> sacrifice elegance of letter form to get greater readiblity or friendliness

In the normal meanings of those terms: yes, almost exactly*. The "authentic" French style for example (like Garamond) is more elgant, but less legible.

* Although "legibility" here, not readability.

> Bringhust mentions that the "Baroque" types have an irregular stress.

Yes, like this:
http://www.themicrofoundry.com/other/odium3.gif
BTW, my initial analysis of Fleischman is that he was a readability demigod, and those "inconsistencies" were careful functional effects. Notice for example in that sample that the x-height is also inconsistent. Of course, that's the type of thing that people like Bringhurst would never accept... It's so unpoetic!

> I find attempts at classification like Bringhurst's a bit misleading

Yes, to some extent. Especially for "living" work.

hhp

"I am wondering whether part of this might be because it is more forgiving to irregular inking, pressure, wear and tear, etc. - following up on your 'workhorse' idea."

The only significant technical development that I am aware of that may be at play here would be the "improved" wooden common press that had been developed by the Dutch. It may have been more revealing in terms of presswork, and required or allowed a different approach, but this is speculation. However, prior to the art movement phenomenon of the late nineteenth and early twentieth century, technological advancements did tend to be of some significance in altering the type road.

>those "inconsistencies" were careful functional effects

Yes, Hrant, that's what I'm interested in - what functional effects, if any did they have in mind? You who know type history or who have studied these faces, do you have thoughts on a 'method in the madness'?

Did they have a specific concept of how variation in x-height or stress might help? Do any of you detect or have read of a system or theory behind it? Or did they just put one letter next to another, and weren't bothered by random variations?
[What is your GIF of, Hrant - old printing?]

> what functional effects, if any did they have in mind?

First of all, there's no "they". It's just Fleischman. And nobody before him or since has had the balls to do what he did (that I've seen).

Second, I'm no Fleischman expert - in fact the font in that sample (#65 in the Enschede inventory, and the most focused expression of his particular "style") was brought to my attention by Andy Crewdson. A number of people have made Fleischman revivals, but none of them have dared to incorporate his structural innovations. Even the most daring interpretations have restricted themselves to superficial aesthetic funkiness - that misses the whole point (except as yet another source to milk) in my mind.

Now, Fleischman was too far in the past to have benefited from the findings of Javal concerning how we read (and certainly not those of Bouma), so he couldn't have known any of the empirical things we know today. On the other hand, he could very well have been smart enough to figure things out -at least on a qualitative level- on his own. It really doesn't take much brains to realize that information arises from contrast, and the part of the lc "d" for example that conveys the most information is the ascender - for that reason you would make the body smaller (as he has). And you might give the "o" a different stress. And you might make your head serifs larger than the foot serifs. Things like that.

But of course I can't be too sure about what's in the head of a man who worked three centuries ago...

hhp

>the findings of Javal concerning how we read (and certainly not those of Bouma),

Where can I read Javal and Bouma?

Well, all I've seen of Emile Javal is indirect, referenced by pretty much anybody interested in reading. As for Herman Bouma, his research is readily available through books and papers, if you have access to a good library (especially one with ILL capability). Unfortunately though, Bouma is now retired, and he never distilled his findings into one neat package. For a superb single work, check out Taylor & Taylor's "The Psychology of Reading", 1983.

Mainly:
- Towards the end of the 19th century, Javal established that reading proceeds in saccades (jumps), not a flow.
- In the 1970s and 80s, Bouma clarified the ways in which we generally read word-shapes (or actually, letter-clusters), as opposed to complining words from recognized letters.

What this means for instance is that an x-height can indeed be too large for optimal readability (even though it might be great for seeing an individual letter).

Also: http://www.themicrofoundry.com/ss_read1.html

hhp

Hrant writes, 'In the 1970s and 80s, Bouma clarified the ways in which we generally read word-shapes (or actually, letter-clusters), as opposed to complining words from recognized letters.'

In light of which, this session at the upcoming ATypI conference will likely be of interest:

Kevin Larson. The psychology of word recognition.
There is a misperception by some in the typographic community that words are recognized by the outline that a word makes. Kevin Larson presents evidence, much of it from detailed measurements of eye movements while reading, demonstrating that we don

Make sure you bring at least one guy from the remaining 90% of reading researchers who would disagree heartily. BTW, this would include MS's own Bill Hill, who [re]invented ClearType based on the bouma model. Note how small Berling's x-height is, for example. But really, if you want a [non-retired] guy who really knows his stuff, try to get O'Reagan. I have to say that giving only the minority view to ATypI attendees does a deep disservice to the craft. The last thing we need now is a return to obese x-heights!

If Larson is basing his conclusions on measurements concerning reading on-screen, then I would simply point out that speeds are much lower there, exactly because it's a lot less immersive! It's like that joke where the researcher concludes that when you cut off somebody's legs, he loses his sense of hearing, because he no longer obeys the "walk" command... But maybe Larson's view is currently convenient for MS... :-/

Although I've done no emprical research myself, the truck-load of research that I've read over the past five years leaves no real credibility for the "letter-compiling" model. Sorry.

hhp

Kevin Larson was hired by Bill Hill, so I would be very surprised to find their views so at odds as you suggest.

I think you are, typically, reacting too quickly to a blurb for Larson's presentation. Why not wait until you actually know and understand what he is going to present and have seen the evidence on which he is basing his conclusions? I suspect, even in reading his blurb, you are probably missing the subtlety implied in what he says. It seems to me that he is looking at the question 'How do we recognise words?' Which is not the same thing as questioning whether we recognise words and groups of letters as the primary units of reading. He is investigating how we recognise those primary units and his evidence suggests that it is not by the outline of those units but by their components. I don't know any more than you do what the details of this presentation will be, so why don't we wait and see? I simply brought the presentation to your attention because it is relevant to the discussion that has evolved in this thread. I'll be perfectly happy to discuss it in detail after I've heard it. One thing I'm pretty sure of though is that Larson is not advocating a return to massive x-heights.

You're right - I'm jumping a bit. One reason is that I really wanted to come to Vancouver, but I won't be able to make it, unfortunately - so this might be my only chance to disagree! I did two conferences last year, and just came back from an overseas vacation, so this year will have to be conference-free for me. :-( Best of luck with it, anyway.

Just be careful of people whose job might depend on what they advocate... On the other hand, maybe I should be happy people are thinking more analytically; really, anything is better than the Familiarity escape clause used for the past 15 years to shirk responsability, both by the PoMo Hooligans and their arch-enemies, the Chirography Grovelers. It's funny - they remind me of the villages of Homs and Hama in Syria...

hhp

I have studied a bit more the Adobe Caslon vs metal Linotype Caslon Old Face, composing in in 11/13 Adobe Caslon an identical paragraph as in the book I have. What is more noticable than any difference in letter forms is: a. greater thick-thin contrast; b. more apparent irregularity in x height; and c., the most noticeable, there is less space between letters and words in Adobe Caslon than the Linotype version.

Has spacing tightened up in digital type generally? Is this affecting readability? If the shapes of individual letters are perceived, as the announced talk seems to suggest, then having looser spacing might help ease of reading.

William

Adobe would have attempted more harmony, less contrast due to the many variabilities in digital output devices. Linotype only had to be concerned with their own linecasting machine. I've used Adobe Caslon letterpress and it worked out well, but it just seemed to lack the character of metal type.

Is the Linotype variation on the x height due to sizing changes or baseline shifts? The latter were quite characteristic of smaller text sizes in machine composition, both Linotype and Monotype. Never did find out the rationale for this, whether it is a concern for readability or a technical situation.

There would have been a natural restriction as to how tight a Linotype could space due to the natural boundary of the metal matrix. Adobe didn't have this restriction; most digital type seems a bit tight. Problematic when you go to letterpress with it as you risk "ink bridges" (between very close characters) developing during the editioning.

>lack the character of metal type.
Thanks, that confirms my suspicion that there is a design issue here, not just different printing methods. I agree that ACaslon is excellent, but it just doesn't have the spirit of the metal.

>variation on the x height due to sizing changes or baseline shifts
There seem to be both mechanical and design issues going on. When I get it figured out, I might try to do a trial face with it.

Looking at laser printouts of the PDFs of Founders Caslon 12 from Fontexplorer.com, it (at 12/14)does capture the 'softness' of the metal, but I don't think gets the same overall effect.

>most digital type seems a bit tight
Ah ha! My unease with something in Minion (pre-open type version) was what brought me to post a question at this site. Now I'm thinking that the problem might be slightly too-tight spacing. But I'm wondering about this in general with digital type. I remember that when photo type came in, my late Uncle J. Ben Lieberman, author of 'Types of Typefaces' (mentioned above) hated the 'close but not touching' fashion that had become possible, and which almost all ads used. He thought that excessive tightness was messing up the shapes of the words and readability.

Thank you Gerald for your very helpful comments!

> the problem might be slightly too-tight spacing.

Note that the actual point size used has
great bearing on optimal letterspacing.

> 'close but not touching'

Which was the second worse thing about the 70s,
after obese x-heights. Oh, and disco. And yes,
I know it's no longer fashionable to hate disco.

hhp

gn

Significant changes in technology tend to allow for different "creative" approaches. If the Linotype technology restricted the ability to set characters tightly and the photofilm technology of the 60s, 70s, and 80s allowed one to set extremely tight; these approaches need to be understood as historical developments and not taken out of context. Digital technology follows a different path and yields a different aesthetic. If work from the 70s appears "fresh and exciting today" that is simply because digital technology allows us to mine the past. This is less a retro thing than the result of technological change; we've seen these before: with the introduction of wood type and the changing of the way type foundries marketed their wares in the nineteenth century, with the introduction of machine composition and the revival movement in the early twentieth century, with photofilm at mid-century.... typefaces from different times and different places proffered up together. Part of the confusion here about Caslon is that the various permutations of it are being taken out of historical context. You really can't qualitatively compare a letterpress designed typeface and a digitally designed typeface, for instance, without a consideration of how they were to be used or are to be used.

Good point(s), but:

> these approaches need to be understood as historical developments and not taken out of context.

That's only half the story, the "Art" side.
Since typography and type design have an essentially functionalist foundation (to me), there's also the issue of how people read, and deep down that only changes when the species mutates! So very very slowly...

One can only hope that when somebody is setting [text] type, he's more minding the subvisible, functional stuff, and less trying to express some ephemeral personal cultural preference. Specifically, too-tight spacing destroys the balance of black and white space essential to making out boumas.

hhp

Isn't a bouma one of those double humped camels? :-)

No, but come to think of it, a bactrian does make a much better bouma than a dromedary!

hhp

>actual point size used has
great bearing on optimal letterspacing
Hrant,do you think that in text size, 10-12 point, the ideal change in spacing is sufficiently non-linear that this should be taken into account?

Gerald, one more question about ink spead in letter press. Is this primarily due to the roller getting ink slightly on the sides of the face, and then getting picked up when pressed into the paper, or is ink actually squeezed out from the face? The reason I ask is that if it is only the first, then I would expect the spread to be even around the face, whereas if the second, then the thick strokes would have more ink spread than the thins.

> do you think that in text size, 10-12 point, the ideal change in spacing is sufficiently non-linear

I think there's a difference in optimal letterspacing between 10 and 12, but you could fudge an average in that range any few would mind.

BTW, InDesign's optical spacing algorithm is bi-linear: no change below 4 point, a straight line between 4 and 12, a separate straight line between 12 and 72, and then no change above that.

hhp

William

To some extent you are encountering both. Each time you pull a print you are taking ink off the form but not all of it. Increased impression beyond what is adequate for the inking of the form, incorrect roller height adjustment, natural variability in ink charging, accumulation of ink deposit on rollers and type, all contribute to the problem. Differring paper surfaces will contribute widely to the problem as well.

Fine printers might clean the type form of accumulated ink fairly regularly during an edition run but this would be a rare occurrence in commercial work. Unattended, the problem will increase during the run. Since the ink is quite viscous it will draw other ink to it, so any ink that begins to accumulate outside the definition (outline) of the letterform will increase. Sometimes this ink is being pushed there by roller action, other times by the incorrect application of ink by the operator (too much ink for the form) or a combination of both. The ink is clinging not only to the edge of letterform but also is drooping somewhat down the bevel. Once this happens there is little recourse except to clean it off. Though metal typefaces often have ink drains built into the bevel, this would not really help that much with sloppy presswork.

Specifically, if the rollers are incorrectly low they will push down too hard on the type and add ink to the outline but this might not be uniform as it would have to do with the relative weight of the letterform (serifs might take the brunt of it), the actual configuration of the form (the lone edges of longer lines or isolated words might be hit harder), as well as the rolling direction of the rollers (vertical vs horizontal movement causing a distorted accumulation).

Did this address your question?

Gerald, what I gather is that the distortions from ink spread can be very inconsistent, so that you can't just say it happens one way. Thanks.

William

Yeah, that's kind of the short of it.

William

One last point to this perhpas. I recently scanned an obscure metal typeface with the intent of digitizing it and in the process I have noticed something that might not have been dealt with.

Chank or someone has a tutorial on typeface design that essentially reveals that pieces of letterform constructs are used over and over again to form the composites of the various letterforms. A time saving device that I assume, as I get into this now, must be a fairly standard practice in digital typeface design.

Problem is, this metal typeface, which has some pedigree, doesn't follow that technique. Every character has a bit of slight alteration to it. No cut and paste technique available then I guess.

So, maybe that is another reason why you see a nice variability to older metal based typefaces that you do not in digital revivals???

Gerald

Just as it makes the manufacture of fonts much easier on the whole, digital design is a particular facilitator of copy-paste modularity. But whether in metal or bezier, this is a Bad Thing, at least in the realm of text faces. Literal modularity is anathema to readability.

hhp

>Every character has a bit of slight alteration to it. No cut and paste technique available then I guess.

I also observed this in my scan of metal Linotype Caslon Oldface. I think this is unquestionably one of the factors in the different look of metal and digital.

I think there is some valid analogy here to a tune in music. I read or heard somewhere that music has to be something like 50% repetition and 50% variation. (I may have the percentages wrong.) The idea is that without the repetition, there is just a string of notes, and no tune. With too much repetition, it becomes boring like a TV jingle.

I think that digital type - and you interestingly identify an source in 'cut and paste' - may have become too repetitive.

The challenge will be to have the greater variation and still keep the unity and balance of the design.

William

If you are looking at the same Linotype Caslon I am it is a great cutting. However, have you viewed the italic?

No so great!

I seem to have missed much while I was moving and not online. Don't want to change the subject but have you seen our new page.

If you don't like ornaments don't bother!
http://www.lanstontype.com/LanstonPriceList.html

Gerald speaks of many ink aspects of type. There are, as he says, many conditions ready to alter a type face. The ink on the first impressions will not be the same in the middle of the run, the ink on the end of a run once again much different. Worn type, battered type, unclean type, linty papers, faulty make readys, soft packing etc.

Lets not forget work ups. Not only spaces did that!

Roller syndrome, sheeps foot, winter and summer gelatin rollers.

But no mention yet of "type stretch". I hate to toss this into the equation. Did you know on a cylinders press you can measure the type form and then the printed text. The text may be a pica bigger??? Plays hell on close registration!

A properly imposed hand set or monotyped page will tend to be slighty taller in proportions. Linotype properly imposed slightly wider?

I do not agree that printers did not care about how a type face looked. A good craftsman, and there were many, cared a great deal how a typeface looked.

I hired many printers and found them to be an extra-ordinary learned lot.

Even in the work horse environment.



Gerald

>If you are looking at the same Linotype Caslon I am it is a great cutting. However, have you viewed the italic? No so great!

If it is Linotype Caslon Old Style, we probably are looking at the same thing. I also don't like the italic, which is hard to read. As I am still trying to figure out what exactly is causing the look I like so much. As you like it too, what do you think it is? How would you compare it to other Caslons, including your 337?

I wish you had some alphabets & PDFs on your web site!

> 'close but not touching'

>Which was the second worse thing about the 70s

You must be joking Hrant!

There was some great work done by Herb Lubalin, Seymour Chwast and others in the 70s employing 'close but not touching' that still looks fresh and exciting today.
Many times they touched and overlapped just for good measure.

I feel there was a more creative approach taken to headlines in the 70s, than what is being composed on computers today.

gn