OK, I don't know if I get it.
I am not sure if Native American and lithuanian uogonek describe the sime sound.
The only uogonek I know is the lithuanian one. It seems to be attached the same way as aogonek (you can find examples at home, in food packaging ingredients typography). The possibility of meeting oogonek (vert. centered) and uogonek (right stem) in lithuanian equals zero I guess.

hi, yeh it looks ok both letters, both is understandable as u-ogonek, but the right one looks more like lowercase letter it gives some sence of unusualnes, its something like big lowercase letter :)

and one more thing, here is letter from typeface "palemonas" developed by lithuanian government specialy for lithuanian language:

Centered alignment of ogonek in above E is an effect of keeping an angle (30'?) between right serif and the end of the tail.
So the example is rather a proof of evolution of a letterform to reach the beautiful optical balance than a rule. (When writting by hand, we never return with the pen so far.)

Ah, sorry, I misunderstood the meaning of your post, Aszszelp. In my opinion Uogonek must not have the foot. Otherwise it looks weird and "cyrillic"; too similar to russian "c" (Ц - U-0426).

I think, in typeface where upercase U has foot its ok tu attach ogonek to the foot, it does not look to similar to russian "c" becouse the shape of ogonek is different, and russian "c" is solid letter (without ogonek) an uogonek is U + ogonek its diferent concept fells different at least for me

Szabolcs,

The problem with the ogonek positioning in the Junicode font is that somebody thought they are centered, with of course isn't true. The top of the ogonek has been positioned at the geometric centre of the base glyph, but of course _optically_ the ogoneks are positioned in the left part of the letter, which looks absurd. And it's the optical impression that counts.

It is quite easily fixed and does not require redrawing the U, and also solves the O problem (my changes are on the right):

Now, the ogoneks are slightly to the right. One could also try attaching them really in the optical centre if one wants (i.e. just slightly more to the left than what I did, but definitely not as far to the left as in the original glyph which is shown in the middle):

But the most important thing is that you _cannot_ just stick a composite ogonek onto O or U or e or even a, for that matter. You need to redraw the junctions every time.

Adam

Marcin wrote:

> I am not sure if Native American and
> Lithuanian uogonek describe the same sound.

That doesn't matter. The German "w" [v] and the English "w" [w] don't describe the same sound. The German "z" [t͡s] and the English "z" [z] don't describe the same sound. The Spanish "r" [r] and the French "r" [ʀ] or [ʁ] don't describe the same sound. Damn, even within the same language one letter can describe different sounds ("u" in English words "luck", "put", "mule" and "fur" certainly describes different sounds each time).

But this does not have any influence on the fact that those different sounds can be written using the same letters.

Adam

Hello,

Even though I like Junicode generally for quick tweakings, examples and fix-ups (I'm not involved in its developement or enhancement), ('cos it's a freely available nice humanist face--- though it has its flaws, like a ridiculously wide capital T) I see now that in this very case it was a poor choice insofar as in determining the initial assumptions.

On a side-note, the Oogonek-examples you gave give the impression as if the ductus was starting at the bottom of the O and would advance _clockwise_, which is unusual for the latin script and is slightly disturbing to me.

"But the most important thing is that you _cannot_ just stick a composite ogonek onto O or U or e or even a, for that matter. You need to redraw the junctions every time."

_That_ was clear :-) This goes for most good diacritics, (I prefer different (mostly: different wide) circumflexes (circumflices ;-) ) on i (e.g. Turkish) and w (e.g. in Welsh language). It is however twice as true for attached diacritcs: ogoneks, cedillas and horns (in Thai).

While looking at it, I'm mysself less and less convinced of a legged variant of U for attaching an ogonek, you did circumvened directly addressing the question regarding its viability by merely pointing out (completely right) that it was unnecessary.

Bruzs's "centered" ogonek on E in the Lithuanian-govt.-sponsored font was insightful as an alternative, maybe less traditional, more avantegarde solution.

Clearly, the sound value of a letter is completely unimportant for glyph design. The complementary distribution of oogonek and uogonek, however, were important for my above considerations, if both were to appear in the same language, my suggestion would have lost validity immediately.

A last question to the explicitly Lithuanians here: do words start with uogonek? I have only met them at the end of words. Would capital Uogoneks appear e.g. in names (not necessarily personal), or capitalised due to the beginning of a sentence, or only in all-caps settings (e.g. if they only happen to appear in medial and final positions)?

Szabolcs

A legged U is historically correct in some lettering styles but not in others. I think in principle, it would be possible to mix a round U and a legged Uogonek in one typeface, but this would need to be done carefully. Some readers may get confused as they might think that the alternate shape signifies some additional difference. Also, the legged U would need to be drawn properly, with appropriate weight distribution on the stems, and with a complete redrawing of the curve. I realize that your own sketch for Junicode is just a rough illustration of the concept, but at the same time it clearly demonstrates that this is not a quick-and-dirty job. This brings us to a final point: in many cases, it would be uneconomic to redraw the base letter so drastically for the sake of Uogonek. I fear that it would often be done incompetently, as diacritic extensions are often done by people other than the original designers. One for sure — I do not see the legged Uogonek to be the default approach to solve the problem. I think it can be considered a valid alternate approach for some, very specific, typeface classes.

A.

Thanks, that's an indeed clear opinion :-)

/Sz

This is an old thread, but a PDF on Junicode has the following explanation:

Because Junicode has been designed for medievalists, some of the default
letter-shapes are wrong for modern languages: for example, eogonek
(ę) is correctly shaped for medieval Latin but looks poor in Polish.

I don't know if the shape in Junicode is indeed correct for the 'e caudata' of Latin palaeography, but at least that explains why the glyph doesn't work very well for the Polish 'e ogonek'.

This is another example of superficially similar diacritics that are unified in Unicode having subtly different shapes in different traditions. The 'o with a hook' used in Old Norse again differs from a Polish-style ogonek, as explained in http://www.decodeunicode.org/u+01EB:

Unlike the Polish ogonek, which goes all the way down to the bottom line (like “p”), the Old Norse ogonek should not be drawn much lower than a cedilla.

Junicode's explanation is poor. The best Renaissance typefaces such as Garamond's include a Medievalist e caudata, and it looks just like it should look for Polish, while it does not look anything like Junicode's:

http://www.flickr.com/photos/adamt/sets/72157619801458510/

Cheers,
Adam

Thanks for the link, Adam. Now that you've drawn my attention to it, I do think I remember seeing similar medievalist e caudata designs in traditional typefaces (funny how memory works). That will teach me to treat assertions like that with a grain of salt, especially if it's something that can be easily checked.

I wonder if there are typographic examples of the 'o with a hook' that show what the traditional preferences are as well.

The ę character is certainly the oldest among the letters with an "ogonek". E caudata was the only such glyph used in medieval Latin (interchangibly with æ for the same sound).

Polish adopted the letter ę for the nasalized "e" sound, probably because it just "was there" and was not used for anything else. Then, Polish borrowed the ogonek shape to form the letter for the nasalized "o" sound, which became "ą". From Polish, the shape was borrowed by the Lithuanian, which in addition to ą and ę also have an i and a u with the ogonek (Lithuania was in union with Poland for several centuries; Poland was mostly the dominant part in that union).

My guess is that the o with ogonek as used in Old Norse is an independent development. However, Decodeunicode's description of "the Polish ogonek, which goes all the way down to the bottom line (like “p”)" is not correct. Depending on the structure of the typeface, the ogonek can be quite large or small. The larger the intended size for a typeface, the smaller the diacritics (that is a general rule). In typefaces intended for small sizes, the ogonek needs to be quite large so it remains legible. In typefaces intended for display sizes, the ogonek and the other diacritcal marks are usually much less pronounced.

There are some images on my
http://www.twardoch.com/download/polishhowto/ogonek.html
but I hope to update this page with lots of newer material in some near future.

Cheers,
Adam

IMO, the ogonek should harmonize with the other accents, so should compare to the acute accent in the same way that the circumflex does.

If you are going to put a legged Uogonek in your typeface, you should make the standard U have a leg, as in Perpetua or Giovanni.

Nick,

I don't think so (regarding your second paragraph). U with ogonek without the leg works quite well. You have a nice round smooth surface at the bottom of "U" that is very similar to that of an "e". Anything that works for "ę" also works for "Ų".

Cheers,
Adam

You misunderstood me, Adam.

I was disagreeing with your comment that, "in principle, it would be possible to mix a round U and a legged Uogonek in one typeface, but this would need to be done carefully."

I think it would always be a mistake to have a round U and legged Uogonek, no matter how carefully it were done.

I could be wrong, but you'd have to show me the evidence!

Oh, I see. Well, I said "in principle". You know, "in principle", I believe every letter with a diacritic could look differently from the base letter, just like you could have several stylistic alternates for each letter. But as the general rule, I think Uogonek should use the same shape as U, so we are in agreement. "In principle" ;)

Cheers,
Adam