.

Only a period as answer?

I always use option #2: I don't know if it's technically correct but, to my eye at least, it's more aesthetically pleasing and it minimizes any possible confusion of the accent with an apostrophe...

Check out the http://Diacritics Project.

The diacritics project entry for caron doesn’t answer this particular question.

Check out the typefaces at Peter Biľak's foundry, Typotheque.

>>Check out the typefaces at Peter Biľak's foundry, Typotheque.

They are expensive :)
I mailed a slovakian colleague to know her opinion on the subject... I will later share her answer

They are expensive :)

"Download PDF"

Preferably (1). You want to maintain clean ascender line (as one would do it similarly with the x-height). The second case (2) could happen in handwriting, though. You might also need to do it in small cap /lcaron since there may not be enough space in bold. I also do it with Lcaron so all "vertical carons" in the font are at the same vertical position.

I do not understand how the second example aligns with acute (acute for lowercase letters or uppercase?). The reasoning behind this kind of caron is not to put the accent above the letter since there used to be limited vertical space in metal typography. At least such is the common understanding. So the accent should not be above or diagonally to the right top, but to the right from the character.

All of this very much depends on the x-height which I cannot see. If the x-height is tall it will force you to push the caron up, esp. in black designs. Regarding the apostrophe: it well may be and perhaps that is actually quite common that apostrophe turns out to reach above the ascender line (?).

IMHO the distinction between apostrophe and vertical caron needs to be doneby changing the size and shape of the marks. The contemporary solution is to make the caron wedge-shaped or more shallow than the apostrophe which typically takes comma-like shape. It seems more secure than relying solely on the positioning. You can read more on the topic here: http://ilovetypography.com/2009/01/24/on-diacritics/

Cheaper models :) to look at are Brousil's (Suitcase) typefaces or TypeTogether's (Maiola or my Skolar of course :). Just use the font testers or PDFs and it should give you enough idea.

Hope this will be of some help.

--
http://rosettatype.com

David, do you think that a more "traditional", comma-shaped vertical caron has any merit in a retro face, or authenticity in a revival?
Admittedly, it may be too much like an apostrophe for present day taste, but if it is quite different in size…

I wanted to add: in the end it is not such a big deal. Both designs (1 & 2) are still readable, are they not?

Nick, I have not done any research in this, but I think it WAS originally an apostrophe used for a different purpose. Then there is a question of what sort of revival it is, if it “revives” (contemporarizes, makes more usable for current typography) or if it is “digitalisation”. You can also consider something in between, some kind of drop. There are many faces which do not try to deal with the distinction between apostrophe and vertical caron and they are being used commonly.

--
http://rosettatype.com

To me, #2 is the option that looks more like an apostrophe, and #1 is the one that would avoid the confusion. So, clearly I must be very confused.

There are many faces which do not try to deal with the distinction between apostrophe and vertical caron and they are being used commonly.

I would imagine that in the past Slovak typographers would compose the vertical caron characters from the plain character and apostrophe, if using "loan fonts". Similarly the Catalan ldot would have been composed from l + period, or more recently, l + periodcentered.

And I suspect that practice still occurs: are there specific keys for the vertical caron characters on the Slovak keyboard, or is it easier for unsophisticated users to type a plain character followed by a hash mark (quotesingle)?

Typing is lcaron or dcaron is very simple. Typically, you key -caron- + -L- and you are done. There is no special key for vertical caron. Note it has same meaning as the classic one.

The ascending lowercase letters with caron (ď, ľ, ť) have adopted a different form of caron clearly for technical* reasons: having a tall letter with a mark *on top* of it was considered unpractical, since it enlarged the vertical height of the line. Especially if the ascenders were long. So a mechanism of "vertical compression" was employed, and the mark was squeezed into the "body" of the lowercase, i.e. was simplified and put next to the letters.

* I mean "technical" not necessarily because it was implied by the printing technology, but also when it came to the "technological effectiveness" of handwriting. Maybe the word should be replaced by "practical".

My observation is that for those lowercase letters, the main point is that the "apostrophic caron" should be designed in such a way that when typeset (e.g. loďka), there should be no visible word gap where the diacritic occurs. So the space should be as sparse as possible, yet the caron must be distinctly visible as well.

This can be explained in an opposition to the English usage of apostrophe: words such as don’t, it’s or let’s have no problem with the apostrophe appearing in place of a word gap. This is largely to the fact that the apostrophe does, in many cases, indicate a contraction of two words (don’t = do not, it’s = it is, let’s = let us). Leaving the Saxon genitive aside (though very down in its roots, it also is a contraction), the reader of English has no problem accepting that don’t is something semantically between one and two words. So the somewhat larger gap that the apostrophe creates is fine.

With Czech and Slovak, it is very different. The words that use the apostrophic caron should not be "disrupted" visually in the middle. So typically, the apostrophic form of caron undergoes a formal reduction: it does not need a fully-qualified comma/apostrophe shape (with a ball and all), but instead, it works well with a reduced form that resembles a nearly-upright acute. The apostrophic caron can be very straight (only slightly tilted to the right), or very very slightly curved (not as curved and round as comma).

The question of the uppercase Ľ is funny: because of its shape, there is lots of space there so the letter does not actually need this sparse, or compressed form of a diacritic. In theory, it could well do with a classic caron placed on top (L with ˇ on top). After all, all the other capital letters use the classic form (I'm mixing Czech and Slovak here, so apologies: Č, Ě, Ň etc. but also those that in the lowercase form take the apostrophic form: Ď, Ť).

However, since Slovak uses both the letter L with a caron and with an acute, for lowercase, the PLACEMENT (not the form) of the marks between ľ and ĺ was apparently a sufficient indication of difference. For uppercase, the same differentiation between Ľ and Ĺ has been made.

This fact CLEARLY shows that the apostrophic caron in Ľ should not be raised too much above the top of the uppercase L. Perhaps its top can be even aligned with the top of the L. Also, in the same spirit, the acute over Ĺ should be placed quite to the left, i.e. its "start" (bottom) should be horizontally aligned with the stem of L. Also, for that reason, the uppercase acutes for Czech and Slovak should be quite flat, while the apostrophic caron needs to be fairly upright.

For some reason, inserting images into Typophile does not work for me :( I wanted to illustrate this with examples. Since this did not work, please visit these links to see the difference in two of Peter Bilak’s typefaces, Greta and Fedra Sans:

• Greta Text
• Fedra Sans

To sum it up: an important point when designing diacritics is not to view them in isolation, but always to design them in context, especially in comparison with other diacritics that occur in the same natural language. To be very explicit: the minimal-pair difference between Ľ and Ĺ should ideally be as clear to the reader as the difference between E and F.

Best,
Adam

And in just general spirit: in the evolution of the letterforms, minimal-pair differences always played an important role. Whenever any two forms were considered confusingly similar, some details became more and more pronounced over time to differentiate them. This is very clearly visible in the uppercase Roman alphabet: the additions that had no predecessors (J, U, W) as well as the later adoptions from the Greek (Y, Z) were chosen or invented in a way that made them distinct from the other letters, and those distinctions grew over time. This is evident in the "quest for distinction" of the letter J, where different solutions have been tried over time: curving the tail more, elongating it below the baseline, placing a bar on top of the letter etc. Similarly, the case can be made that "ſ" was dropped because it was too similar to "f" in the end, so the more distinct form "s" prevailed.

In the diacritic design, the task is quite similar: choose forms that are in harmony with others, yet sufficiently different so that they are not confused with others.

Since different kinds of diacritics are used in various languages, but it's not like all diacritics are used at once, the largest importance is to pay attention to potential differences within one natural language.

A few examples:

• É and È need to be sufficiently different because they both appear in French (both for upper- and lowercase).
• Ľ and Ĺ must be sufficiently different because they both appear in Slovak (both for upper- and lowercase).
• Ö and Ő must be sufficiently different because they both appear in Hungarian (both for upper- and lowercase).
• Ż and Ź need to be sufficiently different because they both appear in Polish (both for upper- and lowercase). An interesting case here: in some display typography, the Polish uppercase Ż is depicted as Ƶ to make it very different from Ź, but this has never become a universal norm.
• The lowercase ł needs to be sufficiently different from l and t, because all three appear in Polish.

So when designing a typeface for small-size pan-European use, designers should first identify the key minimal-pair differences, then get them right, and then apply the design solutions consistently to the rest of the diacritics that are in no conflict to others.

Adam

Thanks Adam for the additional comments, however much of what you just said was said before or in referenced articles (ILT & Diacritics).

Few remarks:
- handwritten caron next to -d,t- is often written as -v- shape. I think the standard adopted the vertical form (e.g. Comenia script is using it), but it was not always like that.
- I have seen remarks about typefaces and theit loosely spaced apostrophe in English. Therefore, I would not go as far as “has no problem accepting”. There are certainly people who consider the gap as something distracting. Besides that you are giving a nice interpretation to “apostrophe is punctuation and as such it deals with words where caron is a diacritical mark and belongs to the lettershape in order to represent a sound” (mentioned in the article referenced). Indeed, two quite different roles.
-note that -É- (Eacute) and -Ě- (Ecaron) are aso used together and no one is considering that the caron in the second letter should be any different or positioned differently. The accents are quite distinctive. Hence using “vertical caron” * in -Ľ- (Lcaron) can be considered a matter of convention. The letter -L- simply allowed the same treatment as the lowercase -l- which is not the case in -D- or -T-.

*) I intentionally avoid to call it apostrophe-like, because then people might think it should look like an apostrophe.

David,

you’re right, the phrase “apostrophic caron” is misleading. I’ll adopt phrases such as “v-shaped caron” and “upright caron”. Does that sound good?

> -note that -É- (Eacute) and -Ě- (Ecaron) are aso used together and no one is
> considering that the caron in the second letter should be any different or
> positioned differently. The accents are quite distinctive.

I completely agree. Acute and the v-shaped caron are distinct enough in their shapes. So they consistute a less-than-minimal pair by themselves. There is no need to pay attention to the details of their placement to make them appear more distinct.

The minimal-pair examples I’ve cited were more tough: the dieresis and the double-acute (“hungarumlaut”) in Hungarian can be misleadingly similar if badly designed, the upright caron next to “L” may be misleadingly similar to the acute which also appears with “L”, the dot and the acute over “z” in Polish can also be too similar, and if the acute and grave are designed very upright, they may be confused in French when placed over “e”. So these are good examples of minimal pairs.

In other words: the acute and the upright caron are a classical example of minimal pairs: they’re VERY similar in isolation, and this is why in their case, their placement, when used with “L”, is of extreme importance. In case of the Slovak “L”, it’s not the shape of the mark itself that makes a distinction between Ľ and Ĺ: especially in small sizes, it’s both the shape and the placement.

I don’t think I’ve seen the aspect of minimal pairs raised in the discussions about diacritics (but maybe I’ve overlooked something). The concept of minimal pairs is a widespread concept in linguistics (http://en.wikipedia.org/wiki/Minimal_pair ). They’re widely used in phonology, but I think the same concept can be well applied to graphology, or typography.

Please note that my point is not that the caron over “Ě” should be any different from the carons over other letters. I’ve said:

So when designing a typeface for small-size pan-European use, designers should first identify the key minimal-pair differences, then get them right, and then apply the design solutions consistently to the rest of the diacritics that are in no conflict to others.

In other words: if there are minimal pairs in any natural language, look at them first when designing diacritics (or characters, in general), and make sure they are not too similar. In other words, when designing acute, the v-shaped caron, the upright caron, grave, dieresis, hungarumlaut, dotaccent etc., make sure that you look at test phrases such as ÉÈĚĽĹÖŐŻŹ, which consist of letters that occur as minimal pairs in natural languages, and make sure that this “looks right”. It’s not a great discovery I’m making here, it’s quite an obvious observation and many people work that way — but I thought it’s still worth emphasizing.

Finally: of course the apostrophe (for the use in English, for example) is sometimes done as too loose. Then it is too loose :) Yet, as you say, it is a punctuation sign to due to its nature, it does need a little bit of space that gives the sign “some space of its own”. But of course it should still be spaced carefully, and, again, “look right”.

> much of what you just said was said before

Believe me, in my experience, stating the obvious once is often not enough. On many occasions I’ve found myself needing to state the obvious again and again. :) Plus, I do have a short memory. I’ve recently been searching for online reviews of a Polish film that was released some 13 years ago. I did find one extensive review, started reading it, and though “Oh, I quite like this one, a good review”. Imagine my amazement when I read it to the end and found my name at the bottom. I wrote it in 1998, and completely forgot about it :D

Cheers,
Adam

Very useful. Thanks, Adam and David.

The term "vertical caron" seems clear enough to not assume an apostrophe shape yet makes it clear what the object is and how it relates to others. I find the Apostrophe-like name confusing as well as the tendency to confuse it visually with the true apostrophe. I certainly admit that I don't speak or read Slovak (but some of my best friends do ;-)
I must admit that kerning any of the glyphs with vertical carons is more taxing than typical kern pairs and causes me to make kerning exceptions more for them than anything else :-)

ChrisL

I notice in the "Greta Text" example Adam linked to, that the vertical caron has overlapped the serif on the following ascending gliph. Is this accepted or even preferred to seeing some space between the two?

Hey Adam, your posts are overwhelming!

Just a little note, you posted the other comment while I was writing my comment to your first post, so I was not actually responding to that.

Regarding the naming: I know I used it myself, but perhaps v-shape is also no good. Best would be to avoid any shape comparison, but that is what actually makes the names more useful. What about caron above and (next-to-)ascender caron? But hey, I am fine with anything as long as there is no word suggesting it relates to punctuation.

Regarding the -v- and caron. I have the feeling that designers do not always realize that caron is not a small -v-. For two reasons: a) the contrast can be quite different b) there are no terminals (namely serifs) in diacritics. Similar case -ring- and -o-.

I like your example of minimal pairs. Well, I do like bringing linguistic models to type design, but it is also useful to put it this way. btw. you get even more fun when designing for academic linguistic texts with an extensive use of diacritics to describe pronunciation in etymology (surprisingly IPA is not as adopted as one might think). Basically any kind of combination can appear there.

Your last paragraph made me laugh loud. :) I guess it is better to say more than less (which is generally my preference in writing).

Regarding the kerning: I feel your pain, sorry! Few years ago I asked few Czech and Slovak designers what is their approach to this. Generally there are two approaches:
1. space -ľ- similarly to -l- (i.e. the caron is overlapping the right sidebearing) and use positive kerning to solve collisions with following ascenders (b,k,h, …), accented letters (á, é, …) and punctuation (brackets, quotes, question mark, … [remember that Czech and Slovak use 66 quotes at the end]) as clashing is very NOT nice. This produces less kerning and it is quite manageable, but if there is no kerning, there will be clashes.
2. space -ľ- so there is enough space on the right side of the caron and it does not overlap with following ascenders and use negative kerning to shift letters closer if possible. The amount is at maximum the space you added because of the caron, however see the note below. You also need to shift a word space a little! More confusing, but less vurnerable – works even without kerning.

Note that the bolder the caron is and the higher the x-height is the more you need to consider the caron shape when spacing. So for example when it comes to approach [1] and typeface with a medium x-height, the -ľ- and -l- spacing in the black weight will be quite different on the right side while in the regular it can be practically the same.

It may be a problem to find all the critical pairs you need to check. If you are unsure, simply check Slovak and Czech alphabets.

When picking the approach, you need to consider your customers and software you are designing for. I guess most of the independent foundries designing for graphic designers using Adobe software can make their life easier and go with [1].

Recently, I wonder if the gaps which necessarily appear in a not-kerned typeface which used [2] are not as annoying as the clashes.

If any of this needs some clarification, let me know and I will prepare a picture.

Cheers,
David

p.s. Heh, my comment is not much shorter. :)

--
http://rosettatype.com

David,
I use the: "1. space -ľ- similarly to -l- (i.e. the caron is overlapping the right sidebearing) and use positive kerning to solve collisions" method since it involves far fewer kernings. I just have a class that includes all ascenders on the left and then I only need deal with butting diacritics from neighboring glyphs. I also adjust the lcaron to space kern to compensate for the open feeling the hole makes.

Thanks for all of your insights, Adam and David!

My practice is to make all accents similar in weight, so that the sprinkling of typographic confetti which surrounds letters in accented languages has an even colour.

Here's what I've done, different strategies to suit different typefaces.
Bodoni Egyptian (vertical caron by Ján Filípek) is the most successful, in my opinion.

BTW Adam, to post images: only PNG or JPG formats, and no spaces in the file name.
Also, keep to a maximum of 588 pixels width to avoid scrolling kicking in.

Nick: Bodoni Egyptian is nice, but to me the bottom left is the most optimal when it comes to distinction between caron and apostrophe. Of course you cannot always go for the most optimal, e.g. for the sake of aesthetics, stylistic consistency. Note you need to be able to distinguish the mark when it is set alone. The mark does not need to be only recognized, but also distinguished (unique). I guess it is clear, but still I felt it needs to be articulated. So testing with two words, e.g. «rozhoď» vs. «rozhod’», is quite useful.

--
http://rosettatype.com

David, is disambiguation between vertical caron and apostrophe really so important?
After all, the vertical caron characters appear extremely frequently in Slovak text, and the problematic rozhoď/rozhod’ very rarely.
Surely it is better to aim for good, even colour in text -- with the two critical areas being:
1. the space between vertical stem and caron
2. the space between vertical stem and following character

It seems to me that the thin vertical slash treatment that many typefaces now employ creates a nasty narrow "bottleneck" between accent and stem.
What should the space between dcaron and lcaron, and the following character look like, as a percentage of the word space?—
1. 0%
2. 10-20%
3. 33%
4. 50%
5. 67-100%

> to post images: (...) no spaces in the file name.

Ah! That's it! Thanks. Forgot about this one. No spaces.

A.

Would anyone suggest the overhead v-caron for these letters in cases where line spacing is not an issue?

In most cases, line spacing does not represent the same problem nowadays. From a typographic point of view, I think the traditional caron give better results when the character goes inside a word. But considering cultural elements, this probably wouldn't be a choice by now. Maybe in the future this could change due to the facilities offered by digital typesetting.

Don't you think it's interesting to include alternate |d|l|+caron glyphs in a font, using the traditional caron above?

Note that in the Latvian alphabet there are several "under comma" letters, but for the "g" the accent is moved up, rather than sideways.

Nick: that is question I am asking myself all the time this thread is active. I do not know. But note, Czech (and Slovak is similar in this) is very phonetic in the writing, so if I confuse the marks I have no idea how to pronounce it right away. Context helps, but it actually happened to me several times that I had to read the sentence again to figure out. rozhoď/rozhod’ is a Czech example. Not sure how frequent it is in Slovak. Sorry, should have pointed it out. I am unifying the two languages quite a lot when it comes to this diacritical mark.

The distance: I would make the gap visually around the thickness of the stem (geometrically thinner).

Bendy:
The answer is no. :)

Freiberger:
Don't you think it's interesting to include alternate |d|l|+caron glyphs in a font, using the traditional caron above?

Same answer. It is considered incorrect in typographic letters. I do not think there is any movement to simplify diacritics in Czechia or Slovakia, but I might have missed something. Also, I would be careful to call one kind of caron traditional and the other one not. Perhaps that is my mistake when I called the v-like one "classic". I am not very well informed as of how the accents changed during the history to be able to say which form is older. It may not be simple.

--
http://rosettatype.com

In most cases, line spacing does not represent the same problem nowadays.

Actually, it still does. You cannot extend line spacing as you wish if only for economical and aesthetical reasons. The letters still need to stick to some kind of average (naxinun) height. Czech is typeset with larger line-spacing (ca. 1 point more) than English with the same font, simply because there is so many diacritics. And now consider newspapers with the new "improvement". That could a big deal.

>Bendy:
>The answer is no. :)

I like this answer. :)

David:
>[remember that Czech and Slovak use 66 quotes at the end]
And what at the beginning?

Bendy:
English “quotes” (6699), česky a slovensky: „uvozovky“ (9966)

--
http://rosettatype.com

Very interesting. I wonder how they evolved that way. Are there any other styles of quotation marks other than the «guillemets» that I should be aware of?

http://www.typophile.com/node/39799

Perfect. Thanks, Riccardo.

Thanks to everyone for the comments so far. They are vey illustrative.

What about k+caron and h+caron letters? Should they have a vertical caron? It seems logical, because l and d have. But all the references that I've seen are k and h with a traditional caron/hacek.

Looks to me that k and h have enough space to the right of the stem for a regular caron while d and l do not.

But usually, what I see anyway, the caron is riding high

http://www.fileformat.info/info/unicode/char/21f/index.htm

http://en.wikipedia.org/wiki/Caron

Is the purpose of the comma-caron to keep the setwidth constant, or to not interfere with descenders in the line above?

Charles, I believe the Czech and Slovak postscript caron it is simply a scribal form, adopted into typography from handwriting.

{Wrong thread, sorry!}