{Changing my torn-up clothes...}

JF, it's great to see "linguistic sensitivity"!

In your nice page with the illustrations there are some interesting things about letterform width with respect to language. I also wonder about two other things:

1) Variations in vertical proportions.
2) Variations in spacing.

For #1, I would point out your figure #5, and note that German relies more on extenders than either French or English. Does that mean German needs a smaller x-height? Especially when you also consider that German words are longer (and a smaller x-height would allow for less overall width of words)? This can also be considered for Latinate languages that use a lot of accents - which is one reason Bilak's Eureka has a small x-height.

For #2, I would point to your figure #2, and note that German seems to have more x-height verticals in proportion to rounds - does that affect "default" spacing?

(Man, I'm actually glad I didn't have to type that over 20 lines in a chat room...)

hhp

Bonjour Jean-Fran

Start with German versus French. In 1997, I realized that German use more variety/various letterforms than French. More ascenders glyphs in various combinaisons, when french use most of the time similar letters, with in some respect, more descenders, in fact very flat language physically speaking. So, German type need to be more straight, more regular to control all the letters togethers. The various forms achieve the rhythm, when in French, the same typeface can quicky be very boring, because French is naturally too flat

Hrant, I don't think Greman should have smaller xheight? Why you want to emphasis ascenders? i will do the reverse, because there is so much, its not necessary to "destroy" more the regularity.

German is lively in the forms used, when french is flat in the forms used.

So, typefaces need to be designed in reverse. Lively for french, flat for german.

My question was a little more precise than JFP's summary. I know from experience that it is possible to design good typefaces for scripts that one cannot read. But I'm wondering how far, within a script, a designer can go to addressing the particular nuances of the ways individual languages use that script?

Ironically, I see in JFP's illustrations the beginnings of an analytical, empirical approach to the question. Look, they seem to say, you can collect data, you can observe different frequencies of different kinds of letters in different languages, you can measure (optically if not mathematically) the impact of these frequencies on text. And you can do all this without knowing how to speak any of the languages.

[Of course, this ignores all the purely cultural aspects of typography: the things that make French books look French even when they contain English texts.]

I find image no. 5 fascinating. I've never thought of how similar letters are used differently in different languages.

What is the difference between the right and left columns of image no. 5?

English use many round letter forms when German less, Joe. I always found English/Us typefaces too round: too much emphasis on round forms. very large bdpq, o compared to n,m etc.

Nathalie, the classification is built in fonction of their various main forms. To try to discern some rules in regard of the language used. ok?

Its not very scientific, its just a way to show something for which I have some feelings that I can't explain. Also, because of all the old French guys say for long about "Latinit

Ahhh...thank you Joe for your observation on the columns. It is obvious now isn't it?

> German type need to be more straight, more regular to control all the letters togethers.

Interesting!

> Why you want to emphasis ascenders?

To improve readability?

But after what Joe pointed out, I might instead ask: does French need a larger x-height? But this contradicts Mandel's statement that culturally authentic French type must have a small x-height. What do you think of that?

Also, what about making room for accents?

--

> you can collect data, you can ....

Exactly! Wonderful stuff. Although it might be too deterministic to be "true science", it's a very powerful sort of tool, not least because its visible.

hhp

Jean-Fran

John, You right, but in context of non-latin versus latin if you see what i mean.

Your analyse of how we can use these elements of comparison is pretty sensible.

But, I think theses differences are built by designers without noticing what they do.

I imagine Zafp designing a new family. He dream as all of us dream letterforms in our language, we saw words in our brain composed in an imaginary new and perfect letterforms (another story when we try to put on screen the perfect imaginary forms... hum)

Different topic:
You've said (most recently in Print magazine, was it?) that Dutch type design is derived from the French. Do you possibly see any "linguistic evidence" of that?

hhp

Can you explain Image no. 1?

Xheight is difficult question Hrant. I don't thinks its a key thing. no?

In French, perhaps interesting, but more important the various widths, what i try to show with the only "french font" I have at the time. Perhaps its more obvious with Sabon Next? The image 5 show that French require strong difference between round letters (o...) and straight letters (n...). because they are combined all the time in left side.

On this aspect, English is little similar. When German require nothing in particular as the natural difference between small letters and ascenders do the job (they naturally enough different?)

The result is a combination of many factors, no?

(where are all the others people on the chat?????)

Hrant: on Print? When? I never saw this article?

Image 1: Its just some typical words of each language, a sort of synthesis of the observation made on others images.
just compare palatino and LM livre in german. Palatino rhythm seems more better for german, the emphasis seems natural. Not LM Livre, e and c are too narrow and don't bring enough light.

I think theses differences are built by designers without noticing what they do

Right, but recognising this is the first step to noticing and controlling what you do. I'm sure, if one only ever looks at one language in print, that is all you will ever imagine. But I think it is enough to look, not necessarily to be able to read, at a diversity of languages in order to begin to appreciate the differences and to respond to them.

An interesting thought: if a client came to me and said 'I want you to design a typeface, and it will only ever be used to print text in German', what would I do differently? I think I would begin by going and looking closely at a lot of German text, so the approach to designing for individual languages is quite similar to designing for different scripts.

Of course, most of the time, we're not designing for individual languages, but in the reasonable belief that a typeface is going to be used for multiple languages. This is even more the case these days, with Unicode and OpenType encouraging extension of Latin script typefaces for central Europe, Turkey, Africa, Vietnam, etc.

So here's a question: do type designers try to make 'neutral' multilingual fonts, try to compromise to accomodate the features of different languages, or do type designers each have a kind of personal platonic language in their heads for which they instinctively design?

Oddly, if I have such a language in my head, for which I instinctively design, I'm not sure that it is English. Maybe Latin.

I think Hrant is referring to the interview with Petra Cerne-Oven, in which JFP describe Dutch typefaces as French types with all the elegance taken out of them. It's a very funny interview.

> (where are all the others people on the chat?????)

I'm sure they're listening, you just can't "see" them!

--

I'll have to find exactly where I read that (it was about 2 months ago), but I do remember it vividly.

John, so where was that article?

--

John, in fact I think "detachment" can bring a certain objectivity.

hhp

John: I pretty sure you make your own analyse of the focused language.

Take Erik Spiekermann types. There all are quite narrow, square with very large s. large top e counters, large a counter, narrow n, etc. I'm pretty sure he don't see all of "his mistakes" (the features I see as mistakes because I'm french! not because I better than him on typeface design!!!!). Remember the number of time when you say that my a are too narrow and French... hum (quiet perfect to me!)

I refer to something beyond your control. its not scientific, its just because we are human with difference.

Little in same way as a Sumner Stone typeface is very round and soft, quiet (as him, soft, quiet guy!).


Univers has been designed to fit German Italian French English perfectly. The result is the page is quite boring! Just look at his books with three columns three languages in same Univers.

What can be fun is to apply languages features for OpenType with various letterforms depending of the language used!!! WGL4 with this feature (who make the double work -- without considering the size feature) can be very fun to do ;-)

The interview was originally printed in Hyphen, the Greek type journal from the University of Macedonia in Thessaloniki. Recently it was republished (slightly edited, I believe, but I have not compared them) in one of the main US design mags. It might have been Print, or perhaps Communication Arts or HOW.

(where are all the others people on the chat?????)

Bonsoir Jean-Fran

I have to go back to work now, but since this is no longer live chat, I presume this discussion can continue over the next few days.

> Univers has been designed to fit German Italian French English perfectly.

Is that even possible?

hhp

Jean Fran

JFP: can you explain your "Image #5" diagram? Are you illustrating the frequency of various characters, depending on the language?

Re: Image #5. Now that I see it, it's obvious.

The left column shows the freqency of characters without extenders (ascenders) and the right column shows the frequency of ascenders for each language shown. (The first row is English, the second row is French, and third, row is German.)

French and German seem to differ the most. English and German seem very similar.

You talk about the tendency toward 'flatness' of French versus 'livliness' of German.

Can you point out some faces that perform well in these three languages?

JFP, I'm sure there all here listening intently.

Can you point out some typefaces that perform well in the three languages you've analyzed?

so, we can't try to start again here! John Hudson posted a question on the chat: Do we need to speak the language to design a good font or something like that?

I think so in some way. As we dream our letterforms in word of our own language, I'm pretty sure we don't see letters in same way, for their proportion and so on.

This remind me a personal experience. When I have designed Bienvenue, I never realised that the a is pretty narrow, because the combination of it started with teh word France Telecom, and naturally the fitting was good until, couple months later, I realised that it was too narrow. I dreamed too much of the a in a context.