Mean and Puritan

quadibloc's picture

In my opinion, although it is very early in its lifetime to say this, I suspect that the popularity of Times Roman bids fair to endure nearly as long as that of Caslon.

Of course, there was Stanley Morison's own famous comment on that font:

As a new face it should, by the grace of God and the art of man, have been broad and open, generous and ample; instead, by the vice of Mammon and the misery of the machine, it is bigoted and narrow, mean and puritan.

Yes, it has short descenders and a large x-height, but not so much so as Corona, for it was designed originally for use in a newspaper, albeit one printed with more care, and on better paper, than the typical North American newspaper.

An alternate version of Times with long descenders was offered for a while by one founder (I think Monotype itself); it was made to match regular Times in size, which meant that leading was obligatory, as it was still cast on the body of its nominal point size - the long descenders became vertical kerns. I don't know if it was ever digitized.

But my question is this: given the popularity - and evident merits - of Times Roman, why hasn't someone (or has someone, outside of my meagre knowledge) produced, as it were, a "Times Book" in which not only the descenders are lengthened, but so are the ascenders (slightly), with the lower case reduced slightly in size relative to the upper case so as to moderate the x-height to a "normal" value, and the stroke widths then also adjusted (made heavier in the lower case, or lighter in the upper case, or both, so that the ratio between them remains at the appropriate value)?

Times is a beautiful font, but if those characteristics of it which save on paper (the vice of Mammon perhaps becoming the virtue of wise stewardship of our forests) can fairly be recognized as a problem, why can't this problem be solved, so that the fullness of the beauty of Times may be appreciated - and the spirits of both Stanley Morison and Victor Lardent find peace?

hrant's picture

Being clinical about this spoils the case I want to make.

But clearly seeming clinical about this helps some people
maintain their veneer of expertise. Nick, move on? You mean
become like you? As long as there are people who want to cast
this craft as mere art, and use language as a tool of deceit to
do so, I will have to maintain this role. It's not self-righteous,
it's out of love for something that cannot defend itself.

--

Saying that Times and Minion have similar vertical proportions
is not OK, and reveals a deep confusion that should be cleared,
not least among people here still learning the basics. And writing
up a convoluted, illogical storm to distract people from realizing
all this is not acceptable to me - I cannot simply watch it happen.

hhp

nina's picture

I think one thing this thread has established is that existing terminologies and definitions are at the very least, confusing. A part of that is that there obviously is more than one factor/element to this complex equation, and we obviously can't change that. But maybe better words can help untangle this.

For example, how about distinguishing "vertical proportions" (inherently relative: how does say x-ht relate to asc height) from "vertical metrics" (which could be taken to mean x [or other elements] to the em)? Is this commonly done?
Maybe we could even find something like "absolute" vs. "relative x-height" to talk about, the former being compared to the em, the latter to the extenders (sorry Nick, I'm still not seeing how cap height could possibly be relevant here). Which would be related to Hrant's "Depth" term.

William Berkson's picture

>not seeing how cap height could possibly be relevant

Nina, Cap height is relevant because it affects how closely you can lead lines that have a lot of caps, such as in classified ads, notices of performances, etc, and anything with a lot of proper names. Also large caps may be more of an issue in setting German vs English. The height of numbers can also be relevant. You can get away with tighter leading with only ascenders because they are fewer. But if you also have a lot of caps, and they are unusually tall, such as Adobe Caslon, then it's more of a constraint.

All this is why I still think that the best thing is just to give the x-height as a percentage of the em. Then you can separately talk about other ratios: x-height to ascenders, caps, etc, as needed depending on the point you want to make.

Terminology is not right or wrong, it's convenient or inconvenient, clear or confusing. Defining x-height as relative to the em is certainly more clear, and that's a big advantage.

Speaking of 'absolute' x-height is confusing because in digital type there is no such thing. Speaking of 'relative x-height' raises the question: relative to what? You may want to discuss any number of relations--x-height to overshoot, to ascenders, to caps, etc., so to be clear you need to specify further.

Trying to restrict the meaning of 'relative x-height' to caps or ascenders or bounding box just adds needless confusion, it seems to me.

Also why not just "call it what it is". In digital type, the x-height is specified in em units, end of story. If you want to talk about other ratios, then say what ratio--it only takes two words to be clear--and go merrily along.

nina's picture

"Defining x-height as relative to the em is certainly more clear"
For the purpose of talking about how large the x-height will appear on the page at a given point size, yes. But for the purpose of discussing optimal x-height as a design feature of a font,* and its effects, what's relevant is how the x-ht compares to other factors.
* It changes the "nature" of a type design much less if every glyph is scaled to be proportionally bigger on the em (thereby changing "your" definition of x-ht) than if the extenders are lengthened or shortened (thereby changing the relative definition of x-ht to ascender).

And of course, in this thread, the point of why x-height came up in the first place was the latter: Relative x-height as an aspect of the design.

(And of course I realize cap height is more relevant for say German than English, but I still think its relevance to reading (and typesetting) pales in comparison to the ascenders in terms of sheer frequency.)

nina's picture

[sry double post]

Nick Shinn's picture

Hrant, this is about you:
You derogate the artistry in type design, because you are not an artist.
You misrepresent the culture of commercial typographers, because you have no practical experience in it.

So I'm duty-bound to protect the tradition of the type craft from people like you (sound familiar?)

However, apart from the peculiar circumstances of this one post, I don't believe that gives me the right to discredit your arguments with ad hominem attacks, and justify that by casting myself in a heroic role--which is what self-righteous means.

Nick Shinn's picture

Nina, about cap height, I wondered:

Which is the most consistent means of comparing the x-heights of different typefaces?

To answer this question, the x-heights of thirteen Regular fonts were measured:
Aldus, Bembo, Dolly, Electra, ITC Garamond, Gentium, Minion, Nimrod, Perpetua, Scala, Scotch Modern, Times, and Utopia. These were chosen to represent a variety of text styles, with variety of x-height-on-em, and variety of ascender overshoot above cap height.

Then the values of their x-heights were calculated three ways:
(1) % of em, (2) % of cap height, (3) % of ascender height.

For instance, Minion:
(1) 43.4, (2) 66.8, (3) 61.1
and Times:
(1) 44.7, (2) 67.6, (3) 64.4

The standard deviation amongst these three groups, for 13 fonts, was:
(1) 4.4, (2) 4.5, (3) 4.3

From this I conclude that the three methods have similar consistency.
In other words, a typeface with a small x-height according to one method will have a comparably small x-height in another method. For purposes of comparison it doesn't matter that the "% of cap height" method ignores the effect of ascender overshoot on x-height proportion.

But what of x-height appearances at specific point sizes?

Next the em-square percentage was used as a benchmark, assuming that this represents the "actual" height a font's x-height will appear at a given point size, and the difference was calculated between this value and each font's x-height as first (a) % of cap height, and then (b) % of ascender height.

So, the difference between "actual" x-height and (a), then (b),
for Minion: 23.4 and 16.7
for Times: 22.9 and 19.7
etc.

Next all values of (a) and (b) were totaled, and the standard deviation calculated:
For (a) 3.3, (b) 2.9

Very close, within half a percentage point. Again, I conclude that the amount a font's ascenders overshoot cap height has no bearing on whether one compares x-height as a percentage of cap height or of ascender height. To repeat: if a font has a small x-height in one method, it has a comparably small x-height in another.

The advantage of the cap height method is that X compared with x gives a stronger visual indication of cap-height line and median line than, say, l and x. The letter x was chosen because it suggests the median line.

hrant's picture

> The standard deviation amongst these three groups, for 13 fonts, was:
> (1) 4.4, (2) 4.5, (3) 4.3
> From this I conclude that the three methods have similar consistency.

That's an entirely unsound conclusion, because it doesn't account for which direction the deviation is for a given font.

Furthermore, why are you using Standard Deviation? Typographers choose fonts, and a given font is one unique case. It's almost like you're saying Times and ITC Garamond have similar vertical proportions!

> Next all values of (a) and (b) were totaled, and
> the standard deviation calculated:
> For (a) 3.3, (b) 2.9
> Very close, within half a percentage point.

?
This conclusion also makes no sense.

Anyway, there are more than 13 fonts out there. What you describe obviously does not work for fonts with short caps (which are not at all rare these days). No obfuscating math is needed to figure that out.

Why use an inferior method, even if you might get lucky with it?

hhp

Nick Shinn's picture

I'm not going to debate your comments, as you don't understand the premise of this analysis.

hrant's picture

I have a Math minor - specialization: Numerical Analysis. Not that most people need any such background to grasp that some rudimentary averaging* is useless when comparing individual fonts. You need only the most basic math. The difference in vertical proportions between Times and Minion is about 3%. That's 30 units in an Em of 1000. I don't know about you, but when I define my vertical proportions, I worry about units in the single digits. Furthermore, and this is just as significant, the difference in widths seems to be about double that. Ergo: the two fonts do not have similar proportions. And you need no math at all to see this - one simply needs to look.

* Hey, remember when you harped on Raph for using math
to understand fonts? Except his work was computationally
comprehensive and methodologically sound.

hhp

hrant's picture

But anyway all this is moot, right? Because with time we can
read text in FF Extra just as well as text in Minion, right?

--

Reprising something that was avoided earlier:
> it is now understood that he swiped the design from
> Starling Burgess and Victor Lardent did all the work.

Can you either retract this or back it up? Saying "it is now
understood" is a mischaracterization to the point of fallacy.

An apparent dislike of Morison (hey, I'm no big fan myself)
does not make it OK to be conveniently unfair to his legacy.

hhp

Nick Shinn's picture

If you're so math savvy, please deal with the subject. It's spelled out in the question at the beginning.

This is not about comparing two fonts using the same x-height definition, or by eye.
It's about calculating which system of x-height definition -- percentage of em, percentage of cap height, percentage of ascender height -- is the most consistent over a number of fonts.

Isn't that knowledge desirable in assessing systems of x-height definition?
Typographers have to deal with many potential candidates, as you stress.
If one system had significantly more or less standard deviation, then it could be inferred that there was a difference in accuracy. But the way it stands, for the most significant figures -- where the standard deviations are 3.3% and 2.9% -- there is only a .4% difference, which corresponds to 4 units of em.

What makes you think that increasing the sample, or changing its members, would impact the outcome significantly?
Bear in mind that there are already four short cap fonts in the sample: Aldus, Bembo, Dolly and Gentium.

*Re Raph, was that when he was scanning and blending letterpress outlines? Or when he was popquizzing which waveforms were more attractive? I questioned the aesthetic validity of both those. As for Fourier transforms, interesting stuff in theory, but again, better to rely on one's eyes to appreciate text colour.

As you are no doubt aware, I think the best way to set type is to try many fonts in a layout, play with size, leading etc., and work out what looks to be the best design solution. Not much theory there. I don't believe that by pursuing numeric analysis of x-heights I'm contradicting my preference for using taste as the arbiter of design decisions. This is pure research--whether x height is better defined in terms of cap height or x-height has no bearing on whether one chooses Times or Minion, and how much leading you give them, because one will try both, with various leading, and see which is most readable.

Nick Shinn's picture

Can you either retract this or back it up?

I'm convinced by Mike Parker's theory.
And I suspect it has a lot of currency, having been promoted by Font Bureau in its Starling typeface material.
But you're right, there is no way of knowing what the general consensus is.
As for Lardent doing all the work, that's well-known enough, isn't it?

hrant's picture

1) The data shows that Times and Minion do not have similar vertical proportions. Not among people who worry about single units in a space of 1000 (or sometimes even 2048).

2) Any analysis that goes against the basic realization that 5% of text cannot have the same relevance as 95% of text is flawed analysis. Specifically here, your conclusion that x-to-cap is just as good as x-to-ascender cannot be arrived at from any Standard Deviation calculation. The results of such calculations cannot be applied to gauge specific fonts. This is critical because typographers do not make decisions based on aggregate statistics, and rightly so. Typographers need to figure out, for example, whether Times and Minion have comparable proportions. If they look at x-to-cap to do this, they will be confused; their results will be inferior; they will be poor typographers.

> there is no way of knowing what the general consensus is.

In fact from what I've seen, most people
don't believe Parker's Burgess connection.

hhp

Nick Shinn's picture

Typographers need to figure out, for example, whether Times and Minion have comparable proportions. If they look at x-to-cap to do this, they will be confused;

Why? Both x-to-ascender and x-to-cap calculations show that these two faces have comparable proportions within the group of 13. In other words, their values are close to the mean, less than the standard deviation. (Although I didn't mention the mean values in my 4:01 post, it does show that the difference between these fonts is less than the standard deviation for the whole group.) This is not to say that the difference between them isn't significant, it's just not big. So that's one conclusion that can be drawn from the data, that these are two faces with "average" x-height, however measured.

But the purpose of the exercise wasn't head-to-head font comparisons, it was to analyze the quality of different methods of calculating x-height, as applied to a large set of fonts.

nina's picture

Nick (since you addressed that to me)

"It's about calculating which system of x-height definition — percentage of em, percentage of cap height, percentage of ascender height — is the most consistent over a number of fonts"

Well previously, the discussion was about which "system of x-height definition" makes the most sense. :-\
The idea that it statistically doesn't make a huge difference whether one chooses to compare x-height to cap or asc height, or to the em, doesn't nullify the relevance of the logical (and actually, practically relevant) difference in choosing one definition over the others.
And I'd agree with Hrant that a difference of 3%, in this discipline, really sounds like way too much for the results of this little investigation to be all that significant.

If you're so sure that x-height, per se, doesn't matter too much, just say so. Switching gears from jumbled logic to empirical analysis, and then again to broad statements of typographer-centric relativism, doesn't exactly enforce your point.

Nick Shinn's picture

I'm not making a point, I'm sharing the results of some statistical analysis, which I undertook out of curiosity, not to prove a point.
The results weren't what I expected.

And what's wrong with putting a fence between pure research and practice? Theory can too easily become dogma.

3% is a lot when you're designing a typeface, trying to hit the sweet spot. But not so much when you're comparing fonts.
Comparing Times and Minion, they are remarkably similar at the same spec, in terms of copy-fit and proportion. Times is a bit stronger, due to its thicker strokes and more contrasty fit, more so than its larger x-height (1.3% of em bigger). So to get a similar heft of text block, you would drop the Times point size, say from 12 to 11.8. The difference in ascender length doesn't come into play after adding a couple of points leading--it's such a negligible proportion of that amount.

I wouldn't say that x-height is unimportant, but for comparing average x-height types like these, where the difference in their x-height is so small however it's measured, it doesn't matter too much.

Nick Cooke's picture

Hey, why don't you two just acquire a pair of duelling pistols and have done with it?

Nick Cooke

hrant's picture

> Both x-to-ascender and x-to-cap calculations show that these
> two faces have comparable proportions within the group of 13.

Not really - that's a false conclusion.

> Times and Minion ... are remarkably similar

Nick, don't you realize this doesn't befit a type designer?

For one thing, if that were true, why would anybody buy your fonts?
It's telling that most people don't buy fonts; to them 3% is nothing.
But there are enough people that care about things that seem tiny for
you to stay in business. And there's a lot more than 3% difference
between staying in business and not.

hhp

Gary Lonergan's picture

Hope this is not too off topic

Times has lots of faults which have been dealt with extensively
elsewhere. But it can still look great

http://www.seedconference.com/seed.php

Nick Shinn's picture

Not really - that’s a false conclusion.

In this statistical comparison, it states that values within the standard deviation are comparable.
That's a true conclusion, based on the data.

Based on the same data, the four short cap fonts in the sample fall outside the standard deviation, as do the "news" fonts, Nimrod and Utopia.

Times and Minion ... are remarkably similar...

"...at the same spec...in terms of copy fit and proportion."

Why don't you include the whole sentence, or would that upset your argument?

My conclusion, after looking at the data, is that Minion and Times have x-heights that are close to the average, to the extent that x-height is a minor consideration when assessing their relative merits--other factors are far more important, such as the relative sharpness of Times' serifs, and the heaviness of the horizontally-stressed stroke on lower case "e " and "c".

Also, it seems to me that you are underestimating the significance of ascender overshoot with regards to cap height.
Certainly, capitals occur far less frequently than ascenders in U&lc text, but cap-lowercase sequences such as T_h and W_h play a huge role in determining the personality of a typeface, in English.

Nick, don’t you realize this doesn’t befit a type designer?

What would really befit a type designer is avoiding being dragged into the mud in a Typophile discussion with Hrant Papazian, which most have the good sense to do.

hrant's picture

> x-height is a minor consideration when assessing their relative merits

Maybe.
But their vertical proportions are not similar, to a good type designer.
And -going back to the origins of this discussion- vertical proportions
remain central to a proper evaluation of the bookishness of a typeface.

> personality

Entirely moot in this discussion.

But: Are you now saying that sequences of x-height and ascending letters
(like "fi") don't contribute much to a font's personality? Mud? More like
quicksand. Thrashing doesn't help.

BTW, it remains that you are the only person in my 12 years online
who has requested that I not participate in threads he has started.
And I honor that request.

hhp

Nick Shinn's picture

...vertical proportions remain central to a proper evaluation of the bookishness of a typeface.

I don't think so.
And that is a conclusion supported by my statistical analysis.
Using your preferred definition of x-height, classic book faces that I measured such as Aldus (56%), Bembo (59%), and Perpetua (56%) do indeed have low values, but there are other perfectly serviceable book faces with "normal" x-heights, such as Dolly (61%), Minion (61%) and Scala (61%). And this distinction is reinforced by the calculation of a Standard Deviation. How do you explain that?
What would constitute, in your opinion, a convincing sample group of text faces, and where would you draw the line between bookish and non-bookish?

I suspect that one's selection of text faces might skew the results, depending on whether one included faces such as Utopia (66%) in the sample group.

What do you make of ITC Century Book (68%)? It is not a classic book face, but it is named and published as a book face and many books are printed in it. Are they all garbage?

Are you now saying that sequences of x-height and ascending letters (like “fi”) don’t contribute much to a font’s personality?

I said that the combinations T_h and W_h contribute a lot to the personality of types. That's where ascender overshoot is most noticeable, and I assume that typographers use mental images of these combinations to help conceptualize, categorize and remember typefaces.

quadibloc's picture

> What do you make of ITC Century Book (68%)? It is not a
> classic book face, but it is named and published as a
> book face and many books are printed in it. Are they all
> garbage?

I don't know if the line between classic book faces and other faces is really drawn between Minion and Times Roman.

And I wouldn't call Della Robbia or Typo Upright book faces either, so I'll definitely agree that proportions relative to x-height can be dwarfed in importance by other factors (which usually aren't varied for non-display faces as much, though).

If I believe the statements

"Minion is a valid example of a classic book face."

and

"Times Roman varies sufficiently in its proportions from Minion that it does not quite qualify as a classic book face."

that does not imply that I have to think that books set in Times Roman, Utopia, or even Corona are "garbage".

Corona and Antique Olive can be quite readable. If someone sets a book in Corona - and at least remembers to use leading - I might be somewhat dismayed that he seemed to have used whatever happened to be lying around inside his Linotype at the time, but slightly substandard typography isn't going to make me sneer too loudly at a book if I find its content valuable.

People should care about typography. But you can care about something and still keep it in perspective.

hrant's picture

> Utopia (66%)

Tellingly, Utopia is a newspaper face.
Times BTW is ~68%. And Minion is ~62%.
So my initial measurement was way off,
in your favor however.

But anyway you don't believe vertical proportions are a big deal in
evaluating whether a font is fit for a book or not... This, BTW, even
though you wrote:
"Times has almost the same vertical proportions (cap to x-height ratio)
as Minion, a face that has become quite popular and respected for book
and text work in general."

> I don’t know if the line between classic book faces and
> other faces is really drawn between Minion and Times Roman.

Same here. In fact Minion is a bit too large on the body for that.
So it can work for a book at slightly smaller sizes, but it would
be more versatile (for book work, but tellingly not overall*) if it
were slightly narrower and had a smaller x-height.

* Noting that Adobe has historically liked to please the most people.

Gary, I also find Times formally attractive - like a nice building, or ship.
But to me that actually confirms one of my axioms: that a true text font
must necessarily look at least a little bit ungainly when viewed large.
This is because scale has a very large say in how letterforms look and work.
So in effect Times is too much of a display face to do the heavy lifting of
extended reading. Minion is far less so.

hhp

Nick Shinn's picture

Times BTW is ~68%. And Minion is ~62%.

Times: x-height 916, ascender height 1422, percentage: 64
Minion: x-height 434, ascender height 710, percentage: 61

Times: x-height 916, cap height 1356, percentage: 68
Minion: x-height 434, cap height 650, percentage: 67

But anyway you don’t believe vertical proportions are a big deal in
evaluating whether a font is fit for a book or not... This, BTW, even
though you wrote:
“Times has almost the same vertical proportions (cap to x-height ratio)
as Minion, a face that has become quite popular and respected for book
and text work in general.”

What's your point? I was pointing out that if one accepts Minion as a book face, then one should not discount Times as a book face because of its x-height.

hrant's picture

My calculations were based on your graphic of November 27. Was that flawed? BTW, I think your ascender numbers are not taking overshoot into account; so the absolute ascender numbers should be lowered (thus increasing the percentage difference). And more significantly the head serifs in Times (like at the top of the "el") are steeper than the ones in Minion, so the ascender value for Minion should be lowered less than the one for Times, thus (further) increasing the percentage difference.

Anyway, 3% is still significant (I'm dismissing the x-to-cap codswallop) especially when coupled to the width difference. I think in practice you secretly agree, because I really can't imagine that when determining your fonts' proportions once you come within 30 units you say: "Yeah, close enough, whatever."

The thing is, I don't grasp why you're nitpicking this point, since you've already stated that Times is really a more readable font than Minion -presumably for any kind of setting- simply because it's seen by readers much more frequently. Proportions, shroportions, right? BTW, do you tell your potential customers this sort of thing? I mean, you might be forgiven for not volunteering such insight when a juicy payment is imminent, but if somebody asks you point-blank "Is this as readable than Times?" I guess you just have to say: "Not even close! This font was just released last month. It would have to be the default MS Word font for two decades to become as readable as Times!"

> What’s your point?

My point is that you can't combine:
- Times is a book face because it shares vertical
proportions with Minion, an "uncontested"* book face,
AND
- vertical proportions are not central to bookishness.

You honestly don't see this?

* After all, the Mage of Canada used it for his Bible!

hhp

Nick Shinn's picture

My calculations were based on your graphic of November 27.

That's sloppy, working from a screen grab.
My calculations are derived from the digital font BCP coordinates.
Don't you have Times and Minion on your hard drive?

the head serifs in Times (like at the top of the “el”) are steeper than the ones in Minion

So, you admit another problem with the ascender method -- the need to make a subjective decision as to where the "optical extreme" of the terminal is.

since you’ve already stated that Times is really a more readable font than Minion

What makes you draw such bizarre conclusions?
Do you honestly think you can pass "Readability Relativists" off as absolutists?!

My point is that you can’t combine:
- Times is a book face because it shares vertical
proportions with Minion, an “uncontested”* book face,
AND
- vertical proportions are not central to bookishness.

Logic Hrant, logic.
If a and b are both members of set c, it doesn't follow that a = b!

hrant's picture

I trust your raw numbers - I don't need to dig up a copy of Minion.
But I do think that image of yours might be a bit off.

> the need to make a subjective decision

Is this Nick Shinn talking?!

Anyway it's not that subjective. And even if we would put the optical line a slightly different place, my point was that at the very least you can admit that the more slanted the top (like in Times) the lower the optical line is. This increases the difference between Times and Minion.

> What makes you draw such bizarre conclusions?

?
You say that Times is a fine book face simply because a lot of books have been set in it. So since far more books have been set in Times than in Minion and the vertical proportions either don't matter or are close enough anyway* then Times must be more readable.

* It's sort of stopped mattering what you actually believe...

The only way I can think of that that's not what you actually mean is if readability is essentially irrelevant. This it doesn't matter how easy/hard a text is to read. That the only important thing is how it looks! (But I guess I shouldn't put that past you.)

> If a and b are both members of set c, it doesn’t follow that a = b!

This is all quite delusional...

hhp

Nick Shinn's picture

You say that Times is a fine book face simply because a lot of books have been set in it.

You're putting words in my mouth. Again. Please stop.

What I actually said was, "...why you would consider the most-read book face ever to not be a “book face”.

And again,
"We’re talking about a face, designed for an upscale newspaper, that has become the most-used, most-read face in book publishing. It seems absurd to not consider it a book face, amongst its many other uses."

And again,
"I’m saying that Times is a book face through widespread usage as such..."

Did I ever qualify it as fine or good?

...it’s not that subjective.

...give or take 3%?

Since far more books have been set in Times than in Minion and the vertical proportions either don’t matter or are close enough anyway* then Times must be more readable.

Sorry, I don't subscribe to your notion that vertical proportions determine the readability of a typeface.
What proof have you of this?

It’s sort of stopped mattering what you actually believe...

Good!
I'm hoping to discover something from this thread, not bully across my prejudices.
At one point I stated that I suspected x-to-cap heights of a group of faces to be less consistent than x-to-ascender, then I did the math and concluded that isn't so. Don't concern yourself so much with trying to guess what I believe, or whether a recent post of mine contradicts a previous one, but try for new insights, such as your observation about the difficulty of accurately measuring x-height--even if you find you are undermining a previous idea you had held.

hrant's picture

As we've known, I concern myself with things that don't suit your agenda. Such as what "book face" might really mean - beyond some banal, vapid consumerism. Apparently to you technical concepts are only useful as tools of subterfuge; and when you are called on an untenable statement, are in fact incapable of "discovering" anything at all.

hhp

Nick Shinn's picture

...banal, vapid consumerism

Spoken like a true elitist.

Apparently to you technical concepts are only useful as tools of subterfuge...

Why do you feel threatened? Aren't you the proponent of reason over intuition?

...when you are called on an untenable statement

If you repeat that it's untenable often enough, do you think that people will believe you?

quadibloc's picture

My opinion of Times Roman is that it is a fine and good typeface.

But it would simply be wrong to say that it is an oldstyle face; it's transitional. And Bodoni is fine and good too, if suitable for fewer purposes than Times; it is neither oldstyle nor transitional, it is modern.

I think that Times Roman has been used to excellent effect in the typesetting of books. However, in the English language at least, phrases don't always mean the sum of their parts.

So when someone says that a typeface is a "book face", or, better yet, a "classical book face", being sufficiently naive that I don't expect that if I am not familiar with a technical term used in typography, then it isn't a technical term in typography... I assume that the speaker (or writer) is indeed using a technical term.

If Corona is a newspaper face, and Caslon, Bembo, and Baskerville are book faces, then I would suppose that Times lies somewhere in between. It certainly wouldn't be legible under the kinds of demanding printing conditions that made the resort to the newspaper legibility faces necessary. But its design was affected by having been aimed at an application in which legibility in small sizes was important.

I think it is useful and meaningful to have such a technical term, and that its existence doesn't mean that it's bad typography to use a typeface which isn't a "book face" to typeset a book.

There's even a simple definition of a "book face" which I can offer that illustrates why x-height is an important factor without having to define a new kind of x-height, or argue why ascenders are more important than cap height. A book face might simply be a face that a book may be well typeset in... without any leading.

Then everything becomes clear and simple. Yes, Times Roman has been used for the text of many beautiful books - but not unleaded. It can be used, and used well, in a book - but because that is not its native element, it needs a little help.

Nick Shinn's picture

What is the amount of leading appropriate for book typography?
Stanley Morison's didactic work on typography (First Principles, 1937?) is set in Bembo, with solid leading--with the result that ascenders and descenders frequently entangle visually, almost touching. Solid setting is not much used today. I believe that most book designers now give Bembo "a little help" in the leading department. The idea that text type needs extra leading is entrenched in the Paragraph defaults of Word, Quark, and InDesign, and in every font that doesn't have its TypoLineGap value set to zero.

In Scotch Modern, my revival of a mid 19th century Modern, I set TypoLineGap to zero, so that (many) web pages set solid in browser display. I think it looks cool and have no trouble reading it, but most people would think the readability is poor. Nonetheless, I think it's appropriate for such facsimile revivals.

(Discussion of leading probably warrants another thread.)

If ... Caslon, Bembo, and Baskerville are book faces,

But were they designed as such? That could be argued, as there was very little commercial or jobbing typography at the time. Nonetheless, Caslon in particular has proved extremely adaptable--for instance as a display face in 20th century print advertising, e.g. for Honda in the 1980s.

I think it is useful and meaningful to have such a technical term, and that its existence doesn’t mean that it’s bad typography to use a typeface which isn’t a “book face” to typeset a book.

I'm not so sure. A lot of books have been set in digital Bembo and Perpetua in recent years, because that was a safe bet, as they are classic faces used in letterpress book printing. However, the digital versions lack the heft of the older technology, with the result that there is a lot of anemic-looking book type out there. Not just an aesthetic problem, but a practical problem for boomers and other demographics with less than perfect vision. I would prefer a splodgy, overinked paperback in Times over a finely-printed work in Perpetua or Bembo, because unless I put on my best reading glasses, turn up the light bright, and hold the book at just the right distance, it's harder to read those classic faces in their toney settings.

...what “book face” might really mean...

Don't fall for that. What Hrant really means is that he would like to impose his narrow conservative taste on such matters as dogma, supported with "logical" arguments. Now that's agenda, as practiced politically by the American right.

quadibloc's picture

I don't know about Perpetua and Bembo being anemic, but I have often felt that way about Caslon.

I was just planning, shortly before seeing your post, to start a thread on leading - with a twist.

As for "narrow conservative taste", I would expect technical terms to have been in use for a long time, and thus to reflect well-established practice. So the fact that we call a font with attributes X a "book face" doesn't have to mean that we have to be so narrow and conservative as to believe it is only appropriate to set books in traditional book faces.

Naturally, letting one side control the meanings of words can be dangerous, but I don't think that Hrant is out to get us to view Times Roman as doubleplusungood.

Nick Shinn's picture

I have often felt that way about Caslon.

But not William's Caslon, I predict :-)

hrant's picture

> What Hrant really means is that he would
> like to impose his narrow conservative taste

What happened to this?
"You’re putting words in my mouth. Again. Please stop."

I'm really not very conservative.
But most of all, I am not a phony.

hhp

Nick Shinn's picture

What happened to this?

What's the matter, you can dish it out but you can't take it?

I’m really not very conservative.
But most of all, I am not a phony.

Somehow, those two statements don't go together.
But don't worry, it's not about you.

typerror's picture

Well Nick... you brought it down to the gutter by injecting (Edit: your) politics :-)

Michael

Syndicate content Syndicate content