Cyrillic lowercase shha (һ) - ascending vs non-ascending

Kent Lew

Just casting about, I am seeing what seems like a fairly even distribution of interpretations of the Cyrillic lowercase shha (һ) U+04BB in fonts that support extended Cyrillic — some are basically a reduced, x-height form of the capital Shha (Һ), while others are a duplication of the Latin h.

This letter is part of several Cyrillic-based orthographies: some Turkic languages and some Mongolian ones.

What drives these two different forms? Are there some orthographies where an ascending glyph harmonizes better than in others? (Given the paucity of ascenders in Cyrillic script overall, the h form seems generally a less comfortable fit, to my eye.)

Are there local preferences in some traditions that lean toward one form over the other?

Kent Lew

{bump}

No perspectives or insights on this? Anyone?

Ray Larabie

I looked into it recently. I made a list of languages which require һ and Google image searched to get a rough idea of which form of һ is probably more common. I rarely encountered x-height һ form. That doesn't mean those readers prefer the ascending h style but it's a good indicator that they're familiar with it.

It was almost always the ascending h style. From looking at fonts that include һ or h, there seems to be no pattern. I think the ascending form looks better in context, especially in condensed heavy weights. I'm interested in hearing more about this mysterious character.

Igor Freiberger

Although I collected a lot of information about Cyrillic, I found nothing about the shape of һ. Due to the number of fonts using the h form, it seems acceptable to chose this, especially considering that most ParaType fonts use it.

Anyway, the x-height һ is being adopted in recent releases which seems based on good research – revised system fonts on Mac and Windows, and releases from quality designers, as FF DIN Pro, Iskra and Wayfinding. So the doubt remains.

I will try to dig some native speakers of the languages using һ and get safer information. By now, the better I could get is this MyFonts search where one can also note that һ, ԧ, and ћ aren't related in fonts with the x-height һ.

[Deleted User]

The user and all related content has been deleted.

Kent Lew

Thanks for weighing in, guys.

James — Did Maxim consult on those extensions, by any chance?

Igor — The presence of the x-height һ in more recent and presumably well-researched designs is what makes me wonder. Especially since, as you point out, this is contrary to most ParaType fonts employing the Latin h form.

Please let us know if you get any input from native speakers from any of the relevant languages.

kupfers

Kent, how about you contacting some native speakers? Like Maria Doreuli, Krista Radoeva, Ksenya Samarskaya etc. These three also run “Cyrillic advisory services”, on Twitter and elsewhere.

John Hudson

how about you contacting some native speakers? Like Maria Doreuli, Krista Radoeva, Ksenya Samarskaya

Native script users, but not necessarily speakers of the languages that use extended Cyrillic characters such as һ. So this could be like asking a native English speaker how to design ß.

Kent Lew

What John said.

Not that I don’t have plenty of respect for Maria, Krista, and Ksenya, and I would welcome their perspectives. Thanks for the reminder. I know Ksenya and will reach out to her by e-mail. But I’m not on Twitter or elsewhere (for better or worse).

kupfers

Fair enough. I was replying to Kent’s “Please let us know if you get any input from native speakers from any of the relevant languages.” I guess I misunderstood “several Cyrillic-based orthographies: some Turkic languages and some Mongolian ones”. So only Turkic languages and Mongolian ones? In any event, it’s fairly easy to contact people who live or extensively lived in Cyrillic-using countries. Maria and Krista ran this site http://cyrillicsly.com There is contact info on there, too.

It’s not that likely that a qualifying native for such specialized cases will just randomly read your thread on Type Drawers, that’s why I suggested you take some further actions. You can also @mention people’s handle here to send them a notification. I think only very few read this forum on such a regular basis that you don’t have to bumb up the thread every few days.

Kent Lew

Indra — Yeah, thanks for the tips. I hope I didn’t sound dismissive. If so, I apologize.

I am, believe it or not, rather shy and reluctant about “cold calling” people I don’t feel like I know in some way. It’s much more comfortable for me to post such questions generally on a public forum.

But you’re right, for specialized insight like this, it may be more productive to find other individuals to contact directly.

I’ve emailed Ksenya (whom I know somewhat) and hope to hear a reply.

If @Krista Radoeva happens to be around, I would very much appreciate her perspective as well. I don’t think that Maria Doreuli is on this forum.

Kent Lew

Oh and yes, to clarify: only non-Slavic languages.

To the best of my knowledge the Һһ belongs only to alphabets of Bashkir, Buryat, Kalmyk, Kazakh, Kildin Sami, Tatar, and Yakut (Sakha). Plus the [former] Cyrillic orthographies of Azerbaijani, Uyghur, and Kurdish, which I believe all officially use other scripts now.

George Tulloch

For what it's worth, here's the list of non-Russian Cyrillic characters given in a Soviet type specimen book (Fotonabornye shrifty: katalog-spravochnik, comp. G. I. Kozubov & V. V. Efimov, Moscow, 1983).

[Deleted User]

The user and all related content has been deleted.

George Thomas

I've noticed that quite a few fonts use the x-height ascender version rather than the cap-height ascender version. SIL's Andika font as well as some others in their catalog have both glyphs, with the x-height version as a variant.

So I asked them about it and they responded with this: "According to the developer, ‘It's not a contextual alternate. It's a Character Variant and a person would have to select it to use it.  It was requested for two languages in Azerbaijan, where some people wanted it in that style.’ "

I have been unable to find any reference which indicates the x-height variant is preferred in a particular language so it appears to me this has to be considered a personal preference until further evidence surfaces. It isn't that difficult to include both versions in fonts going forward.

Hrant Հրանդ Փափազեան Papazian

Sounds like a nice rare opportunity to not fret over relative precedent merit and make one's own considered decision.

Stefan Peev

Well, I think that the answer could be seen in PT Astra Sans and PT Astra Serif - two font families by Paratype, which must be a replacement of Times New Roman (so, we could think about them as a kind of standard or the point of view of the Russian typography). If we seek for some logical decision, then the x-height ascender version is more logical because the shha form with x-height ascender is more closely related to all other Cyrillic forms in the alphabet unlike the Serbian script where Cyrillic forms live together with purely Latin forms).

Kent Lew

Stefan — You provide two contradictory viewpoints. On the one hand, the PT Astra family could be viewed as a kind of standard to follow (where the shha has a full ascender, basically a Latin h); then in the next breath you state that it could be more logical to have an x-height ascender, not following the PT standard.

So, unfortunately, that doesn’t provide any practical answer. But thank you for joining the conversation.

André G. Isaak

If there's no authoritative data available on this point, then why not simply provide both? In absence of evidence, we don’t even know if the different languages which use this prefer the same form.

André

Hrant Հրանդ Փափազեան Papazian

André G. Isaak said:

why not simply provide both?

You still have to choose a default. (And let's leverage the freedom from precedent.)

Stefan Peev

Kent Lew said:

Stefan — You provide two contradictory viewpoints. On the one hand, the PT Astra family could be viewed as a kind of standard to follow (where the shha has a full ascender, basically a Latin h); then in the next breath you state that it could be more logical to have an x-height ascender, not following the PT standard.

So, unfortunately, that doesn’t provide any practical answer. But thank you for joining the conversation.

That's it, Kent. There is a contradiction. That's what I wanted to point out. But the contradiction is not in my statement. The contradiction is in the life. I personally choose the Paratype's decision. Because PT Astra families are designed to be fonts for Slavic and non-Slavic people who use Cyrillic script. And they will be widely distributed as such typefaces. So PT Astra family will design the standard for a shha glyph. Even if they are wrong. I am also a linguist and I know that language never evolves according to the rules we are trying to draw. Often a massive repetition of a language error leads to the formulation of a rule that normalizes it as a norm.

Kent Lew

Thank you for your clarification, Stefan. I tend to agree with this.

Stefan Peev

And here is a very interesting example for a repetion of a language error in written Bulgarian language (very common!). So please be informed that there is no way in written Bulgarian text to have situation like [space]й[space]. The glyph uni045D (ѝ) with space before and after it denotes the word "her" ( her book = книгата ѝ ). This mistake is so widely spread that a foreigner could decide that it forms some kind of a rule.



I'm afraid that the situation with the shha (uni04BB) is the same. I'm not convinced that the choise of Paratype is correct. Look at uni04B6, uni04B7, uni04B8, uni04B9, uni04BA. All these glyphs are based on Cyrillic Che (uni0427), che (uni0447). So it's strange the lowercase shha (uni04BB) to be based on the Latin lowercase "h". But the distribution of the PT Astra can turn what seems illogical into a rule. Meanwhile, look in the font Roboto where the shha (uni04BB) is drawn precisely from the logic I describe (in connection with uni0447 - lowercase che). That's why I argue that the right decision is in Roboto, but the final winner will be probably PT Astra (because this family must replace Times New Roman and Arial in Russian written communication). It's why I said in my previous post that "repetition of a language error leads to the formulation of a rule that normalizes it as a norm".

André G. Isaak

but the final winner will be probably PT Astra (because this family must replace Times New Roman and Arial in Russian written communication).

Sorry, but I’m unclear about the word “must” above — how do we know that PT Astra will come to replace TNR?

Hrant Հրանդ Փափազեան Papazian

André G. Isaak said:
how do we know that PT Astra will come to replace TNR?

Do a search on "paratype astra sanctions"...

André G. Isaak

Thanks — I was unaware of this.

However, my question still remains: Even if Astra Linux uses PT Astra Serif in place of TNR, does this necessarily guarantee that this will become the “new normal” for Cyrillic typesetting in Russia? Will existing Windows users in Russia not continue to use TNR? And it sounds like Rosa Linux will be using Libre Serif in place of Times. If sanctions continue for a significant period then there is a possibility that the Russian government might switch from recommending TNR to PT Astra Serif, but I don’t see this as fait accompli.

André

[edit: of course, since TNR also uses the h-like shha, this probably doesn’t matter with respect to the issue under discussion].

Paul Miller

So ... if I'm designing a Cyrillic character set for an existing font should I use the x-height һ or use a 'h' for character U+04BB ?

Kent Lew

An interesting avenue of research, with regard to precedent, would be to try to identify what was used in the earliest days, when Cyrillic orthography was first imposed on these languages, in terms of the typography.

What I mean is: in all likelihood a new sort was not cut for the need initially — either a Latin 'h' was pressed into service, or a 'ч' was rotated in the line.

Also, I have tried to find examples of handwriting in any of these languages, so see what users use when not relying on prefabricated letters, but without success.

Stefan Peev

Even if I don't like this answer, here is what is written in Unicode 10:

04BB һ CYRILLIC SMALL LETTER SHHA

• originally derived from Latin “h”, but uppercase form 04BA Һ is closer to an inverted che (0427 Ч )

• Azerbaijani, Bashkir, ...

→ 0068 h  latin small letter h

So...

Paul Miller said:

So ... if I'm designing a Cyrillic character set for an existing font should I use the x-height һ or use a 'h' for character U+04BB ?

The answer should be... use a 'h'... to stay in accordance with what Unicode 10 prescribe...

Hrant Հրանդ Փափազեան Papazian

Kent Lew said:

An interesting avenue of research, with regard to precedent, would be to try to identify what was used in the earliest days, when Cyrillic orthography was first imposed on these languages, in terms of the typography.

What I mean is: in all likelihood a new sort was not cut for the need initially — either a Latin 'h' was pressed into service, or a 'ч' was rotated in the line.

This makes it interesting like the Hindenburg disaster... For usability, not so much.

Kent Lew

Yes, Hrant, I’m well aware of your disdain for history and precedent.

But honestly, I’m unclear which form you are advocating for. Is it your view that the ‘h’ form is too Latinized and a turned ‘ч’ harmonizes better with the rest of Cyrillic script? Or would you favor the ‘h’ form because it introduces more differentiation with its ascender?

Care to take a stand?