Whether they be Roman-Greek-Cyrillic, Hebrew, Arabic, Chinese-Japanese, various South Asian, or whatever, the individual characters of the world’s writing systems, be they alphabetic, syllabic or ideographic, tend to have aspect ratios, on average, of between 2:3 and 5:6 and a portrait orientation. Pourquoi?
I don’t see data that expressly notes any anisotropism in the layout of the photoreceptors of human eyes, although there is an implication of a triangular packing of them here, which, depending on its orientation, might induce a difference in spacing of photoreceptors horizontally versus vertically.
Or could this aspect-ratio bias have to do with the direction one scans such characters (but consider the conventional Chinese direction of scan)? Does a difference by direction of saccade capability get involved in this question? And thinking of a bias in favor of horizontal saccades, since humans figured out what kind of eyes to have long before they got literary, would such a bias relate to the most common movement of other animals (both up and down the food chain), as being most often in a horizontal direction? Would such a bias make the effective resolution of human eyes higher in the horizontal direction, despite an isotropic spacing of photoreceptors in them?
Or is this orientation and aspect-ratio bias a result of how images are mapped onto the particular part of the brain most optimized for visual pattern recognition, that part which recognizes human faces. . .which are oval to some similar aspect ratio and orientation (in their usual tête-à-tête confrontational situation)?
Ray (A rose by any other aspect would be a tulip.)
Nice jargon. I mean, really, it’s quite lovely.
Perhaps the aspect and orientation has more to do with the making of the symbol than with its later visual apprehension-- i.e. the hand held in a writing position has an easier and wider up and down movement than side to side (only uses the wrist and hand muscles rather than the lower arm as well, as in side to side motion).
Well, I never claimed to be brilliant and in this case I have absolutely no idea what you’re talking about. I do, however, read and write Nepali, so if information on Devangari script would be useful to you, clue me in as to what it is you’re asking.
“I should not take bribes and Minister Bal Bahadur KC should not do so either. But if clerks take a bribe of Rs 50-60 after a hard day’s work, it is not an issue.” ----Krishna Prasad Bhattarai, Current Prime Minister of Nepal
Lucky, ‘aspect ratio’ just means the ratio of width and height. In this case, of each character. At least I think that’s what he meant.
My theory would be that characters tend to be squashed in the direction they are written, i.e. sideways in European languages and vertically in Chinese. Each character in Chinese or Japanese is actually almost 1:1 aspect ratio, at least on the books I have around. If English characters were wider than they were tall, you’d have to move your hand faster and your fingers less when you write, which would slow you down. Better to slow down the hand movement at the cost of more finger movement. Similarly if writing vertically (e.g. Chinese), you want the characters wider for this to be true.
When you speak of “writing”, I believe you’re referring to the cursive script of handwriting. But roughly block letters came first, right? And they were taller than they were wide – before cursive handwriting, I believe. And since these days the production of these characters – in documents presumably optimized, at least somewhat, for fast reading – I’m sure if the characters were optimized before for fast writing, and the result was not optimized for reading, it would’ve been by now – independent of how fast the characters could be lettered or handwritten. I hold fast to the thesis that the aspect ratio relates to reading (human visual), not writing/printing/whatever.
Another question, as to lower case letters: Do consonants, in the Roman alphabet, tend to have ascenders and descenders, while vowels do not, because consonants have more information – and thus it is more important, when reading fast, to be sure to catch the consonants. Consonants do contain more information (meaning-distinguishing capability) than do vowels, in, I think, just about all languages, and vowels are often not written in Semitic languages.
Original “block” letters that were carved were not designed for fast reading, but for formal documentation and ornamentation. The shape of the forms weren’t optimized for fast eye-brain recognition, but to look good artistically.
When stuff was written down on tablets, etc., it was probably geared towards quick writing, not reading. Take a look at some cuneiform writing sometime. There also was a tendency to make it hard to read, to make sure only certain people could learn.
When you get into paper writing with ink, the situation was pretty much the same as today- individual handwritting tends to vary a lot, and for some people, is geared to quick writing, and for others, for pretty writing.
Another factor is that most ancient people read aloud (even to themselves), which would limit the top reading speed regardless of the alphabet.
This question becomes more valid when you start talking about different typefaces. There are thousands of widely different typefaces for the “Roman” alphabet, some are easy to read, and some are almost impossible. I have heard of studies which claim that Times Roman is particularly easy, but of course I don’t remember any references ;).
Well, I don’t see that what you say here really refutes at all my stance that present block letters – or whatever letters you want to call those composing the fastest-readable typefaces, e.g., Times Roman font – tend to have a certain narrow range of aspect ratio, because that ratio enhances reading speed and/or acuracy.
While cursive fonts became popular in medieval times, apparently, when monks thought they needed to speed up writing a little, because they didn’t have movable-type printing yet, it’s clear that some time after the Gutenberg press, Europeans went back to using block letters as “capitals” and straightened forms of cursive letters as other letters, most certainly due to their greater readability over the previous scribes’ hand-written letter forms. And both of these tended to evolve to the aspect-ratio range I mentioned. Certainly, long dissertations then produced by printing presses were not composed of letters primarily designed to be esthetic (save perhaps chapter- or paragraph-initial illuminated or otherwise elaborated letters), but rather of letters designed to be easily distinguished from each other.
How in the world did you determine that? Did one of them write such a statement down? Now, you know you can’t always believe what those guys said. They could well be putting you on. But, then again, maybe you were there in your prior life.
It is, of course, irrelevant to my thesis, that there are propagated today a huge number of ornate alphabetic font styles, many of which are very difficult to read. Certainly, people do not use those to print the text in large books. . .or even small books.
I guess there is a claim made that simple, slightly serifed type, such as Time Roman, is the easiest/fastest to read. I’m not a fast reader, but I think I prefer certain modern sans-serif types. There are clearly several features in type styles that are balanced in different ways in different font faces to maintain efficient reading. It appears to me that the two most important are 1) a set of letter shapes that are the most different from each other that they could be and yet retain simple shapes, and 2) something to maintain the reader’s tracking on a line of print as (s)he traverses a page, as opposed to inadvertently skipping to the line above or below during this traverse. Block-type letters without serifs optimize ‘1)’ (ignoring certain problems in the stereotypes of Roman letters that don’t optimize distinction between some of them apart from implementing differences in their serifications), while horizontal serifs tend to optimize ‘2)’. Given the results of most OCR programs on quite legible type, it is apparent that the Romans could’ve chosen letter shapes just a bit more carefully, in view of the forthcoming digital world. Beats Chinese, Japanese, Korean and all those other schemes by a long shot though.
Ray (may talk to himself on a occasion but doesn’t read aloud)
Nanobyte says, “How in the world did you determine that? Did one of them write such a statement down? Now, you know you can’t always believe what those guys said. They could well be putting you on. But, then again, maybe you were there in your prior life.”
I’m still not sure where I was in my prior life or lives but remember Jerome? The guy that translated the OT & NT into Latin? Nice guy. He was supposed to be the person who figured out that reading silently could still be called reading. Up until then everyone read aloud even when they were alone.
Oh, I’m gonna keep using these #%@&* codes 'til I get 'em right.
Jerome who? You may have guessed that I never followed those OT and NT too well – in any language. I’m sure that whoever believed in the OT and NT would certainly believe whatever Jerome said.
Ah, but Times Roman that we use today is pretty much exactly what was used 2000 years ago. I don’t think the Romans spent much time optimizing their letter shapes for maximum readability, utilizing subtle differences in eye-brain pattern recognition. However, I’m willing to believe that the Times Roman typeface itself has persisted because it is (slightly) easier to read.
Many ancients wrote down statements about reading aloud. It is fairly well known that most people read aloud then. The idea of writing was to reproduce the spoken word. As Jois mentions, Jerome is a good example. He obviously knew his linguistic stuff- he translated the Bible from the original Hebrew and Greek into Latin.
Another factor that would impede the max reading speed in ancient times is the fact that they used scrolls instead of books.
Was this your original thesis? If so, then I agree!
It’s probably hard to really determine what’s easy to read, since it depends a lot on what you’re used to. Those cursive manuscripts from the middle ages are almost impossible for most people to read, until you’ve worked with them for a while. After that, they’re not too bad- what used to look like a continuous stream of cursive letters turns into distinguishable words. For a fair comparison between Times Roman and some wacky typeface, you’d have to give the person some practice in the new typeface. Another factor (which I think you’ve aluded to already) is that people don’t read by seeing individual letters, but groups of letters, so you’d have to take that into account also. And don’t forget the spacing! Proportionally spaced fonts are easier to read.
I think the fact that OCR programs have trouble with easy to read type is because the state of the art for OCR is light-years behind human ability, not because the typeface is particular hard to distinguish. To give OCR an easy time, we should print our books in barcode format! Easy for them, hard for us.
Hey, you’re supposed to be from the same side of campus as I am, according to your profile. Did you start on the other side, or is there a religious factor in your position? The engineering way, in anything (well, within limits), it would seem to me, would be to try to improve that thing by looking for universals in it, not just by relegating its nature to one based on cultural preferences.
Well, our caps are pretty much what they used, but they didn’t use our lower-case fonts, did they? Such appear to have evolved from the medieval scribes’ scribbling, and these do make reading faster than would be the case for all caps. However, the improved effect of the small letters is probably mainly a result of their ascenders and descenders in changing word shapes.
Lesson 4 here includes kerning, that idiotic cultural game played by typesetters with nothing better to do – merging 'f’s with 'i’s and '[l’s and all that sort of thing. So what happened when we went from printed hard copy to pixelated screens, we immediately couldn’t tell ‘rn’ from ‘m’ in many type fonts, and still have a problem with this in some fonts. Certain kerning may improve word-shape recognition, but in general, I think it only improves esthetics at the expense of reading efficiency. I do not believe the statement made in this stuff that proportional spacing improves reading speed.
Yeah, and many people wrote about famous people discovering things like the world’s being round, while common people hundreds, if not thousands, of years before knew of such things. (Check through your nearest channeler. ) Do you really believe any of these writers, who “wrote down statements about reading aloud”, went around snooping at what even the few who were literate did in their spare time with respect to written materials? Do you believe that every time even some Phoenician looked at a log of what he put on his boat, or whatever, he repeated aloud an oral rendition of each item on his list? Phooey. What is this magic that important authors in history have that makes people believe every last silly little word they come up with? And what about the non-most people back then?
Not always. Consider kanji (Chinese words) in respect to their rendition of spoken Japanese.
Does that make him an expert on how people read in their bedrooms behind closed doors? Maybe they didn’t vocalize there everything they read, because the neighbors might not understand. I never heard of the guy, but I presume that, if he translated religious stuff he was probably a monk, who didn’t get around all that much and maybe didn’t know at all how alchemists read.
Was what my original thesis? If 50,000 kindergarten classes invented 50,000 new type styles, it wouldn’t affect my claiming that our type styles most commonly used for body text is no doubt somewhat optimized for reading efficiency. The fact that one can build airplanes that don’t fly, doesn’t mean that the most common ones aren’t optimized for flight. . .besides just looking like airplanes. Let’s get off the ground.
Who said science and engineering were easy. If you just want to be cultural, as indicated in that sentence, you stay on the other side of campus and become a psychologist or sociologist, I’d say. But you never invent anything that way. BOOOOORRRRRing. Granted that people get used to a lot of weird things and can deal with them pretty well in a lot of cases, even though they aren’t close to being optimized. But why not push the envelope and make those things more optimal?
Mais oui! ‘Fair’ is our middle name. The object is to find the optimal combination of raw tool and trained usage on that tool, is it not? Who’s faster someone trained on a QWERTY keyboard using that keyboard, or one trained in Dvorak using one of those? That is the type of question.
Yeah, I guess I should’ve put that in my list of two features above that optimize lines of text for reading efficiency.
I challeng that, although it appears to be generally claimed. Although that tends to hold words together mor tightly, I think that’s mainly an esthetic quirk of typesetters, along with kerning.
But OCR shouldn’t need anything like the computing power humans can bring to bear on the task, at least with some reasonable type-font modifications literary/journalistic typographers are apparently unwilling to compromise on – modifications that would no doubt improve human readability of the same texts. . .yes, after retraining.
Is that a means of getting the right drink set before you. . .or is that what the legal hounds speak? We should probably hold up on the bar code until we get lasers implanted in us. I once thought computers would force the issue on irregularities in English and other cultural nonsense, but you can see how things have moved in the opposite direction.
Speech recognition is difficult and really not all that much in demand; but what’s with OCR gurus? With today’s cheap memory, simple brute force should do better for them, even with some sort of general-purpose adaptive software reading all those typewritten pages those other monkeys are pouring out, trying to duplicate the British Museum.
Maybe, more than HDTV, we need vastly improved alphabets and language(s) that work exceptionally well on computer screens and as read by OCR. . .and then Silicon Valley can replace all those neobiblical scribes that have only turned in their quill pens
What? Are we now going to have lawyers and this seudo-one, who proclaims himself a “deep-thinker”, decide which threads began as other than requests for information? Just because he has no relevant info on the subject, we should listen to him because he puts ‘Esq’ after his name?
Ray (Wasn’t it OH where they tried to legislate pi as equaling 3? OK, maybe it was PA or TN, somewhere back there where attorneys tried to reshape the earth in accordance with their highest count of brain cells.)
Okay, I tried to write a thoughtful answer to this, but I have NO idea what you’re talking about. What is an aspect ratio? I’d be happy to discourse on writing styles in Hebrew if you can explain whatever the hell you mean.
Scr4, above, explained aspect ratio. I thought everyone who watched TV or looked a computer monitor knew what ‘aspect ratio’ meant – simply, in these cases and in the subject case, all two-dimensional ones,it is the ratio of the height of a rectangle to its width or vice versa – in the subject case, the term protrait designating a height to width ratio that divides to a number greater than one, while the term landscape disignates one that comes out less than one. The corresponding three-dimensional parameter is usually termed ‘form factor’.
Looking at typical Hebrew letters, I think they generally are just a little bit higher than they are wide, but most of them tend usually to be closer to 1:1 than do Roman and Greek letters. But, hey, your musical symbols, at least notes and clefs, tend to be “portrait”-biased, though not, of course staves and some rests.
And, oh yeah, Ct. Jester, love that I’net piece! So the actual attempted legislation of history was in IN. Well, that’s right next door to His Esquireship.