I can't say I've got any experience with it, but wouldn't it make more sense to use U+0242 (with uc U+0241)?

I always saw U+0294 as for almost exclusive use in IPA encoding, and U+02BC as just that, a modifier letter (which means it forms an unbreakable semantic unit along with the letter(s) it modifies).

U+0027 and U+2019 just seems to be taking glyphs instead of characters. Of course I suppose if they REALLY want to encode it one way or another you can just take your prefered glottal stop character and (license permitting) place it in U+0027.

U+02BC is the correct character for the apostrophe-shaped letter used in many orthographies to indicate a glottal stop. Here is the description for 'modifier letter apostrophe' from the relevant Unicode Chart:

* glottal stop, glottalization, ejective
* many languages use this as a letter of their alphabets
* used as a tone marker in Dogri (swarbal chinha)
* used as a length marker in Bodo (gojocoma)
* U+2019 is the preferred character for a punctuation apostrophe

Unicode including this as a 'modifier' letter is a bit unfortunate, but the above description makes it clear that this is the character required.

U+0242 and U+0241 don't apply here as they are different letters altogether. They are letters that frequently represent the same sound as U+02BC, i.e. a glottal stop, but in different orthographies with different casing behaviour (U+02BC doesn't come in separate lowercase and uppercase versions). They are different letters in the same way that 'thorn' and 'theta' are different letters although they generally represent the same sound.

Returning to the original post, Charles, could you clear up for us whether the author is insisting that the U+0027 glyph is used, or that the U+0027 code point is used?

The author is insisting on a straight up & down mark. I'm taking this to mean U+0027. BTW, I'm not in contact with the author. The chain, typical for a compositor, is comp > designer > editor > author.

* * *

I believe I put the original question poorly. At this point, what I'm asking is really a language question, a *glyph* question, rather than a purely type or phonetics question. I posted to this forum because there is a tremendous level of expertise on several levels. Originally, I though the question similar to "how to draw a proper ogonek in Polish" that Adam gave us a while back, but I see that is perhaps wrong.

Wrong because the proper glyph to signal a glottal stop varies across languages. In Polynesian, U+2018 shows the proper glyph. In Apache, Navajo, etc., U+2019 is the character that shows the proper appearance (even if not the "proper character"). Occasionally, the "phonetic" glottal stop glyph is used in the language itself, not just in analysis of the language.

But things are never simple, what I want to know is "what's the proper character (glyph) in Mayan if you have the whole typecase to choose from? Keep in mind that the character U+0027 is erroneously used for the glottal stop generally on several internet sites. For instance,

http://www.omniglot.com/writing/apache.htm

Notice that the glottal stop seems to be the "vertical apostrophe" U+0027. But if you poke around on the site -- fairly valuable, generally -- you notice that U+0027 is all there is. It's used for "English" apostrophes; U+2019 never occurs. Same problem with Wikipedia.

I have never seen a scholarly use of U+0027 for a glottal stop, and I've seen quite a few varying orthographies over the years. What I have seen is indigenous people pick up information from various places, and with the advent of PostScript, create their own fonts. For example, in the early days of Fontographer, the Navajo schools made up a Navajo font. Good for them. But somebody had described the "hook" (AKA ogonek) as a "reversed cedilla," so they took a cedilla & reversed it.

Who am I to tell the Navajo they're wrong about their own language? (But I did notice that they've now redone the font, and the ogonek is used instead of the reversed cedilla.)

Or, Mayan writers may have decided that they want control over their language, disregarding what scholars have done. Their right. What I want to guard against is the use of a glyph because it was the only thing available to an author, or because an internet site such as Omniglot gave the wrong information.

The book is to be set in an Adobe font, which I can modify. Since I'm the end user, there is essentially the whole typecase available. Anybody know what's the proper glyph in Mayan?

Thanks, Charles. That's much clearer.

It's always confusing when it comes to the typographic treatment of the apostrophe, the prime mark, and other characters that are superficially similar when written by hand.
Because of the ease of input, they are typically substituted by U+0027, especially on the web.

For the apostrophe in English and other major languages, we can confidently say that U+2019 is the right character, and that the raised comma is the correct shape. If you received a manuscript in English that used the 'vertical apostrophe' throughout, you wouldn't think twice about changing it to the correct form.

With something like the glottal stop in Ixil Mayan, it's more complicated, especially if there is no real established typographic tradition. However, I think that as the usual apostrophe glyph has been used in the past for Mayan languages, there is no reason to use a 'vertical apostrophe' glyph unless there is an explicit indication of a change in practice. I'd ask a Mayan linguist, but there's no guarantee that you'd find an expert on Mayan typography.

The images http://here (p3) are unfortunately a bit low-res for seeing punctuation marks clearly... but they do look suspiciously like 0027 don't they? (Alongside definite curly apostrophes in the English text)
_______________________________________________
Ever since I chose to block pop-ups, my toaster's stopped working.

Thanks all.

I did get an email, then a phone all, from someone who had done extensive work, including work on a scholarly level, with Ixil Mayan. According to him, the preference is indeed for the straight up and down mark, typical of U+0027. He did indicate that if, for some reason, U+0027 is inclined in the font, then another mark should be used.

I'm certainly no expert, but with this, I'm satisfied that the straight up & down mark is acceptable for Ixil. There may be competing orthographies with other scholars, but that happens.

* * *

In theory, there is no difference between theory and practice. But in practice, there is. -- Yogi Berra

Let's hear it for Yogi Berra.

I was worrying about whether I'd been handling the glottal stop in "Hawai‘i" correctly -- the U. of. H Press occurs frequently in bibliographies I set. According to charles_e and Jeongsong, I'm probably safe using the single open quote char. (U+2018, a.k.a Alt-0145 in Windows). Guifa's note sent me scurrying to the Unicode charts to confirm that the IPA U+02BC isn't in the fonts I tend to use, but its reverse U+2BB sometimes can be found; Unicode's description includes "used in Hawai`ian orthography as `okina (glottal stop)." So maybe I should have been using that?

Checking the UH Press website, I see that they use both U+2018 and U+0060 (grave) on a single page. And to get really picayune, "Hawaiian" is an English word, so it doesn't get the glottal stop (apparently rendered in the Unicode description with a grave).

David

David,

What you are doing is fine. Polynesian -- which of course, is the source of the word *Hawai‘i* -- uses the single open quote *glyph* to signal a glottal stop. Strictly, from a *character* point of view, it should be U+02BB, for "turned comma" rather than U+2018. Of course, the origin of what we now call "single open quote" (AKA U+2018) was -- you guessed it, a turned comma.

Gufta (Matthew)'s comments apply more to the analysis of language, rather that what various languages use to signal things like a glottal stop.

* * *

We use to set books for the University of Hawai‘i Press when the text had Chinese before Unicode and OpenType CJK fonts became so common.

I, too, did several things for them back in the old days. The one I have at hand is from 1996 and the title page shows it came after the press changed its name from the University Press of Hawaii; it also has the glottal stop mandated by the state legislature.

One worry I have -- perhaps because I don't know any better -- is that if I use U+02BB ("turned comma") for jobs that end up being distributed electronically it could screw up searching and indexing on account of its relative rarity. The U+2018 convention seems safer -- not to mention more in line with your Yogi Berra practicality.

You can cut it seven ways from Sunday. If you wanted to search for all and only glottal stops in a work with Polynesian, using U+20BB would make that easy. The whole point of using the right character comes from future "electronic repurposing." The glyph for the Latin capital "A" is the same glyph as the Greek capital Alpha. For printing only, you could use either one. Syntactically they are quite different.

Having said that, guess which character we use for the raised or turned comma to signal glottal stops.

Part of the reason is I can count on the thumbs of one foot the number of competent book fonts that have any of the spacing modifier characters. I guess I could put them in, kerning should be a bit different for an apostrophe than for a glottal stop. But (1) I'm lazy, and (2) all the manuscripts come in with either 0027 or 0218. Nobody's going to pay us for doing this search when the manuscript is labeled "editorially correct.

"Nobody’s going to pay us for doing this [re]search..." The editor's expertise lies elsewhere, the author isn't sure (and is clueless about design), meanwhile the head of production is more interested in moving jobs through the system. What's a conscientious compositor to do?

“Nobody’s going to pay us for doing this [re]search...”

Actually, I'll usually do the research, no charge. I find it interesting, and look at it as an investment. What is learned carries over to the next time the subject comes up. That's different from performing hand searches in a particular manuscript, no leveraging there.

As to which character to use, you can get a difference of opinion, so I'd not change the expected 2018 without clearing it with the publisher. That you can argue the publisher is often wrong is a different matter.

What started this thread wasn't that I would have insisted the publisher or author was wrong, but that they might want to consider a "Note on Orthography" if there were varying opinions, or if scholars tended to use a different orthography than "indigenous people." And, if a usage were odd enough, and there was to be no note on orthography, I might want to remove our name from the "composed by" line on the CIP page.

That's about as conscientious as I can get.

By the way, I see you're a compositor working in the University press market. "New Haven" makes it sound like we might be competitors.

Here is something I noticed: in the very same Wiki table presenting the Modern orthography of Mayan Languages, U+0027 is used in the ALMG (Guatemalan Academy of Mayan Languages) orthography whilst the corresponding IPA representation uses U+2019. [But is is possible that the cooccurrence is unintentional, the U+0027 may have been typed, and the IPA may have been cut and pasted.]

As I see it, composition is where the rubber meets the road in publishing. The customer is always right, though sometimes it is worth the effort to educate him or her about less-than-obvious details. And as you say, any innovation needs to be cleared. I like to know how things ought to be as well as what's practical now, hoping to keep the files I submit today from becoming out-dated too quickly. Some publishers appreciate that, and I think it helps me retain customers in a competitive environment. One can't charge for researching arcanities, they're just another "loss leader". This week, for example, I get to make a few modern forms based on archaic Chinese "Stone Drum" inscriptions -- about as obscure as you can get (though I see there is a wikipedia page).

That is an indication of the small niche in which I operate: almost all my work involves Chinese, sometimes very old. So I'm more a student (thanks to your contributions on these forums) than a serious competitor. As for New Haven, well, like a lot of folks here I came for Yale and and stayed for the town. In my case, it wasn't the design program but rather the anthropology department on the far side of the campus. Back in the days when IBM compatibility meant something I bought a PC clone that couldn't print from a Chinese add-on to DOS, and had to learn all about interrupt 17h. That proved more fun than finishing a dissertation on Chinese archaeology.

I had a look at the ALMG (Academia de Lenguas Mayas de Guatemala) site and, after selecting "Publicaciones > Alfabetos", I was redirected to a pdf document including the following alphabet for the Ixil language.

The word "saltillo" (also used for all the other mayan alphabets in the document) caught my attention. I first found the Saltillo Wiki entry, where I learned that "saltillo" (meaning little skip) refers to a glottal stop, that it is a straight apostrophe-like symbol that has often in the past been substituted for by letters such as U+02BC, MODIFIER LETTER APOSTROPHE and that the codepoints for the saltillo are U+A78B LATIN CAPITAL LETTER SALTILLO and U+A78C LATIN SMALL LETTER SALTILLO. The corresponding unicode chart gives the following example glyphs:

I also had a look at the Summer Institute of Linguistics (SIL) in Mexico Electronic Glossary of Linguistic Terms where I found this definition of saltillo: A (usually enlarged) straight single quote or dotless exclamation point, often used in practical orthographies to represent a glottal stop. The name "saltillo" (possibly used only when this symbol is used for languages of Mexico) is borrowed from a Spanish word meaning 'a little skip', a description of what the voice does in pronouncing this sound.

Also on the SIL site, I found this ch'ol dictionary that uses a font called CharisSILAmArea, that encodes the "saltillo" by a private codepoint, U+F21D; here are two words from that pdf file with that "straight apostrophe" (I copied them from the pdf, pasted them side by side in textedit and selected CharisSIL, but they look just like in the pdf):

If I look at the CharisSIL fontfile with FontForge, I see that the character at position 0xF21D is nothing but uniA78C. Things seem to fit together.

Michel

Hmm. I must confess that the word ats'am in the dictionary is better spaced than what I get in textedit.

Thanks for the research, Michel. So that should confirm that the 'straight apostrophe' shape, preferably an enlarged one that could be described as a 'dotless exclamation point', is desired for Ixil Mayan, at least until we find another authoritative source telling us otherwise.

However, the same PDF document looks like it presents the saltillos at least sometimes as raised commas, although it's hard to see with the low resolution. Other PDFs from the ALMG site use the raised comma for the saltillo as well. A result of automatic 'smart quote' substitution, perhaps? Who knows. One wishes they would be more careful.

Brian, you must have spotted the quotes in running titles. The chꞌol dictionary contains indeed 185 occurrences of U+0027 and 182 of them are in the word CH'OL in running titles; there is also one occurrence on page 67 in the word ch'ol and two on page 27 in 'asi debes decir'. There is no other "apostrophe" in the text.

There are three occurrences of U+2019 (single quote) for a quote inside a quote in the bibliography.

There are 12231 occurences of the minuscule saltillo uniA78C and 16 occurrences of uniA78B (capital saltillo), all of them in the word CHꞋOL.

Just a note: U+A78C isn't even in the Unicode Standard (5.0) book. I did find it on the web in Latin Extended D. Anybody want to take bets on whether or not it is in any commercial fonts used to typeset books?

This sort of thing does leave one in a quandary. U+0027 will be most useful for later "repurposing" as such occurs in the book world, but it does look like U+A78C would be the proper character. Of course, I can put in in the font for the printed book, but whoever "repurposes" those files will be in for a bit of fun.

Charles

I am not familiar with the editing tools you are using. For someone like me, working with a utf-8 text input (for XeTeX of instance), "repurposing" would be quite easy with the saltillos in the source file encoded as U+A78B (for the larger one) and U+A78C (for the smaller one): anyone in the future wanting the saltillo to be encoded with U+2019 would then simply apply a global substitution on the source file, substituting U+2019 for both U+A78B and U+A78C; similarly if U+0027 is then preferred.

Michel,

OK, makes sense. Especially if we were doing any of the "repurposing" work. The real situation is we typeset the books, they are printed, and the publisher doesn't look at the text files again until there is a need. We give the publisher at least one PDF file (for printing), and the InDesign files minus the fonts -- it would break the EULA to give them the fonts.

Lord knows who will be doing the work for any subsequent "repurposing." Likely not us.

I could find the proposal for the unicode points U+A78B and U+A78C here on SIL site (with capital and small saltillo switched with respect to what unicode adopted). It is argued that in the Huasteco mayan language U+02BC is used for glottalized consonants whilst a vertically symmetric glyph (I guess that is part of what is meant by "straight") is used for the glottal stop proper, giving :

They also have a FAQ on how to encode the glottal stop.

There is another “straight” modifier letter available in Unicode: U+02C8, used in Skolt Sami.

The most interesting approach I saw to Mayan glotal stop was made by Steve Ross in Yukatek font. You can find information about it in pages 25-26 of his paper.

Ross's Yukatek innovation is awesome. We need to apply such bravado much more broadly.

Igor & Hrant ~
Please confirm the link for "Steve Ross in Yukatek font" . . . ("his paper" is now a broken link)
~ thanks!

I put a copy on my own site-to-be. There are both the Yukatek paper and specimen.

Looks like you need http:// to make the links work.

Thanks! Links are correct now.