A variant of a with ogonek/stroke

[jsbien]

Before a with ogonek became generally used for a nasal sound in Polish, a with stroke was used instead. In a diachronic IMPACT corpus a with stroke has about 30000 occurences (while a with ogonek has about 40000). In the corpus the character was represented respectively by LATIN SMALL/CAPITAL LETTER A WITH STROKE (U+2C65 and U+023A), so it can be considered a glyph variant of this codepoint; however using U+2C65 and U+023A is not a generally accepted standard, so treating it as a variant of a with ogonek also has an advantage. Reportedly the stroke had different shapes, but supporting only one variant illustrated below is in my opinion completely sufficient.

[psb1558]

Do you have any images of the capital? (And, btw, should there be capital versions of the ø variants?)

[jsbien]

The first hit in the IMPACT corpus: I will look for more examples.

[psb1558]

I really like that! Would be interested in any other variants.

[jsbien]

Another volume of the same work: Another book: And a strange ogonek-stroke I noticed now for the first time:

[jsbien]

As for the capital versions of the ø variant, there is no real need for it, but they may be added just for completenes.

[psb1558]

Do the variants of lowercase ą with the line through and the line above and below ever occur in roman style, or just italic?

I don't believe in case pairing for its own sake; if uppercase variants of ø would get no use, I won't add them.

[jsbien]

Unfortunately the IMPACT corpus is now not fully functional and I cannot find easily specific examples. Definitely the character is used not only in italic style, but the other style in that time was often a kind of fraktur, which is now rendered just as a roman style. So, with this qualification, I think the answer is yes.

[psb1558]

Thanks—I'll run some images by you later today and get your advice.

[psb1558]

A first attempt at lowercase roman.

Suggestions?

[psb1558]

I feel more confident about the uppercase, since the models are very clear.

[jsbien]

Uppercase is OK. As for lower case, I think the crossbar should moved a little to right and upwards, the lines should cross, not just touch. I uoload another example, but I haven't found yet a really good model.

[psb1558]

This is a big help, but other examples would be welcome.

[psb1558]

Like this?

It's tempting to widen the lower-right curve just a little bit to accommodate the diagonal stroke (though I don't think whoever cut the original type did so):

[jsbien]

I think the stroke can be shorter and not so steep. I will look for a good model, but I want to make haste slowly and try first to circumvent some technical problems with our corpus,

[jsbien]

This looks quite good, but it seems to me that the stroke is more to the left than usual.

[psb1558]

I think it's pretty clear that these old punch cutters were working with existing matrices for a, and we're seeing various responses to the fact that the lower-right curve of a is not wide enough for a crossing stroke. The stroke in the "skąpey" example is quite steep, about 20°, which is one way of dealing with the problem, but even so ink spread has made it quite indistinct. In the "skuteczną" example the stroke is a more comfortable 40°, but there's no way to make that cross only the curve. I can widen the curve a bit and make the angle 20°, so

If I make the angle 30° I need to nudge the stroke just a little bit left:

At 40° it starts to look awkward (to my eye, anyway)

unless you take the approach of the ą in the "skuteczną" example:

I think the 30° stroke with the widened curve on the a looks best (the 2nd image), but I'm not thinking about it historically.

[jsbien]

From an aestetic point of view the version from the first scan above seems quite nice, but your proposal is also OK.

[psb1558]

I like that. Position, angle, wedge-shaped stroke. I'll work on this later today.

[jsbien]

FYI, it comes from https://cbdu.ijp.pan.pl/2910/.

[psb1558]

The a is very different in this font, and I have to stick with the one in Junicode, but I can reproduce the position, shape and size of the cross-stroke:

[psb1558]

FYI, it comes from https://cbdu.ijp.pan.pl/2910/.

It's a beautiful piece of printing.

[jsbien]

The shape is very nice.

[psb1558]

The four masters of this font:

[psb1558]

This same work has a very nice eogonek, which I think as I'll use as a model for the shape and position of the stroke in that other issue:

[jsbien]

OK. BTW, this is an ephemeral print, I wander whether anybody knows which printing house it comes from. An interesting topic of research :-)

[psb1558]

Second alternate:

[jsbien]

OK

[psb1558]

Can you find any more images of the italic aogoneks? It's very hard to make out what's going on in the one on the right: Which ot the three is most common?

[jsbien]

These are all the examples I can provide now. Yesterday I made a systematic search in our corpus and other examples I can find only incidentally. In your post the first letter on the left is a standard Unicode a with stroke. It would be nice to have the variant for the one in the middle, but it can be either approximated by the left one or perhaps rendered with a combining stroke. The one on the right is the most important, it seems also most common.

[psb1558]

I think I can more or less see what's going on. I've added U+023A and U+2C65 to the roman font, but I've echoed U+2C65 as aogonek.alt2 and added it to cv02. (cv02 should be used to produce this shape, since it then remains searchable as aogonek U+0105.)

Is a shape matching U+023A ever used for Aogonek?

[jsbien]

Nasals in Polish never start a word, but upper case versions may appear e.g. in capitalized titles. So the answer is yes.

[psb1558]

Italic.

[psb1558]

Italic caps.

[jsbien]

OK except the lower case on the right. I don't think the stroke should be broken. BTW, a general remark. Contrary e.g. to Adam Twardoch (https://junicode.sourceforge.io/ecaudata.html) I don't believe in one canonical shape and position of ogonek. There are so many variants of it in the texts that the choice which shape use in a font always will be more or less arbitrary.

[psb1558]

Thanks--I'll work on the last lowercase shape.

Interestingly, Adam Twardoch contacted me years ago about the position of the ogonek in eogonek and talked me into moving it. But I also kept the older shape, and that's why there has been an alternate eogonek in Junicode for so long.

[psb1558]

A simpler approach.

[jsbien]

Sorry, but I think we don't need a glyph like that. In all a with stroke I know the a is crossed by a single continuos stroke. Your glyph is similar to "o with horns" from issue #3, but we don't need "a with horns".

[psb1558]

I thought it was like the second of your three italic examples in your first post. I suppose I misinterpreted it, and will delete this.

[jsbien]

The print quality of some texts in the corpus is horrible and identifying a character can be quite a challenge.