There Are More Than 26 Typefaces in Existence

quadibloc's picture

Lots more than 26, in fact, so presuming someone could find a set of 26 typefaces such that each one of those typefaces could be readily distinguished from all the others in that set on the basis of a single character...

instead of saving ink and paper with such minor steps like having really short descenders and huge x-heights, each letter could now encode the letter following it in its typeface, thus cutting the length of texts almost in half!

I'm assuming that for readability, words would still be separated with spaces, and so one might have a 27th neutral typeface for the last letter of words with an odd number of letters.

hrant's picture

Huh?

hhp

quadibloc's picture

Well, I admit this proposal was made in jest.

Introducing a set of syllabary symbols into the Latin script similar to the Japanese katakana would be a simpler way to increase the density of texts.

The idea (if it wasn't obvious from the original post) was that, in one possible implementation of the scheme, a letter in Artistik is that letter followed by an A; a letter in Arnold Bocklin is that letter followed by a B; a letter in Cheltenham is that letter followed by a C; a letter in Lydian is that letter followed by an L; a letter in Optima is that letter followed by an O; a letter in Peignot is that letter followed by a P; that sort of thing.

However, since this requires recognition dependent on fine details of the letters, the obvious refutation of this proposal is to present the clearly better alternative of reducing the type size by 30%.

bojev's picture

Like Hrant said Huh! or maybe Duh!

Martin Silvertant's picture

There's little information in this post I can actually follow but I'm really curious as to what you're proposing. Could you try to explain it once more? The third is the charm...

quadibloc's picture

Ah. I had thought that my joke was clear and obvious.

But I will be fully explicit.

I am proposing leaving out every second letter in every word. But the information is not removed - because the letters that precede each omitted letter can be in one of 26 different typefaces - which typeface depends on the identity of the omitted letter, and is used to communicate it.

So, since the use of different typefaces is an available degree of freedom, I exploit it to convey actual textual information.

Martin Silvertant's picture

I have problems interpreting sarcasm correctly. I guess I'm still taking some of the things you said too seriously.

What are you proposing this for, and why would you only omit the second letter? How does this cut the length of the texts almost in half?

Nick Cooke's picture

Hilarious.

LexLuengas's picture

Take it a step further and encode every word in the dictionary as one revival of the Garamond types. Since clearly we only need one glyph from each of those typefaces (say, the letter ‘a’), we get what is perhaps the most space-saving system of all time.

quadibloc's picture

What are you proposing this for, and why would you only omit the second letter? How does this cut the length of the texts almost in half?

Well, if you have 26 different typefaces to work with, then each typeface can stand for a different letter.

So if you have a letter, the letter it is stands for one letter, and the typeface it's in stands for the letter following.

If I omit all the letters, how can anyone see what typeface they're in? Or perhaps you meant that you thought I was omitting only the second letter of the word, instead of every second letter - omitting the fourth letter, because its identity is given by the typeface of the third letter, omitting the sixth letter, because its identity is given by the typeface of the fifth letter, and so on, also.

Take it a step further and encode every word in the dictionary as one revival of the Garamond types.

There may be enough of them, but would they be easily enough distinguishable for the system to work?

Thus, there are even more than 676 typefaces in existence, but using the typeface of a letter to indicate the two letters following is clearly not practical.

Now, another degree of freedom would be to use the Greek, Cyrillic, Armenian, or Ethiopic scripts to represent letters. Or, to get enough visually distinct objects to represent all the words in the dictionary, there are Chinese characters. But while the Chinese characters map systematically to the syllables of Chinese, how would one map them to the words of English in a memorizable fashion?

To find a high-efficiency writing system for English, one option might be to adapt the Korean script...

Martin Silvertant's picture

Or perhaps you meant that you thought I was omitting only the second letter of the word, instead of every second letter - omitting the fourth letter, because its identity is given by the typeface of the third letter, omitting the sixth letter, because its identity is given by the typeface of the fifth letter, and so on, also.

Indeed I thought you only meant to omit the second letter. I wonder what a page set with this method would look like. It's clearly a ridiculous system, but I might try it out just for fun.

bensyverson's picture

26 is overkill... You could probably get away with two faces: serif and sans. One for missing vowels, and the other for missing consonants. Most of the time you can fill in missing letters based on context even without a vowel/consonant clue, so it would still be pretty readable.

Or you could change the system, and only remove vowels. Instead of a different typeface, you could make consonants with leading or trailing vowels italic.

other becomes thr
three becomes thr
there becomes thr

If a word becomes too ambiguous, and the context doesn't help, you could just drop in the vowel.

Snds lk a vr rdbl nd cmpct sntx. W shld lbb cngrs mmdtl!

Martin Silvertant's picture

Snds lk a vr rdbl nd cmpct sntx. W shld lbb cngrs mmdtl!

Sounds like a very readable and compact syntax. We should —
I'm not able to read the last words. I suppose someone more intelligent would be able to decipher it easily though.

quadibloc's picture

lobby Congress immediately!

... the problem is not so much intelligence as cultural context; that is, familiarity with the legislative bodies of the United States of America.

Martin Silvertant's picture

Why do you think that's the case? Isn't it just about insight and rationality? Well, perhaps rationality is not the issue here since I do understand the decoding method. Still, I probably should have guessed "immediately".

quadibloc's picture

I'm just saying that guessing that "immediately" would be easier for someone who sees the phrase "lobby Congress" enough for it to be something that comes to mind quickly. That's how the brain works, after all.

Martin Silvertant's picture

Yes, and this process of associating is correlated with intelligence. The more associations you make, the more insight you have and the easier it is to decipher. Certainly knowing the context helps in making the appropriate associations. I could have deciphered this if the appropriate associations would have been made, whether it's by recalling memories or by being creative in my thinking—both seem to be linked to the capacity to associate.

I have to admit though—being Dutch—for some reason I didn't even consider trying to fit /y into the words (as it's not a common letter in the Dutch language), which might have given enough insight to guess the words. Although, not being well versed in (American) politics I'm skeptical as to whether I would've guessed "lobby congress". Those are not words I generally use.

Té Rowan's picture

Looks a bit Mark Twain to me...

Syndicate content Syndicate content