Elongation of Hebrew letters at the end of the row

Hrant Հրանդ Փափազեան Papazian

Ori Ben-Dor said:

"you can include it, of course, but you don't have to. Leaving it out won't make your font broken or substandard."

Certainly not "broken", but "substandard" is far hazier...
I feel your use of "obsolete" is culturally limiting.

Ori Ben-Dor said:

All three examples are of serif fonts suitable for long texts. The designers probably thought there's a chance they might be used for religious texts (and in the case of at least two of them, rightfully).

Assuming any religious text (not necessarily the entire Torah for one thing) must be in a highly readable serif font is also limiting. As is assuming that a religious flavor should never be injected into a nominally secular design.

Also, do you know what category of typeface Vasil making? Or really, anybody reading this discussion.

Ori Ben-Dor said:

I live in Israel, I open my eyes, and I have never, not even once, seen any use of the elongated forms of any of those three fonts. That's just a fact.

Does that mean they shouldn't be?
Don't you see the advantages? I mean beyond the literal religious aspect.
Don't you want Hebrew typography to keep evolving?

For decades the Armenian «հ» looked just like the Latin "h" in virtually all fonts. I was guilty of that myself for the first few years of making fonts. But eventually I realized it's culturally impoverished, and started making –and doggedly promoting– a more authentic structure. And this structure is becoming increasingly mainstream in "modern" fonts (such as Noto). I would like to encourage designers of all non-Latin scripts to revive useful and appealing features that have gone dormant, usually for bad reasons, like saving money. I would even like to encourage new features! Like capitalization in scripts that don't enjoy that.

Ori Ben-Dor said:

Hyphen too is a connector, what's the difference? Maqaf is just "the Hebrew hyphen"... Moreover, the role of the hyphen/maqaf in hyphenation is exactly to connect the two otherwise seemingly separate strings and indicate they're actually one word. It's a job for a connector, isn't it?

In parallel to Armenian, I think there are good reasons to distinguish a dash for line-breaks and a dash for compound words.

--

In the end, instead of telling Vasil to simply ignore elongated forms, I would suggest he consider it if he wants to facilitate Hebrew's further flowering.

Ori Ben-Dor

I'm not telling Vasil to ignore elongated forms, I'm telling him it's okay to do so. There's a big difference.

Perhaps my initial response ("nope") read as the former, in which case I regret putting it that way, but what I meant was the latter (frankly, the former doesn't make much sense to me, as—technology concerns aside—the idea of an upper bound on any glyph set in general doesn't make much sense to me, so I didn't think anyone might think I'm suggesting Vasil shouldn't offer elongated forms if he wants to).

You make some valid points, but I think they belong to a separate discussion, one that concerns the ought, while Vasil's question deserved an answer form the is department (or at least that's how I interpreted it).

At the end of day, there's the empirical, ontological question of whether elongated forms are a living part of modern Hebrew typography (even if rarely used) or not. It could be either way, and it's a valid question, coming from someone not very familiar with Hebrew typography, which one it is. (And as you've probably realized by now, I stand by my initial answer: it's not a living part of modern Hebrew typography. I'm a 38-y-o Israeli Jew, I've lived my whole life in Israel, I've been interested in, and therefore attentive to, typography for over 20 years now, and not once—once!—have I seen elongated forms used outside the context of religious texts or imitations of them. If the term obsolete has any meaning at all, this testimony of mine means, to me at least, that elongated forms are indeed obsolete in modern Hebrew.)

I felt Vasil was looking for this specific piece of information, and even if I misinterpreted his intentions, at least we can imagine that's the piece of information he was after. This forum is exactly the place where you'd ask for that piece of information. And under my interpretation, I think it's not entirely fair to confuse Vasil (the real or an imaginary one), and maybe even mislead him a little (out of good intentions, of course), with broader ideas instead of giving him a straight answer. I mean, it's perfectly legitimate to raise those issues and discuss those broader ideas even in a thread started by a more limited-scope question such as the one I'm attributing to Vasil, but then we need to make sure it's clear we don't mean the answer to his question is positive when it's actually negative. Even if we wish it was positive.

==

>> Also, do you know what category of typeface the OP is making?

What's OP?

>> In parallel to Armenian, I think there are good reasons to distinguish a dash for line-breaks and a dash for compound words

First let me point out that there are bigger problems with the use (or lack of) of maqaf. Due to bad influences of technology, it's very common to see hyphens used instead of maqafs in Hebrew nowadays. I believe many, maybe even most, Israeli Jews (at least under some age) don't even realize these are two different entities. They see maqafs all the time, of course, but the difference never registers (much like how people read double-storey g's all the time and still can't pick the correct form from a lineup), and when they write or type they use hyphens, unaware there is (or at least there could be) an alternative.

Anyhow, how would you distinguish between a dash for line-breaks and a dash for compound words? If your idea is to use a hyphen (that is, middle height as opposed to top height) for line-breaks, then I don't like that idea. I think the structure of the Hebrew letters begs for a top height dash.

Ori Ben-Dor

Oh, and as for

>> Assuming any religious text (not necessarily the entire Torah for one thing) must be in a highly readable serif font is also limiting. As is assuming that a religious flavor should never be injected into a nominally secular design

The point I was trying to make is that I'm not sure you can draw from the presence of elongated forms in the fonts Scott gave as examples the conclusion that their designers must have thought that elongated forms had use outside the context of religious texts, as Scott suggested. If those designers thought their fonts might be used for religious texts, as I suggested, then that could explain why they might choose to include elongated forms.

Ori Ben-Dor

By the way, the case of the Alef Alef Alef foundry offers another evidence that elongated forms have indeed become obsolete in modern Hebrew.

Alef Alef Alef is an Israeli boutique foundry. I think it's fair to say that they're by far the most prominent Israeli boutique foundry in terms of marketing and hype (their presence in social media is much higher than any other boutique foundry, their website offers much more than just fonts, they're virtually the only Israeli foundry to sell merchandise, they always keep coming with new projects, etc).

One of their marketing strategies is to pack their fonts with all sorts of alternative forms and other goodies. They would even invent new ligatures!

And yet, even they don't include elongated forms in their fonts (though I've only checked a sample).

Hrant Հրանդ Փափազեան Papazian

Thank you for the detailed, considered replies.

Ori Ben-Dor said:

It's not a living part of modern Hebrew typography.

I actually believed you the first time you said that. But to me "obsolete" means something like "no longer useful", and I think that simply cannot be true here; it does not to me mean "virtually nobody will pay extra for it". Elongation has great functional and cultural potential, and I'd say Alef Alef Alef is a prime candidate to be woken up to that. Maybe they feel it's too old-fashioned. Well that's what many people thought of my Armenian «հ» revival, but their case is much weaker now... Things change. Things can be changed.

Also, I have to repeat that the total separation of religious and secular typography is artificial.

(OP means "original poster".)

Ori Ben-Dor said:

I believe many, maybe even most, Israeli Jews (at least under some age) don't even realize [the maqaf and hyphen] are two different entities.

You're most probably right, but that's where our dark craft kicks in: text fonts especially are not about what people think they're seeing; they work "subvisibly" to create a mood, and affect reading. Our subconscious picks up so much, and can handle a lot of complexity: making the maqaf and hyphen look different* would be enough for them to work differently. Concerning vertical placement: I have no clear opinion on that [yet] except that it might actually be enough distinction between the two.

* As the Armenian hyphen and yentamna look different, although sadly the latter is almost never deployed.

Ori Ben-Dor

I don't know what category Vasil is making, that's true. I should have pointed out the religious texts exception already in my first response.

==

Obsolete doesn't mean "virtually nobody will pay extra for it", I totally agree. When I said obsolete I meant obsolete: as I said, outside the context of religious texts, elongated forms would look weird and out of place. Cultural norms could change, but until they do, that's just the way it is (sad as it may be). A typographer setting a novel in Hebrew today would make a mistake to use elongated forms — they would look weird and out of place. They would make the readers ask "what is this, a novel or the Bible???" and then hit the streets, wearing yellow vests, calling for the sack of the typographer.

There is an exception, as I've already mentioned: in case the typographer deliberately wants to give the text a Biblical look and feel, using elongated forms could be a valid artistic choice. If the novel depicts the lives of ultra-orthodox characters, for example, I can imagine a sophisticated typographer using elongated forms. But then the typeface would have to be "Biblical" too, or at the very least some classic serif, or it wouldn't work.

Of course, if you understand the status of elongated forms in modern Hebrew typography and decide to use them anyway for something other than a religious text or an imitation of one, then that's fine. But you need to understand their status first. You need to understand you're breaking norms and do that consciously. And while you are allowed to break norms, you can't expect type designers to provide you with the tools.

==

As for the rest, I think it's very important to make the distinction between the ought and the is. "Has potential", "things can be changed", etc. — all of that belongs to the ought department. Vasil has a right to be interested in the is.

==

Can you please post an example of the Armenian hyphen vs. yentamna? I'm intrigued!

Hrant Հրանդ Փափազեան Papazian

Good post.

You can see the yentamna versus the hyphen used for line-breaks in the side-by-side images towards the bottom on this page:
http://armenbadal.blogspot.com/2014/03/latex.html
The yentamna is just a bit curvy. However neither keyboard access nor (sufficient) software support is there, so it's basically laying dormant for an eventual spring...

Ori Ben-Dor

Thanks for that link. I can't say I've already managed to establish an opinion...

As for Hebrew, I think the problem with a middle height hyphen is that Hebrew is rich with letters the are open to the writing direction. While in Latin only c, k, r, v, w, x and y are problematic, and out of these only r and c are very problematic (r-? c-? it doesn't look so good), whereas v, w, x and y are less problematic and also relatively rare, in Hebrew some half the letters, including many very common ones, are medium to very problematic:

Vasil Stanev

Ori,
I was under the impression many of these Latin symbols are still used in Africa.

I am not making a Hebrew font, people. I saw some magazine about the "Bible code" or Bible hunters or whatever, and the image on the cover featured a book with elongated forms (that way I knew right away it was medieval, since it was no scroll). That's all. 

(And when I searched on the web first, I discovered "Jew" is a curse word, so much of what I got was........ You learn something new every day.)

Johannes Neumeier

Reading this thread makes me wonder about Variable Font use.

Could you not have a "justification" axis that interpolates from regular to widest form for all stylistically viable letters. Then make use of it in on layout engine level, elongating anything from the last letter to word to last-n-words to entire line to justify the column that way, more or less in automated fashion?

The beginning of this thread seems to suffer from miscommunication. Is the discussion regarding historical forms and their appropriateness, or justification in scripts that have, in calligraphic tradition, made use of elongation to justify text to a column width? Then again, both are interesting and interlinked aspects.

Hrant Հրանդ Փափազեան Papazian

Vasil Stanev said:

I am not making a Hebrew font, people.

Luckily public discourse is not merely about its participants...
Thank you for starting it!

John Hudson

Johannes, using glyph variations for justification has been discussed and demo'd over the past couple of years, mostly in the context of Arabic script, but the principle is generally applicable. Unfortunately, the roadblocks are also generally applicable.

See:

The third section of my TYPO Labs 2017 presentation.

Sahar and José's 'Glyph Extension' variation axis proposal.

Discussion of the axis proposal (in which I try to identify all the roadblocks). 

Sahar and José's TYPO Labs 2018 presentation (includes really nice demo).


The Hebrew extension case is generally a bit more simple for Hebrew than for Arabic because of the lack of connectivity between letters and associated contextual behaviour and GPOS.

Scott-Martin Kosofsky

After I mentioned the possibility of using variable fonts in a post above, I wondered if the current spec could accommodate the positioning of diacritics, which in non-variable fonts is accomplished through the mk and mkmk features, along with a number of composite glyphs and others that are fixed and have their own Unicode numbers.

More importantly, this issue—i.e., Hebrew using both vocalization and cantillation diacritics—is exclusive to the Hebrew Bible, and that’s where the use of extended letters has the greatest tradition. (Liturgical and instructional Hebrew uses only the vocalization diacritics.)  In Bible texts, the compositor applies a sense of literary hierarchy in choosing which words are given the extended letter treatment. That would be way outside any current capability for automation that I know of—someone tell me if I'm wrong. I include a couple of grammar-driven substitutions in the GSUB tables of my fonts, but only one involves diacritic positioning—pretty simple compared to the problem of applying automated variable widths based on an elaborate dictionary.

Here’s an example of some biblical text set in a very narrow column (there is translation across from it, not shown here), justified using extended letters, a number of which carry diacritics, some of them multiple:Including alternates and glyphs that are seldom used is a great achievement of OpenType and they’ve changed the choices that designers make daily. The typographic tradition of extended letters in Hebrew fonts goes back to the 1470s and for nearly four centuries they were a feature of reading the language. They disappeared in cheaply produced books that began appearing in the late 18th century and in Yiddish books of the 19th century, after the old Yiddish semi-script alphabet, Waybertaytsh, was abandoned in favor of the familiar forms. Modern Hebrew, by comparison, has a much smaller tradition, beginning only in the 1870s and 1880s and it took some decades before it began to produce some new types—largely for display, not text, which was dominated by the same middling stuff. It is, in some ways, a rigid style, much of it the result of being a minority language in the era of metal machine composition and later developments, before Fontographer. Now we have the tools to liberate it from these limitations. I’ve looked backward, to the 16th century, for inspiration. Someone else can do it by looking forward.

John Hudson

After I mentioned the possibility of using variable fonts in a post above, I wondered if the current spec could accommodate the positioning of diacritics, which in non-variable fonts is accomplished through the mk and mkmk features, along with a number of composite glyphs and others that are fixed and have their own Unicode numbers.

This would remain the case in a variable font, presuming axes such as weight, width, or optical size: the GPOS data for the diacritic positioning would be interpolated along and between the axes. You could also set up GPOS to vary across the design space, if you had a reason for that.

John Savard

Hrant H. Papazian said:

(like how English sadly lost the Thorn and Eth when importing fonts from the continent),

I don't feel myself competent to say if the horizontal contrast pattern of Hebrew letters makes wide versions a good option for justification, although I suspect that the reading habits of Israelis have been shaped enough by contact with Latin-alphabet languages that they will tend to find wide letters as distracting as we Latin-alphabet users would.

As a native speaker of English, though, I will note that it wouldn't be a bad thing if English had an extra letter for "th". While we're at it, we could borrow "sh" and "ch" and even "zh" from Russian.

But the phonemic distinction between two forms of "th" that eth and thorn represent is well and truly lost from the English language, and there's no getting it back.

For that matter, I've felt that dropping all those letters from Russian at the time of the "new orthography" may have been a bad idea, but Russian script is highly phonetic at present, so it must be admitted those letters weren't needed. And I certainly would not inflict having to put a redundant hard sign at the end of every word ending in a consonant not followed by a soft sign on the Russians.

Still, the lost letters, with a changed sound value, could be helpful in writing foreign words, containing sounds that Russians wouldn't be able to pronounce anyways.

John Savard

Hrant H. Papazian said:

Does that mean they shouldn't be?
Don't you see the advantages? I mean beyond the literal religious aspect.
Don't you want Hebrew typography to keep evolving?

For decades the Armenian «հ» looked just like the Latin "h" in virtually all fonts. I was guilty of that myself for the first few years of making fonts. But eventually I realized it's culturally impoverished, and started making –and doggedly promoting– a more authentic structure.

...

In parallel to Armenian, I think there are good reasons to distinguish a dash for line-breaks and a dash for compound words.

In English, a hyphen is used for both line-breaks and compound words, but the dash is also used as a punctuation mark between words, and it generally distinctive.

While one famous author suggested a reformed alphabet for Hebrew that was heavily Latinized, in general, of all the threats the Jewish people have endured to their physical and cultural survival, an externally imposed change to the appearance of their script was not one of them.

Hebrew typography will indeed evolve, and its evolution will largely be driven by popular demand - and perceived popular demand by advertising agencies, which choose and commission display typefaces.

Ancient Hebrew manuscripts were written without spaces between words, which, even if the absence of hyphenation were not an issue, makes elongated letters almost necessary, and so they've been retained in religious texts out of the intense conservatism often associated with such texts. There is really no reason to expect that they have the kind of cultural meaning to native Hebrew speakers to make their use in modern Hebrew appealing.

It is natural for ethnic and linguistic communities to be protective of the distinctive aspects of their cultural traditions. So much so that they usually don't need any encouragement on this front. Yet, honey is more effective than vinegar, and thus Hollywood and Motown and Coca-Cola have indeed led to cultural homogenization in ways that conquest and oppression could not.

The larger issue, as I see it, is that, yes, it is definitely, therefore, legitimate to encourage the retention of cultural distinctiveness - but as cultures evolve, things in them may fall into disuse for reasons other than outside influences. Elongated letters in Hebrew seem to me to be a possible example of this, and thus placing emphasis on bringing them back will, I suspect, seem contrived. Tilting the balance back against a tide of homogenization makes sense, but such an effort must itself also include balance so as not to seem outmoded or even derisory to the actual members of the cultural group at issue.

Vasil Stanev

Just passing by to leave a note for future me and some of my type friends: there is a good amount of handwritten Hebrew in the 2016 vlog of the Youtube channel "Flux Academy" . I am currently binging on some of its content and the creator, Ran Segall and his team go to some cafe in Tel Aviv in almost every episode of the vlog. You can see interesting handwritten Hebrew on the chalk boards with the menu, and elsewhere, like street signs, posters, ads, traffic signs...
https://www.youtube.com/watch?v=MQSqcp7dkCs 
You can see also the sign for shekel, their currency, before the price numbers.