I can confirm Victor's comment about the omicron. In a low contrast sans serif type, such as the Helvetica Linotype Greek, I've been able to unify the Latin o and Greek omicron, but in higher contrast book types it is seldom an option unless the Latin is particularly condensed. Even if I am maintaining the ductus of the Latin in the Greek (which I am increasingly loath to do, but that's another topic), I generally find that the omicron needs to be narrower than the o. In a typeface I am working on at the moment, the different is only 10/2048 units, but it is enough to significantly improve the melody of the Greek. Note that I say melody and not rhythm: this too is another topic.

Of all the scripts I've designed for to date, Greek is consistently the hardest, but it is also my favourite.

Well, I'm glad you've changed your mind. :->
Next for Cyrillic!

> I say melody and not rhythm

Hmmm.

hhp

I have not changed my mind. I have maintained for a long time that the Greek and Latin lowercase letters, because they developed completely independently of each other, under the influence of different cultures and different writing implements, are only artificially unified. The Cyrillic script is a more complex case, because of the importation of a specifically Latin model, and the long relegation of the historical style to religious texts. I'm not in favour of turning back the clock 250 years in the interests of exoticism. In the case of Greek, there is no clock to be turned back, only a very recent and unfortunate fashion in graphic design that favoured Latinised Greek types.

Come on, John.

And I'm not talking about returning to pre-Petrine forms here - that's a different fish.

The Cyrillic "cognates" need to be different than their Latin counterparts for the same reason that the Greek omicron needs to be different than the "o": language - you know, the stuff our fonts are used to set. The level of historical sensitivity you describe is insular and self-referential. The needs of the living users come first. Plus Cyrillic came from Greek anyway!

Greek has more rounds than Latin, and Cyrillic a lot less. It makes total sense for that to affect the Whites, and if you believe in notan, the Blacks are tied in too.

hhp

The original Cyrillic script was developed from Greek uncial letterforms. Like the Latin lowercase, the Cyrillic lowercase developed in isolation from the Greek lowercase; the Petrine reform Latinised even the remnant Greek forms, which is why Russian ef and Greek phi are very distinct forms.

The Greek lowercase is a Byzantine development: it belongs to Constantinople, Alexandria -- it is as much a product of the Middle East as of 'Europe' (and like the other major scripts of the Middle East, the Greek lowercase has a steep ductus and involves rotation as well as translation.) If it can be said to have developed alongside other scripts, those were neither Latin nor Cyrillic: they were Arabic and Hebrew.

I've designed several Cyrillic typefaces, and have extensively tested all of them, and I have not found any good reason not to unify Latin and Cyrillic cognates (whereas I don't accept that Latin and Greek even have cognates). Nor have any of my more experienced Russian colleagues with whom I have corresponded, and who know a lot more about the needs of living Cyrillic readers than you. Don't think I have not thought about this. What are the cognates? a e o p c y x for upright Russian, plus i j s for Ukrainian and Serbian. The counters of a e o p determine the appropriate width and weight of the other letters in the Cyrillic script, much as they do in the Latin script, and they have done so for 250+ years. In designing Latin type you can easily begin with just o and base the proportions of the other letters on that, and you can do exactly the same thing in Cyrillic.

Now, the vast majority of Cyrillic typeface designs begin with an existing Latin design, so the influence is entirely one directional. My preferred working method is to design Latin and Cyrillic together; in fact, for design purposes, I might as well consider them a single script. So it is not that my Latin o determines the width of my Cyrillic i, but that my letter o determines the width of my letters n and i.

One interesting aspect of this approach is that a letter like ze can be developed in isolation from other Cyrillic forms, and checked against the width of the Latin s, even though the latter does not occur in the Russian alphabet. If properly balanced against s, the ze will be found to also be well proportioned relative to other Cyrillic letters. Why? Because all the letters are informing and influencing each other without regard for nominal script divisions.

Latin and Cyrillic are the only scripts that I approach in this way (excepting the Greek uppercase, which is almost entirely cognate with Latin and Cyrillic). If I'm designing a Latin-Cyrillic-Greek typeface, I always design the Latin and Cyrillic lowercase together; then the Latin, Cyrillic and Greek uppercase together; then the Greek lowercase last, because it is a different writing system and requires a kind of psycho-cultural distance.

> The Cyrillic "cognates" need to be different than their Latin counterparts for the same reason that [...]

I admire your self-confidence.

> The needs of the living users come first.

Whose needs? Mine? I suspect I am a 'living user'. Spasibo.

John, you keep using history as a crutch.

> I have not found any good reason not to unify Latin and Cyrillic cognates

Here's two:
1) Vertical proportions. Which is why when you equate the x-heights Cyrillic seems smaller (ie less important).
2) Cyrillic has more straights. This means the spacing has to be looser. This means the bodies have to be more open (because of notan). For example in Armenian, you shouldn't make the vo the same as the "n"; Cyrillic is essentially a less pronounced case of the same thing.

--

Maxim: I believe that catering to laymen is very different than catering to other designers!

In any craft there is some inertia/conservatism, and certainly this helps hone skills and delve deeper. But change/progress is good too. In the 70s the Greeks didn't see any reason to make the their script different from Latin either, but now they do. Hopefully we won't have to wait until 2030 for the same to happen to Cyrillic!

A great example of progress is Adobe's recent Sava:
http://www.adobe.com/type/browser/P/P_1506.jhtml
It's not a text face, but that "Y" is very encouraging.

hhp

> I have not found any good reason not to unify Latin and Cyrillic cognates

BTW, I've been meaning to ask:
Let's think about the converse: What reasons are there to make the cognates identical?
I can think of two reasons (both not very good, in my opinion), but I'll wait for your answer.

hhp

> John, you keep using history as a crutch.

Great argument. Come to think of it, you're right. John is an expert, and expertise may be considered a handicap. You feel humbled. You think, 'I know that I know nothing'.

Ignorance, on the other hand, is so liberating. It allows you to highlight problems so easily, to come up with broard generalisations, to jump to conclusions, to voice your opinion unhesitatingly, to offer solutions and advice to them Tamils, Tibetans and Tajiks on how to reform their scripts. Piece of cake. Fun too.

'For in much wisdom is much grief. And he who increases knowledge increases sorrow.' Indeed.

> Maxim: I believe that catering to laymen is very different than catering to other designers!

You caterin' to me?

> In the 70s the Greeks didn't see any reason to make the their script different from Latin either, but now they do. Hopefully we won't have to wait until 2030 for the same to happen to Cyrillic!

Are you trying to foster our liberation from Latin slavery? How considerate of you.

> A great example of progress is Adobe's recent Sava: http://www.adobe.com/type/browser/P/P_1506.jhtml. It's not a text face, but that "Y" is very encouraging.

That 'Y'? Encouraging... whom? to what? Take it from a 'living user': Sava is indeed a very nice Akzidenzschrift but its design has no more bearing on Russian typographic conventions than, say, Hammer Uncial on Latin.

I have much to learn, but at least I freely admit it and are eager to advance.

Some people on the other hand are either too set in their ways and/or too afraid of admitting doubt.

Everything is a two-edged sword. Yes, even knowledge - because it's always limited.

> Piece of cake.

No, it's actually pretty hard, when you consider I'm up against the crushing weight of the establishment. It's not easy arguing against you, Maxim!

> You catering to me?

To some extent, yes - although not you specifically, more like the design establishment.
That's because peer acceptance does have value, even though it should always be secondary to catering to the real users. And laymen don't get all this stuff, they only feel some of it, and very subjectively.

> Are you trying to foster our liberation from Latin slavery?

Yup. It's basically a fight against globalisation.
I would do the same for the Greeks, but they seem to be doing OK now.

As for Sava, it's encouraging to me, at least. It indicates that you can make a font that goes against the superficial Modernist fallacy of identical cognates and still get it sold through the most mainstream font house out there. And maybe it's encouraging to Serbs as well? They're rightful users of Cyrillic too.

And nobody can predict the influence of something. Although it's not probable, Sava could possibly be a great catalyst. And maybe so can I.

hhp

Specifically, I'm eager to learn if there are reasons I'm unaware of for making the Cyrillic Veh identical to the Latin "B" for example. Releasing a font sooner is good business, but not good culture. Are there better reasons? To some people the fact that a Tsar messed things up 250 years ago is reason enough, but not to me, and I believe not to laymen either. Maybe in Canada 250 years is a lot, but I suspect not in Russia, and I'm sure not for an Armenian.

hhp

Some people on the other hand are either too set in their ways and/or too afraid of admitting doubt.

I hope you're not talking about me. I try very hard, sometimes against prejudice, to consider the possibility that you are right and I am wrong. This is why I discuss things with you. And when you say that Cyrillic letters should be differentiated from their Latin cognates, I think about it. I think about the Cyrillic typefaces I've designed, I think about Cyrillic text in a variety of languages, and I consider what changes I would make to those designs if I were to completely ignore their Latin counterparts. And, really, I don't see anything.

You accuse me of using history as a crutch, but you seem to have, typically, completely ignored what I've actually written about design processes, the nature of the cognates, their relationship to proportions in both scripts, etc.. History isn't anybody's crutch: history is the explanation of where we are. In the case of the Greek script history explains why attempts to conform the Greek lowercase to the Latin always look artificial and awkward. I don't rely on history to determine how I see: I use history to understand what I see.

Specifically, I'm eager to learn if there are reasons I'm unaware of for making the Cyrillic Veh identical to the Latin "B" for example.

Yes. It's the same friggin' letter! and so is the Greek uppercase Beta. The only reason to differentiate them is if you are designing in different styles. Yes, a poluustav Veh, an Irish uncial B, and an Attic inscriptional Beta all look different because they are in different styles. If Veh, B and Beta are in the same style, they are the same form.

Glad to hear you're listening. So am I. And I'd like to know the answer to my question.

As for your explanation of the "design process", I think I'd need you to elaborate to be sure (and I'd enjoy it to boot), but in toto it seems to me that you place too much value on precedent, on producing fonts the establishment will reward best. I know this is the sort of thing some people (like Maxim) class highest, and certainly such a stance is important to keep the machine well-oiled. But I expect more exploration, even more verbalized doubt, especially from somebody with experience. The more we learn the more we realize there's more to learn. Imagination is more important then knowledge. And all that. :-)

hhp

{Our messages crossed.}

> It's the same friggin' letter!

To whom? Not for your average reader, not at all! For most people one is a Russian (or Serbian, or Hungarian) letter, the other is English (or French, or Dutch). Their similarity is essentially circumstancial to them. You're talking about evolutionary lineage, but who besides us cares -or should care- about that? When people simply don't know the history, how much does it matter? Type design isn't about teaching history, it's about helping to communicate somebody else's message.

> The only reason to differentiate them is if you are designing in different styles.

I cited two reasons to differentiate them above, and I'll add a third:
To increase readability by reducing confusion when the two scripts are intermixed.
Here's my (now-tired) example: http://www.themicrofoundry.com/image/s_rome3-1.gif
In the part below, Meta's Cyrillic (the "normal" way) is simply too homogenous with the Latin.

And why wouldn't your logic apply to Greek or Hebrew - because of precedent? Too limiting.

hhp

Imagination is more important then knowledge.

Ah, we have a fundamental disagreement. That's good to know.

I don't know what I can usefully add to what I have already written: I design for the Latin and Cyrillic scripts as if they were a single script if both are in the same style. I believe there are sensible, historically-informed reasons to do this and, more importantly, I have found that the results work: it is possible to establish proportions, weights and spacing for both scripts around a shared set of cognates.

The same sensible, historically-informed reasoning also tells me that some possible cognates between different scripts appropriately take different forms within particular styles. For example, the Ypsilon in Veljovic's Sava is a form that is appropriate to the Greek uncial style, but it is not appropriate to his Latin design, in which it would appear too mannered. I don't know why you're so excited about the Yu in that typeface: the uppercase form of this letter is never the same as either the Latin Y or the Greek Ypsilon, so the fact that Veljovic uses this form is not innovative or groundbreaking: its just the appropriate form of Yu for the Serbian monastic uncial style. If there is something notable about Sava, it is not that it differentiates Cyrillic letters from Latin cognates (it doesn't: A B E M H O C T X are all the same glyphs in both scripts; only P is differentiated, as appropriate to the style, but interestingly Veljovic has unified the Latin K and Cyrillic Ka, so he is actually unifying letters that are normally distinguished, again, as appropriate to the style), but that it belongs to a specifically Serbian tradition.

In the part below, Meta's Cyrillic (the "normal" way) is simply too homogenous with the Latin.

To you? I look at it and can instantly see where one script ends and another begins. In order to have confusion, you would need to introduce Latin-script words into Cyrillic text that use only cognate forms, but that's what italics are for: to distinguish foreign words in text.

And why wouldn't your logic apply to Greek or Hebrew - because of precedent? Too limiting.

No, not because of precedence. I really don't care much about precedence, per se, I care about where scripts come from and how the normative forms of the letters were established. So I don't treat Greek (lowercase) and Hebrew in the same way as I treat Latin and Cyrillic because they are different scripts without a shared history. Latin and Cyrillic have a shared history: if they didn't I certainly wouldn't take the approach to them that I do, because it wouldn't be appropriate or authentic. But because of that shared history it is both appropriate and authentic. Yes, that shared history is pretty short in the grand scheme of things, but it is the specific history that has produced the modern scripts that readers use. I've never once said that the readers know or need to know this history: the history is for the designer to understand where the reader is at, where the script is at and, yes, also where the script can go. I've said many times that I'm a jobbing type designer -- I'm not an artist trying to create something 'original', and I'm not a reformer trying to give people something innovative that I think will be good for them -- I make type that people can use and read now, that also fits into the cultural framework that they inhabit. For most Cyrillic readers, that cultural framework is decidedly post-Petrine for everything except religion and nostalgia.

That's all for now. Some of us have typefaces to design.

> the uppercase form of this letter is never the
> same as either the Latin Y or the Greek Ypsilon

You can't say never.
But I guess we're lucky van Krimpen never made a Cyrillic...

> its just the appropriate form of Yu for the Serbian monastic uncial style.

Which is a distinction totally lost on the layman! So what most people see is a Yu that deviates from the Latin; and if it's less Latin, it essentially becomes more Cyrillic.

> it belongs to a specifically Serbian tradition.

But you don't think that would prevent its use by non-Serbs, do you? And that's how culture migrates, in this case hopefully making Cyrillic less Latin, albeit in a small way.

But yeah, the Ka sucks.

> I look at it and can instantly see where one script ends and another begins.

I think if you were immersed you'd stumble.

As for using italics:
1) That's limiting. It's better if the two scripts were different outright (for text).
2) What if the main text is in italic? :-/

> they are different scripts without a shared history

None of this history directly affects reading! Some of it certainly does affect reading indirectly, but my point is simply that gently deviating cognates, giving the two scripts slightly different vertical proportions, etc. flies under their conscious sensibilties, but makes for greater functionality.

BTW, doesn't Daniels (The Expert) say that there really is only one alphabet? (Via the Qalam list.)

--

John, I respect that you're so good at making fonts valuable to many people.
But please don't stop there.

hhp