These issues might be caused by poorly designed (usually too small) small caps. They should be less intruding, yes, but when they look plain wrong with a genetive s or if you introduce a capitalised word in the sentence it's nothing but bad design.Some type designers introduce mid caps (3/4 height) as a solution. I do see the relevance of just-above-x small caps in book design, but once you embark on more editorial stuff they're rarely useful.

You are a bit quick with your judgement "it's nothing but bad design", especially when what you criticise is standard practice and there are good reasons for doing so, as you indicate right in the next sentence "I do see the relevance of just-above-x small caps in book design". After all, this is a special case.

Maybe in this case set "DIA" in uppercase but at slightly smaller size, if it stands out a little then be it so. Or, if the author/editor allows and unless "DIA Inc." is the official abbreviation, omit the "Inc." from the abbreviation and use small caps as you would do with NASA.

(Wasn't there an earlier discussion on this with quite good suggestions? Cannot find it.)

I said "when they look plain wrong". If that is standard practice, then the standard needs some reconsidering. Especially for editorial design. This and related issues are not uncommon: Cap vs. SC, SC vs genetive s, SC sticking out when its task is the exact opposite are all potential caveeats.

That said: The designer should, as Karsten said, consider breaking the typography rules when they don't work.

Doesn't Bringhurst have something to say on this in his book?

I agree that any caps that are not visibly taller than the x-height are wrong. Just because it's a tradition does not make it a good idea.

setting them in full size caps makes them really jump off the page.

Depends on the font of course. For something like a novel it's better to use a font with modest caps. It's mostly in reference books (where you have to scan for a proper noun or a date*) that large caps are more effective, and in those cases it's useful to have midcaps for extreme cases, like acronyms of three -or maybe four- letters or more - you probably never want to set "FYROM" in full-size caps. :-)

* Where additionally the numerals should be cap-height (or at least large).

There was a recent thread strongly related to this. But lately even Google can't seem to crack Typophile...

hhp

Thanks for all the feedback. Who needs Bringhurst when you have Typophile! (Kidding; Bringhurst doesn't really address the issue specifically in Elements, therefore the reason for this thread.)

The font is Celeste Pro, so it is certainly no slouch typeface. Its small caps are larger than x-height, though only slightly. Most of the time I really appreciate the subtle qualities of its small caps. It is only rarely (<1/year) that I run into an issue like this, but I've seen it for a few years running now, so I thought I'd ask. The project currently is a newsletter, though I've run into the issue with Annual Reports as well.

I ended up setting "DIA" in small caps, leaving "Inc." capitalized, and scaling up the small caps a bit as a make-do (116% vertically, 108% horizontally). I know it deforms the characters a bit, but I don't have the budget nor time to add a proper 3/4 cap to the font (not to mention getting permission).

Scaling the full caps down to 3/4 caps proportions and setting "DIA" in full caps at a smaller size both left the characters a bit too thin. Scaling small caps up left them a little dark, but much less noticeable than a reduced full cap, which is too light. This also lets me make the change consistent across the document.

A recent discussion:
http://typophile.com/node/89876

frode frank, I agree with you. If the standard practice looks wrong, it's time for a change.

In all of these examples, I would use all small caps. There's nothing wrong with starting a sentence with a small cap. Using a full cap would look stranger, I think. "DIA Inc." With DIA in small caps would look fine to me.

Hrant — As a book designer, I am bothered by small caps that are visibly taller than x-height, and usually consider this a serious reason not to use the font. I will sometimes shrink the small caps by a point to get them down to x-height. There are rarely situations where this is visually or logically ambiguous. If a genitive s is pretty much indistinguishable from a small cap s, that's fine with me.

(Well, most of the time.)

I am bothered by small caps that are visibly taller than x-height

Can you identify why?

hhp

Very good question. I don't know the answer.

But here are some thoughts: I think it's particularly troubling in combination with oldstyle numbers, where I would expect the size of the small caps to match the x-height of the numerals. Since abbreviations and numbers often go hand in hand — 9:30 PM, 640 BCE — this slight difference in height can look haphazard. It might be a similar situation with NASA's or UFOs — I feel like the slight height difference is noticeable enough to be distracting, but not large enough to seem deliberate.

I will say, though, that I do not always use small caps for abbreviations. For each individual project I'll choose to use small caps for some or all or none of the following:

acronyms — UNESCO
initialisms — PTA
abbreviations with periods — U.S.A.
Mixed case — Ph.D. (actually I don't think I would ever used small caps here)
"All caps" for stylistic purposes — I said, STOP!
Headings and editorial emphasis
etc

And which of these I employ will depend mostly on the feel and aesthetic of the piece, and where it falls on the continuum of literary to journalistic/academic. It also depends on what the small caps look like in the particular font.

In some contexts, particularly academic, small caps that are not very small may actually be the least obtrusive, as they more closely resemble what the reader expects to see. In more literary contexts, I like them as small as possible.

Jeeze. This one again. I can think of 1000s of examples where the use of small caps in place of full caps looks awkward. Give me $5.00 each & I'll list them & be able to retire. Old favorites: MoMA. GMbH. The G.I. Bill.

As you've poionted out, any short abbreviation that is part of a word pair where one of the words is a capped proper noun with the rest lower cased letters, (DIA Inc, the G.I. Bill, etc.) Or, any mixed upper & lower case abbreviation/acronym with c, o, s, u, v, or w in them. Or, any mixed upper & lower case abbreviation/acronym with a lower-case letter that has an ascender, or even a descender.

Here's the deal. No matter what you do, there will be situations where either choice will look awkward. You have to make a choice which awkwardness to accept. The absolute wrong thing to do is to blindly follow a rule that doesn't have general acceptance, or appeal to an authority (particularly a disputed one) who represents about 5% of the audience.

How about this. Set a longish text and change all the naturally occurring full caps to small caps. Take a look, do you like it? I'd guess there is a certain percentage of people who would say "yes." What can one say to them? Recommend they take up one of the CJK languages?

Nice post. Thanks!

And Mr Keyser's solution seems to be a good one.

and scaling up the small caps a bit as a make-do (116% vertically, 108% horizontally).

Yes, we use non-proportional scaling to get good (i.e., effective) small caps run into text a fair bit. The problems with doing this in InDesign are two. The first is you can't scale up a letter without also changing it's weight. To my eye, 107% is about as far as you can go. Obviously, it depends on the letter and the font. Since we do this globally as a part of a character style for run-in small caps, "depending on the letter" drops out -- they're all affected.

The second one is a bit more subtle: If there is any kerning, it will be blocked between a scaled and unscaled letter. For example, if you have a scaled-up small cap D and a program kern between D.sc and quoteright, that kerning will be blocked. IIRC, you *can* put it in by hand...

It is for such reasons I like to make the changes to the characters in the font rather than in the applications program. Done in the font, the weight increase can be adjusted, and kerning will remain programmatic. BTW, I have absolutely no problem with a non-proportional scaling of the small caps, depending on how they were originally drawn.

The absolute wrong thing to do is to blindly follow a rule that doesn't have general acceptance

If you're blind, certainly do follow the masses. If you can see, think for yourself.

Set a longish text and change all the naturally occurring full caps to small caps. Take a look, do you like it?

You mean even just single caps for first sentence words and proper nouns? I've never heard anybody [here] advocate that. To me that might only make sense if the font has defectively large caps.

hhp

You mean even just single caps for first sentence words and proper nouns? I've never heard anybody [here] advocate that.

I'm advocating it as an experiment. It will show you how much you can tolerate with small capping.

It mirrors a friend of mine's experiment, which involved taking out all the word spaces and finding he could still read the text. That was suppose to be evidence that very tight setting was a good thing.

* * *

As the man said, "We've already determined what you are, we're just haggling over price."

Slow down & read both posts, Hrant.

I think it's particularly troubling in combination with oldstyle numbers, where I would expect the size of the small caps to match the x-height of the numerals.

This trouble might be better solved with oldstyle numbers with a larger "x"-height than with small caps lining at the alphabetical meanline?

Same debate between Joe Clark and Aegir Hallmundur.

I personally have to agree with Joe here. Every time I see small cap acronyms in a book I get excited like a child on Christmas morning, but they don't actually make things easier to read, particularly in the situations you brought up. Scaling down and tracking out full caps just a tiny bit works fine for me.

You can just forget about the bullshit elite prescriptivist notion that acronyms have to be set in small caps. They don’t. And in fact, they look ridiculous 99% of the time. Then you bump into the inevitable and not even very uncommon edge cases, like postal codes (M5W 1E6) or acronyms that include numbers or punctuation (A380, 3GS, GT-R).

Just type the damned capital letters.

Your solution for “NASA” is particularly half-assed. Even small-caps apologists would lampoon that one, and rightly.

Yeah, and just use damned Times...

Joe, you can't ignore the font at hand (which often one is not allowed to change). If its caps are too large your advice* will often produce lousy results (except if the smallcaps are even more messed up**). That's why many people prefer smallish full caps*** but sometimes (like "FYROM") even those will be too large. Good typography is never "Whatever".

* Which I would equate to saying something like "just set the text at 12 point".

** Which I admit is too often indeed the case.

*** And/or largish smallcaps.

Also, Craig's suggestion of also choosing the numeral style carefully is solid (and frankly should be obvious).

hhp

emkeyser – ... at the beginning of a sentence. ("NASA plans to launch another shuttle in 2020.") For these instances, I will usually set the first letter as a full cap with the rest as small caps.
k.l. – ... and use small caps as you would do with NASA.

I missed the "set the first letter as a full cap" which in fact better be either all caps or all small caps.

I have seen acronyms set in small caps, primarily in books from the United Kingdom. In the United States, however, this is definitely not standard practice.

I don't dispute that this makes acronyms look less jarring, compared to the use of ordinary full-sized caps. But because it isn't normal practice, one shouldn't be surprised that some typefaces won't be designed to facilitate this practice - and so the small caps might not be tall enough to work well in the particular special cases outlined.

Along with mid-caps, another way to avoid capital letters being quite so distracting might be to use a typeface in which the capital letters are somewhat condensed.

Joe - Your lengthy response to Aegir is full of passion but ignores what is usually the problem: using text fonts with small caps that are too small (at x-height).

Stephen, that the typeface in question has too small small caps, as Frode and you claim, is not "the real problem":

1) The typeface's small caps are well sized for book typography, i.e. setting author names in cap plus small caps. Typeface is FF Celeste. I trust that Dr Burke, as scholar-designer, can design a book typeface. I trust that FontShop only adds quality typefaces to the FF library. Funny that you recommend to "buy new type".

2) It is the typographer who chooses and uses a typeface. If the typographer's decisions contradict each other – text having mixed caps/lowercase abbreviation / using small caps / putting mixed abbreviations into the small caps category / using typeface with pretty tall caps – then the sum of these amount to asking for trouble. The typeface's fault?

Alternatives: Choose type with larger small caps. Use caps at reduced size. Use caps. Choose type with smaller caps, then you do not need small caps for abbreviations (mixed or not). There are many possible problems – and solutions. Some involve relaxing the "use small caps" dictum which leads to ...

Mr Clark's blog post is enlightening in that it deals with an education and followership phenomenon by way of a typographic detail: Typographers learn that good typography means adhering to some Bringhurst- or Tschichold-compiled checklist of rules and later continue to adhere to them strictly. (Sometimes not only to make good typography but also to demonstrate that they know what good typography is, or putting it the other way round, in fear to look like an ignorant when not doing so.) I read the post not as being against Bringhurst or Tschichold but against following any one or textbook too strictly.

Sure, small cap height is a design decision and one size doesn't fit all needs. I just happen to prefer small caps slightly larger than x-height for a majority of the work I do, and that happens to be a reasonable solution to the common problem we're discussing.

There are many possible problems – and solutions.

I'm with you. I'm just adding one more solution to the pot. My tweet sounds more absolute than it should have been.

I read the post not as being against Bringhurst or Tschichold but against following any one or textbook too strictly.

That's a fair reading and I agree with the sentiment. I just found it odd that the typeface design solution was missing.

Karsten, fonts don't make mistakes, but people do, and that includes Burke, FontFont, Bringhurst, and you. I don't understand the resistance to appreciating that in some fonts the caps are large enough that setting them in sequence ruins things, and if their smallcaps aren't too small it solves that problem. Maybe this is yet another case of Modernism's fetish for Simplicity rearing its ugly head? Dunno.

hhp

Thinking about this some more, I think there's a better explanation than -what I view as- Modernism's clash with true functionality. An attempt at some cause-and-effect historical analysis:

Reading has slowly been shifting from an act of reference (where you often look up proper nouns and dates) to an act of entertainment (where you read once, linearly). More significantly, acronyms used to be virtually non-existent, but now they pepper most texts we read. These two phenomena have combined to cause the cap-size requirements of text to shift, but many of the fonts we use predate this, which sets up a conflict between conservatism (which is basically following the rules of a previous generation) and progress (making new rules, which is hard to get right) in the realm of "proper" cap size(s).

Type designers concerned with the full-caps of "old-fashioned" fonts being too large (and the smallcaps being too small to adequately sub in) make them smaller (and/or make the smallcaps larger) and like-minded type users choose fonts along the same lines. This is not the same as saying that one should avoid full-caps as a rule (certainly not for grammatically capitalized words). It simply means that a good cap size depends on context. I think this is a nice refinement, even though it increases complexity. Just like making the effort of choosing a font beyond Times and company.

hhp

Karsten: I did in fact not see the image. I did not suggest Celeste was a poor design, but that designing (and marketing) a face for editorial work with such low sc is.

Small caps in book work is IMO a display feature, much more than a text feature: they are intended to mark the beginning of a chapter, contrast with capitals or sit on it's own line for subheads, running heads etc. In editorial design the function is the opposite.

Good historical perspective Hrant.

I would add that when words (not acronyms) were set in small caps, they were letterspaced (the traditional precept of all cap setting), in which situation having them lower case x-height was actually preferable to taller. Similarly, old style figures, tabular of course, were very open in fit.

Now, text is much more tightly fitted—kerned, ragged, and with word- and letter-space justification, and word spaces are smaller. In these circumstances, that spacious treatment of figures and x-height small caps is out of place.

I would add that when words (not acronyms) were set in small caps, they were letterspaced

How about both, in the same paragraph, on the same line?!

Here the titles (Empress, Monarchess, Sovereign) are set in small caps without spaces, and the names of their most august carriers (Yekaterina, Pavel) dutifully letterspaced. The source is [Gavriil Derzhavin.] Obyasneniya na Sochineniya Derzhavina. St.Petersburg: Alexander Smirdin, 1834, p. 1.

I just found it odd that the typeface design solution was missing. [Stephen]

Actually this was the first solution offered, by Frode. Given that the original question was typography related, and its wording implied that typography and typeface are not among the variables, a type design related answer seemed besides the point to me.

I don't understand the resistance to appreciating that in some fonts the caps are large enough that setting them in sequence ruins things ... [Hrant]

Wow. I said that the first answer does not address the question, reminded that smaller small caps do fulfill a typographic purpose, and added that typography offers enough variables to address any problem in more than one way. This is rather the opposite of resisting to appreciating anything.
My own typefaces, by the way, may indicate that I am not a fan of tall caps because they may cause problems.

What I try to resist, in fact, is commenting on otherones' typefaces because (whatever I may think of them based on my conception of type and anticipated use of type) I am certain that their designers had good reasons for making the design decisions they made (based on their conception of type and anticipated use of type).

If I misunderstood you, I'm sorry. There are others [here] however that do clearly seem to think smallcaps are an affectation.

I am certain that their designers had good reasons for making the design decisions they made

To me that's a very peculiar stance, especially when it concerns a craft. Any decision comes in part from misconceptions, and limits on time. For example Mrs Eaves has poor spacing. It's OK to say that. And in your own work, aren't there things you now realize were wrong? In my case I see problems with my work all the time, and sometimes I even have to ignore them. And sometimes others -because they can be more objective, or they have more experience, at least in a certain aspect of the craft- can see such problems before you. Isn't it nice when they help you see them?

hhp

Of course, based on my conception of type there are things that others do that I don't appreciate and things that I did that I would not do again. But first of all I want to understand why things were made the way they were made, be it a piece of design or a text or anything else. Something that appears wrong at first glance may make perfect sense after considering it a bit and putting it in context of the problem it was made to solve. (I got into type because I found it illogical that, in Times, x-height serifs of n vs v started at different height and set out to "correct" that. I prefer to not comment any further on my former self.) That does not mean that everything is right, though ... Discussing work is necessary but, I think, more fruitful when done in private. It allows both sides to be frank, without one side looking like a fool and the other like the expert coming to the rescue. It also allows discussing both sides' conceptions, the critic needs to know the conception on which the work is based if he wants to offer more than an opinion, the one criticised in turn needs to know the critic's conception to make sense of and evaluate the criticism. Unless one of them is an absolute beginner, any clear-cut right/wrong is likely to disappear in the course of this discussion.

I agree totally with most of what you wrote, especially with the need to put each person's personal objectives in context. Please see my second comment here:
http://www.flickr.com/photos/56183109@N02/6964823544/in/photostream/

However I see a great many advantages to doing a critique in public, not least that it's dangerous to rely on too few people; and sometimes the best feedback will come from a totally unexpected, unsolicited source. And if a person is too sensitive to criticism, that's his loss. Survival of the fittest! It's also relevant how useful a critique (or really any discussion) is for the person doing it: it forces him to think more clearly, and quite often he learns something himself! And this is amplified when it has do be done in public.

hhp

Your point is valid, Karsten. I did not need to shout “bad design”. I apologize for that.

However, I do honestly think that what you consider standard practice is a misconception when applied to contemporary editorial design.