I'll bring this thread to the attention of my partner, Ross Mills, who has a lot of experience working with native American languages, but I believe the editor is almost certainly correct. The below macron is a fairly typical mark in the orthographies of many languages.

Thanks, John.

I actually agree it should be a macron below, except for the small problem that some Comanche's apparently don't agree. In the past, I have found myself arguing with the Navajo that their early rendition using a "reversed cedilla" was wrong for the nasalized vowels, and with the Cherokee that Plantagenet Cherokee served them better than the official Cherokee, when mixed with a particular Latin font.

As you can imagine, neither of these is a comfortable place for a Norwegian-American to find himself.

Well, you could argue for either. Typographically, you may want to use the macron beneath an 'i' or an 'I' and the low line beneath wider glyphs. Semantically (in terms of text encoding), one or the other should be used. So, what is the expected form? Generally, I would say it is the wider, or underline, form. The combining lowline is encoded as the combining form of the underscore, and theoretically should join on the left- and right-hand sides if in sequence (which, by logical extension would mean that the combining low line needs to be a bit wider then any glyph it is to sit under). One could say then it is not a diacritic, but that would be wrong, because it has effectively been used as such. The orthographies of Comanche or Pawnee use(d) an underscore to mark short or voiceless vowels and a barred u for a rounded high central vowel.

Ideally, the underline width should optically match the glyph it is sitting under in terms of width. In producing fonts, I am mapping both combining marks to pre-composed glyphs for composition/decomposition because these are fairly ambiguous marks and if someone is making a keyboard that uses them, they may use either 0331 or 0332 (there is no codepoint for the pre-composed glyphs for the bar/line below characters although there is for the barred-u). In reality, such niceties have not seen much use due to technological barriers; rather the characters are represented most often as underlined and struck-though in case of the barred-u.

There is much in some of the more isolated languages and orthographies that can be viewed as typographic aberration. This has to be tempered with users' expectations and is an interesting subject. Were the Navajo you mention actually using a backward cedilla instead of the proper ogonek? It wouldn't surprise me, as many fonts come with a cedilla, but not so many with an ogonek (or a+ogonek with acute, or epsilon with ogonek etc. etc.). Native languages may possess an orthography, but haven't been graced with typographic sensibilities—either in available typefaces or in the production of documents in terms of necessary technological requirements for handling such orthographies. This can lead to a degradation of quality of the typographic presentation of these languages, which in turn becomes the norm, which in turn may become a user expectation. Sometimes this takes the form of the perceived value in that which is degraded, or mutilated.

I wouldn't classify the use of an underline as such an typographic aberration (I'm not willing to argue the linguistic side, as I'm not a linguist) but rather the sort of adaptation that goes on all the time in establishing orthographies.

-Ross

Thanks, Ross, I will pass this thread on to the editor.

As a typesetter, I don't face the problems a font designer does. However the MS comes in, we'll change the character to uniXXXX0332, and I will make up precomposed glyphs for the print edition. Following your suggestion, I suppose the underline should be optically the same width as the character, with a contextual alternate feature for successive underlined vowels, so in those cases, the underlines would overlap. I'll take a look at the MS & see just how many of those we encounter.

In passing --

The Navajo with the reversed cedilla was a font prepared by the Navajo Schools, available for a while (long ago) on the SIL site. As I remember, it was one of the Times Romans, and an ogonek was available in Times. I believe that font has now been withdrawn, or replaced with one using the ogonek.

As I'm sure you know, there can be competing orthographies. Ella Deloria's rendering of Lakota is quite different than the Buchtel (sp) system -- and far harder to typeset, though probably easier to mark with a pencil. And while many orthographies use the raised comma (aka "apostrophe", "single close quote", "quoteright") to signal a glottal stop, Polynesian uses a turned comma (aka "single open quote", "quoteleft") for that purpose. At some significant level, what is right is a matter of convention, not a pre-established system.

Thanks again,

Charles