the concept of an "ellipsis" and the 
abstract character for "horizontal ellipsis" (as currently conceived) 
are not fully congruent. And, as was reported here, there is established 
(and recommended) practice of using character sequences. (Plus there are 
subtle interactions with surrounding punctuation which are beyond simple 
glyph design variation.)

Taken all this together, the preferred action is to document such 
practice, instead of pretending that the "horizontal ellipsis" covers 
every possible expression of the concept of an "ellipsis". It happens to 
work well if you want something that is a single character and has a 
moderately compact representation, but it's not something that works, or 
is even preferable in all contexts.

This is one of those instances when Unicode can and should be 
descriptive, instead of being prescriptive.

As we speak the text of the standard is being updated to counter the 
presumption that it's just a matter of always using "horizontal 
ellipsis" and trusting the glyph design in the selected font.

A./

>
> But I'm still not getting it.  When you say that the AP requires 
> tightly-packed dots because news copy is space-conscious, that has to 
> do with how they PRINT things, right?  It doesn't matter how loosely 
> kerned the dots are in a reporter's print-out, because that's not what 
> takes up space in their columns: it's what they *print* that needs to 
> be space-conscious.
>
> So it sounds like the issue here mainly involves the electronic 
> submission of manuscripts, when you email plain text in to the AP or a 
> publisher.  There's no such thing as "plain text" once it's on paper; 
> print is as print looks.  So in electronic correspondence, the AP 
> prefers you write an ellipsis as U+002E U+002E U+002E, while the CMoS 
> standard says you should use U+002E U+00A0 U+002E U+00A0 U+002E.  Is 
> this right, so far?
>
> In that case, *neither one* seems to be asking for U+2026 HORIZONTAL 
> ELLIPSIS.  This sounds like two standards for "how to write an 
> ellipsis when all you have is periods and spaces."  If you have an 
> actual ellipsis character, then either standard can easily accept it 
> whether it looks all crunched up or all spaced out or like a sparkly 
> unicorn (how it looks is a matter for a font to determine, not "plain 
> text.")  Or either standard can decide not to accept it at all.
>
> To be sure, this makes for some weirdness, when you have a monospace 
> font, which by definition means that "all characters must take up the 
> same width."  If you consider an ellipsis to be a single character, 
> then, yes, you'll get a horribly crunched-up ellipsis no matter whose 
> standard you prefer.  But that's what you get for taking a symbol 
> that's designed to be wide and forcing it to conform.
>
> IOW, I don't see this is something to do with Unicode.  At best it's a 
> glyph variant, if even that.
>
> ~mark
>
> On 4/17/23 15:50, t0dd via Unicode wrote:
>> Hello all,
>>
>> Narrative writers working in the English language, and in particular 
>> the US (I can't speak for the rest of the English-language world), 
>> are generally required to adhere to the Chicago Manual of Style 
>> (CMoS) when submitting manuscripts and screenplays for publication. 
>> News people generally follow the AP (Associated Press) style. The 
>> rub: they each use a different ellipsis. The CMoS requires three dots 
>> spaced apart. The AP, because news copy is space-conscious, requires 
>> dots tightly packed.
>>
>> Other style guides follow one or the other, but most follow the the 
>> Chicago style or they are indifferent. For example, in school many of 
>> you were required to follow the MLA style guide. That also requires a 
>> spaced-out "Chicago" ellipsis (I am just going to call it that from 
>> here on out). Conversely, if you wrote for the Psychology Review, you 
>> follow the APA style which adheres to the "AP" ellipsis.
>>
>> Unicode only supplies one horizontal ellipsis: U+2026. The AP 
>> ellipsis. This ellipsis is constructed via three periods with no 
>> additional spacing: U+002E U+002 EU+002E under the covers. (Spaces 
>> between the codes here have been added for readability.)
>>
>> That construction is not sufficient. Ironically, the most commonly 
>> needed ellipsis is not the one defined by Unicode. The more common 
>> need is for something constructed with three-periods separated by 
>> three non-breaking-spaces. I.e., something like U+002E U+00A0 U+002E 
>> U+00A0 U+002E. Again, treated as a solitary character and 
>> unbreakable. And, of course there are repercussions if it lives next 
>> to a sentence-ending period, or if it is adjacent to a quotation 
>> mark. Etc.
>>
>> What most writers do to get around this issue is find-and-replace all 
>> ellipsis characters with three periods spaced out. But that doesn't 
>> word wrap correctly. Slightly more savvy writers find-and-replace all 
>> ellipsis characters with three periods separated by a non-breaking 
>> space (see above). Or they change the character spacing style within 
>> their word-processing application for their three-period "word". Or 
>> they just use the AP ellipsis and hope no one cares.
>>
>> It should be noted that grammar and spell checkers see these 
>> user-generated constructions as errors.
>>
>> This is ugly. There really needs to be a Unicode character that 
>> supports the Chicago ellipsis.
>>
>> None of the word processing packages builds any robust workaround for 
>> this. LaTeX has an ellipsis package to work around this and the 
>> associated complexities 
>> (https://tug.ctan.org/macros/latex/contrib/ellipsis/ellipsis.pdf is 
>> really worth the read), but that's not ideal. LaTeX is not software 
>> designed for the Everyman.
>>
>> I hear rumor that some typefaces come with stylistic alternatives to 
>> address this, but that's not the case with any typeface that I have 
>> ever had to use as required by a publisher (namely New Times Roman). 
>> Plus, that's . . . kludgy.
>>
>> So . . . please. Someone. Advocate for supporting a spaced-out 
>> ellipsis so that all of us who have to adhere to a standard that is 
>> not the AP Style don't have to do bizarre find-and-replacey things or 
>> other workaronds. Newspapers are dead, haven't you heard? 😉
>>
>> We all have access to an em-dashes and en-dashes and other dashes. A 
>> Chicago-styled ellipsis (for lack of a better nomenclature) is way 
>> way overdue IMHO
>>
>> What think y'all? Note, I just joined the mailing list in order to 
>> voice this. Be kind please. :)
>>
>> Cheers. -t
>>
>>
>> P.S. NOTE: This topic has been touched upon a bit in the past, but 
>> not quite exactly the same ask. (Reference: 
>> https://www.unicode.org/mail-arch/unicode-ml/y2006-m01/0164.html) 
>> That thread devolved into lovely poetry. Worth the read. ;) I digress 
>> . . .
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://corp.unicode.org/pipermail/unicode/attachments/20230418/09446a69/attachment.htm>