Optical correction in Arabic monoline

I examined several Arabic monolines and it seems that most if not all of them reduce the thickness (width) of vertical stem for some reason. An example of this is Tahoma Arabic. What is the reason for this? In unmodulated Latin, the horizontal stem is thinned a bit to achieve perceived consistency in stroke width throughout the glyph. I see the opposite done in Arabic

I understand that Arabic is baseline heavy, but why do we need to further thin out the vertical stems? Is it to make it more Naskh-like in appearance?

Here are 3 different stroke variations on a word from my typeface, do any stand out as being more "correct" than others?


Tagged:
«1

Comments

  • I would guess they're deviating from monoline to keep Arabic's vibe. After all unlike cultural authenticity, monoline is merely a means, not an end.
  • I would guess they're deviating from monoline to keep Arabic's vibe. After all unlike cultural authenticity, monoline is merely a means, not an end.
    This would seem like a logical conclusion. Naskh is thinner on the verticals after all.
  • As a native Persian speaker, I would say definitely the third one is correct! I agree that this might be just to maintain the writing system's perceived correctness! However, I'm really curious to see what Latin speakers think about these three options!
  • Christian ThalmannChristian Thalmann Posts: 798
    edited March 5
    I'm a non-user of the Arabic script but tried to include it in my latest geometric sans typeface. I stuck to the same horizontal/vertical weight distribution as in the Latin, since that's what looked right to me... does it still work? Or does this mainly concern chirographic Arabic typefaces?

  • Bahman EslamiBahman Eslami Posts: 37
    edited March 6

    If you want to bring contrast to your typeface where do you get the contrast in Latin? It's from the tool that writes the script. It could be a broad nib pen or expansion pen. The angle of the pen causes the vertical strokes become thicker in Latin. This naturally happened because the most important part of the construction in Latin is in vertical strokes. If you make the the verticals thinner it would become much harder to recognize the shapes in text sizes. Actually Latin calligraphy starts with practicing lots of vertical strokes.

    In Arabic the angle of the pen causes the contrast to be more horizontally stressed and this happened naturally because the construction of arabic is majorly distributed horizontally or in simpler words Arabic script is mostly horizontal strokes. If you make the horizontal strokes thinner in Arabic then it is Latinized or reversed contrast.

  • John HudsonJohn Hudson Posts: 1,026
    Personally, at this weight, I would go for something between the 2nd and 3rd options, so maybe 76 on the vertical. But of the three presented, the third looks best.

    I think this kind of optical correction addresses a mixture of perceptual bias—a tendency of a person familiar with the norms stroke modulation in the script to perceive monoline as unbalanced—and actual optics. It's difficult to disentangle these.

    I know, as someone familiar with the modulation patterns of a lot of different scripts, that my perceptual biases are not linked to my native script (Latin); that is to say, how I perceive monoline Arabic is affected by my experience of Arabic, not by the modulation of Latin.
  • John SavardJohn Savard Posts: 19
    I don't know a word of Arabic, but to me, the third one looked as though the stroke width was uniform, so the thinner vertical lines may be a simple optical correction for reasons that are not dependent on the script or culture, except as they are a consequence of the visual appearance of the script.
  • Bahman EslamiBahman Eslami Posts: 37
    edited March 6
    @John Hudson
    Although it has something to do with perceptual bias it's also serving a function. In Nastaliq (نستعلیق) for example the farther the construction is from the baseline (major korsi or کرسی) the thinner the stroke becomes. This also could be justified that in Arabic script most of strokes are occurring around the baseline so if we see a thick stroke farther from the baseline it sticks out. The calligrapher tilts the pen with some pressure, this causes the reed pen to bend at its tip and make the stroke thinner at those situations. It looks more pleasing to the eyes but it's also probably a method to spread the black mostly in parts of the construction which is more repeated and less black in less repeated parts, making a more balanced negative space around the words.
  • Hrant H. PapazianHrant H. Papazian Posts: 473
    edited March 6
    No surprise coming from me, but still gotta say it: sorry Bahman, all that imaginary marking device dogma masks what type needs to do.

    @John Hudson
    > how I perceive monoline Arabic is affected by my experience of Arabic, not by the modulation of Latin.

    FWIW my views of the nature of nativity and conscious bias largely preclude such objectivity.
  • John HudsonJohn Hudson Posts: 1,026
    Bahman, the normative modulation of Arabic letters that you describe, arising from the movements of the typical writing tool, are what produces the perceptual bias: this is the modulation that we are used to seeing, and hence deviation from that modulation pattern looks wrong, even in the context of a low-modulation design. I have a similar bias from Latin modulation when I see a truly monoline uppercase A: the left stroke appears too heavy, even though I know it is actually the same weight as the right stroke.

    Yes, there are probably good aesthetic and functional reasons why the traditional modulation pattern was established in the writing of the script—or, to put it another way, why certain letterforms evolved from the moving of a particular writing implement—, but that seems to me orthogonal to perception of forms with low modulation in which, for instance, there is very little difference in the balance of negative space in the three examples shown.
  • @Hrant H. Papazian
    I'm just trying to find an explanation for what calligraphers were doing and why I find it pleasing to my eyes. It's merely a justification not a dogma. If you you have your explanation I would like to hear it too.
  • My point is that where contrast came from is tangential to where it needs to go.
  • I also add my vote to those who found the 3rd adjusted example most pleasing. The verticals being thinner than the horizontals has been a hallmark of Arabic calligraphy for more than a thousand years now, so it is embedded in the visual memory of its people. 

    Only in the early stages of the development of calligraphic traditions after the advent of Islam, we can find manuscripts where the stroke stress is on the vertical, rather than the horizontal in the Kufic styles. But within a couple hundred years, we see the stress shifted to the horizontal strokes. With the wide spread of the cursive styles, more contrast was considered more beautiful. And this continues to our own time.

    The Maghribi styles of Spain and North Africa, are the most uniform in stroke thickness and the contrast is limited to the terminals of strokes in bowls and arcs. Here you will see examples where verticals and horizontals have the same stress and others where the stress is on the horizontal strokes, but rarely where the stress is on the verticals.

    In my view, some horizontal stress should be used in Arabic typefaces whenever possible, including monoline designs as in this font. 
  • John HudsonJohn Hudson Posts: 1,026
    My point is that where contrast came from is tangential to where it needs to go.

    Speaking of dogma.

    Hrant, I think you need to consider the possibility that the scribes who consciously developed particular ways of writing were intelligent people who understood some things about both writing and reading. You sometimes seem to have a kind of deterministic view of tools, looking only at the patterns that arise from the shape and movement of the tool, and not considering the decision making of the person wielding the tool. When there is palaeographic evidence, as Mamoun notes, of changes in the way the tool was held and used, we should consider all the factors that might be involved in that change, including both aesthetic and functional factors. Finally, Arabic scribes regularly employ manipulations of the tool that are mostly absent from European writing, and that you'd probably find interesting to study, since they consciously circumvent the translation stroke model that you think problematic, altering the relationships of inner- and outer stroke edge and, in some situations, actually becoming drawing and filling of an outline.

    One might make the case that Arabic scribes were dealing with notan avant la lettre. :smiley:
  • Hrant H. PapazianHrant H. Papazian Posts: 473
    edited March 7
    The scribes were regular people, some of them intelligent, most of them right-handers who simply needed the ink to flow right, hence resorted to things like the broad nib. The æsthetic appreciation was mostly merely an inevitable byproduct of that physicality. Which is fine. But those tools are now gone from the lives of those we type designers must serve. Is there some metaphysical connection between conventional contrast and what reading needs to be? I'm not one to demand formal proof, but come on.

    Clinging to the byproduct of an obsolete tool is not necessarily harmful, but at the very least it's an arbitrary constraint on notan. You can want to respect the white all you want, but painting the black makes that hypocritical in practice. My favorite parallel here is Thomas Jefferson: he loved Sally Hemings, but she was still a slave.

    I like calligraphers! Yes, especially those who engage in deviation. But calligraphic text type is still like trying to serve soup with a fork. As I said in my ATypI Mexico City talk, this is less true towards the display end of the type design axis. But it remains a truth.
  • So are you saying perceptual bias is irrelevant? Or that it’s merely a nocuous side effect of ingrained standards and conventions? In which case we’re looking at a hen-and-egg situation.
  • Hrant H. PapazianHrant H. Papazian Posts: 473
    edited March 7
    I do believe perceptual bias is relevant, even though it is indeed partly* a side-effect of the physicality I mention. And it's not harmful, in fact it's crucial to good design. My objection here is to the way some people give primacy to the origin tool. This is a romantic historicism that misdirects and/or stunts progress because a tool –especially when it has stopped evolving– is much more of a ball & chain. So freely conclude for example that Arabic should have horizontals thicker than verticals even in a "monoline", but don't keep using the now-circumstantial origin to justify it.

    * But admittedly not even entirely:

  • John SavardJohn Savard Posts: 19
    Optima, Bifur, Data '70... there are plenty of typefaces out there that are not calligraphic in nature. Even a conventional typeface like Bodoni has letterforms that would be difficult to execute with a nibbed pen.

    Thus, I think that referencing a typeface to traditional tools is a choice these days, not something that is imposed on every type designer. That text typefaces for general use are often very similar to previously existing text typefaces, with only slight changes, is a natural consequence of popularity being determined by what people are used to.

    Thus, I don't see the situation as a problem that we can do anything about, except for encouraging type designers to engage in bold and original thinking, to undertake daring experiments from time to time.

    While we already have plenty of daring experiments - many of which appear to be horrible failures no one would want to use - it's precisely the most talented and successful designers that need to be encouraged to experiment occasionally, since the impulse to experiment is often strongest in those who have not yet attained mastery of the craft.

    And, of course, while Optima is made of strokes that could not have come from a quill pen, its proportions were taken straight from the Trajan column, and other aspects of the typeface were also conservative; originality was blended with tradition in that case.
  • John HudsonJohn Hudson Posts: 1,026
    Let me back up a moment, since I'm the one who introduced the term 'perceptual bias' into the conversation, to clarify what I meant by it. I wrote:

    I think this kind of optical correction addresses a mixture of perceptual bias—a tendency of a person familiar with the norms stroke modulation in the script to perceive monoline as unbalanced—and actual optics. It's difficult to disentangle these.

    In other words, what I am calling perceptual bias is a learned bias. It is entirely due to familiarity with stroke modulation that is typical of a given writing system, and independent of — while hard to disentangle on a case-by-case basis — from perceptual phenomena arising from optical and cognitive hardware. When I say 'difficult to disentangle', I mean that when we see e.g. symmetry as asymmetry, both learned perceptual bias and optical illusion may be in play.

    The origins of the typical stroke modulation patterns really seem to me a red herring, in terms of the topic of this thread. What creates perceptual bias — how we learn it — is familiarity over time with that pattern. In this respect, it doesn't matter at all how the pattern came about.
  • John HudsonJohn Hudson Posts: 1,026
    Very nicely put by E.H. Gombrich in Art and Illusion (1960):

    ‘The stimulus patterns on the retina are not alone in determining our picture of the visual world. Its messages are modified by what we know about the “real” shape of objects.’

    [Thanks, Frode for reminding me of this.]

    For ‘real’ in this instance, read common conventional or typical.
  • So freely conclude for example that Arabic should have horizontals thicker than verticals even in a "monoline", but don't keep using the now-circumstantial origin to justify it.
    By extension of your argument the vertical stress in text and common Latin typefaces is also circumstantial to the limitations of hand and the pen. I agree that in a monoline the direction of contrast does not contribute much to balancing the negative space and also not much helping the letters to be recognized easier, but lets say we have a higher contrast typeface which is also more conventional in text sizes and remove the serifs too. Why such design hasn't been considered for text. Is it mere sticking to a dogma or those serifs are serving a function and or the vertical stress is helping to recognize the shapes? A feature that works in higher contrast could affect the reader in lower contrast in a same way but to a lesser extent.

    The Latin calligraphers made a decision to rotate the tip of the pen more in a direction that makes the vertical thicker not only because of the limitations of hand and the tool. Why? Because Arabic calligraphers did the same with similar tools but the contrast is reversed compared to Latin. Calligraphers made their tools, the tools are not circumstantial completely. The tip of Arabic calligraphy reed pen is cut with an angel but in Latin it's not titled. This is a decision that was done to make the pen easier in hand when direction of contrast is in a certain way. I say restrictions could be construction of scripts that changed the contrast direction and some other things, not hand and tools.

    @Nina Stössinger
    I'm curious since you had designed the Nordvest (which I personally admire), have you considered a reversed contrast sans for text purposes in the process? Did you try to see if it works?
  • @Bahman Eslami Yes, Nordvest had a sans companion for a while when I was developing it at TypeMedia; that seemed less unique and interesting to pursue than the serif, but I think it can be made to work fine as long as it’s very subtle. I would say there are actually more “subtle” reverse-contrast sansserifs than serifs out there. I think a rev-con sans can handle less contrast; strengthening the horizontals did not have such an interesting effect as with a serif (where the stronger serifs make for a strongly coherent texture) – it gets spotty easily. Overall though I got the impression that the ingrained contrast patterns can be really fine to question. A lot of this is self-reinforcing convention.
  • > it's precisely the most talented and successful designers that need to be encouraged to experiment occasionally

    Something we're not doing nearly enough of.

    > what I am calling perceptual bias is a learned bias.

    I'm not sure it's useful to exclude the "hardware" dimension from that. But if we must: what to call those?

    > it doesn't matter at all how the pattern came about.

    To readers, indeed. To some creators however sadly it does.

    BTW "what we know about real shape of objects" can migrate between objects, making the conventional less valuable.

    > A lot of this is self-reinforcing convention.

    Yup.
  • John HudsonJohn Hudson Posts: 1,026
    edited March 8
    Nina,

    A lot of this is self-reinforcing convention.

    That's surely true of a lot of stroke modulation pattern, in a lot of scripts. But it's also the case that most of signs that make up a writing system evolved through the action of writing, and hence some signs really presume a particular ductus and modulation, and run into problems of internal space collapse, proportion, or illegibility if one tries to write them with a different modulation pattern.

    As you note, the lower the contrast, the easier it is to subvert typical stroke modulation. I think it is also easier in Latin than in many other scripts. [For several months a couple of years ago, I fell into the habit of writing with my pen much more steeply angled than usual: something close to a typical Hebrew angle. I can assure you that Latin written in this way looks much less weird than Hebrew written with a typical Latin pen angle.]
  • Nina StössingerNina Stössinger Posts: 146
    edited March 8
    John, that’s very interesting. (The investigation that turned into Nordvest started from this same point — my pen was too steep in a writing exercise and I liked the effect.) I have been wondering how much of a connection there might be between the structures themselves and the contrast patterns they were developed with; I was actually surprised to find Latin not “collapsing in on itself” more quickly when changing the contrast pattern, though the familiar rhythm of the vertical stems certainly gets weakened quickly. I would wonder if Arabic (to circle back to the original topic here) is less tolerant of this kind of thing? I’m totally seeing this with an outside eye, but it would not surprise me if with the connecting horizontal seeming so dominant and relevant also structurally, it might not be able to take much weakening versus the verticals until the visual hierarchy of the shape gets really weird.
  • D. Epar tedD. Epar ted Posts: 676
    "Here are 3 different stroke variations on a word from my typeface, do any stand out as being more "correct" than others?"

    Abdullah,
    What size, is this intended for use at, please?
    And what's the far right glyph in your specimen please?

    Thanks!
  • Hrant H. PapazianHrant H. Papazian Posts: 473
    edited March 8
    > some signs really presume a particular ductus and modulation, and run into problems of internal space collapse, proportion, or illegibility if one tries to write them with a different modulation pattern.

    Such collapses are not due to a violation of an imaginary tool, but a lack of imagination of how to go beyond it; it's key to nix the "writing" angle (pardon the pun) and instead think directly in notan. Which is natually much more difficult (especially with chirobaggage) but inherently more fruitful, at least towards the Text (versus Display) end of the reading spectrum.

    > I would wonder if Arabic (to circle back to the original topic here) is less tolerant of this kind of thing?

    Considering perceptual bias (as per John) arising from precedent, that makes total sense.
  • John HudsonJohn Hudson Posts: 1,026
    Such collapses are not due to a violation of an imaginary tool, but a lack of imagination of how to go beyond it

    I don't think in terms of imaginary tools — and as established many years ago on Typophile, I believe type design has always been, at most, para-chirographic —, but I am interested in how the normative shapes of signs evolved and what impact tools have had on those shapes. And if, like you, one rejects the notion of the signs having skeletons, then I think there is all the more reason to be aware of the shapes as being constituted of positive and negative arrangements, rather than as skeletal forms to which, conceivably, any and every modulation of thick and thin could be parametrically applied.

    This notion of a modulation pattern should be unpacked, perhaps. By pattern, I mean a systematic approach to where the thick and thin fall across a set of signs. Modulation patterns can be more or less strict, but generally speaking we expect some consistency of approach; indeed, this is one aspect of what makes a typeface design a coherent whole and not a frankenstein or ransom note collection of random parts. Nina's Nordvest brilliantly applies a pattern of atypical modulation to a Latin serif text face, and it works in large part because the pattern is applied systematically (which is not the same as saying it is applied deterministically. as if the shapes were made by an imaginary tool held at a steep angle; the necessary adjustments for individual shapes, for size, for weight, are part of the system).

    I'm very interested in the modulation patterns of South Indian scripts in their conventional typographic guise, because these signs originated in a monoline manuscript tradition (scratched in palm leaves with a needle). When these scripts were reduced to typography in the 19th Century, they were given modulation patterns, typically following an expansion stroke model, i.e. reflecting the dominant European writing tool of the day. Since the expansion stroke model is based on pressure rather than the shape of the nib, it is much more flexible in terms of deciding where thick and thin fall. So, I wonder, who decided? Where did the typical typographic modulation patterns of scripts like Kannada, Telugu, Tamil, or Sinhala come from? What were the criteria in deciding 'This bit should be thick, and this bit should be thin'?
  • > type design has always been, at most, para-chirographic

    Indeed, although that still spans a very broad spectrum of intent. What's notable though is the persistence of the skeleton whenever you "paint" the black (para or not) which I believe precludes ideal notan.

    I certainly agree about the necessary coherence of contrast (modulation) patterns. And I do believe in considering history, even trying to read the minds of long-dead practioners. But in large part to identify and overcome anachronisms, not merely to play in the sandbox of expectations.
  • Thanks for your responses everybody! I appreciate the insightful discussion this has triggered. I do agree that the third options is most agreeable and the verticals could use some thinning.

    "Here are 3 different stroke variations on a word from my typeface, do any stand out as being more "correct" than others?"

    Abdullah,
    What size, is this intended for use at, please?
    And what's the far right glyph in your specimen please?

    Thanks!

    Mainly for running text on screen at 16pt or less, preferably. The glyph in question is an Alef.

Sign In or Register to comment.