Typophile - Rule or Law - Comments

And if by ’noise’ we

Eben Sorkin — Sat, 3 May 2008 15:41:06 -0700

And if by ’noise’ we include not only variation in letterforms or spacing but all those other factors that diminished the quality of so much printing then the manuscripts are by far the cleaner product.

If you get the objects themselves in front of you you will quickly see just how very true this is. And compared even to the vast majority of modern day printing this is still true. The level of detail possible is absolutely huge.

I don’t think even spacing and color forces a designer to iron out variation in shapes. Indeed variation in shape is often needed to achieve even color, because of the very different shapes of different letters.
Very true. When I talk to people about contextual alternatives in text they often bring up the fear that it might produce a textuta-like over regularity of form. But I don’t think that is remotely likely for the reasons similar to those that Bill gives above. A well tailored custom fit for ’Notan’ will produce increased clarity re: the differentiation between letter shapes. Like an adjustment, an overshoot for example you have to have some judgement about it to achieve that effect, but like Bill I think equating an search for even color with literal regularity is a false or at best a misleading idea.

Bill: I really wonder how

John Hudson — Sat, 20 Oct 2007 19:22:31 -0700

Bill: I really wonder how long the humanist script—with the lower case mimicking Roman capitals—would have lasted had not printing come along.

A good question. On the one hand, one can guess that a major change in script style might not have occured until it did in fact occur with the influence of a new writing tool, but prior to the invention of type the scripts of Europe were constantly evolving and subject to alterations in compression, proportion, etc., and of course there were important regional scripts that gradually disappeared with the standardisation of roman types. Noordzij turns Bill’s question on its head, wondering what might have happened if type had not frozen formal writing into those particular forms: “A stop of 500 years in the development of text scripts is long enough to make one look upon formal writing as static. Against a continuous tradition of 5000 years, however, the interruption has been merely incidental.’

And my impression is that even in formal documents, such as contracts, writing quickly became more cursive after the introduction of printing. Is this right?

In Italy, chancery writing tended to be cursive anyway, but also highly elaborated with swashes and ornamental devices: so cursive in form, but without the characteristic speed of writing we associate with that term. The humanist roman was specifically a book hand, intended for literary, theological and scientific texts. The quickly written cursive, without elaboration, which is the mother of italic type, was a kind of note hand, often found in manuscripts copied for their content by individual scholars rather than by professional scribes (such manuscripts seem to me the very model for Aldus’ small format books in italic type).

In England and some other parts of northern Europe, blackletter remained the norm for chancery documents for a long period, only to be replaced by the split nib scripts.

>writing and typography of

William Berkson — Sat, 20 Oct 2007 13:40:20 -0700

>writing and typography of other scripts.

Ever eager to test my hunches, I was looking over the Rabbi’s shoulder this morning as he was chanting the Torah portion today. I’ve got to admit that the script—a now rare instance of handwriting of extended text by a professional scribe—was admirably lucid and readable, as well as quite beautiful.

My feeling is that the problems of getting even color in Hebrew script are a lot less, because of it being a square script. The problem in latin script is the presence of diagonal and round strokes, and their connections to stems. And the problem is bigger in lower case humanist script than caps. When it is done well it is marvelously clear, as you have greater differentiation between letter shapes than in Hebrew or even roman caps.

I really wonder how long the humanist script—with the lower case mimicking Roman capitals—would have lasted had not printing come along. Did the Carolingian miniscule last very long the first time around? I have the feeling that it is quite difficult to write well and quickly. And my impression is that even in formal documents, such as contracts, writing quickly became more cursive after the introduction of printing. Is this right?

Nick: Firstly, type forms

John Hudson — Fri, 19 Oct 2007 21:43:36 -0700

Nick: Firstly, type forms have shorter extenders and are more compactly line-spaced.
Secondly, scribes introduce much irregularity with contextual forms and flourishes (even modest ones).

Yes and yes presuming we’re limiting the discussion to the Latin script. There are, of course, counter examples in the writing and typography of other scripts.

>I don’t believe that

William Berkson — Fri, 19 Oct 2007 13:16:29 -0700

>I don’t believe that readability is quite so one-dimensional.

I don’t either, as I tried to make clear above. I was just referring to where type has excelled. There are a lot of things that contribute to easy readability of a typeface. Also the merits of a highly readable face can be totally subverted by bad setting.

An example of another factor in readability is darkness of the face. In the testing of Clearview as I remember they found that a medium weight sans was better than a bold for legibility in signs. I think there is also a sweet spot (or range) for text as far as darkness. But darkness is complicated, as perceived darkness is not simply a matter of sheer density. Reabability is quite complicated, I think, and not that well understood. I am not talking here about the obvious, like bad handwriting that you can’t decipher to save your life, but rather readability at the high end—readable vs highly readable.

if by ’noise’ we include

Nick Shinn — Fri, 19 Oct 2007 12:31:25 -0700

if by ’noise’ we include not only variation in letterforms or spacing but all those other factors that diminished the quality of so much printing then the manuscripts are by far the cleaner product.

True, a good manuscript page is then cleaner than a bad printed page. But that kind of cleanliness isn’t really relevant to the superiority of the typographic effect as described by Bill. If the benefit of type is to present more text on a page, within the grasp of an eye’s scan, and less “noise”, i.e. irregularity, then type has the edge, for two reasons.

Firstly, type forms have shorter extenders and are more compactly line-spaced.
Secondly, scribes introduce much irregularity with contextual forms and flourishes (even modest ones).

Of course, scribes CAN produce stilted work which is as mechanically uniform and succint as type, but that is not their professional raison d’être vis-a-vis the printing trade. Their added value is in handmade grace, which goes against typographic readability-value as defined by Bill in mechanical terms of maximizing quantity and uniformity of text.

However, while accepting Bill’s explanation for the superior readability of type in one respect, I don’t believe that readability is quite so one-dimensional. It’s not something that can be reduced to scientific or mechanistic proportions. That’s fast food.

John, you have a good point

William Berkson — Fri, 19 Oct 2007 12:09:18 -0700

John, you have a good point about the messiness of early printing. In looking at some 18th century printing in Caslon I was struck by how much of a mess it is, and painful to read by comparison to printing post-Baskerville, when inks and press techniques were improved. Unfortunately, the ’modern’ high contrast types that became popular had their own problems with readability, in my view.

The benefits of the finer control that the designers introduced was only fully realized later. Still, the breakthroughs were made then.

Peter, I don’t think even spacing and color forces a designer to iron out variation in shapes. Indeed variation in shape is often needed to achieve even color, because of the very different shapes of different letters. For example, the tapering of diagonals prevents clotted joins, contributing to even color, but it also creates more lively, varied shapes in the alphabet. Accomplishing both liveliness and relatively little ’noise’ is what makes great text faces great.

Bill: I think there are a

John Hudson — Fri, 19 Oct 2007 11:49:36 -0700

Bill: I think there are a number of other factors involved, but that being able to get rid of noise is where type has the edge over handwriting. In other areas probably the best scribes were probably just as good for readability.

Again, I think we have to take text size into account here. Where type has the most obvious edge over handwriting is in being able to produce smaller text with the same regularity and evenness of colour and spacing as a good scribe could produce at typically larger sizes.* This is a much more obvious advantage of type than the notion of type getting rid of noise, which I find very problematic, especially if we’re comparing the textual culture of the first 100 years of printing from moveable metal type in Europe. The fact is that the vast majority of early printing was not very good: a lot of type was either over-inked or under-inked, the impressions were uneven (even on the same page), the metal alloys were not well developed and the type wore quickly, there is often a lot of extra-textual noise from ink spots. Even some of the most well-respected names in typographic history produced a lot of poorly printed books — I know, I own some of them —, not just the justly admired masterpieces that get reproduced in books on type history. By comparison, the scribes — seeking to compete in quality where they could not compete in quantity or economy — were doing their best work during the same period, using highly prepared parchments and papers, well-maintained reeds and quills, carefully prepared inks. And if by ’noise’ we include not only variation in letterforms or spacing but all those other factors that diminished the quality of so much printing then the manuscripts are by far the cleaner product.

[The preparation of the writing surface in the Islamic scribal tradition is particularly important, because the reed pen is pushed so often when writing Arabic, and the sized papers they use are generally smoother than anything produced for either manuscript or print in Europe. The cleanest, sharpest letters I have ever seen were handwritten Arabic script.]

* There are traditions of very small writing — micrography — in some scripts, notably Arabic and Hebrew. These are most often employed in decorative ways: you see a picture or a pattern, look at it more closely and you see that it contains a text in tiny letters. The Arabic term for such writing is ghubar — literally ’dust’ —, and a special style of writing was developed for it. The typical alif height of ghubar is less than 2mm; sometimes as small as 1.3mm. Because ghubar was sometimes used to write Qur’anic passages, legibility was, of course, extremely important.

Bill, if reading affordance

enne_son — Fri, 19 Oct 2007 05:13:36 -0700

Bill, if reading affordance is a plateau with thresholds and ranges and heights, I might even become compelled to concede that not only ranges and thresholds but also heights are susceptible to adjustment beyond writing.

But I’ll only accept ranking comparisons if they acknowledge: 1) that over time, type has probably re-tuned our internal maps so that we’ve become separated from the readability of writing to a degree; 2) (if they acknowledge) that through the broad-nibbed pen applied to a carolinian separated script, western writing had, by the time type arrived, carved out for its script a criterion-level or benchmark-level affordances plateau; 3) (if they acknowledge) that readability, or affordance in reading, is not a simple equation but a densely multifacted, delicately balanced thing.

The delicate balance thing is shown by the fact that removing noise in the optical-grammatical part of the equation might depress character in the gestural-atmospheric part of the equation. It might even be the case that some noise is necessary for the eyes to become and stay locked-on in visual terms to text, or for visual wordform resolution to stay robust, even though noise has a computational cost. Beyond a certain point, transparency might be a counter-productive chimera.

>Type may be made to be more

William Berkson — Thu, 18 Oct 2007 21:03:10 -0700

>Type may be made to be more noise free, but is that equivalent to more readable?

I think there are a number of other factors involved, but that being able to get rid of noise is where type has the edge over handwriting. In other areas probably the best scribes were probably just as good for readability.

Oh on the ’quirk notes’ idea of Hofstedter. Where we do have enough information to automate things is when you have two weights of a design that resemble each other in the designer’s eyes, then computer interpolation create other weights that are very ’family’ looking. So as of now it seems to take two fonts with similar characters with a different values of a parameter to actually have the needed information to carry through the ’quirk notes’ idea..

Bill, this probably needs a

enne_son — Thu, 18 Oct 2007 20:44:30 -0700

Bill, this probably needs a longer reply. But for now, think of noise first in terms of ’computation costs’ at the visual cortex level. Also think of ‘noise’ as productive in gestural-atmospheric terms. Noise is the basis of inflection, and inflection provides ’character’ and character has appeal — another factor operating in our everyday experiential anecdotal judgements of readability. I still think your ranking judgement is too easily made. Type may be made to be more noise free, but is that equivalent to more readable?

First of all, just to be

William Berkson — Thu, 18 Oct 2007 17:52:03 -0700

First of all, just to be clear, I am not disparaging the accomplishments of scribes. There were over 3000 years of scribes writing alphabetic scripts and now something over 500 years of type. The scribes already did a great deal to make writing readable, including not only letter proportions, but also key matters of layout, such as amount of letter and word spacing, line measure, space between lines, and margins. And these, as Giampa points out, are also keys to readability. The scribes had already worked out much of the rules of readable text that we still follow.

However, there is a remaining area of difference of opinion I have with John and Peter, unless I have won Peter over in a discussion off-line. I think that evenness of spacing and color provides a relatively neutral framework or grid, within which the differences between letters are easily detected by the eye. When color or spacing is uneven, or there are other extraneous variations, this constitutes ’noise’ that interferes with the detection of the ’signal’ of letter differences.

That is why I still think that good type, well set, is more readable than good writing. Good writing still has more inherent ’noise’ in it than good type, because of greater variability. The dimension of readability I have in mind here is how cognitively taxing it is to read text. This aspect of readability has not yet been tested, to my knowledge, but I have several times advocated such testing here on Typophile. I think that when good writing in familiar styles easily decoded by us—eg not Fraktur for most of us—is read it will prove more taxing than good printed text. And the more florid writing, with extraneous swashes etc, the more taxing it will be.

If this is a general psychological principle, we can reasonable apply it to readers in the 16th century, who didn’t have greatly different DNA than we do.

This is also not to say that writing is less beautiful. In the hands of a superb calligrapher, it can be more beautiful—though less serviceable for long passages of text.

Peter: So to me this says

John Hudson — Thu, 18 Oct 2007 16:52:01 -0700

Peter: So to me this says the greater feature-manipulational flexibility in punchcutting, could initiate a consolidation of the readability affordances of writing beyond writing...

This is better stated and doubtless more accurate than my attempt to get at this idea by looking at feature manipulation as an extension of the stroke logic inherited from writing. I like this idea of consolidation of a process that is already taking place in writing. It fits well with examples from a number of different scripts, in which one sees formal styles of writing develop that begin, sometimes in individual features or letters, to break with ’the single track of a tool’. [It is because of such instances that I don’t fully agree with a definition of writing that completely rules out occasional stroke repetition, build up or outline.]

...though not without a cost. The cost is growing distance from the readability and character of first-rate scribal writing, especially when you factor in the unfamiliarity of the foreign or earlier mother tongue.

This is an observable cost, and it also has parallels within typography as in e.g. the loss of ability of many people to read fraktur with any ease or comfort. I wrote earlier in this discussion that we could not say whether e.g. Griffo’s types were significantly more readable than the scribal book hand for people who were used to reading that hand, i.e. the contemporary audience for both kinds of books. Type could make consistently readable text at smaller sizes, and hence had an economical advantage, but at equivalent sizes I’m not convinced that there was any significant gain in readability. The ’consolidation of the readability affordances’ were expressed in a reduction in the size of text. I wonder if any records survive indicating a rise in the number of lense grinders working in Venice at the same time?

William: “Also, just to be

Giampa — Thu, 18 Oct 2007 08:40:47 -0700

William: “Also, just to be clear, I am not saying that all decisions that historical creators of type made were based on readability. What I am saying is that some of them have been, and those concerning even color and good rhythm resulted in type being more readable than handwriting.

And that is not because type creators they were any better workmen or artists than the scribes. It is because the medium offered greater control over outlines, and in a more limited way, spacing. And some superb type creators, such as Garamond, were able to take advantage of this greater control.”

As we all know, control over outlines does not always result in superior outlines. Look at the state of type design today.

Many think that type well set is of minor importance compared to type design in the determination of readability. I care to differ. I could take Garamond and make it very difficult to read even at 12 point printed crisply. No, I do not mean by removing anyone’s reading glasses.

Also, some pain from “unreadable type faces”, such as Bodoni, can be eased by careful setting.

Quickly handwriting became more of a “hangover and not so much a model for type design”.

Tools do affect outlines, not to mention right handed people have dominated left handed. Counter punches compared to gravures, the two schools of thought? Now the cold mechanical feeling of digital type. This could be a lengthy discussion but I am going to spare us all.

The church found Gutenberg’s Bible of minor importance compared to the lavish financial profits from the indulgences he printed. The business of printing dictated many different objectives. As I mentioned much earlier, Aldus Manutius invented the pocket book. There was great interest in making types readable in smaller sizes. This saved not only paper but allowed for more pages in the chase thus reducing press time.

Is handwriting more readable/legible, or even possible at six point than type? I think not! I predict, if it is possible to handwrite that small, (avoiding an embarrassing surprise) it must be hard to read at all. Besides I don’t think they made a goose feather small enough. For the record, I believe the smallest hand set type ever made was 2 1/2 point, probably commissioned by an unscrupulous lawyer.

You may have noticed in an eye examination letters are used. Well, they aren’t in Finland. Still, keep in mind the conspiracy between French eye glass manufactures and printers. They both had much to gain by smaller type sizes. Not to mention, the consumer was advantaged financially. Well, they all were.

I’m having a little

enne_son — Wed, 17 Oct 2007 20:03:51 -0700

I’m having a little trouble letting this thing sit just yet, so I’ll give it another go.

After some more reflection, here is what I see now through my ‘perceptual processing in reading’ lens. (It probably brings me closer to what you, Bill, feel to be the case, but this way of framing it eases my reluctance.)

I’ll start with this: Gerrit Noordzij sees the middle ages as the period of the consolidation of the word.

In writing with the broad-nibbed pen the consolidation of affordances approached a kind of stall-point or highpoint or limit around the time type made its entrance. In type the consolidation dynamic could continue apace concurrently or be productively resumed because of the less constrained feature manipulation it allowed. Feature manipulation overtly addresses optical-grammatical (‘look’) issues (like regularity, or evenness of colour) and gestural-atmospheric (‘feel’) wants (like, greater angularity or cursivity or varied stress), all the while — just under the surface of conscious awareness — re-mapping visual-wordform resolutional affordances, quietly adjusting its thresholds and ranges, and basis.

Because it re-tunes the basis, it requires a small-incremental reconditioning of it’s victims (a pick-up phrase originating in some other context from some of my conversations with Noordzij). But the re-tuning of the thresholds and ranges are real readability gains.

So to me this says the greater feature-manipulational flexibility in punchcutting, could initiate a consolidation of the readability affordances of writing beyond writing, though not without a cost. The cost is growing distance from the readability and character of first-rate scribal writing, especially when you factor in the unfamiliarity of the foreign or earlier mother tongue.

Rule or Law

enne_son — Thu, 20 Sep 2007 13:39:04 -0700

That’s the title of a “sharply critical” Gerrit Noordzij essay Robin Kinross has recently put up on his Hyphen Press site to accompany its recent realease of the Christopher Burke book on Jan Tschichold. You can find the essay here

As Kinross says, it “tells a large truth about how teaching can happen, and how learning can happen.” It probably also has something important to say about what typophiles might strive for.

Peter Enneson