Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

John Savard


John Savard
Last Active
  • Re: Will this work for Cyrillic?

    3. But you can’t simply take some of them out…
    Why not? If they are superfluous?
    Here's a point that can be answered.

    It's not like a typeface is a collection of independent attributes, any one of which can be removed like a pair of socks from a drawer.

    Or maybe the definition of "superfluous" needs to be examined more closely.

    If a sans-serif typeface like Univers can work successfully for text, then serifs are not absolutely needed for a text typeface - and neither is stroke contrast.

    But Times Roman is a typeface which was designed as a whole. One couldn't just take Times Roman and remove the serifs, but change nothing else, and expect the result to be attractive in appearance, readable, or legible.

    Univers still has a lot of structure and detail. It differs a lot from Futura, for example, which is a geometrical sans-serif. So in that sense, it isn't "minimalist", yet it does not have certain things that Times Roman has.

    So maybe one could say that serifs and stroke contrast... are superfluous for a readable body text typeface, but not superfluous for the specific typeface that is Times Roman, as one way to make the distinction.

    You can't simply take these things out, but you can leave them out in a new design based on doing without them.

    TNR (or other typical text faces) are systems of parts designed to work together.

    To say a typeface with a lot of parts means those parts are superfluous is like saying only minimalist fixie bikes are ideal—bikes with shifters, 2 brakes, lights, fenders and reflectors are worse because they have too many parts. The fixie might look clever but the other bike is a more comfortable ride.

    It's interesting that I started with the same premise as you did, but came to what appears to be the opposite conclusion.

    But just because I think Univers is workable as a text typeface does not mean that I prefer it to Times Roman, so I don't disagree with your conclusion. Here, there is a different issue; the difference between "not absolutely needed" and "not beneficial".
  • Re: Will this work for Greek?

    Not being a native speaker of Greek, I hesitate to comment in this area, but it seems to me that the change in direction of the left descender of the lower-case lambda as it joins up with the main stem is visually distracting.

    I've done a quick check of a few Greek typefaces, and in particular I looked at Greek lettering for drafting as well, and I didn't see any where the lambda had this characteristic.

    However, in your body copy samples, I can see that this characteristic does perform one useful function - it causes the lowercase lambda to relate well to the x-height, the way all the lower-case letters with ascenders in the Latin alphabet do.

    But Greeks may not need to have the lambda do that.
  • Re: Eth and Thorn - historical forms

    Quite often historicism is merely an easy way to feel confident, which is of arbitrary relevance to users.
    Although I disagree with your views on this issue in general - and they don't seem to have gotten much support from other posters here either, I think you have made a valid point in this sentence.

    Why should I expect that a modern user of a typeface is going to care about how faithful it is to a 15th Century model, of which that user is unlikely to be aware?

    Maybe this will help the type designer avoid being laughed at by other type designers who do research into such obscure matters, but that serves a selfish interest on the part of the type designer, not the users of the typeface - or, at least, that certainly could be argued.

    I can think of one coldly practical reason, though, that a type designer might want to see what eth and wynn and thorn looked like in the 13th Century or whenever. If all the designer knows is what they look like in Times Roman and Palatino... then he could conceivably be laying himself open to a copyright infringement lawsuit! (Yes, I'll admit that prospect is remote, since filing off the serial numbers isn't that hard...)

    Also, to neglect historical background would naturally be unappealing to type designers who fancy themselves to be taking pride in their work.

    But I will return to what I see as valid in your point; the word "nebulous", appearing in the next sentence, indicates what is going on.

    A historical model, one can assume, is something that has undergone a long period of organic development. Thus, its rough edges should have been smoothed off, and likely the letter forms will have reached a peak of optimum harmony and legibility.

    Ahem. From one of the posts above: "be careful not to confuse thorn and wynn". The Hebrew square script in its usual typeface representations has several letters distinguished only by tiny serif-sized details.

    Assumptions... sometimes turn out to be mistaken.

    I think that I do assign awareness of historical models a considerably higher value than you at least are appearing to here, but I fully agree that they still need to be viewed with a critical and skeptical eye despite their potential value. Sometimes, the answer you've looked up in the back of the book isn't the right one after all.
  • Re: Eth and Thorn - historical forms

    Why not unfetter them from historicism in favor of the needs of living users?
    Nothing wrong with that.

    But if you're designing a typeface, the first thing you do need to know is what the letters of the alphabet look like. What makes an A, a B, a C, and so on. That's inferred from looking at other typefaces, at hand-lettering, and so on.

    If you're going to include eth, thorn, wynn, and so on, in a typeface, then, one has to know what they should look like. And if one's only reference is, say, modern typefaces for printing Icelandic, or academic printing of Anglo-Saxon... well, there's always the possibility that these modern sources have gotten it wrong.

    Although the fact that some of the characters involved are still in use in Iceland at least means there's one living tradition to draw upon.

    After one has found out what those characters looked like when they were in active use as part of a written language, then one has a starting point to work from. With that starting point, one can still choose not to slavishly follow historical models.

    But at least one is not working in ignorance, and one isn't bound to repeat errors that may have crept in to the common forms of those characters as they've been conventionalized by type designers without any feel for the scripts and languages that used them.

    Yes, therefore, to making eth, thorn, wynn, and so on work and fit with whatever typeface they're added to, to changing them to serve contemporary readers. But a resounding no to ignoring history, and therefore to being ignorant of the earlier forms and the development of these characters.

    They should not be fettered to historical forms, but the historical forms should be a base on which they can stand.
  • Re: Will this work for Cyrillic?

    The typeface DIN was precisely the sort of thing of which Neutor reminded me. So I am glad that I gauged your intentions in this regard correctly.

    I tend, however, to agree with the conventional wisdom that serifs, stroke width variation, and so on, are highly relevant, and indeed almost essential, to the function of readability in body text.

    And yet, sans-serif typefaces can work in that role. Perhaps it would be worthwhile to study why, for example, Univers works much better than Helvetica in body copy - even though in display sizes, Helvetica is usually considered to be the more beautiful of the two.

    Then there are the "Egyptians" - monoline typefaces with serifs. Some of them are passable or better in body copy, though seldom used.

    One can draw a typeface that looks like DIN, and hope that people will use it for extended text. But one is likely to be disappointed. Perhaps what is more worth doing, if one's goal is to achieve minimalism, is trying to see what it is about typical text typefaces like Times Roman that is superfluous and can be discarded - and what is not.

    After all, the true goal of minimalism should be to discard everything that is not necessary... but to retain everything that is necessary. Otherwise, if the superfluous is not sorted out from the necessary... the new stripped-down typeface, missing some of the necessary items, will not be used, so no progress towards minimalism is made.