This thread is not just another Notan / Bouma / Readability ramble. Instead it's a place to let us post visual examples of excellent, ( or vile ) Notan to discuss. Samples could be from a font, be just a glyph pair, or come from non-digital sources such as stone carving, calligraphy etc. Who's first in line?
Typophile
The good ship Notanic
This copy was truncated by the archive that captured it, so the thread may be incomplete.
Chris, what observations would you like to make about Dead History? When I look at it I see an experiment with pushing boundries between type styles & merging them. Certainly it's more successful than you might guess it would be & that took some skill. But I don't know that it was an experiment that emphaized with Notan. Maybe I 'm wrong. What do you think? Why?
BTW here is a sample I got from Font Font of 'Lance' which is, I think, the font Hrant is referring to:
Lancé's Japanese/Latin face Shirokuro doesn't begin with R, but it's in Language Culture Type and the figure/ground play is striking. Not to mention Shirokuro means "black and white" in Japanese. The other faces in the book are Nichiyou Daiku, Pesaro, and Shuriken Boy, just in case one of those helps to jog your memory.
As a relative newcomer, I've recently been rummaging around the archives doing notan research. I'm especially interested in concrete examples, so this thread is of special interest to me as well.
David
Hrant: is shirokuro the one you mean?
http://kamedesign.com/e00jan_shirokuro.html
Perhaps I aught to examine it more closely but my initial impression is that it's very much about playing with duality of black & white within the context of the glyph's em box - not about creating blanced glyph realtionships across a word. So I don't get why it's useful in a text face context; however successful it is in it's own dualistic context. Hrant, if this is your example I am not sure what you want to get across with it. Maybe you can help me see what I am missing... It is an interesting extreme example of playing with b/w or figure/ground though. I would have placed an example image for my fellow typophiles but the site asks for NO reproduction so I respected that. Use the link. Hrant, if this isn't the one you mean you really are going to have to fetch the thing yourself!
RE: walk the walk.
Yes yes. I am hastening slowly. I want to post something I can get behind. Something really useful like your Wh example or Egypt. Those were great.
Chris: Okay okay. I guess I will have to remain puzzled for now.
So Chris, "decent type designers have been giving equal importance to both blacks and whites for centuries", but you come up with Dead History, a black-centric exercise if there ever was one? Fusing the blacks of two very different fonts to get a new font. Right.
To me you count as a "decent type designer", so you must have at least been keeping an eye out for exceptional instances of notan, right? Even if you might not have made anything notanically worthy yet (which is actually the case with me). Don't you go looking around now - you should have a few example sitting right there in your head...
> It’s your thread — walk the talk.
I think providing one example as a starter might have been a good idea, but I actually advise people starting a new thread to give it some room/time to breathe, so you get more of what's in people's heads instead of putting things in their heads. In fact I thought Eben's second post was way too anxious.
--
Yes, Shirokuro is it. Is it heavy or what? And bi-script on top of it.
BTW, I will note that it has two "r"s in it, just not one at the beginning. ;-)
Thanks David, and thanks Eben for digging up the link.
> it’s very much about playing with duality of black & white
> within the context of the glyph’s em box - not about creating
> blanced glyph realtionships across a word.
Check out the last word in my first post.
This goes back to my opinion that there are different types of notan, for different purposes, most centrally display versus text. BTW, display notan is one place I think chirography can be very useful; and that's why I once paid money for Ex Ponto (and I still don't regret it). Digressing dangerously.
> I am not sure what you want to get across with it.
It's an example of a designer consciously and formally aware of notan.
Unlike what some people would have us believe, that is very rare.
hhp
Hrant,
I will quote myself, perhaps you missed it the first time:
"I never said Dead History was a notanic study."
Look at the Dead History font with fresh eyes and look for something else operating there. There is a point to it.
I notice that you have not posted an example either? I don't mean your own work. I won't post mine because, in all honesty, I never understood the meaning of notan in the way you describe it--I understand more now after the recent thread prior to this one. I would only say that I attempt to have figure and ground in a harmonious relationship which locks them together. If this is called notan or not is not of much concern to me--I am just trying to make good type. Perhaps Froggy may be part of the discussion (not to say it is notanic) after someone takes a harder look at Dead History though. That is by no means to say that I think I have been successful in the execution, it just means I strive to be.
If you want examples of type that I feel works towards equal importance of figure and ground, I would say Frutiger. I don't really "look around" much. I prefer not to be influenced any more than I can help. I will also name one of the faces you hate, Bodoni--not because it is easy to read, it isn't. Just because of the positive and negative form. Legato is another one which works for me. Most of Dwiggins stuff as well. I doubt if he ever thought about it but it was just within him. Some of the blackletter faces and Constructivist faces have it as well. I would say that Zapft's work generally does not, it is more linear an gestural.
ChrisL
> I notice that you have not posted an example either?
I'm the one countering your claim that it's common, remember?
Anyway Legato and Shirokuro are the two best -and nicely "opposed"- examples that I know of - and I for one have been paying attention*. And I think it's very important to pay attention to what others are doing - that's a difference between Design and Art. Being influenced by good things is good for one's own results. It's good for the user.
* And if you haven't, how do you know that “decent type designers have been giving equal importance to both blacks and whites for centuries”?
> I won’t post mine because, in all honesty, I never understood
> the meaning of notan in the way you describe it
Hopefully these discussions aren't merely a means to acquire new terms, but to understand new things, to see them in new ways. When you say "I am just trying to make good type" I would reply that yes, of course you are, good for you, but to make better type you have to keep looking, listening and thinking. Maybe you are (sufficiently) notan-conscious, or even notan-savvy; but those who are not have a lesser chance of making good notan. True, awareness is not an absolute requirement (with Dwiggins a great example of that) but it can only help.
> I would say that Zapft’s work generally does not
I agree.
And one would have to suspect that chirography has something to do with that...
Anyway.
> Puzzled ... puzzles ...
What's with the Zen master routine? Man, you can be worse than me!
Is relevance completely moot, and admitting a mistake unthinkable?
hhp
Shirokuro is pretty artsy for you no? It is also not very readable and it's a display face--All surprising given your disdain for dislay faces. I thought the holy grail of notan was in increased readabilty? Legato is at least trying to go that way. I don't think it is any more readable than any decent sans. I like Legato and it reads fine but I see no quantum leap there. I would be curious if you know of any unbiased testing of readabilty between it and say Formata, Frutiger, Thesis, etc.?
"is relevance completely moot, and admitting a mistake unthinkable?"
I have no idea what that is all about. Care to explain? The only person I know of around here who thinks admitting mistakes is unthinkable is Bush--and perhapes some others? :-)
ChrisL
RE: What’s with the Zen master routine? Man, you can be worse than me!
Is relevance completely moot, and admitting a mistake unthinkable?
Even if I did live in a Zen temple in Japan for a while I am not making that claim! No, I am just saying that I don't get what Chris may be challenging me to see. And that I am willing to suspend judgement. I have to admit I feel dubious about the example. But I will keep thinking about it anyway. Chris, how about a hint?
I am almost ready to give my own visual examples to chew on. If you want to get started early - my example is not a Font, instead, inspired by Breim I am going to look at a pair of glyph combos across a series of typefaces. They are 'La' & 'ra'. Now to find a nice pre-digital Bembo or Caslon where the flag on the r is not quite so tamed...
Chris, I am beginning to suspect that you associate Notan with a kind of tricky suspension between views. Eg - it's a drop/it's a Q. A related kind of suspension seems to be present in Shirokuro and it Dead history too. Am I getting warmer?
Here is my sample. It's derived from online renderings, scans & assorted other sources. It is imperfect. Let's try to use it anyway.
1. is a bembo revival about which I make no claim except that it has a Bemboish r. Note the 'flag' shaped flag or arm.
2. is an 'ra' from metal 72 pt Caslon.
3. is Big Caslon
4. is Univers 65 Bold
5. is Scala
6. is Lucida console
7. is Patria
Let's start by looking at the 'ra' in rahja.
The problem it presents as I understand it is that the flag or the r and the top of the a may look as if they are going trhe clash unless you make arrangements to avoid it. When I look at the older typefaces aI see that the 'r's arm is bigger & it is the 'a' that accomodates it. When I look at Univers this sitiation is reversed to a some degree. The 'a' becomes stronger & the r's arm is shorter. After that even in serif faces such as Scala this trend continues ( obviosly not all typfaces but you get my drift I hope ) despite the fact that the shape Scala's a is quite bembo-ish. Then there is the Monospace. I threw this in because it illutrates an modern but opposite condition where the r must become more assertive to help fill the monospace. In fact it has to get a vertical serif to appear balanced! Thinking about Courier ( not shown) you have a less asserive 'r' made possible by long serifs. And finally there is Patria. In Patria I see a synthesis of a strong but not wide r with a 'shy' but un'deformed' a. Each of these is a pretty succesful example of managing a tricky balance of forms to create Notan that works. As you compare solutions you are looking at Notan in play.
I have more to say but I will leave it at that for now. Hrant thanks for rendering Patria for me.
Eben,
Think of a continium where the Water Quality logo is at the right end and Dead History is at the left end. Shirokuro and Legato are between the 2 extremes but I am not sure which is closer to Dead History and which is closer to Water Quality although my gut says Shirokuro is closer to Dead History.
This continium is not a better or worse thing. It is just more clearly one thing than another. Nothing in this string is either good or bad for this purpose.
ChrisL
Chris, are you contrasting glyphs that have many identifying cues ( maybe too many really ) with those that barely have enough to read as the glyph they semi-represent? That's the only sense I have been able to dig into so far.
BTW - the reason I built my example in the way I did to show that Notanic consideration isn't the province or flavor of 'special' fonts but relvant to all type design - especially text and workhorse faces.
"are you contrasting glyphs that have many identifying cues "
No, it is much simpler than that--the real clue is the salient point Hrant makes in his description of Legato.
Think about it til tomorrow--at which point I will spill the beans. It is nowhere near as complicated as you seem to be seeing it.
ChrisL
Eben you wanted some scans of original Caslon 'La' examples.
Here are some from Lady Willoughby's Diary. This is 1844, so it's 19th century ink, but I think the original type, as this was the first revival.
These are both large sizes. The first is I think the Double Pica, a title. The second is Great Primer.
I'll post some 'ra' combos tomorrow.
The issues here are more what I, following Doyald Young's usage in his 'Fonts and Logos' book, would call 'color', rather than 'notan'. Specifically, here the Double Pica size here has too long a leg on the L, and the spike a bit large. The Great Primer colors a lot better, I think.
Note that the a on the Double Pica is wider, but the d narrower.
Placed in this context Patria looks pretty darn good. The visual--or, as I like to say: optical-grammatical--integrity of the wordforms or boumas the letters make is really quite remarkable in comparison with some of the others, or are some of the others just badly kerned?
I think the Patria setting shares this with the Scala.
As Eben has hinted, it seems to have a lot to do with the black of the letters having been set up with an eye to the consistency of the white.
I'm reminded of Gerrit Noordzij's remarks about the 'consolidation of the word,' and the white being 'my only holdfast.'
Having said this about Patria and Scala, I'd also say that they exist in different 'notanic [or dark / light] registers'
Chris, I think you mean this: Hrant - 'disposing of the linking between the two edges of the black' I think you are saying that the outlines don't follow ductus... it sure as hell can't be while still appearing conventional to the reader now can it?
Sorry that's the best I can figure out. So spill those beans if ya got em!
Peter, I am not 100% sure that the kering values are correct for all of these. I don't own all of them all & some of them were rendered on the web - sub optimal for sure. Actually if folks have sample of these or other faces they want to compare - please do.
Re: Patria. I agree Patria does indeed have more interplay than we see in the other post Univers modern faces. Caslon has more than I realized & Bembo too. All that weird figgiting. Those odd details. But hey, they are DOING something. In some ways Patria is even weirder than Bembo & Caslon. In other ways it is tame next to Caslon. What I think is striking is the way that modern faces make the letters feel so isolated and unitary in comparison. I think this effect lingers in Scala and most of the faces being made. There are exceptions by degrees and they feel somehow a bit more magical and pleasing because of it.
Eben, greater visual (optical-grammatical) integrity is something different than 'interplay'. I was thinking of a quality of the whole. It's something about the object-like word image.
Also my "different ‘notanic [or dark / light] registers’" might be just another term for colour 'registers'. This part was not to signal a better / worse comparison.
Not to hijack the thread more a personal effort to comprehend notan, my current understanding is that these Tokyo pictograms represent good (or considered) notan or am I fooling myself?
Is the best place to look for good notan in the base character or in the character in its environment (as per William’s samples but maybe not cleaned up) or does one naturally follow the other?
Tim
{edit: what's going on with the leading here?}
Ok Eben, here's the ra combination in two other Caslon sizes. These are scanned from James Mosley's facsimile edition of Caslon's 1766 'Specimen of Printing Types'. The top is from the Pica 2 and the bottom from the Long Primer 1.
Here is a scan of some of Shirokuro, referenced by Hrant as an example of good use of notan.
First things first:
There is very much good and bad notan, but it depends on the objective (like any aspect of Design). A given notan "scheme" can be good for aesthetic enjoyment, or immersive reading, or making people nauseous, or whatever, but it is always good or bad for various things. Everything does matter, but in different ways.
--
> Shirokuro is pretty artsy for you no?
> ... your disdain for dislay faces.
> I thought the holy grail of notan was in increased readabilty?
Are you maybe starting to realize that you don't really know what I've been talking about? To me it's now clear that you don't. And considering how much typing I've been doing here lately I think the ball is in your court to understand more than in mine to explain.
But hey, what the heck:
1) There is no holy grail.
2) Notan is a thing, and it's always there. It doesn't have a goal.
3) A designer can leverage notan in different ways depending on what he wants to do, for whom. A good designer does this awarely, a bad one does it Vegas style.
Concerning readability (which is only one place notan is relevant) my contention has been that the two halves need to be treated equally (and that chirography/painting inherently goes against this). Readability-directed notan is NOT the case in Shirokuro, and that's wonderful. The only think I have disdain for is pretending (to understand/leverage notan, in whatever way).
> Legato ... I don’t think it is any more readable than any decent sans.
Pure conjecture.
Plus you're firmly in the Emigre/Shinn camp of
"no inherent readability" so what's the point.
> I like Legato and it reads fine but I see no quantum leap there.
Sometimes a quantum leap is the subtlest thing that can be. To see Legato's leap you'd have to be acutely aware of notan, how letters form it, and what can be good about it. Otherwise it's just another flavor-of-the-week that might sell well or not.
> I would be curious if you know of any unbiased testing of
> readabilty between it and say Formata, Frutiger, Thesis, etc.?
Of course not. But this is not about the numbers anyway. It's about thought: seeing the good kind, applauding it, and hopefully acting on it.
>> is relevance completely moot, and admitting a mistake unthinkable?
> I have no idea what that is all about. Care to explain?
Dead History is irrelevant here; and I think you came to realize that.
(Eben, the Zen comment was directed at Chris.)
> I have never been an artist
Each of us is partly an artist. As you were saying about "notan", we shouldn't get hung up on terminology. It's not what your boss or your wife call you, it's what you are that counts.
> Think of a continium where the Water Quality logo is at the
> right end and Dead History is at the left end. Shirokuro and
> Legato are between the 2 extremes
I'm not seeing any such thing.
--
Eben, your comparison can be revealing, but there are some textbook monumental obstacles here, such as apparent size, intended point size, lo-res bitmap fidelity, etc. Frankly I don't envy you trying to do this! :-)
> In Patria I see a synthesis of a strong but not wide r
I hope it's actually slightly wide - which is in fact a conscious sacrifice of readability for the sake of reinforcing the style. This is a balance central to Design, and foreign to Art.
--
> Placed in this context Patria looks pretty darn good.
I'm very glad to hear you say that. I guess you've gone back on your former agreement with JFP that the forms don't hold together.
The thing is, Patria is still way too black-centric. If it does enjoy good (text) notan that's most probably due to a combination of dumb luck (hey, I'm entitled to some too you know), a black-directed minding of spacing (a rejection of the form->space->kern idiom) and a concerted effort at functional divergence.
> Patria and Scala ... exist in different ‘notanic [or dark / light] registers’
Could you explain?
BTW, I really like the "La" in William's "Lady Willoughby’s Diary" samples. Color is one thing, but the interletter notan seems great.
--
>> "disposing of the linking between the two edges of the black"
> I think you are saying that the outlines don’t follow ductus…
> it sure as hell can’t be while still appearing conventional
> to the reader now can it?
The insides and outsides (so to speak) of the black have been disjoined (which is something more than saying they don't follow something or other). This is in direct violation of the moving front.
As for appearing conventional while doing this, the way this is possible is that there's an essential difference between conscious expectation and subconscious effectiveness. Unless you meant that fonts that follow ductus can't appear conventional: I assume you mean those that follow it slavishly (for example giving the "Z" a thin diagonal) and would point out that it doesn't have to be slavish to be relevant, and a font certainly can appear conventional while "alluding" -even strongly- to chirography.
> In some ways Patria is even weirder than Bembo & Caslon.
Thank you! :-)
And (I hope) it's a "good weird", not merely an eccentric one (those Brits :-).
--
> these Tokyo pictograms represent good (or considered) notan or am I fooling myself?
They are "considered" notan (see how there are white spaces conveying things). Whether they are good or not can be debated - I think some are good and some not. Overall it feels like exhibiting notan has been made into an objective there, which is too much Art ("Look at meeeee!") and not enough Design ("Look, and use.")
So it follows that the answer to:
> Is the best place to look for good notan in the base character
> or in the character in its environment (as per William’s samples
> but maybe not cleaned up) or does one naturally follow the other?
is: it depends on what the font is for. Notan is there for individual letters as well as letters in context; the effectiveness of a given notan scheme depends on what the font (or shape, or collection of shapes) is intended to do.
hhp
Chris, I prefer the metrics of the original Univers, without the kerning to the right of the r.
I also think the r_a in the Big Caslon is too tight.
I used to go for the tight, even look that can be achieved with kerning, but now I am more interested in the unkerned look of letterpress, especially for traditional serifed faces. I just did an OpenType version of Paradigm, which I designed in 1995, and have revised several characters and redone the metrics. In 1995 I was going for irregular shapes and reducing "excess" white space between characters by negative kerning. Now I've reined in the more frisky letter shapes (r and g no longer protrude above the x-height) and left more "interest" in the spacing. Above: 1995. Below, 2006.
> I also think the r_a in the Big Caslon is too tight.
Which might however come back to my warning about scale/intent/viewing.
> I am more interested in the unkerned look of letterpress
What do you think the readers of your text faces benefit from?
> I just did an OpenType version of Paradigm, which I designed in
> 1995, and have revised several characters and redone the metrics.
I wish Emigre had done that with Mrs Eaves.
In Paradigm (which looks like a cool design btw) I do think the new spacing is better by being looser, but I think "interest" for its own sake is anti-text. Like when you have such an accomodating "a" it's a shame to let the "r" cause its classical issues (thereby causing the "ah" to look too tight in comparison) when it doesn't need to. The degree to which one let's a character/glyph "be itself" is certainly not limited to the black, but -in a text face- bouma cohesion is critical (just like how loose tracking can decimate readability).
hhp
"Pure conjecture."
Everything you say is pure conjecture with a bit of "The Emporer's New Clothes" thrown in to boot. The only concrete thing you have said is, "The insides and outsides (so to speak) of the black have been disjoined". That is the exact relevance of Dead History (Eben, yes, you got it!).
"> I would be curious if you know of any unbiased testing of
> readabilty between it and say Formata, Frutiger, Thesis, etc.?
Of course not. But this is not about the numbers anyway. It’s about thought: seeing the good kind, applauding it, and hopefully acting on it."
If you believe that then you are far, far more in the "Emigre" camp than you try to falsely place me.
What you have said clearly is that "Chirography and painting" are wrong, bad evil and need to be fought and exterminated. That hardly is showing the way to something constructive. If 10% of what you type were of a constructive nature rather than a destructive nature, you might have come up with something by now other than disdain for anyone who does not bow and scrape to your dismissive utterances. Try saying something constructive, it may just get you more of the approval you crave.
ChrisL
What do you think the readers of your text faces benefit from?
Well, if they are reading traditional serifed forms, I think they will benefit from something closer to the unkerned metrics associated with the old faces, because such metrics are integral to the synergy of those type designs.
“interest” for its own sake is anti-text.
What I mean by where the "interest" is: it's the more active dimension of the design. It's kind of like choosing the sweet spot between making the letter shapes more diverse or the spacing more diverse -- striking a balance.
Chris, the difference between my saying that Legato approaches an ideal of readability and your saying that it's not measurably more readable isn't that one of us has Proof and the other doesn't, it's that my strance is based on an idea while yours is based on a lack of ideas.
> What you have said clearly is that “Chirography and painting”
> are wrong, bad evil and need to be fought and exterminated.
Maybe you're the one who should have replaced Rumsfeld.
You're only listening to yourself.
> That is the exact relevance of Dead History
No. Dead History is a formalistic hack job of two black-centric fonts. It does not break any sort of order in the black, not for the sake of the white nor anything else. It breaks some -imagined- rules concerning classification, and only for the sake of breaking them. Comparing it to Legato is ludicrous.
> ... such metrics are integral to the synergy of those type designs.
I don't think that's how readers function.
> it’s the more active dimension of the design.
Why?
hhp
… such metrics are integral to the synergy of those type designs.
[Hrant said] I don’t think that’s how readers function.
Well, you can play Bach on anything, and irrespective of how the listeners function, the preference seems to be for acoustic instruments. In other words, there is a sympathetic relationship between form and the original technology for which it was created. I'm not proposing that one takes no advantage of new technology, just don't overdo the kerning.
Also, people who make stuff, whether artists, designers, engineers, whatever, often have the wackest reasons, and can be totally f*cked up, but that doesn't have much bearing on how well their creations work. The Garamond typeface was based on some kind of metaphysical theory of I and O (and the Godess Io?) apparently, what was no doubt considered serious thinking at the time, yet the type still rules long after its theoretical criteria have become obsolete. Go figure.
it’s the more active dimension of the design.
[Hrant said] Why?
I'm really trying to get a different balance between lettershape and spacing. Before the letters were a bit wonky and spacing pretty slick, now I am trying to make the letterforms better behaved, and the spacing more varied. So I described more variety of spacing as making that dimension of the typeface more active.
> ... the preference seems to be for acoustic instruments.
1) That's conscious, while immersive reading is not.
2) Maybe it's because of what people are used to hearing Bach in. If you wrote a new classical composition (equivalent to making a new "old style" font) and broadcast it played with a synthesizer for a few years, they might not like the acoustic version when subsequently introduced.
3) Maybe people prefer acoustic instruments in general.
> there is a sympathetic relationship between form and
> the original technology for which it was created.
This is plausible.
But how to decide where a new font fits?
This seems to be wading into the font classification quagmire...
In the end, to me faking bad spacing is just as bad as faking rough contours.
Not that it's useless, but it seems too style and not enough substance.
> the wackest reasons, and can be totally ----ed up, but that
> doesn’t have much bearing on how well their creations work.
I'm not seeing the connection here.
But to your point: I think a lot of that is post-rationalization (or convenient rationalization). Like how when Hangul was invented, Sejong had to make up a story of how it preserved the divinity of Chinese (which was preferred by the social elite at the time). Also, no matter how theoretically/spiritually inclinded a designer is, the need to make something useful to the mainstream usually pushes the innovation/idea down the Subtle Zone, making it non-obvious. Think about how Grandjean "tamed" the RdR.
> I’m really trying to get a difference balance between lettershape and spacing.
That can be an interesting angle.
hhp
Eben, so can you explain to me what is your concern here with the La and ra?
As I said, to me the issue is one of even color, meaning visually the same average density of black on white.
Do you know Doyald Young's 'Fonts and Logos'? He has wonderful studies of the evolution of his logos, starting with hand drawn versions of standard typefaces, and then adjusting them, partly for even color of the given combination of letters. The story of the evolution of the Prudential logo is particularly revealing.
At any rate, to me 'notan' makes more sense for individual letters, where you have closed or nearly closed counters. The way Frutiger talks about it, if I remember correctly, is how the white in the inner counter relates to the white on the outside of the line. What is at issue here is whether the whites and black 'lock' as Chris put it, or 'click' as I put it, or 'become part of the page' as Cyrus Highsmith put it.
And 'color' has more to do with how the letters in mass, next to other letters. Here both the design of the letter and the spacing affects whether you succeed with even color.
Do others see this distinction between color and notan?
Eg, the 'notan' in Nick's two examples is the same; the color in the second one is better, to my eyes. In my Caslon La scans, all the letters have good 'notan'--they become part of the page, with b & w vitality--but the color is much more even and better in the lower example.
Eben,
The continuum I spoke of earlier is my projection of figure/ground (or what I might call inner and out space). The continuum starts resolved yet disjoined (but not harmoniously locked) as in the Dead History far left example (this is our “not enough to be good” extreme). It end at far right with the Water Quality logo which is two forms locked in a syntactic relationship which goes beyond their form in an attempt to be a referent to a meaning outside of them—Water Quality as opposed to just Q and Drop (this is our “too much to be good” extreme). Both of the opposite polls operate by having disjoined figure/ground. This does not operate in all of the glyphs in Dead History, not the "o" or "a" for example. The positive and negative disjoin by being 2 different fonts interjected on each other. There is no specific meaning attempted though. It is a formal relationship that still just refers to the letterform. Dead History "H" is just trying to be an H. There may be some connotation perceived by someone looking at the type but this is not consistent or intended. The Water Quality logo, on the other hand, is intended to have a connotation other than just Q and Droplet. The point of the continuum is to place items between the 2 poles to see where they fall in relationship to others. If we, for the sake of a beginning argument, agree that the typeface Legato has something of what we are looking to define (call that notan only if you like) and place it between the outer poles, what would go after it and what before? Something after it would be the desired state we seek. Something beyond it to the right would be “too much” of it to remain positive. If I am testing items which might be too much, I might postulate that “meaning outside of or greater than” the glyphs would be too much—hence the Water Quality logo, You can replace the Water Quality logo with whatever you might deem “too much” as well to come up with another in-excess quality to create an ordinal system with. By the same token, you can replace the Dead History face with some other item which has “too little of” the desired trait. In my initial post in this thread, I said that the Dead History face was “one that should evoke some debate”. We can use this technique to test where we place visual examples fall next to each-other in relation to any 2 variables. We can even place a very chirographic sample at far left and something antithetic to it at right and again place Legato as our zero point. The point of this is to have then a collection of variables and ratings. The value is in seeing what variables are in operation as well as having something physical and visible to compare and describe rather than just philosophical terms. Think of it like discussing what might be a recipe for good soup. You can name ingredients but you have to determine relative quantities before you can test the soup meaningfully. You might also find out that a particular ingredient needs to be left out completely.
ChrisL
William: Thank you again for posting those scans!
RE your question: I do think there are many differences between 'Color' & 'Notan'. Here are a few ways they differ:
Specificity
To begin, I think Color is a very broad almost sweeping term meant to encompass many variables. It is substantiually vague & hard to understand precisely because it sweeps over so many aspects. It's not a useless term to people who spend lots of time with type but it does require experience to begin to be comprehensible. Notan is superficially much simpler but also more specific or precise & therefore maybe a bit more useful. It's just about the relationship of lights & darks. 'Relationship' is key to that phrase.
Ironically Notan is even useful for thinking about bright yellow text on bright blue or vice versa where as our Typographic term Color is not. That's because unlike 'Type-Color' which is rooted in specific type history, Notan is just about lightness & darkness which is an aspect of Color - in the common, or even scientific sense of the word. Notan is relationships in one Color Axis: Light Dark. Chroma, or Saturation or color intensity is another. And Hue is the last one - the color frequency eg red blue green etc. Ironically we can talk much more readily about the Notan of Bright blue text on Yellow because both colors have light & dark values along with their Chroma & Hue - so you can talk about the ways the relationship. In contrast, talking about the 'color' of colored text is messy for obvious terminological reasons ( which 'color' was that?) but also because of the relative imprecision of the term itself.
Implied Scale
Color & Notan also differ in that one implies something at the level of scale the other does not. Color is something you tend to be aware of or notice more broadly - like at the level of a paragraph. A single line of text cannot really be said to have color I think. Do you disagree? Maybe you can see it in a word. I am pretty sure I cannot. In contrast, Notan occurs at the glyph level, the inter-glyph level, the word level & all the way up & down the size scale. In type design practice I think we are most concerned with the glyph, interglyph, word & paragraph levels. Graphic designers & typographers are likelier to be concerned with Notan at the line, paragraph & layout levels.
Descriptive of Magnatude ( or Not )
Type-Color also refers to a degree of darkness or lightness on the page. It is to some degree quantitative. In contrast, Notan also does not refer to the degree of grey. I think it would be a misuse for instance to say: 'that has dark Notan'. Notan is not descriptive of Magnatude but to something superficially simpler - the relationship of the whites & blacks.
Descriptive of Quality or Superiority ( or Not )
The phrase 'good Color' is essential to any understanding of what is meant by 'Color'. It is not helpful to say 'Good Notan' because it implies that one relationship or arrangement is superior. Notan is not about that klind of judgement. Similarly you can't quite say 'good type' outside of a context : use, paper culture, audience etc. Notan refers to a way of looking - not a result or a judgement of that result.
Sequence
Philisophically you might say that Notan is higher up ontologically - that it to say it comes first or put another way 'Color' results from Notanic interactions. Or if you prefer you could turn the order around could think of Color as being the observable phenomena and of Notan being a theoretical construct to explain an aspect of that phenomena. Either way they are not sequentially non-contemporary.
Usefulness
Notan is more useful if for no other reason than because Notan is less vague than can the term 'Color'. I also think that new insights and progress in the design of type can be leveraged from observing and thinking about Notanic relationships.
--
Nick: Thanks for posting your example and your desription of your thinking about the choices you were making. I wholheartedly agree that the 2nd one is a BIG improvment.
Peter: I like your distinction between 'visual (optical-grammatical) integrity' and ‘interplay’. I think an awareness of Notan can be used to either one or both and that they are indeed quite different. I can of course not prove it but I am wondering if visual grammtical integrity & cue values in the bouma are enhanced by the successful notanic interplay between glyphs. Maybe not. Maybe interplay just gives the eye something more compelling & comfotable to spend time with. But my gut says - no. My guts says this interplay is where the big gains are to be had and it's where I think the best new faces get their power from.
Hrant & Chris: Let's make an effort to keep the signal to noise ratio reasonable! ;-)
Chris: It was fun trying to figure that one out. I am almost disappointed to have gotten it!
Hrant: I agree this isn't easy. But I think examples and discussion of what we are seeing in them especially as they relate to Notan is the best way of helping a deeper, more useful understanding of Notan as it applies to type to emerge. Because Notan is so precise and can be seen at so many scales levels and is so superficially simple it is also hard to get a brain around. It's almost like an object you need tweezers to pick up. I hope our examples can be those tweezers. Eventually I hope we can just say Notan. Saying Notanic relationships is a little like saying 'Safe Safety'. It's weird.
>A single line of text cannot really be said to have color I think. Do you disagree? Maybe you can see it in a word. I am pretty sure I cannot.
Yes, I can see even or uneven color in a word, I noted that the L in Caslon Double Pica gives the word 'Lady' relatively uneven color compared to the harmony of the L in the word 'Lads' in Caslon Great Primer below it.
If you read 'Fonts and Logos'--which has a lot about evenness of color in logos--you will also be able to see color in a single word and will know what I am talking about. Evenness of color is one of the keys to readability, because you need a relatively uniform background matrix. If the background has uneven color, then it randomly adds information that interferes with the information you need to detect in the varying features of the letters--what Peter calls 'role architectural features and evoked forms' in the letters.
To me 'good notan' does have a meaning in a glyph, at least for text faces. To me it means what I think Chris meant when he talked about in 'locking' the black and white together, or what Cyrus Highsmith said about 'becoming part of the paper'. But it may well be a quality that requires the glyph to be in context of others in the face to really see it. Perhaps that is because the other letters create boundaries and so define the outside space.
Because evenness of color involves how the blacks of letters relate to other blacks in other letters, as well as the proportion of black to white, I think it is not a subset of 'notan'.
Furthermore, I would say that evenness of color is more important in a text face. Lacking that, its usefulness is really compromised. Good notan adds a handsomeness of character which makes it a really good face. So you need both.
So I can see the conceptual framework I have been using here is different than yours--which is why I haven't understood you. I do think the terminology of color is rooted in type design realities, as is shown by its long use and the in depth illustration and demonstration of it by Doyald Young.
By the way, the way notan is used in Shirokuro is where whites become foreground, competing with and here cleverly complementing the blacks. This damages readability--though clearly that was not the goal in this display face. This would be very bad notan for a text face.
Thank you Eben, for inviting me to your thread. Happy to find out that Typophile has a nice description of traditional Notan in its Wiki, but I think humans have come upon the concept outside of Asia too. I guess I came to a similar approach from a different angle: As a teenager I was given Paul Klee's "Pedagogic Sketchbook", where, in chapters II.23 and .24, he explains "non-symmetrical balance" (and expands it interestingly in III.26 and IIII.40) – a concept which, to me, explains modernist composition quite nicely, and how effects of vivid tension can be achieved.
A few words about Shirokuro – Akira Kobayashi just interiewed me for the Japanese magazine "Design no Genba" (I'll put the entire text on my website sometime soon) –
"During my studies at the Basel School of Design we spent much time thinking and working in figure-ground shapes and exercises, which I enjoyed a lot. In type design class we also often dealt more with white space than black form. When I began designing typefaces in 1993, I therefore also played with Romaji whose shapes were composed of white as well as black shapes, but did not get far since I found too many possibilities. Years later when I learned Kanji, I remembered again and thought, in most Kanji I might be able to separate the radical (hen) and the rest of the character in a positive-negative treatment? Here too there were too many variations at first, but in the Kana I soon found the best design rule: On the upper right often appeared ten-ten or maru, and these had to be positive shapes, so they could simply be added separately. Hence the negative part could only be located on the left or at the bottom. Thus, Kanji with a 'hen' went fairly quickly, but simpler glyphs were difficult – i needed to 'break' them in a way. Back then I still did not understand much of good Kanji formgiving, and hid my ignorance behind a playfully-random papercut style. When my Japanese design was finished, I was able to create the Romaji with the same rules very easily – in only two evenings. Many designers hate rules – I in turn often search for them, as I like using them like toys."
Correct, Shirokuro is not easy to read, actually in more ways than mentioned. Readers of Kanji actually appreciate uneven color in their text, because it assists in recognizing the characters also by their color, i.e. amount of strokes. Especially the Japanese who mix them with Kana, can easily identify the lighter-looking inflections, particles etc. in Hiragana, or foreign words in Katakana. And that is where Shirokuro totally "fails", because every character has almost equal amounts of black and white, which results in rather even color, and the simplicity or complexity of a character can only be perceived indirectly through the coarseness of its details. Still, they seem to like my idea a lot, since they told me they are trained to see Kanji only as collections of strokes, not as inside and outside shapes. Which is evident when you see how even the comparably few closed shapes are built from several strokes placed one by one, not as continuous ovals and connectors like western writing.
And yes, Shirokuro's notan is on a glyph-by-glyph basis only. So where is "interlettral / panglyphal" notan? In Blue Island? Or in Ed Interlock? :-)
Joachim, thanks for that full elaboration (good move, Eben). Some very interesting and good points there. I'd love to get into a discussion concerning certain points in the interview, but... see below. My own admiration for Shirokuro is based on its "proving" that letters are not necessarily black - that we can decipher them much, much more cleverly than the classical views would have us believe. The PoMo exercices from the 90s were just the tip of the iceberg!
> In Blue Island? Or in Ed Interlock? :-)
Good examples!
--
Speaking of examples, I have a suggestion: we typically spend a lot of time opining and discussing, probably because we love doing that. But I think this love might be diverting us from a good opportunity: to simply observe, at least to start. As much as Eben's robust elaboration of notan as a term and William's presentation of his views make me want to type up a storm, I think this thread has great "passive" potential - White, or Yin potential shall we say! :-)
So I suggest that we hold off on the ASCII and go heavy on
the imagery, for the time being. Just a suggestion of course.
hhp
Joachim: Thank you for telling us about your exprience & ideas. I know we all appreciate that very much! I hope you will stay involved in the thread and on typophile. I am sure we can benefit from your perspective. Cheers!
RE: the Examples Chris posted & images generally: When I have been looking at Type examples I have been preparing them in photoshop cs which has several different rendering schemes. And then there are the differences widely observed in the rendering in various OS. No doubt all of these variables will alter the subtle relationship of the glyphs. Does anybody want to suggest a standard way of doing this so that are comparisons have a rendering model in common? Maybe this is fruitless...
Hrant & William: I agree with Hrant's point about sticking to observations. My ideas about Notan are highly provisional and it's time to rub them smooth by testing them on real type. That is indeed what this thread was for - creating an opportunity to share real observable examples for mutual benefit.
There are so many other interesting points left but that's all I can get to just now. And it's probably better to keep these posts shorter anyway.
Captured from commoncrawl.org on 26 Apr 2015. original URL