Jordan, this is a great line of thought.
Since '98 I've been working on an Alphabet Reform project where dissimilarity (and specifically in the realm of boumas -word shapes- more than individual letters) is emphasized. Because information comes only from contrast. The rest is convention, and malleable (although you have to bend it not break it)

I think the trick to optimal readability lies in making the glyphs are different as possible while limiting errant saccades (eye fixations to the wrong part of the text). But the tresholds still escape me. And actually, a font system like you describe could actually be a useful tool when used in empirical testing (which however is very difficult/expensive to carry out properly).

hhp

Not that I have a lot of expertise or evidence to support my posting here, but a few things come to mind in this general realm of alphabet reform.

One is the relationship of letter similarity/dissimilarity and dyslexia. The fact that there are shapes that are reflections or rotations of eachother is a major tripping point for dyslexics. Some people offer certain typefaces and rules (sans over serif, differences in descenders, etc) to fight the problem, but I think an alphabet reform would cure it. (http://www.dyslexic.com/database/articles/fonts.html)

Of course, when it comes to designing a font over designing a typeface, there are issues of uniqueness to a font, character, and so on, that seem to warrant a more radical and fundimental change. I think it's fun to think about the idea of slowly rolling over from one alphabet to another, easing in to it.

Ah, but how much of a difference in this would could a new alphabet make? I imagine that, beyond dyslexia, it could perhaps benefit literacy on a whole, but this is speculation. Are there any sets of research to support such a theory?

Most importantly: Is our alphabet broken? Do we read best what we read most, or is there possibility for a better possible system?

> http://www.dyslexic.com/database/articles/fonts.html

Cool resource, thanks.

> Is our alphabet broken?

Is my BMW broken? The front brake disks are warped, so the steering vibrates when I brake hard on a downhill; the driver's upholstery is falling apart; one of the tail lights is out; etc. But it mostly does its job, valiantly. Would I like to fix it up and make it better? Duh. But it costs money, and I have more important thing to spend money on. What does Alphabet Reform cost? The kind I'm talking about, nothing. Unless you count [unhealthy] pride.

hhp

Sorry to hear about your BMW, it's a shame.

Jordan, not to hijack your awesome thread, do you have any sketches? I'd like to see what you're thinking so far. Same goes for you, Hrant, I remember once seeing a little something of yours in another thread. I just want to see what we're thinking here.

Not to be a crazy theorist, would this be a revolution of the type designers? It seems to me like the alphabet could use some modernizing, and who better?

Should we also be looking at Caps? Or, maybe the variety of shared parts in traditional caps are a good place to start for the lc. It strikes me that there should be some kind of real similarity between lc and UC that doesn't exist now (Aa, Qq) while some letters are nearly resized (Vv, Xx, Zz) or others are well translated (Bb, Mm, Tt).

Also, as we are moving to an increasingly typeset culture, can we ditch structural relationships derived from script handwriting? (that ought to make you happy, Hrant.)

K, sorry about the free flow of thought, I just type it as I think it.

Aaron

Aaran: Ever since I started seriously thinking about designing type, and actually experimenting, it's always bugged me how some letters are (relatively) easy to create once you've got a few basic shapes down, and others a real pain.

I hate drawing z, g, and most of all k. They always seem so disharmonious with the rest of the typeface -- even in the most wonderful types it's just a matter of them being squashed in better than in lesser faces.

At the same time, it always amused me how traditionally, the clearly identical forms of b, d, p and q had these subtle little divergences (presumably I though, just to bug us designers). The flick on the q's tail, the square serif on the bottom of the b, the opposing serifs on a d and the 'backwards' serif on the p tail. Just so they can insist they're unique!

I've not got any sketches or anything, it's just something I was thinking about. I'm drawing a very simple face at the moment, and I kept realising there was this crossroads I came to with each letter -- it could be conventional, 'boring' and readable, or unconventional, unique, but difficult to read. (Even when only using simple basic forms.)

Then I thought that surely differently shaped letters should help readability, not hinder it -- but that didn't seem to be borne out when trying to read a face with unusual letterforms (like the faces that use serif splicing experiments such as alita postmodern).

maybe it's a beauty thing -- it's said the faces (as in, human faces, not type-faces) that most people find most attractive are those that are symmetrial but with slight inconsistencies (beauty spots etc.) Maybe we're too preoccupied with the 'beauty' of a typeface (or more specifically, it's individual glyphs) to appreciate that it's function may be served better by having more than slight inconsistencies in it's alphabet -- and we visually reject 'odd' conceptual alphabets on their superficial lack of beauty alone; where the beauty spot has become a boil!

I want a b that looks like a lower case delta, a w that looks like an omega, a q that curls up underneath it's neighbouring u, maybe even joining with it, and an l that's more than just a vertical line to be confused with a 1, an I and a |.

I just wonder if it'd 'work' or not. No doubt it would take some getting used to, but it would be interesting to test.

Incidentally, Hrant, I've enjoyed reading your work on Boumas and revolutionary alphabets (I confess that it was re-reading a post of yours when you posted your experimetally revised alphabet that encouraged me to put some words down on my thoughts). One thing about Boumas -- given that there are an (effectively) infinite number of shapes that words can form, and that you believe it's the word-shapes' divergence that leads to reading ease, do you think that our mind forms a 'bouma alphabet' of some kind -- maybe even a 'visual-syllable' alphabet?

Obviously, it would be impossible (or at least, very improbable) for our mind to catalogue every Bouma, and that many very different words will inevitably have similar shapes, how do you think we can store the information in a bouma without first focusing on individual letters first? maybe we just focus on one 'key' letter in each word that leads us to recognise the rest of it by it's shape -- there's probably a good analogy for that but I can't think of one....

Jordan

Sorry for the delay.

> Sorry to hear about your BMW

Don't be, it's still a nimble -if ragged and aging- tiger in this world of lumbering hogs. And there's nothing like a beat-up convertible: part tanning salon, part pickup truck. :-)

--

> sketches?

Here's something:
http://www.themicrofoundry.com/other/maslucida.gif

> would this be a revolution of the type designers?

Yes, actually (which is why so many people hate it), but a "subvisible" revolution, because for something like this to succeed you can't rely on re-educating people, so you have to "fly under the radar" sort of and aim for the subconsious. People can't notice, otherwise they'll reject. This does not make improvement impossible however, because reading is essentially a subconscious task anyway.

> Should we also be looking at Caps?

Eventually, yes - but it's less than 5% of (non-German) text.
As for the relationship between caps and lc, I think they need to be diverged, by the same logic as above. People can handle much more than a few dozen glyphs - just look at Chinese.

> can we ditch structural relationships derived from script handwriting?

That's already happening, of course, very naturally. But we can still try to accelerate it. Why? Because it's an artificial constraint.

--

> They always seem so disharmonious

Or they're part of a deeper harmony...
Morse Code is highly harmonious, but also highly unreadable.

> it could be conventional, 'boring' and readable

And legible, maybe. Readability is anti-boring!

> Maybe we're too preoccupied with the 'beauty' of a typeface

Indeed. Probably because tradition can only be learned/taught consciously. People don't like to be told that half of human reality is beyond their appreciation. But it is, and that's why a "real" text face cannot be evaluated with superficial formal rules (like ductus).

> do you think that our mind forms a 'bouma alphabet'

Yes, in the same way that you learn the letters, you learn the shapes of [frequent] cluster of letters. But generally you learn them out of practice, not some schoolteacher telling you. Over time you build boumas in your head, yes, compiling them via the individual letters when you run into an unfamiliar one, but gradually less and less.

hhp

check out www.readregular.com

So ...

Needless to say my idea never got going, things to do, people to see, you know how it is. But I have been thinking about this again, and I think a place to start would be to try and keep common word clusters' boumas similar (to their existing shapes), while diverging and experimenting with individual glyphs.

I've been reading a little about Cryptography recently, and thought a great place to start may be looking at the shapes of the most common letter combinations. I found this page:

http://www.prism.net/user/dcowley/analysub.html

Eureka! I think.

> keep common word clusters’ boumas similar (to their existing shapes),
> while diverging and experimenting with individual glyphs.

I think this would backfire, but maybe I'm misunderstanding you.
Could you please elaborate?

hhp