Angle of the tonos?

Nick Shinn's picture

In 1980 the Greek government decreed that the polytonic alphabet would be replaced by the simplified monotonic system.

Prior to that, there had been a popular/commercial movement towards this, featuring various forms of the simplified tonos accent -- dots, vertical dashes, and triangles.

In 1982 the Greek government (perhaps in a move to combat the tackyness of many of these accents nouveaux) decreed that the tonos would be angled, like an acute accent. This has subsequently become the Unicode standard.

HOWEVER

The tonos in Lucida Grande (Apple OS font) is a small vertical dash, quite different from the acute accent.
The tonos in various Adobe Pro fonts, such as Minion and Myriad, is slightly angled a few degrees off vertical, again quite distinct from the acute.

SO

Is this like the Euro symbol situation, where a typographic standard has been set, but found wanting and freely interpreted in practice?

dezcom's picture

I always thought the Tonos was part way between the angle of the acute and the vertical. Hopefully the Greek authorities will chime in on this one (I don't mean the Junta :-)

ChrisL

dezcom's picture

Gerry Leonidis spoke highly of the new Adobe Garamond Premium Pro in terms of Greek.

ChrisL

John Hudson's picture

The monotonic tonos is identical to the ploytonic oxia, and should have the same form. Essentially, it is an acute accent (indeed, the Unicode combining diacritical mark for the oxia is the combining acute), but in traditional Greek types it tends to be steep and often taller than the Latin acute.

Nick Shinn's picture

Well yes, but this Unicode standard doesn't appear to have much of a leg to stand on if it is being ignored by Apple (Lucida), Adobe (Minion, Myriad, etc.) and Microsoft (Tahoma)!

gerry_leonidas's picture

Nick, I see you are dabbling in Greek ; good! I will add my two bits in relative haste, under great pressure of time. (I've got it also on my to-do list to respond to the thread about caps, but boy do you and Hrant make it difficult to keep up...) So, here goes.

Firstly, the Greek government would not give a monkey's for the aesthetics of the monotonic accent. The passing of monotonic in 1981 was one chapter in a very long debate going back several decades, but precipitated by politics and concerns by the local publishing industry, in the face of imminent capital investment and associated human resources costs. With hindsight, keeping in mind that this happened only three years before the DTP explosion (read "me and my dog in my living room are a service bureau") which was built on 8-bit fonts, this paved the way to font piracy and localisation of the most abysmal quality.

All the Greek reference works have stated, from day one, that the monotonic tonos is identical to a polytonic acute. Any dots, triangles, dashes, and such are wrong, ugly, and inappropriate for the formal characteristics of the script.

In reality, this must be tempered by the context in which it was expressed. Therefore, it is acceptable to design the tonos at a slightly different angle, length, and weight from the acute, since the tonos sits alone over vowels, and does not need to accommodate breathings -- but must always be right-inclined. My personal view is that diacritics should be designed to relate in stroke thickness and modulation to the lower-case glyphs; in other words, make them loud and clear.

So, Unicode is correct in this. Unfortunately, Lucida Grande is as poor in the Greek as it is good in the Latin, so Mac users do not reference out of the box.

You're doing well to be looking at what others have done, but you should be casting your eyes on a wider range of printed material, and -- why not? -- even writing examples.

Nick Shinn's picture

Thanks for your input Gerry, and respect for your work (with Mr Hudson et al) on the ClearType fonts -- the language support of the project, to me at least, overshadows the technological aspect. I'm sure it will provide a huge boost to Greek and Cyrillic typography, by opening up the design space available to typographers there (without meaning to be patronizing to local type designers) with new fonts with italics, bold, and small caps. I've also been impressed with Adobe and Typotheque's committment to supporting polytonic.

While I had originally intended to produce only a monotonic font, I've been persuaded to go the whole hog.

Would I be correct in assuming that, in creating a polytonic font, I can use Gentium and Minion Pro as models for the unicode values and accents?

The typeface I'm working on is a revival of a roman didone, from c.1870, featuring a strong vertical stress and absence of pen-written ductus. The Greek characters are new, as I was unable to discover any precedent that I considered to be sufficiently "modern". It seems to me that the correct tonos for monotonic use here is vertical. That would be no good for polytonic, as the tonos there must be angled, working in conjunction with the grave. Presently, I'm leaning towards the idea of making the text faces polytonic, and the display faces monotonic, with vertical tonos. A bit of a shame, as I liked the look of the vertical tonos in text. Is it really so inappropriate?

http://www.shinntype.com/Albany.pdf

John Hudson's picture

Well yes, but this Unicode standard doesn’t appear to have much of a leg to stand on if it is being ignored by Apple (Lucida), Adobe (Minion, Myriad, etc.) and Microsoft (Tahoma)!

The tonos marks in Minion and Myriad differ from the oxia accents, yes, but they are right-leaning, which is the main thing. What I think is a mistake is the fully vertical form.

It is worth bearing in mind that Chuck Bigelow, Chris Holmes and Matthew Carter all made their first Greek types back when things like the triangle form in Helvetica Greek were all the rage. A lot has happened in Greek type design since then. I've recently been collecting samples of various demestica versions of Helvetica from Greece, and have noticed that most of these have rejected the old triangle form in favour of something more acute-like.

gerry_leonidas's picture

The typeface I’m working on is a revival of a roman didone, from c.1870, featuring a strong vertical stress and absence of pen-written ductus. The Greek characters are new, as I was unable to discover any precedent that I considered to be sufficiently “modern”.

Nick, thanks for the peek. I only will apologise in advance for saying what I have to in a few sentences only, and hope it does not come across too dogmatic.

I think you are taking "Modern" as it applies to the Latin script and imposing it on the Greek script, ignoring typographic interpretations of the script at the time. The Didots' contribution to the development of Greek typeface design is seminal, and very much not in a "strong vertical stress and absence of pen-written ductus". If you are not in a hurry, I could send you some images of the real thing in the next few days.

As it looks on your PDF, the Greek looks like an update of the misguided genre set out by Times Greek in the late fifties. I would be very interested to see what you would come up with if you loosened up your take on the script. To borrow from Dwiggins, have a couple of tumblers, then write with a soft-ish nib without resting your arm on the surface.

gerry_leonidas's picture

Forgot to say: it is great that you will be producing polytonic!

Nick Shinn's picture

Gerry, perhaps it is presumptuous of me to have my own ideas of what a Greek type may be, not being a Greek, but I have studied a wide range of interpretations, including the Didots', and have, as I said, been unable to find a "modern" precedent fully suited to match the formal qualities of the mid 19th century modern roman, although Bodoni's Greek of c.1790, which does in fact have a pronounced vertical stress, led me to believe that my direction was not completely anachronistic (GB being more of an inspiration than the Times).

BTW, the italic of Albany will feature the more traditional pen-driven "Baroque" (to use Bringhurst's term for 18th century italics of varied stress angle) ductus. This will establish a considerable contrast between the upright and slanted styles, the lack of which is presently IMO a weakness of Greek serifed typography. One of the riches of serifed western type is the stylistic contrast between roman and italic. In fact, many people (although not I) consider slanted roman to be a second-rate form of oblique emphasis, a failed modernist experiment (note Optima Nova Italic). I should point out that different italic forms do exist in the traditional Monotype Greek 90, they are nicely done, but quite subtle. This is also the case with Minion Pro. The ClearType Greek italics have alternate more-cursive forms of beta and theta (although not the curly rho), but are basically stylistically slanted versions of the roman glyphs. So this stiff-arsed Albany roman serves to set up the loose italic, which I'm working on.

I approached the design of this typeface with an eye on history and culture, but also philosophically as a problem in pure design. At its most basic, my design argument is this: if the Roman lower case, derived from pen writing as surely as is the Greek, can exist with or without serifs, why can't the Greek? I'm not taking it to that extreme (yet), but that's the couple-of-tumblers, "why not?" logic at back of it.

I'm not concerned that modernizing, or borrowing typographic techniques first invented in the West, is somehow the dreaded Latinizing. After all, if Western concepts like a slanted emphasis style, small caps, and bold face are OK, the cat is out of the bag, ennit? And thanks for the sans.

There seems to be a confusion in Greek typology between alphabetic form and typeface style. The common typeface during the 20th century was known as "Plain", or by its Monotype designation, Greek 90. In the West, a similar situation had existed, prior to the 1890s, after which typefaces began to be known by their designer's name. A sense of historical relativism structured this understanding, and led to an explosion of stylistic diversity, fostered by eclectics such as Morris Benton and Stanley Morison. To categorize those Greek faces which have provided stylistic diversity, such as Times Greek, as "misguided", is restrictive. Sure, the Elzevier types, of which Times is an example, aren't considered as posh fonts, but not every typeface has to be an aristocrat. In some circles, ITC Garamond is considered a travesty, but it has proven a successful and useful face, and while some may consider it misguided, it is not wrong in the sense that you are implying a post-ductus Greek lower case is.

To me, the "gestural" quality of the Greek alphabet which underlies the ClearType Greeks is a myth. I don't deny that it has great historical authenticity, and provides a marvelous corporate stylistic consistency, but it is not the only legitimate road forward.

I take it then, that given the "misguided" nature of my typeface, an equally misguided vertical tonos would be quite appropriate?

John Hudson's picture

Nick, you've hit the Greek upright vs. italic problem pretty squarely. I described it this way to Gerry when I was working on Constantia: in terms of increasing cursivity you get

Latin upright -> Greek upright -> Greek italic -> Latin italic

If you follow what you've characterised as a 'gestural' approach to Greek, you end up with a design that is more cursive than the corresponding Latin. As I've stated elsewhere, I think this is appropriate. The problem then is how to distinguish the italic, other than by slant alone. I think there are things that can be done in the italic itself -- as opposed to pushing the upright to be less cursive --, and I'm quite impressed with what Robert Slimbach has done with the Garamond Premier Pro upright and italic Greeks in this regard. That said, I'm all for typeface design as a laboratory, and look forward to seeing the Albany italic.

Nick Shinn's picture

>more cursive than the corresponding Latin. As I’ve stated elsewhere, I think this is appropriate.

You've demonstrated it perfectly in Constantia, with sufficient heft to not be outweighed by the Latin.

And I agree with you that there is much that can be done to differentiate the italic, but I think it may also be appropriate to take an alternate route with upright Greek, in a face of extremes such as this kind of modern.

>Garamond Premier Pro upright and italic Greeks

I haven't checked out the Italic yet, but I just had a look at the PDF for the upright, and noted that the monotonic characters have a vertical tonos, whereas the polytonic versions of the same character have an acute tonos. I don't have the font, so I'm wondering how that's handled. A stylistic alternate?

dezcom's picture

Nick,
I think you have done a terrific job! The fit with the latin and the color is excellent. The text at the bottom where the 2 scripts meet is telling. The proof is in the pudding and you have shown that. The only thing that bothers me is the latin M as the Greek cap Mu and that is a small thing and just me. I feel like the strokes should not be vertical in theMu and that it looks a bit too condensed for the rest of the Greek. Also, the vertical tonos creates too much symmetry for me. It becomes too much part of the glyph below it. This may be because I just am so unaccustomed to seeing it vertical and it might grow on me.
I like the modern didone feel! I don't think that the calligraphic model is the only one to be followed in Greek or in any other script. We don't make roman capitals to look like they have been chiseled into marble either. I think you really have something here.
But what do I know, I am just one of your Souvlaki eating buddies:-)
ChrisL

Nick Shinn's picture

Thanks for the props, Chris -- I'm thinking Albany and Leporello Black would team up nicely in Greek. Care to swap?

I am going to go with the angled tonos, because I've realized that it's an integral part of polytonic (or "Regular Greek" as Yannis Haralambous calls it). Also, without spacious leading the vertical tonos melds into descenders on the above line, so an angled tonos allows tighter leading. Doing a separate version for polytonic and monotonic in the same font, as per Adobe Garamond Pro, is a bit of a stretch for me. But I will make the display version monotonic with vertical tonos, what the heck.

As for the Mu, this goes back to another thread, concerning whether to have Greek caps that are different from the Latin, for better balance between the cases. The sample text I used here is from an online translation lesson (except I changed the Latin name from Carrefour to Morisons), representing a typical piece of commercial writing, a PR release, mixing languages. The situation here would be strange with different Ms for the different languages, as the sound is the same, so that led me to stick with the common glyphs. I hear you, this is an odd Mu, but it is also a bit of a strange M in the original face this revival is based on, so I'm afraid I can't change it (doing a revival makes those decisions for one).

dezcom's picture

"...I’m thinking Albany and Leporello Black would team up nicely in Greek. Care to swap?"

Sounds like a plan! I was always a sucker for a boney Didoni "I've got a girl named Boney Didodoni, she's got Tonos like a stick of old Marconi..."

ChrisL

John Hudson's picture

I haven’t checked out the [Garamond Premier Pro] Italic yet, but I just had a look at the PDF for the upright, and noted that the monotonic characters have a vertical tonos, whereas the polytonic versions of the same character have an acute tonos. I don’t have the font, so I’m wondering how that’s handled. A stylistic alternate?

The tonos isn't totally upright: it leans a little to the right. This follows Robert's approach in Minion and Myriad also (I'm guessing he was largely responsible for the Myriad Greek). As I've said, I don't see any reason for the tonos to differ from the polytonic oxia. If it is going to differ, as it does in these Adobe designs, then it should still resemble the oxia/acute in form -- i.e. not be a triangle, circle, square, etc. -- and should still lean to the right.

A nice, authentically renaissance touch in Garamond Premier Pro: the Greek accents are not at the same height on all letters so, for example, the tonos on the alpha and eta sits a little lower than one the epsilon and iota. The vertical placement varies relative to the to the height of the letter at the point of horizontal alignment.

John Hudson's picture

I feel like the strokes should not be vertical in theMu and that it looks a bit too condensed for the rest of the Greek.

Ironically, I recently found myself having to straighten the sides of a Greek uppercase Mu, contrary to the gently sloped left side of my Latin M. The reason for this is that the font is designed for scholarship and I needed a clear distinction between the Mu and the ancient letter Shan, which is very similar to the Mu. After some discussion with Gerry and another font developer who has done a lot of work for classical and Biblical scholarship, I decided that the sloped sides were appropriate to the Shan and that the Mu should be straighter for contrast.

dezcom's picture

"Ironically, I recently found myself having to straighten the sides of a Greek uppercase Mu, contrary to the gently sloped left side of my Latin M"

John, I had no scholarly reason. I was trying to distinguish between the M and the Mu as well but felt the angles sides would feel more comfortable with the general tendency of Greek to be more irregular looking (less rectilinear?) than latin. Nick mentioned he is trying to be true to the revival and that made the decision for him.

ChrisL

Nick Shinn's picture

>I would be very interested to see what you would come up with if you loosened up your take on the script. To borrow from Dwiggins, have a couple of tumblers, then write with a soft-ish nib without resting your arm on the surface.

Gerry, I'm not doing a script font.
And spending the time to take my rusty Greek hand beyond its arrested development, into cursive, is not something I can afford to do.
Even so, I'm not sure writing, or drawing even, has a necessary involvement with the process of type design these days.

However, I do have a sense of where you're coming from, and of the importance of the written form to Greek text.

Look at it this way: imagine a culture takes the best available technology and mechanizes writing. It happened during the incunabula. But forget about that, and all the baggage of the western typographic tradition. Look at Greek writing now (both classical and vernacular), and the OpenType format. Start from scratch. Use contextual alternates at the level of sophistication created by Adam Twardoch in Zapfino Extra Pro, to capture the full cursive quality of the writing. Either work subjectively, and base your typeface on your own hand, or go into the field and research how people write today in Greek (and I mean a proper ethnographic study). I'd be very, very, interested to see what you would come up with.

John Hudson's picture

Either work subjectively, and base your typeface on your own hand, or go into the field and research how people write today in Greek (and I mean a proper ethnographic study). I’d be very, very, interested to see what you would come up with.

Segoe Script Greek.

Nick Shinn's picture

>Segoe Script Greek.

Where can I see this?

John Hudson's picture

I have not seen any sample of the Greek online. Perhaps if Si Daniels is reading this he could point us to something. Carl Crossgrove gave a presentation on the Segoe Script OT Layout stuff at the ATypI type tech forum in Helsinki, which included an overview of the Greek. It is based, with Gerry's help, on fairly typical contemporary Greek handwriting.

Si_Daniels's picture

Sorry, only just spotted this thread. Nick please send me your sample text and I'll set it using the latest version of the font and send back an image.

Cheers, Si

Nick Shinn's picture

To conclusively update my explorations and experiments:

1. The vertical tonos (in monotonic setting) was abandoned, for the primary reason that in an upright font, tightly leaded, such a tonos becomes too easily entangled and confused with the (serifless) descenders of the line above.

2. There are several logical reasons for a vertical tonos, but they are all trumped by the above consideration. Most obviously, the reform movement sought to replace all accents -- left, right, etc -- with a generic accent, so the vertical form (as expressed in the triangle) made sense, being neither sharp nor flat. However, it's worth considering Hungarian, Polish, etc., which have only one direction of accent, acutely angled. A vestige of the written form? Perhaps, but also typographically functional as per point (1) above.

3. The steeply angled tonos in Adobe Greek fonts (Minion, Myriad) is not a problem there, because I do not believe they are designed for tight leading (Utopia -- a "news" face -- is much bigger on the body). However, it is a parameter of Albany, the face I'm developing, that it should be able to be set solid with extenders almost touching, and a vertical tonos does not work well in that situation.

4. I found a good source of polytonic text on the Internet (Smyth's Grammar), but it uses the monotonic versions of the tonos-accented characters, rather than the proper polytonic versions. How do I know? Because in fonts such as Minion Pro, the polytonic tonos is angled more vertically, whereas the monotonic version is more like an acute. So, it makes sense to use the same-angled tonos for monotonic and polytonic, rather than create unnecessary sophistication, which may not work in real-world situations like this.

5. Some "cheating" of the tonos angle is appropriate in the compound accents.

John Hudson's picture

I found a good source of polytonic text on the Internet (Smyth’s Grammar), but it uses the monotonic versions of the tonos-accented characters, rather than the proper polytonic versions. How do I know? Because in fonts such as Minion Pro, the polytonic tonos is angled more vertically, whereas the monotonic version is more like an acute. So, it makes sense to use the same-angled tonos for monotonic and polytonic, rather than create unnecessary sophistication, which may not work in real-world situations like this.

This is sound analysis.

Note that this is not necessarily simply a case of a user choosing to use the tonos diacritic characters to encode the oxia diacritics: alpha with oxia (U+1F71, ά) has a canonical decomposition to alpha with tonos (U+03AC, ά); i.e. Unicode considers them to be identical (which they are, of course, grammatically and historically), and normalisation may result in decomposition of oxia diacritics to tonos diacritics. Note also that tonos and oxia share the same combining mark encoding (U+0301, the combining acute accent*), which means that any design distinction between the two will simply disappear if fully decomposed normalisation is applied. Since text may be subject to different kinds of normalisation, you have no way of anticipating how the diacritics will be displayed unless you make the tonos and oxia identical.

* Note that in a bi-script, Latin and Greek font, it may be necessary to include two forms for U+0301 (and other common combining mark characters), appropriately linked to each script/language using the 'locl' feature. This is one of the reasons why I am shipping a separate SBL Greek font in addition to including Greek support in SBL BibLit, so I can directly encode the combining tonos glyph as U+0301, the combining varia as U+0300, etc., in the former.

Syndicate content Syndicate content