Random thought: Ideal shapes for reading

Frode Bo Helland's picture

Isn’t it strange that the pinnacle of type boils down to the traces of a tool thousands of years ago? As interesting as it may be, I am not convinced the blooming/recent trend of reversed contrast is much of an improvement. It is, however, a necessary first step. Bla bla bla.

hrant's picture

Type does not boil down to any physical tool, or even a set of such. Those who have based their lives on that illusion –because that makes them feel better– want us to believe that. Expose them, for the love of the craft, and the love of mankind's cultural progress.

Reverse-contrast is indeed not an improvement per se, but yes, it can be catalytic in the eventual abolition of the illusion. BTW we should not use "reverse" because that reinforces the skeuomorphic illusion; I recommend "vertical" and "horizontal".

There are about half a dozen typeface designers that are on the right track. Or at least off the wrong one.


Frode Bo Helland's picture

… traces of a tool *set* thousand of years ago. Please forgive my poor English.

quadibloc's picture

Unlimited freedom in design often leads to chaos.

It happens, fortunately, that the characteristics of the nib of the old-style pen led to subtleties and details in letter shapes that improved legibility and reading. Thus, type design from calligraphic roots provides a beneficial discipline and framework.

Never the less, you are absolutely right that putting ink on paper can be done arbitrarily, and there is no particularly good reason that type designs should be calligraphically-based. One useful means should not be confused with the end.

In subtle ways, ever since the pantograph freed type from the technical limitations of punch-cutting, type design has included non-calligraphic elements. Think of the subtle ink traps placed within some newspaper typefaces.

And what of Bifur, or Peignot (Peignot Bold being its most famous weight) or the typefaces based on MICR to give things a computer look? Display faces such as this show that the non-calligraphic is hardly unthinkable.

Today, the designer intending to design a typeface for the ordinary reading of body copy generally includes in his consideration how people are creatures of habit, and thus includes familiar calligraphic elements in such designs. And there is still room to do new things in mixing elements from Jenson, Aldus, Garamond and Caslon or reviving the work of other type designers from that period or thereabouts.

But even though you may be disappointed in the pace of progress, sans-serif faces like Helvetica and Univers do not look to me as though they were written with a pen.

And then there is Optima.

If people experimenting with abandoning the tyranny of the pen take their experiments seriously enough to aim at rivalling these successes, I may applaud, but if they just design bizarre-looking letterforms that have no chance of acceptance as anything but attention-getting display faces, I am not going to accord such things any significance.

Theunis de Jong's picture

"Abandoning the tyranny of the pen"-related:

Wim Crouwel's "New Alphabet", still looking surprisingly modern and technical.

Designed to "embrace the limitations of CRT screens", rather than trying to fit round characters in square (pixel) holes.

Frode Bo Helland's picture

I wouldn’t surprise me one bit if the typeface best suited to our eyes and brain looks unfamiliar – even hideous – by today’s standard.

Nick Shinn's picture

Reading is a learned skill, and culture-specific.
The shape of alphabets has evolved as part of an inter-related ecosystem of reading, writing, typography and technology.
The eye and brain are trained to read.
Reading is also reader-specific.

Therefore there are no shapes that are ideal for our physiology, all that can be ascertained are averages of what works best for a particular demographic group of people, educated in a particular manner, at a particular time and place, using a particular technology.

Frode Bo Helland's picture

That is certainly part of the equation, Nick, but I’m not convinced it’s the whole answer. In any case, moving further is in the best interest of type designers.

Frode Bo Helland's picture

If I sound like I’ve got the answer, that’s not the case. I’m just interested in what you think.

quadibloc's picture

@Frode Bo Helland:
I wouldn’t surprise me one bit if the typeface best suited to our eyes and brain looks unfamiliar – even hideous – by today’s standard.

Certainly, in theory, that is entirely possible.

But if it were true, and someone even designed that very typeface in all its optimality, how would we test that typeface in order to discover it was better than what we had now? If it's unfamiliar and hideous, people would find it awkward to learn and read, so a test comparing it to what people are familiar with would be biased against it.

So in exploring unfamiliar regions to find optimality, one is wandering in the dark, striking blind. Instead, to find what our eyes and brains need, it seems to me we will have to take baby steps to make progress.

It may also be noted that it doesn't seem that a technical improvement in typefaces is really needed, although the limitations of the dot matrix screen encourage some reaction. Do Garamond and Baskerville really need to be improved on?

Of course, they were improved, and what we got was Corona (and Aurora and Ionic No. 5)... so we already have typefaces that are highly legible, despite being set at point sizes that constitute cruelty to the human eye. Which, of course, raises the question of the desirability of conniving any further in such unbridled excesses of capitalism.

On the other hand, they're trying to come out with smart watches, and if new typefaces could let them display more information, I suppose it would be of some use.

@Nick Shinn:
Reading is a learned skill, and culture-specific.
Therefore there are no shapes that are ideal for our physiology,

I think that conclusion is unwarranted, despite the fact that I believe (as noted above) that such shapes would be very difficult to find.

After all, the world has many different writing systems in current use. Studying the reading performance of Koreans and Chinese and Armenians and Israelis and Inuit and Georgians and Tibetans and Ethiopians with different typefaces for their languages, for example, might yield some insights into the right direction in which to move.

Also, it might be noted that in mathematics, I am a Platonist. I think that a peek-a-boo picture of reality which says that what we cannot see cannot exist is inherently inconsistent.

However, when it comes to the classic philosophical conundrum

If a tree falls in the forest, with no one to hear it, does it make a sound

I would note that the careful definition of words is of assistance:

If by sound you mean the qualia associated with auditory experience, they have not been excited or manifested in any human mind by the event;

if by sound you mean longitudinal pressure waves in the air, clearly they have been created by this tree falling as well as they would have by any other.

Of course, someone might say that I'm just following the example of Noah Sweat: "If when you say whiskey, you mean", but I think the distinction is valid in my case for determining the domain in which the question is asked.

russellm's picture

Unlimited freedom in design often leads to chaos.

You say that like it's a bad thing. :o)

J. Tillman's picture

Re: change and testing. The change will be evolutionary, and the testing will be done, as always, in the market place. And let's remember some schools are not even teaching cursive any more.

The change that I am seeing is increased angularity (less flow) in type. Here I'm thinking of Skolar or Turnip. And my eyes see this as more readable on paper and on screen. Is this angularity happening in non-Latin languages?

Will the changes in type influence the way we print? I think, over time, yes, this influence will be reversed. And let's remember (part 2) that throughout history the dominant language in an area has always been the language of the powerful nation state. Whatever changes happen, they may not be in the Latin script.

Maxim Zhukov's picture

After all, the world has many different writing systems in current use. Studying the reading performance of Koreans and Chinese and Armenians and Israelis and Inuit and Georgians and Tibetans and Ethiopians with different typefaces for their languages, for example, might yield some insights into the right direction in which to move.

In the fall of 2007 Gerry Leonidas held a seminar at the Type Directors Club. I was there as Gerry’s guest. His seminar turned very successful: it was most informative, educational, and fun. Like all participants, I had learned a lot. I had a special angle on the seminar’s content, agenda, and methods used: earlier that year, also at TDC, I conducted a similar seminar on Cyrillic type design.

The relationship of Cyrillic and Greek scripts, their similarities and differences, were always of great interest to me. What caught my attention at Gerry’s seminar was that in the beginning he asked all present to try their hand at writing Greek letters; he commented on the participants’ worksheets, explaining the finer points of glyph construction.

I asked myself, why couldn’t I have used the same exercise to kick off the discussion of Cyrillic letterforms? Why didn’t I even think of it? It felt so natural, so obvious. Upon reflection, it dawned on me that the reason could be in the fundamental difference in genetics of the two scripts, and the development of their respective typographic forms.

Unlike Greek script whose roots go back to North Semitic writing, or even earlier, Cyrillic did not originate as a product of natural evolution. It was invented, custom-designed to record Slavic speech, and much more importantly, to spread the word of God. Yes, Cyrillic writing derived from Greek writing, but in the grand scheme of things this circumstance is rather collateral. At the time it just couldn’t have been based on any other contemporary script—even if several Cyrillic letters derived from Hebrew letters.

For quite a while—actually, for nine hundred years—Cyrillic letterforms evolved naturally, moving farther and farther away from their Greek progenitors. Following the universal pattern (Schrift kommt von Schreiben), the glyphs of the fonts used in early Cyrillic print, with few exceptions, had been fashioned after written letterforms (vyaz, poluustav). The Latinised forms of some West- and South-Slavic scribal hands, printing types (e.g., Bosančica) or monumental lettering had little bearing on the mainstream trends in Cyrillic typography.

A really big change took place in the beginning of the 18th century when Peter the Great decided to modernise, or to be more accurate, to establish, Russian secular printing. At his initiative a new printing type had been developed, vaguely based on the Western forms. The glyph set of the new, ‘civil’ type was different from the one used before: it was adapted for the typesetting in contemporary Russian, not in Church Slavonic. Since 1708–10 the development of Cyrillic typography has taken a totally new direction; the forms of the post-Petrine printing types, not only Russian, can be traced back to Peter’s Civil Type.

All things considered, the new, ‘westernised’ forms of Civil Type had little precedent in Russian writing, and leave an impression of an arbitrary, whimsical refurbishment. Peter’s typographic reform was indeed a triumph of the will. It took a while for the new letterforms to get broken in and integrated into a harmonious ensemble, and for the everyday writing to finally catch up with the reformed printing type.

There are reasons to suggest that in Cyrillic type the chirographic aspects take a back seat to invention and engineering. I find it significant that Peter’s grazhdanskii shrift had been consciously and purposefully designed to imitate the forms of printing type (one hundred years later, in a curiously similar way, the shapes of the newly-invented Cherokee letters had been informed by the forms of printing type, Caslon).

It looks like the familiar adage of Schrift kommt von Schreiben does not quite apply to post-1710 Cyrillic type. If Cyrillic did indeed come from writing, then it’s come a really long way; so long that its connection with writing has become hardly traceable. To a considerable extent contemporary Cyrillic presents a piece of conceptual design where the convention in the shaping of glyphs takes precedence over the aesthetics of writing.

quadibloc's picture

It's true that Peter the Great took his design for the Russian alphabet from type for the Latin alphabet. But I'm not at all comfortable with your conclusion that this means that the Cyrillic script is divorced from its connection with writing.

The reason I am not is that to all appearances, the Cyrillic script is quite thoroughly copied from the Latin script - and, thus, if its model, the Latin script, is not divorced from writing, then neither would the Cyrillic script be so divorced.

Russian handwriting and Russian italics also exist. But from Peter the Great's Bodoni-esque Cyrillic, one can directly proceed to Cyrillic Times Roman, Cyrillic Jenson Oldstyle, or Cyrillic hand printing. Nothing got so redefined or reinvented in the copying that there are any Russian letters for which one can't do anything one can do with a Latin letter.

This may be bad for the Cyrillic alphabet as an independent object of study, but Cyrillic type design and Cyrillic calligraphy should have no problems.

It's scripts like those for Armenian and Greek, which are under pressure to copy from Latin models, where those models are not perfectly suited to the nature of the script, but resources are lacking to do an equivalent indigenous development, that have problems.

Russian type designers are contributing new type designs to the pool available for the Latin script, so it is not as if the Cyrillic script area is reduced to copying; in fact, perhaps the best way of looking at this is to view the Cyrillic script area as a part of the Latin script area which just happens to be somewhat differentiated as to character set. Like German, which has the eszet that other Latin-alphabet languages do not, only a bit more different.

Maxim Zhukov's picture

Dear John, I did mention that after a while the new letterforms introduced by Peter the Great got broken in and integrated into a harmonious ensemble through careful fine-tuning. It is a fact, however, that while the construction pattern of Western italic, which was informed by handwriting, is consistent with the upright forms, post-Petrine Cyrillic italic came out thirty years later than roman, and was actually an afterthought. This can explain many puzzling and confusing inconsistencies in letterform construction of Cyrillic roman and italic:

Without doubt, Cyrillic type design is not “reduced to copying”, and designers do their very best to preserve the visual identity and the design integrity of multilingual typefaces across various scripts, with regard to the style category, and/or the aesthetics of the historical period, if/as applicable.

Cyrillic script is quite thoroughly copied from the Latin script - and, thus, if its model, the Latin script, is not divorced from writing, then neither would the Cyrillic script be so divorced.

Okay, let’s agree that modern Cyrillic type is not fully divorced from writing, but is related to it through Latin, so that cousinage is, say, once removed.

William Berkson's picture

This has been a running debate here on Typophile: to what extent is readability a matter of familiarity and to what extent the demands of the eye and brain? In the recently published Typography Papers 9 Peter Enneson and I have a new paper called 'Readability: Discovery and Disputation' about the research work of Matthew Luckiesh, done in the later 1930s in collaboration with Linotype.

Luckiesh found that blink rate increases more over time with difficult reading conditions, when other variables are controlled for. He used the term 'readability' to indicate ease of reading extended text. And he said that he research show that it is inverse to increase in blink rate over time.

Luckiesh found that in particular boldness of type is a variable that influences readability, with a medium weight being better than very heavy and light. —He tested four weights of the slab serif Memphis. Frutiger has also argued, independently, that there is an ideal black:white ratio. And, indeed, I think you see this across writing systems, even based on different instruments, contrasts, etc.

W.A. Dwiggins agreed with Luckiesh about there being an ideal, but added a number of other variables that he thought influenced readability. None of this has to do with the degree of differentiation of letter forms, which has for long been a focus of interest.

As to the issues of the relationship to the moving front of the pen, and contrast, I think this is complicated because the shapes traditionally drawn by the pen and brush I think are influenced by the demands of the eye. For example, the vertical emphasis of latin type and horizontal emphasis of Hebrew I think relate to the design of the letter shapes in their respective alphabets, and this in turn is a response to the demands of the eye and brain. So it's the letter design + contrast that is at work.

I'm hoping that Peter and I will have a presentation on this at TypeCon this summer. Meanwhile, you can rush out and buy Typography Papers 9 :)

quadibloc's picture

After my post, I felt I left one thing out. Since I said that Armenian and Greek have a problem, but they have the option open to them of Latinizing their typefaces, I shouldn't have said that Cyrillic has no problems. To the extent that Peter the Great replaced a valid indigenous model for Cyrillic script with one copied from the Latin script world, there was a loss in the cultural sphere.

I'm not sure, though, that such a loss did occur at this time, even though you and Hrant are agreed that one did. No doubt that is simply due to ignorance on my part, however. The reason I say this is that the difference between pre-Petrine and post-Petrine letter forms appears to me to be simply one of typeface style; the letter forms were put into Western dress, but they weren't altered.

Russian display typefaces based on pre-Petrine models exist to this day.

So the Russian alphabet hasn't been destroyed the way the Greek (lowercase) alphabet would be if someone were to Latinize Greek thoroughly by using Cyrillic as a model and the result were to be adopted in Greece to the exclusion of the previous form of printed Greek.

Cyrillic may have suffered a cultural blow, but because its present form is so close (not identical, as you point out) to the Latin, the kinds of problems I was saying it didn't have are typographical ones. If it can be done in Latin, it can be done in Cyrillic. (Well, small capitals might be confusing in Cyrillic, admittedly.) Without, as in the case of Greek and Armenian, paying a cultural price - because the price has already been paid.

Nick Shinn's picture

Sure, there would be an ideal—if everyone were the same.

William Berkson's picture

And individuals vary, but within limits. For example, optical illusions work pretty much for everyone. And that's an example of the real constraints, and their influence.

The ideals for readability I think have ranges, and they allow for endless variation and creativity. However, the constraints are real. Of course also ease of decoding is not always paramount, either. But in extended text, generally it is.

Maxim Zhukov's picture

So the Russian alphabet hasn't been destroyed the way the Greek (lowercase) alphabet would be if someone were to Latinize Greek thoroughly by using Cyrillic as a model and the result were to be adopted in Greece to the exclusion of the previous form of printed Greek.

The Russian alphabet has not been destroyed because it did not happen to exist before 1708–10. For written Russian Old Church Slavonic alphabet was used, which was as clumsy and inefficient as, or even worse than, using Greek alphabet to record Slavic before the invention of Cyrillic.

It is interesting to compare the typographic activities of two European monarchs, Louis XIV of France and Peter I (also “the Great”). The Commission Jaugeon appointed by Louis was just modifying the shapes of the typographic script whose letterforms were already quite well established—for H.M.’s own imprimerie. Peter was altering the very complement of the alphabet itself, for the whole Russian Empire—not only revising letter shapes but dropping the letters that were of no, or very little, use for printing books in the new secular Russian language of early 18th century.

In his book The Petrine Revolution in Russian Culture (p. 276) James Cracraft wrote:

Not only a newly streamlined, Europeanized Russian alphabet but a newly streamlined, Europeanized Russian language were among the outcomes of Peter’s drive to renovate his realm—here, its architecture—along contemporary European lines.

Just look at how impatiently, obsessively and adamantly did Peter cross out in that font proof of 29 January 1710 not only the glyphs that looked old-fashioned but also the characters he thought useless for his neue Typographie

quadibloc's picture

Well, I cannot fault Peter the Great for making it possible for the Russian people to write their own language in an alphabet suited to it.

But how closely the post-Petrine script was connected to the original Slavonic script, and what was lost, I will admit, I cannot judge.

Nick Shinn's picture

Bill, I don’t think you can compare readability to optical illusions.
The whole idea of readability is that there are optimal values for physical properties of type letter shapes.
I would compare it more to chairs or stairs.
There are best averages for the height of a chair seat, based on the average length of shinn bones, no doubt.
Similarly, steps are sized and shaped to fit adults, not children or those in wheelchairs.
Why would readability not follow the same principle?
i.e. what is ideal for one person is never ideal for another.

Damn standardization; the average is not ideal.

William Berkson's picture

I don't agree, I think optical illusions are very germane. They actually are part of the story—e.g. the desirability of overshoots. I think the ideals have to do with ratio of black to white, contrast, rhythm, relation of spacing counter, etc. The ideals allow for different writing systems, for a start. Some people have a hard time reading—dyslexia, visual impairment—but I think for those who can read fluently the variations aren't as big as in size of people.

Nick Shinn's picture

Sure, optical illusions are relevant, I wasn’t debating that.

But just because they fool most people doesn’t mean that there are ideal letter proportions.

Do you really think perfection is possible? —because that’s what ideal means!

quadibloc's picture

I think it may not be possible for us to find the "perfect" letter proportions or shapes. But if I was to say that they didn't exist, their nonexistence would have consequences for the consistency of typographic theory. Basically, I don't think mathematics would benefit by regarding existence proofs as invalid, and I apply the same reasoning to logical thinking in general.

Not that I'm excluding local minima, or even multiple characteristics of a typeface interacting in a nontransitive manner.

Nick Shinn's picture

…the consistency of typographic theory…

Don’t you mean reading theory?
Typography is a practice.

William Berkson's picture

Nick, I'm not talking about perfection, I'm talking about a range of, for example black to white ratios within which reading is most comfortable. Though, as Dwiggins said weight variable has to be "hitched" to others, and they all influence each other. Perhaps "optimal range" for reading comfort of text is a better terminology.

Syndicate content Syndicate content