Somebody Agrees with Hrant about Black-Letter

...about something I had imagined that hardly anyone would agree with him about. (His views on the importance of cultural authenticity, I believe, are widely shared these days.)

And it isn't just anyone who is in agreement with him. It's Stanley Morison, no less.

Here's the quote, from his First Principles of Typography:

"It does no harm to print a Christmas card in black letter, but who nowadays would read a book in that type? I may believe, as I do, that black letter is in design more homogenous, more lively and more economic a type than the grey round roman we use, but I do not now expect people to read a book in it. Aldus' and Caslon's are both relatively feeble types, but they represent the forms accepted by the community; and the printer, as a servant of the community, must use them, or one of their variants. No printer should say, 'I am an artist, therefore I am not to be dictated to. I will create my own letter forms', for, in this humble job, no printer is an artist in this sense."

While he is not at all sanguine about the prospects of shifting the general preference, and in this respect he may differ from Hrant, he is in agreement that black letter is "better", in whatever sense one wishes to take that term.

I had thought it to be obvious that black letter was objectively less legible (and, indeed, far less legible), and one could demonstrate that by, for example, subjecting specimens of black letter and Roman type each to a two-dimensional Fourier transform, and observing the much greater high-frequency content in the former.
«13

Comments

  • I've heard the argument before that blackletter is supposed to be more legible because it has less self-similarity than antiqua. I don't buy it. It just has more visual noise, and some letters are actually more prone to confusion in blackletter than in antiqua, such as /c/ and /r/, or /t/ and /k/ in Fraktur. Also, many blackletters go to great lengths to achieve homogeneous texture, which works against recognizability of individual letters (probably less of a problem for book blackletters, though).
  • John SavardJohn Savard Posts: 1,088
    edited March 2019
    I should admit, though, that while I'm no fan of blackletter, it does astonish me that the private press art printers seem to all agree that one of the earliest Romans, that of Jenson, had achieved perfection - and it is also astonishing that the more conventional world of printing moves just slightly forwards in time, to the Roman of Aldus Manutilus - from which nearly all subsequent Romans are but slight variations.

    Both of those typefaces are very good and beautiful, with this I am in full agreement.

    But for a beautiful and readable letterform, I think we can do more than just experiment in the direction of, say, Helvetica and Optima. I think that history leaves us with a great number of roads not taken that are worth exploring.

    Thus, I think the various typefaces of the rotunda and related styles are worthy of study and of further development; from them, perhaps something could be derived, similar to the Roman in its merit, but different and independent.

    Of course, we already have trivial variations available that are underused. Clarendons, Egyptians (monoline slab-serifs) and Latins (wedge serifs) are perfectly capable of being made in forms suitable for extended reading as body copy, and indeed (at least, to my knowledge, in the first two cases) they are already being so made, but to a terribly limited extent.
  • Thomas PhinneyThomas Phinney Posts: 2,730
    Latin serif faces have been made for body text. Frutiger’s Meridien for one. I set a friend’s master’s thesis in it. (The thesis was on avant-garde film, so it seemed apropos.)
  • I vaguely remember Morison thinking highly of blackletter's functionality. I think I might have blocked it from my memory because I dislike most of his typographic views...  :-)

    Such a discussion is only useful if we can separate mere instances of blackletter (which can very much suck) from what makes blackletter blackletter. Which is not easy to do.

    My own epiphany concerning blackletter came when I bought a discard from the UCLA Research Library, about 20 years ago. It was an early 20th century German novel. Hundreds of pages. Set with unequivocal good taste and high craft. And it struck me that it must have been easy to read, contra our official-party-line "enlightened" prejudice. When I looked closer, I realized why, and it just makes sense.

    Blackletter, in its spectrum of convention, incorporates two important advantages over whiteletter: because it can make curves into verticals it's horizontally more economical, something the Latin writing system needs help with, being out of harmony with the "geography" of our visual acuity; and it tolerates, even encourages, much more divergence, especially in its extenders, which are crucial for bouma decipherment. Yeah: http://typedrawers.com/discussion/2285/brain-sees-words-as-pictures

    For example if you make your whiteletter "h" descend, you've pretty much killed it as a text font for most customers. In blackletter, it's supposed to have a descender.

    Concerning the overly-homogenous texture contention, many instances of blackletter (in fact, entire styles of blackletter, such a textura) very much suffer from that. But all I have to point out is the fraktur "o", with its resplendent internal divergence, even post-modern hybridity. You can't do that in a mainstream whiteletter.

    Now, all this does not mean we should use a traditional blackletter for much more than middle-school Participation certificates. For one thing, the caps are illegible! But we can, and should, try to assimilate the above advantages into our whiteletter spectrum of convention, broadening it, giving it more freedom, not least the freedom to be more readable.
  • John SavardJohn Savard Posts: 1,088
    Thank you for this clarification. Now I have a clue about where to look to sort the good from the bad in blackletter.
  • BTW, I highly recommend this affordable and accessible compendium on blackletter, which includes some great articles as well as eye-opening examples of hybrid approaches (my favorite being Fraktoer by Hans Heitmann, 1996).
    https://books.google.com/books/about/Blackletter.html?id=E4c3sJhNVqkC&source=kp_book_description
  • My own epiphany concerning blackletter came when I bought a discard from the UCLA Research Library, about 20 years ago. It was an early 20th century German novel. Hundreds of pages. Set with unequivocal good taste and high craft. And it struck me that it must have been easy to read, contra our official-party-line "enlightened" prejudice. When I looked closer, I realized why, and it just makes sense.
    Do you have a photo of that?
    As for your Fraktur /o/: It does use less space, but isn't Garamond's wide stride and decadent extender length what makes it so comfortable to read?
    And are you sure you're actually widening the shape inventory of the typeface by turning /o/ into this convoluted Frankenshape? You've removed the circle — one of the simplest and most recognizable shapes imaginable! — and replaced it with something containing a vertical stem, making it blend in with the picket fence of the other letters.
    Fraktur shapes might be extra-special, but they are extra-special in an almost obsessive-compulsive way, rendering them confusable again.
  • The core of my readability belief is exactly that we can benefit from more complexity than we're afforded (especially in this age of Modernism) specifically much more complexity than whiteletter likes. A great piece of evidence is that the Chinese are not exactly illiterate...

    Furthermore, Latin is too wide for optimal reading, and the blackletter approach can help mitigate that. When it comes to extenders, there's no reason blackletter can't have them long.

    I'll try to dig up the book from the gara... I mean, archives :-) but in the meantime here's a cheesy old comparison using a scan from it:

    (The background glyphs are from my Brutaal design.)
    Do note that they're the same width there, but the blackletter has greater apparent size (plus has –more readable– looser spacing) so could be scaled narrower/smaller.

    BTW at ATypI 2000, I gave a talk about all this. But back then nothing was recorded...
  • notdefnotdef Posts: 168
    Latin is too wide for optimal reading
    That’s a big claim. Where is the evidence supporting it?
  • No evidence (that I know of, yet) "just" logic: the fovea is circular, and in the parafovea we quickly run out of lateral acuity. The way Latin sets, this ends up meaning we have to perform extra saccades, instead of taking advantage of acuity in the vertical dimension.

    To be fair, most writing systems have this problem; Chinese (especially when set horizontally) and Hangul are notable exceptions. BTW vertical scripts (such as Mongolian) have it much worse because they go exactly against our acuity geometry (which has resulted from biological evolution).

    BTW this is why the idea behind Hangulatin is so fascinating:

    And from my original Alphabet Reform essay:

  • Ugh, that gives me a headache. It forces my eyes to jump around in two dimensions rather than one, and requires additional CPU time to figure out the order in which to parse each block. I'm sure it would become easier with time as you memorized more and more blocks, but I doubt it would pay off.

    It's particularly uncomfortable for English, with all its dead letters and reduced syllables.
  • Ergo: "such a modified script would be .... too much to expect in terms of reader effort."
    Although Anita did such an amazing job I've since become less certain of that...
  • John SavardJohn Savard Posts: 1,088
    The Koreans seem to be quite satisfied with their script, not least because it is quite phonetic. If the entirety of a syllable block can be included in the foveal region, it might not be needed to move the eyes in two directions.

    Of course, in Korea the rationale was to mimic the appearance of the Chinese script, because of the high degree of respect the Chinese civilization enjoyed. Thus, it's unclear that there would be an actual rational reason to modify the script for English in this way. One obvious loss would be bouma, as every syllable would have the same shape, a square block.

    I approve of Hangulatin as a fascinating experiment. (Apparently, the font includes an English dictionary, so as to avoid forcing the user to space after every syllable, which is sort of a cheat...)

    But while I think that avenues for further improvement in the Latin script should continue to be explored, it seems like its users are the most fortunate among users of existing writing systems.

    There is more to memorize to write Chinese or Japanese.

    Amharic, Korean, Devanagari... are all more complicated.

    Hebrew and Georgian just seem visually less readable.

    Latin, Greek, Cyrillic, and Armenian, however, are about equal.

    Appearances, though, can be misleading. In a way, I'd like it if there were a more mysterious and profound way to write... and yet, mechanical simplicity of typewriters, universal literacy even when there is still significant poverty, these things have a great value.
  • Of course, in Korea the rationale was to mimic the appearance of the Chinese script, because of the high degree of respect the Chinese civilization enjoyed. Thus, it's unclear that there would be an actual rational reason to modify the script for English in this way. One obvious loss would be bouma, as every syllable would have the same shape, a square block.
    I believe the –superficial– mimicry of Chinese was a ruse to appease the ruling class; in fact the original design of Hangul was surprisingly geometric; also note the –eventual– prevalence of the circle shape, which is notably absent in Chinese. But that has no bearing on Latin's evolution anyway. Your point about boumas does require careful consideration however.

    Latin is easier to learn, but then limits you, both in terms of reading speed and cultural lyricism. Alphabets in general are better for business than culture...

    --

    BTW apparently @Dan Reynolds gave a talk about blackletter hybrids at last year's ATypI! Worth a watch for sure, and hopefully will serve to inspire future efforts.
    https://www.youtube.com/watch?v=CV6qNHdtiWo&list=PL0oMAzSh5W9q64iHo9pMa0WdgUdz411hD&index=52&t=0s
  • BTW Reynolds seems to have put forth "Midolline" as a term to describe such hybrids. I was thinking... "grayletter".  :-)

    I started a poll:

  • Thomas PhinneyThomas Phinney Posts: 2,730
    edited March 2019
    Hrant,

    Chinese readers use MUCH shorter saccades (as do Japanese readers when reading kanji as opposed to katakana). It doesn’t seem to be an issue of visual arc and acuity, it appears to be information density and info-per-saccade.

    Don’t get me wrong: I am sure we could have a typeface or writing system that exceeds a critical threshold, such that people could no longer get maximum info per saccade. But I am pretty sure whiteletter/antiqua is not there, in typical text usage.

    I believe experiments with text size tend to bear this out. Reading speed only starts to drop when text gets a LOT bigger!
  • John SavardJohn Savard Posts: 1,088
    Latin serif faces have been made for body text. Frutiger’s Meridien for one.
    This sent me on a search for information about Frutiger. A few days before, while looking for information on Excoffon, I noticed that Ondine was designed by another famous typeface designer, but then forgot who.

    In this search, I encountered some information about his work with IBM on adapting Univers to the IBM Selectric Composer - and in that information was a proof from which I was able to derive the unit widths for the various characters provided when using the Composer for the Russian language. This happened to have been very helpful for a part of my web site that I was working on.
  • My point is the density gain is in the vertical dimension.
    And shorter saccades might be an indication of non-immersion.

    I think in proper readability testing (where boumas come out of hiding) speed drops shortly above typical text sizes.
  • Christian ThalmannChristian Thalmann Posts: 1,938
    edited March 2019
    Not sure whether TNR should represent antiqua in such a comparison. Garamond offers a greater variation of letter shapes and proportions, for example:

    And note how the fraktur absolutely distntegrates into a block of noise at smaller sizes:

  • Vasil StanevVasil Stanev Posts: 759
    edited March 2019
    This whole meandering and year-long debate about legibility, and how objectively something is this or that, is unsolvable and prone to heated flamewars until SCIENCE can understand the workings of the human brain and how it relates to the mind much, much more fully. As far as I know, the same part of the brain that humans use for speech is the one other great apes use for gestures, so one can argue blackletter is more legible or not based on human gestures. And there are also other parts of the brain for reading and writing, and gestures are a book unto itself, some universal, other culture-specific. And the ductus of the hand is also a gesture. And both letters and gestures mutate over time. And some writing monks used sign language because of abbey rules. And medieval Latin built MUCH on Roman one. See how complex the matter actually is? We go ever deeper in the sea and the deeper it becomes. Stay dry on solid ground, perhaps.
  • The best term to use for hybrids, in my opinion, is … well, hybrids.
    Of course you'd have to indicate what it's a hybrid of, which makes it so wordy as to discourage use.

    A new word is only naff if it's unnecessary. For example taking a selfie might be naff, but without a word for it how could we concisely express that opinion? :-)  All of us use words that were invented after we were born.
    In the end, every hybrid typeface published in Germany in the 19th and 20th century remained an experiment.
    ....
    the text that people read most in the 21st and 22nd centuries is not going to look like a German hybrid from the 19th or the 20th.
    Of course they will look different, thankfully.
    But as anybody (especially a revivalist) should admit, inspiration can morph something dead (actually, dormant) into something useful for the living. This is why I'm so glad you gave that talk, not because of historicism for its own sake.
  • Not sure whether TNR should represent antiqua in such a comparison. Garamond offers a greater variation of letter shapes and proportions
    Comparing fonts for readability is tricky business... For one thing, apparent size cannot be ignored, and a Garamond would need to be much larger (hence wider, a problem) to match the letterwise legibility of Times. Anyway the casual analysis in my image intends only to demonstrate divergence, and there Garamond is the same as Times.

    I chose Times because it's ubiquitous (hence an easy reference point) if also because it has a reputation for high readability (which it frankly mostly deserves). Burgess did a good job... ;-Þ
    Christian Thalmann said:
    And note how the fraktur absolutely distntegrates into a block of noise at smaller sizes:

    Our screens must be doing a different job, because I'm seeing the opposite (at least at the bouma level)... Anyway they all fare too poorly, because they're not designed for that size.
  • Stay dry on solid ground, perhaps.
    "A ship is safe in a harbor. But that's not what a ship is for."
    — John A. Shedd

    To put it more clearly: lack of convincing empiricism does not relieve us from having to take action with whatever justification is available.
  • John SavardJohn Savard Posts: 1,088
    lack of convincing empiricism does not relieve us from having to take action with whatever justification is available.
    That's true.

    But what does "convincing empiricism" do for you? It compels universal agreement.

    I'm happy to accept that Fraktur and other blackletters have aspects worth borrowing, because personally I think that the rotunda typefaces deserve further study - that there may be other ways for roman to have developed in addition to the Jenson-Aldus path that would have been equally valid.

    So one could produce something that is legible by conventional measures, but which also includes, from calligraphic roots, extra individuality in the letters, thus enhancing readability on another level.

    That any innovation, to be accepted, needs to be hardly noticeable as innovation - I think it's also Morison who said that - though, seems to me to be almost inarguable. But Optima, certainly noticeable as an innovation, was immediately accepted, even if it didn't "take the world by storm" in the sense of causing a large fraction of new designs to belong to its category.

    I had no trouble reading Hangulatin, but then my reading skills have been developed to a high pitch through years of alphabetic reading. What about children in school learning to read? Not everyone reads easily and well: that is a major problem which needs to be urgently fixed.

    Can we dumb down our writing system much further than by using an alphabetic system? For English, of course, there's spelling reform. But I can't think of anything simpler.

    The Houston Chronicle's experiment of using a Jenson adapted to newspaper printing was interesting.

    Corona, incidentally, since its very large x-height still leaves some room for ascenders and descenders, is evidence that bouma has value.
  • Many interesting points.
    John Savard said:
    What about children in school learning to read? Not everyone reads easily and well: that is a major problem which needs to be urgently fixed.
    That's Hangul's most powerful feature: you can slowly compile a syllable from its constituent alphabetic letters, but once proficient you can read it as a cluster.
  • notdefnotdef Posts: 168
    edited March 2019
    If you want to make a narrower Latin alphabet, you can always remove the serifs …. However, that style was never widely adopted for long-form reading. Counter to @Hrant H. Papazian, I suspect the main readability benefit of the serif (in wide measure book typography) is precisely its width: serifs allow elements to move farther apart, without falling apart visually. A “normal” width serif is bigger – horizontally – than a “normal” width sans serif of the same size.
  • I think serifs help letters be less themselves and more parts of boumas; and serifs or no, generally the more you can fit in the parafovea the faster you can read.

    The thing is, if boumas don't exist and we only read via foveal letterwise compilation, letterspacing is moot. And you still go faster by fitting longer words in the fovea!
  • A friend of mine claims to be able to speed-read (Latin) by scanning forward between two lines and then backwards between the next two lines, boustrophedon-style, allegedly with funcional reading comprehension. Make of that what youy will...
  • notdefnotdef Posts: 168
    edited March 2019
    Adding to my hypothesis: Serif types with contrast (thick verticals, thin horizontals) also allow for a more regular stem interval.
Sign In or Register to comment.