User talk:CoryCohen/Old English Letters

Definition from Wiktionary, the free dictionary
Jump to navigation Jump to search

As promised, I've scanned in a couple of samples of the text I was reading for my entry "ʒere". I think we'll all agree that there are both ezh and yogh characters. The flat top on the ezh is very clear, as you can see below.

After reading the excellent discussion Muke pointed me to at http://www.evertype.com/standards/wynnyogh/ezhyogh.html, I spent some more time looking at my source. I think I've resolved the issue now -- it is essentially a typographical problem. There's several different font faces, which as far as I can tell are identical, except that the sizes differ. For reasons that still elude me, the ezh is only (and always?) used in the larger face, and the yogh is only (and always?) used in the smaller face.

The images were also selected for their proximity to some other interesting glyphs, that I mentioned earlier. Here's the first image:

File:Scan 01.png

This is the quote that I used for "ʒere", I selected it pretty much randomly as the first occurence of an ezh that I came across. It has three thorns, an ezh, three superscript lowercase letter Es, and one of those crossed double Ls. I now believe that the ezh should have been typeset as a yogh, and was not because the larger point size apparently had no yogh. I'm very curious about the use of the superscript letters. In "þe", it's no great surprise, as the letter at least completes the word. But in the other two cases, it seems to be a replacement for "th". Muke wrote in the beer parlour:

Actually I checked the Unicode charts and these characters do exist, in the combining diacritical marks block in a subsection for "medieval superscript letter diacritics". There is e (yͤ) as well as a i o u c d h m r t v x, the range is from U+0363 to U+036F. —Muke Tever 00:18, 15 Sep 2004 (UTC)

This list closely matches what I'm seeing. I've got examples of:

  • e - the, ye, and used for ordinals.
  • o - ano, ao, xxo, xxio, iio sacks, almost always after two (ii).
  • c - for Roman C or 100, as in xiiijc lxxxiiije for 1484.
  • m - not in this text, but I've seen it in others for the Roman M or 1000, iijm for 3000.
  • li, s, & d - rarely for currency, xijli xvijs iiijd.
  • r - mister (mr).
  • x - a pair of superscript x's indicates that the number below is in scores, iiijxx iij for 83.
  • l - usually as ml, indicating a thousand I believe, I'm unsure how it differs from the superscript m. I'm fairly sure it's a an ell and not a one, since it's occasionally crossed with a tilde like the li for[[ pound.
  • th - in some cases as th, e.g. xiith & xxiijth.
  • is - unknown, ... by yis presentes, yat...
    That last would probably be "this" and "that", by use of y for þ. —Muke Tever

That adds s and l, (not present in Unicode?) which Muke didn't mention, and leaves a u & v unexplained. For e, o, is, and possibly l, I suspect some obscure Latin term. Any suggestions?

This image includes a line of the "smaller" font face, for comparison to the larger size that makes up the majority of the sample.

File:Scan 02.png

In the this sample, I've got example of "ȝere". It's clearly a round topped yogh, and not an ezh. There's another occurence in "ȝerre" a little later. Plenty of thorns, and more occurences of the unexplained tilde crossed L. Note that it is crossed on the double L at the ends of Ball, Aprill, and cristall, but not in the middle of William. I've not found it used except at the end of words ending with a double L. The single l crossed with a tilde is used consistently in this example for li. The curly nature of the cross is quite clear, and it's not like two crossed L's from the Unicode character set.

Finally, I have one crossed (ħ) and one crossed (ƀ), both occuring in typical contexts. The ħ at the end of a word, usually following a t, and the ƀ in oƀ. Could the tħ be related to the thorn some how?

Also, regarding the swash S variant: Thanks a bunch Muke! That's exactly the insight that I had missed despite spending a good while looking at it. The is a small difference in the glyph for the ſ and f. In the typface I was looking at, it was pretty subtle. And seeing the sans serif version finally made the connection with the very tall swashed italic S I've seen frequently.

I'm clearly not using the right tools to find and use the full range of unicode characters. How were you able to locate the superscripted letters, and how would I locate the tall swashed italic S that I'm looking for. I've been to the unicode.org website, but the character charts are in 3 dozen different files. :-( Is everybody else in the same boat I'm in with a bunch of crappy Microsoft fonts that support most but not all of the Unicode standard? I've downloaded the Code2001 font, but it seems to have a very limited character set.

Actually I ran across the superscripted letters entirely by accident—I have copies of most of those three dozen different files on my computer and was browsing through them while sorting them to their proper folders. I do have a decent range of fonts though my browser (Opera) is not yet capable of choosing Plane 1 fonts (the only font it decides to display for me is for Linear B) but use of those is generally marginal, so. Code2000 isn't so much limited as it is out-of-date—when it was first released it had everything, except for a few Indic scripts, I think. Now a lot more characters have been added in more useful blocks and I don't think it's kept up. Code2001 is just for Plane 1 and shouldn't really be relied on to have any of the characters relevant to this discussion (which are all in plane 0). —Muke Tever 17:58, 16 Sep 2004 (UTC)

Any insights would be greatly appreciated. I realize that this isn't really "dictionary" related, but I haven't found anyone else who cares or seems to know anything about these issues... -- CoryCohen 03:49, 16 Sep 2004 (UTC)

Code2001 may have a small range, but it covers some characters that no other font yet does. It, Andagii and Penuturesu are the only fonts for plane one characters.
I recommend downloading BabelMap, which gives you access to all (well, most) unicode characters, whether your fonts can display them or not. --Vladisdead 04:34, 16 Sep 2004 (UTC)
Wow. Awesome tool! Thanks a lot. That's almost exactly what I was looking for, and has solved a lot of my complaints with charmap. I'm still a bit confused about why there's not a free, good, fairly complete unicode reference font. At least now I can evaluate the fonts I do have for which parts of unicode they do/do not support. -- CoryCohen 06:23, 16 Sep 2004 (UTC)
There's projects that have tried (there are fonts "freesans" and "freeserif" floating around out there) but the problem is, Unicode is huge, so either you have to pay for good work, or end up with incompleteness or dodgy quality. The best you can hope for now is to have good fonts for particular ranges of Unicode, and then use a modern browser which lets you choose which font you want for which range, and let it put it all together for you. —Muke Tever 17:58, 16 Sep 2004 (UTC)