The Type Directors Club is having a seminar on non-Latin scripts next year. You might check their website.

http://www.tdc.org/events/education/non-latinweek-ends/index.html

ChrisL

unfortunately, most design info for cyrillic is in Russian. i've been meaning to attempt to translate some of the literature to English, but haven't got around to it yet...
this has some info that may be helpful (i hope):
cyrillic how-to

the last article listed, in particular, has several illustrations that are quite helpful, even if you can't read the text.
http://www.prodtp.ru/index.php?act=recipes&CODE=03&id=13

The essays by Maxim Zhukov and Vladimir Yefimov on Language, Culture, Type: International type design in the age of Unicode edited by John D. Berry (2002).

I bet the articles by Vladimir Yefimov at Paratype E-zine would deserve an English translation as well : http://www.paratype.ru/e-zine/

“A Guide to using computer in Russian studies” :
http://www.yt.cache.waseda.ac.jp/fontguide.html

Maxim Zhukov: "On the Peculiarities of Cyrillic Letterforms" (Typography Papers 1. 1996, Reading University Press). It is possible to obtain an electronic copy (in mediocre-quality PDF scan) from http://www.bl.uk/services/document/orddocs.html

A.

Try to get a copy of Paratype's recent massive
specimen book, and observe observe observe.

And as you're understanding Cyrillic, please try to also understand
why it's become too Latinized, and how to make it less so in the future.

hhp

And as you’re understanding Cyrillic, please try to also understand
why it’s become too Latinized, and how to make it less so in the future.

Short Answer: Because Peter the Great wanted it to be.
Long Answer: Read the article on Civil Type in Language, Culture, Type

By "why" I didn't mean the history, but the functionality, as in what's bad
about it being so Latinized. On that, there is no book -or even article- yet.

hhp

.

I think that latinization is the expression of modernism within the realm of writing systems. In the 20th century, type designers were fond of chopping of serifs and simplifying forms, and also unifying the formal language of writing systems -- mostly by imitating the Latin forms. To me, modernism has always been an echo of the 19th-century fascination with the machine.

My idea for the 21st-century type design is a search for more natural, but also more ad hoc forms -- in all writing systems. Neither directly derived from the writing hand nor overly simplified. The technical limitations of type design are less and less existant, it's possible to draw thousands of glyphs, create glyphs made of components, engineer automatic, contextual substitutions. Typefaces need to find a balance between hand influences, mechanical influences and electronic influences.

A.

> any new type design in any script.

Uhhh, except making a font in the Latin script Latinized is sort of OK.

> surgically delete the individual’s imagination.

Equally bad is the desire/attempt to allow
individuality to subvert the needs of a society.

> latinization is the expression of modernism within the realm of writing systems.

Latinization and Modernism are indeed strongly linked.

> Neither directly derived from the writing hand nor overly simplified.

A big hug to you.

Everything has a role to play, and the better we understand a
given factor's stengths and weaknesses the better the Design.

hhp

> Uhhh, except making a font in the Latin script
> Latinized is sort of OK.

That goes without saying.

What I'm saying is that all writing systems have undergone "modernismazation" attempts in the 20th century, including the Latin script. For non-Latin scripts, latinization was one of the aspects of "modernismazation". I believe that the drive for modernism is no longer a good idea, which includes dropping the latinization trend for non-Latin scripts, and dropping other modernism-driven trends such as simplification, for all scripts.

A.

Uh uh, that sounds like stirring up a hornet's nest... :)
Maybe I'll wait a while and just observe (or hide my own attempts long enough).

Thanks for the links and tipps so far. And of course I have a follow-up question: What are the (some) »excellent« cyrillic typefaces (if this is possible to tell), I should have a close look at?

Sebastian

.

Sebastian,

Matthew Carter’s Georgia is most likely the best Cyrillic typeface designed by a non-Russian, and one of the best Cyrillic faces ever. It’s perfect in every sense.

The work done by Tagir Safayev and by Vladimir Yefimov for ParaType is seminal.

Especially check Yefimov's own Octava, his Cyrillic renditions of ITC Charter (great!), Kis (great!) and Raleigh as well as Safayev's Rodchenko typeface and his Cyrillic versions of ITC Officina Sans (great!) and Serif and of ITC New Baskerville.

Later check Yefimov's Cyrillic versions of Futura, Futura Futuris, New Standard, ITC Avant Garde, English 157, and Safayev's Cyrillic versions of ITC Kabel, Humanist 521 (Gill Kayo), FF OCR-F, FF Meta. The ParaType version of Humanist 531 (Syntax) is also very good, by Isay Slutsker and Manvel Shmavonyan.

Some of the new Microsoft ClearType fonts that ship with Office 2007 (Calibri, Candara, Consolas, Corbel) have very decent Cyrillics. I’m not very fond of Constantia and Cambria as designs altogether, but I’d say their Cyrillics are also not worse than their Latins. However, in general, Luc(as) de Groot, Jeremy Tankard, Gary Munch and John Hudson have a fairly firm grip on designing Cyrillics, while Robert Slimbach’s Cyrillics (Minion, Garamond Premier Pro) are more controversial.

The work of Alexander Tarbeev and his students is also worth looking at.

A.

"I would say that if you go with your imagination, and design an original, new typeface, and make the “ef” work smoothly within that context, you will have a better chance at transcending the weight of history."

I think this is the best approach to any type design as long as you are not attempting a revival where the historical part is your driving force.
Regarding history, it has been written, interpreted, revised and reinterpreted many times. If we always base our design logic on this shifting sand, it will always stand the chance of missing the intension. Better to design in the present for the present and allow your new design to be fulfilled in its own terms. Whatever you do, there will certainly be those who decree, "but it isn't...[siting the historical references which they most agree with]" so why not just let it be its own cosmos. It is probably more efficient to design something and reflect on its plus and minus characteristics than to do nothing because of an endless debate that stops you from ever getting started. There are those who will say it does not follow tradition and there are those who will say it is the same old stuff as we always see. The world has an ample supply of critics.

ChrisL

Robert Slimbach’s Cyrillics (Minion, Garamond Premier Pro) are more controversial.

Are there on-line critiques of these, I'd like to attempt to read them, if i can.

Also, the work that Alexei Chekulayev has been doing for Linotype in recent years is quite good: Frutiger, Sabon, Univers, Wiesbaden Swing, San Marco.

BTW, Chekulayev recently extended Frutiger for Linotype while some years ago, Tagir Safaev designed ParaType's own Cyrillic version of Frutiger called FreeSet. Both are good, though I prefer Safaev's version. I think it should be interesting for you to compare these two.

A.

Paul,

I don't think there are any critiques available, online or offline – especially in English. I believe that the Russian type designers who would be able to do such a critique are far too polite and modest to actually publish anything that would criticize fellow type designers (who otherwise do excellent work, like Robert Slimbach). However, if you track Russian typographic publications, you will notice the trend that the work of certain designers is regularly praised and awarded, while others are left unmentioned. Also, typographic conferences such as ATypI are often a good occasion to sit down with some Russian designers and talk with them one-on-one.

For anyone interested in Cyrillic type design, I recommend the TDC Cyrillic Weekend which will be held on March 2-4, 2007 at the Type Directors Club in New York. Maxim Zhukov’s expertise in the field of Cyrillic type design is unquestionable, he was the consultant who helped Hudson, Carter, de Groot, Tankard, Munch and many other Western type designers to get their Cyrillics where they are now. Maxim’s approach on Cyrillic is quite orthodox, with certain rules and schemes. As with anything in typography, these rules are very practical to learn even if at some point, you decide to break them.

Adam

> make the “ef” work smoothly within that context

The context of the users is more important than the context of the designer.
Especially for text.

And the necessity of going beyond precedent does not
mean the designer's whim can be given run of the house.

> Slimbach’s Cyrillics (Minion, Garamond Premier Pro) are more controversial.

Interesting.
1) How so, exactly?
2) Do you think this was intentional?

> the Russian type designers who would be able to
> do such a critique are far too polite and modest

Damn shame.

hhp

.

> that is best served by internal consistency within the typeface

No.

hhp

Interesting. I think the Cyrillic alphabet is really beautiful, Cyrillic text looks more balanced than its Latin equivalent, due to its lack of descenders and ascenders. More like Latin small capitals.

I'm able to read it (quite slowly, however) and to write it, but I wonder what the regular pitfalls are for a 'Latin' writer, trying to 'control' Cyrillic lettershapes.

Or could it also be an advantage, not knowing too much about history and tradition? You might get very close to the shapes themselves.

> Cyrillic text looks more balanced

In the same way that Latin caps look more "balanced" than Latin lc. But there's a
price to pay - and that actually alludes exactly to why Nick and Chris are wrong.

> could it also be an advantage

There is indeed such a thing as freedom from formal education.
The trick though is to maintain tons of self-doubt.

hhp

There is indeed such a thing as freedom from formal education.
The trick though is to maintain tons of self-doubt.

Self-doubt which is, if you cannot judge your own design properly in terms of legibility or suitability, close to not-knowing-what-to-do. I guess. Like writing a novel in a non-native language.

Here's an example of what I mean. The Cyrillic ZH is agreed to be one of the
trickiest to "balance". This is an unconventional ZH "solution" that I drew
during a (very short) trip to Russia in late 2004:

I think it's as sweet as Russian pastry. But I'm 100% ready to dump if the right
people give the right type of discouragement, no matter how well I might think it
satisfies some sort of "internal consistency". And this is something that somebody
who is essentially an artist would be qualitatively less likely to do.

hhp

Hrant, I'm not quite sure why you're so against "internal consistency". Aren't you the one who considers Baskerville's slightly swashy italic "N" to be obtrusive?

Internal consistency as I defined it, "the individual gylph fitting well within the context of the other glyphs," isn't the idea the same, for the individual glyphs to mesh into the whole?

Hrant: wow, that's really really nice.

I think I know what you mean. But then it's more about having the 'right' people around you (people who have grown up with that specific kind of alphabet), than knowing everything about its tradition and conventions yourself.

Internal consistency as I defined it, “the individual gylph fitting well within the context of the other glyphs,” isn’t the idea the same, for the individual glyphs to mesh into the whole?

But how do you know if a typeface is internally consistent, or is considered so by a group of users, if you're not aware of the grammar of the alphabet? I mean, it could look good, esthetically, but if it's not read well, then you're back where you started.

it could look good, esthetically, but if it’s not read well,

If it looks good, it will read well.
There are many well-established Latin typefaces that have a character or two with an odd alphabet-form, which nonetheless harmonize visually with the rest of the font. You could look at the bottom bowl of Baskerville's "g" and say that's not right (as some revivalists have), the bowl should be closed, but it's really not a problem. Souvenir's "g" is even stranger, and yet it has become a well-loved face.

If it looks good, it will read well.
There are many well-established Latin typefaces that have a character or two with an odd alphabet-form, which nonetheless harmonize visually with the rest of the font.

That's an interesting point of view. But that would mean that, for example, I could design a suitable Japanese textface just by studying the shapes, and the shapes alone. (supposing I don't have a clue about Japanese language but yet about esthetically good design)

But that would mean that, for example, I could design a suitable Japanese textface just by studying the shapes, and the shapes alone. (supposing I don’t have a clue about Japanese language but yet about esthetically good design)

You might be able to. You would need a long list of popular fonts to compare variations of form, an idea of how mainstream each one is, and fat helping of artistic talent.

Choz
!Exclamachine
.~

p.s.
If there is anyone out there who is interested in a giving feedback on the Cyrillic component of !Disc Inferno Internationale, I'd be happy to send them a custom beta copy to play with. Eben's feedback really helped improve ease of reading from the free version, with just a few simple tweaks. Keep in mind this is a theme-based display face, I'm just worried about serious showstoppers. Message me if interested. Thanks!

> is considered so by a group of users

First one would have to actually care about users...

> ... if you’re not aware of the grammar of the alphabet?

In fact "internal consistency" precludes any real sensitivity to
"the grammar". Essentially, it's about the self of the designer.

> If it looks good, it will read well.

Codswallop.

hhp

.

Choz,

I would gladly take a look (adam at twardoch dot com).

A.

Adam, what about that ZH?

hhp

.

The fact that it's easier should raise the warning flag even higher.
In fact I advocate making the "identical" letters not at all.

hhp

Hrant,

I fully agree. If you read Maxim Zhukov's paper, you will find that he firmly believes that some Cyrillic letters *must* be identical with their Latin "relatives". I don't think it's reasonable.

Most ParaType cyrillizations, and most Cyrillic typefaces altogether, suffer from the fact that Latin and Cyrillic share the same letter proportions. I think the Cyrillic lowercase could be slightly taller, and have larger sidebearings -- that'd be the minimum. I don't see why some forms couldn't be a bit different altogether.

This is wonderfully proven with ITC Officina and FF Meta. Of both Spiekermann's typefaces (both adapted for Cyrillic by Tagir Safayev), Meta actually works better in Latin text, but for Cyrillic, Officina is far superior. Officina used to be very popular in the West but when Meta was released, it gradually replaced and suppressed Officina. But not in Russia. Officina Cyrillic is still one of the most popular modern sanserifs there, and Meta is hardly present.

Officina has looser spacing and a higher x-height than Meta: for Latin, it is a bit over the top. For Cyrillic, it works perfectly. I don't even think it's an ad-hoc hypothesis: I believe that the intuitive choice of typographers in both script rooms and their "natural" preference for Meta over Officina or the other way around, delivers actual proof.

Adam

.

> it breaks down in one particular circumstance,
> which is where the two scripts appear together.

On the contrary, that's exactly where it's most potent.

hhp

Nick,

both in Renaissance and in modern fonts, Latin and Greek differ noticeably, and there is no "problem". The Greeks had their latinization phase but they're over it now. The Russians don't necessarily have the confidence yet.

Also, nobody said that the letterforms have to differ *noticeably*. "d" looks like a "p" turned upside down, and the top "B" counter looks like it's the same size as the bottom, but we both know that neither of that is true. Harmonization of forms does not mean making them geometrically identical.

In fact, the current Cyrillic faces often suffer from the problem that if you set two columns of text, side-by-side, one in Russian and one in English, with the same point size and leading, then either the English is too tight or the Russian is too loose.

In my opinion, the principle of script harmonization within multilingual typefaces is simple: when set side-by-side in two columns *and* when mixed together within one column, the scripts should *look good*. Now, what "look good" exactly means is up to the type designer to decide. I'm sure you'll agree that it doesn't have to be simple geometric copy-pasteness.

As for the Cyrillic "ef" in New Standard, it fits. It is a reminiscent of the era, and is probably a way of reaching out to the "greekness" of some Cyrillic forms rather than following a Latin model. In Cyrillic, the double and the reverse letters (ef, zhe, ya, yu etc.) are among the most unorthodox ones.

I recommend that you order a copy of Yuri Gordon's book (http://store.artlebedev.com/catalog/books/design/kniga_pro_bykvy/ ). It shows some nice examples of Cyrillic letterforms in different writing styles, and also their extreme treatment in logos -- which can be helpful in exploring the "acceptable" borders for each character.

Regards,
Adam

.

> either the English is too tight or the Russian is too loose.

And the Russian appears smaller.

> when set side-by-side in two columns *and* when mixed
> together within one column, the scripts should *look good*.

This is a tall order - too tall for many and maybe most combinations of two* scripts. This is why Nour&Patria is a multi-lateral system. Yes, the archetypical Modernist decries it as Too Complicated, but as far as I'm concerned that's more of a reflection on him. :-)

* And when you get into more than two (which is rare, I admit) then it crumbles.

> I’m sure you’ll agree that it doesn’t
> have to be simple geometric copy-pasteness.

You are? :-/

Wow, Yuri Gordon's book looks scary-good.
Somebody should write something like that for Latin. :-)

hhp

> I won’t be convinced until shown an actual setting.

I don't think even then. For one thing, you don't see a problem with the mainstream scheme, so the foundation isn't there. I'm sure your "internal consistency" crosses script boundaries effortlessly.

hhp

I'm most curious to hear what is "controversial" about Robert Slimbach's Cyrillic designs.

FWIW, Maxim Zhukov has also been a consultant to Adobe on our recent Cyrillics. (Though this should not be taken to mean that he approves of all of Robert's choices.)

Cheers,

T

Too Complicated

Art Director A to Art Director B, "Is that the Latin Cyrillic, or the Cyrillic Cyrillic?"

a way of reaching out to the “greekness” of some Cyrillic forms

Possibly. My interpretation is that it's more an expression of the appropriateness of symmetry in a "modern" (didone) typeface.

I recommend that you order a copy of Yuri Gordon’s book

I downloaded the sample pdf a while ago, and was mightily impressed, especially with the serifed text face it's set in. But is it published yet?

> Art Director A to Art Director B, “Is that the Latin Cyrillic, or the Cyrillic Cyrillic?”

Cyrillic Master, Latin Master, Subordinate Latin, Subordinate Cyrillic.
I can understand if that's too complicated for Art Directors - I just hope
they don't find out what happened to Garamont.

hhp

> especially with the serifed text face it’s set in.

Hmmm, look at the lc "ef": what do you think the chances are that its vertical proportions are "internally consistent" with its (prospective?) Latin companion?

hhp

.

> Your multi-lateral system would have to be widely adopted

Yes, Comic Sans is the epitome of typographic quality.

You might ask Emily Artinian why Nour&Patria was the only
type system on the planet that did what she needed done.

Related: http://typophile.com/node/29461

--

BTW, it turns out that Artemius -Gordon's font- is a bi-script system. You might want to check out the terribly internally inconsistent vertical proportions...

hhp

.

Is Yuri's book only in Russian or is there an English translation as well?

ChrisL

Chris: I asked them per e-mail four or five days ago, but got no answer yet...

Thanks Sebastian. Let me know what they say?

ChrisL

I downloaded the sample pdf a while ago, and was mightily impressed, especially with the serifed text face it’s set in. But is it published yet?

Is Yuri’s book only in Russian or is there an English translation as well?

Yes, it's available, but the shipping is steep and it's all in Russian. It is very impressive and the illustrations are very informative. As for Russian the Wiley / Katzner Russian English dictionary is the only one I've found that has any sort of typographic terms in it. Anybody know of a better one for design and type?

George
I felt bad because I had no shoes, until I met a man who had no Bodoni

> Any plans for a Cyrillic along the same lines?

It depends. For Cyrillic I'd need plenty of "native" advice though. And as I've explained in my Spatium/Hyphen article it's more work adding a new script in a multi-lateral system: you have to make a new master with subordinates for the pre-existing scripts, plus you need to make a subordinate for the new script for each pre-existing script. The good news is that a lot of the process can be mechanized, like how Patria's Roman and Italic sprouted from the subordinate Latin, and how the subordinate Armenian came from the Armenian master.

> I would expect it to accentuate the effect, in parallel
> column setting of Latin and Cyrillic, of the Latin being
> more tightly leaded.

An effect which, especially being "pre-existing" (I mean in the conventional scheme) is best alleviated by giving the Cyrillic a larger "x-height" (which is more significantly justified by the issue of apparent size). On the other hand, there are much worse things that different apparent leading, or even misaligned leading between two columns - for example a Cyrillic that's harder to read because the proportions are off.

> Artemius’ descenders, x-height, sidebearings, and common
> glyph-shapes appear to be the same across the two scripts.

Which coupled with the difference in ascenders makes
the whole too subtle, like a mistake, in your view, no?

hhp

> Anybody know of a better one for design and type?

ATypI was working on a multi-lingual type glossary - dunno where it's at now.

hhp

> best alleviated by giving the Cyrillic a larger “x-height”

Or on the typesetting side by setting the Cyrillic slightly
larger in point size (with the same line-to-line leading).
This does introduce color issues (assuming a color-balanced
Latin/Cyrillic system) but not much. In fact color differences
can be used to compensate for script-inherent apparent size
issues (like how Arabic always looks smaller than Latin).

hhp

"For Cyrillic I’d need plenty of “native” advice though"

I like this notion. Should that advice come from the Bulgars for whom it was originaly intended? the Moravians who first received it? The Bulgars who exentually received it or the Russ who eventually made it there's. Like, there's only one Cyrillic native?

.

Should that advice come from the Bulgars for whom it was originaly intended? the Moravians who first received it? The Bulgars who exentually received it or the Russ who eventually made it [theirs].

All of the above.

David, I guess pity those who need native advice for Latin. :-)
Or maybe, we all need it.

You could pick your advisors based on potential sales, or number of users, or some other scheme. Fortunately type design is non-deterministic - there's never really a "solution", so you find the best advice you can from as many people as seems reasonable; the best you can do is never better than what your circumstances allow for. As I like to joke: there's always room for improvement... so stop trying! :-)

> Giving the Latin ascenders extra height is a nice touch

This does not jibe {edit - was "jive"} with your previous
view that differences should either be none or obvious.

> In a mixed setting like this caption from the Gordon book, set in
> Artemius Sans, surely having different x-heights for the two
> scripts, and different letter-forms or proportions for the common
> letter shapes (B, C, e, a) would be an unnecessary distraction.

I agree, but only because it's a caption: deliberative reading.
For immersive reading (much better with a serif anyway)
there is no distraction; the subconscious doesn't care.
And the benefit comes through.

hhp

.

While frivolous theory can play havoc with other people's practice.

--

BTW, are there other Cyrillic (text) fonts where
the descenders are longer than the ascenders?

hhp

.

ROFL!!!

ChrisL

You two should make a movie.
http://www.imdb.com/title/tt0109686/

hhp

.

>Bill

You rang? Yeah, theory should *adapt* to observation, just as in science. The eye is the final arbiter in type, not the theory--even though I think theory can be a big help. When theory becomes the--ahem--absolute monarch it can do a lot of damage.

> theory should *adapt* to observation

Nah, theory should adapt to the needs of the artiste, like
his need not to have his opinions exposed as fraud in public.

In The Official Thesaurus of Shinnistan, "flexible",
"convenient" and "self-serving" seem to be synonyms.

--

BTW, for the record, I really LOVED "Dumb & Dumber".
But it certainly helped that it wasn't cast as a documentary.

hhp

name-calling not needed, let's stick to the topic, hrant.

A question, for Adam and of course others with good
insight into Cyrillic, and the perceptions thereof:

Do natives feel that the lc looks like a mish-mash of UC and lc forms?* Maybe if only as a result of exposure to Latin? Or is this perception (which I very strongly have) merely an artefact of Latin nativity (and lack of Cyrillic nativity)? And if it's only the latter, how much does that matter? Lastly, if this sentiment is there, even mildly, has anybody done anything about it?

* Not that the two Latin cases don't have that issue - although in Latin
there's the real distinction of case - you don't see much intermixing.

hhp

.

Chris,

Yuri Gordon's book is in Russian only but it's useful even if you don't read a single word in Russian. The illustration and the drawings included there already make it a good investment.

David,

> Should that advice come from the Bulgars for whom
> it was originaly intended? the Moravians who first
> received it? The Bulgars who exentually received
> it or the Russ who eventually made it there’s.
> Like, there’s only one Cyrillic native?

The Russian Cyrillic tradition now has the position of, say, the American Latin tradition. The Bulgarian Cyrillic tradition could be compared to perhaps the Norwegian Latin tradition, or something. I mean, regarding influence.

You're American, David, so you should know that you're the most important guys out there. Same for the Russians in the Cyrillic domain :>

(Only half-kidding)

But it is true that the Russian type designers or consultants, such Vladimir Yefimov or Maxim Zhukov, are often more rigid and orthodox in their views than, say, Jovica Veljović, who is a Serbian native. There was an interesting discussion between them at the ATypI Prague conference where Jovica showed his typeface in progress that showed some unorthodox Cyrillic forms.

Jovica Veljović's Sava Pro is worth looking at. It's small-caps-only but shows some wonderful "brave" Cyrillic capitals.

Adam

Even more than bravery I think Sava is notable for its sensitivity. The "Y"-shape variance between the three scripts it supports encapsulates what I've said about "anti-identicality".

http://www.adobe.com/type/browser/pdfs/SAVA/SavaPro-Regular.pdf

hhp

Dez, would you stop that, please. I mean, at least put some caption(s) indicating who's who. Otherwise it's just gratuitous stooges, (or is that what you mean)?

Paul: All of the above.
I've seen that spec. The mystery left in this work, (regardless of script) is how one weights the choices: geographically, by population density, to a historical time or technology, or etc., so when one making the gross and then detailed decisions, one is targeting something less than all, or looking long at the end of ones own tail. I think, in general, if you want to serve them best you have to design their core glyphs first with a mind clear of your own. I.E we design around our use of our l.c. e.g.

Hrant:David, I guess pity those who need native advice for Latin. :-)
Either I don't understand, or the word pity translates to serve if I say the same thing?
The point is, Latin has been hyper-styled for obvious reasons (massive cultural and ethnic absorbsion plus geographic spread, so, as other scripts are freed to do so, it's of benefit to examine what Latin's done, how and why. No?

> how one weights the choices

Non-deterministically. Non-Modernistically. As best one can.
Which is always better than either dismissing the issue or running away.

> design their core glyphs first with a mind clear of your own.

Clear? Never! :-)

Concerning Latin what I meant was that if one needed "native" advice the way you worry it needs to be done, one might be in worse shape than for Cyrillic.

If you mean that Latin is more free than Cyrillic nativity-wise, I would agree on a certain scale, but never completely, and because it's used so much more broadly the net issue is at best comparable. For example note the issue of acute/grave accent angles in Polish versus French, or the particular -if nonformalized- preferences in certain African lanaguages that use Latin.

> it’s of benefit to examine what Latin’s done, how and why.

Certainly. Study, but not emulate, not per se.

In the same way that type should study chirography and take away things like stroke contrast, but not because of where it's coming from, but because it's good.

And these two can be combined: in Thai the "traditional" style is monoline, but looking at Latin and being inspired to introduce stroke contrast (as is now done) is a good thing. But being inspired because it's from Latin (and/or chirography) can only bite you in end.

hhp