Zompist Bboard Again

bradrn wrote: ↑Thu Aug 15, 2024 9:21 pm From my observations, the major scripts of the area evolved in generally different ways. Impressionistically, Latin scripts tended to be regularised; Greek letters tended to take on more fluid forms; Cyrillic stayed more or less the same, with a tendency to exaggerate the long strokes of the majuscule script. Gothic could take any of these approaches — or something different entirely.

I once explored a (temporary, visiting?) exhibit about medieval Germans who had traveled to and from the land of Georgia...so maybe also look at one of the scripts for Georgian? ie,https://youtu.be/bUSy13xBPbk?si=wk55yUL79-OChYuY for references.

{I forget which museum in Germany I was visiting at the time, sorry}


Also, maybe combine features, like a little bit of Greek's fluidness, Latin's regularizedness (which appears evenmoreso in one of Georgian's scripts), etc.

My advice would be to get a calligraphy set and play around with it. It’s great fun, and gives a much better idea of how these scripts evolve.

definately do this - its quite very fun indeed, i agree.

keenir wrote: ↑Fri Aug 16, 2024 1:02 pm

bradrn wrote: ↑Thu Aug 15, 2024 9:21 pm From my observations, the major scripts of the area evolved in generally different ways. Impressionistically, Latin scripts tended to be regularised; Greek letters tended to take on more fluid forms; Cyrillic stayed more or less the same, with a tendency to exaggerate the long strokes of the majuscule script. Gothic could take any of these approaches — or something different entirely.

I once explored a (temporary, visiting?) exhibit about medieval Germans who had traveled to and from the land of Georgia...so maybe also look at one of the scripts for Georgian? ie,https://youtu.be/bUSy13xBPbk?si=wk55yUL79-OChYuY for references.

{I forget which museum in Germany I was visiting at the time, sorry}

As far as I’m aware, Georgian (and Armenian) have had little if any influence on the other European scripts. Even Cyrillic shows no signs of Georgian influence. It would certainly be interesting to see a Georgianised Gothic script, but I leave it to Emily to decide if that’s plausible or not — I don’t know the history of these Goths.

(The converse isn’t necessarily true: Armenian in particular has obvious European influence, up to adopting lettercase. I’m not sure why Georgian never went down that path, but it hasn’t. Nuskhuri does have similarities to blackletter (oddly enough, a Western European style!), but Mkhedruli looks quite different to any other neighbouring script.)

All this work is very nice Might give this thread a second read when I have the spoons to actually process more of the technical details but I'm a fan so far

I don't have a lot of conlang experience and I'm not the best at making them sounds naturalistic so I am DEFINITELY taking notes on how you used palatalization for the gentive

thank you that's the kind of detail that can go a long way: a coincidence creates something that's almost a pattern, so the speakers turn it into an actual pattern. but i may need to do some more tests with it to see how feasible it is to retain it by the end of phase 3, where the palatal consonants have largely merged and, say, /ʃ/ could be the palatal form for any number of sounds. would the potential ambiguity be so great that speakers would ditch it? only precision laboratory testing will tell

month names
got the final versions of the month names worked out! i had come up with the original names a while back, but only recently have i worked out the details (including inflections) through the last phase

the only possibly attested month names in gothic are naubaimbair (november) and jiuleis ("yule"), both of which are problematic. (the word "jiuleis" appears in something that may be a phrase "fruma jiuleis", that is "before yule", or alternatively may be a compound word "frumajiuleis".) i ended up deciding that "frumajiuleis" would be a compound word used for november. i set jiuleis as december and "aftumajiuleis" as january. in retrospect i think it probably would have been more accurate to have frumajiuleis as november and aftumajiuleis as december, and then form a different compound for january. but for the time being that stands

at least three early germanic languages had their own calendar of months, but they may or may not line up with julian/gregorian months and at any rate are not at all consistent between each other. so i used them more for inspiration than exact derivation. in addition to early germanic calendars, i also looked to slavic month names (the ones not derived from latin, obviously). again, they aren't necessarily consistent between languages, but they do have more overlap than the extant germanic ones seem to. the final list i came up with:

• yule months:
  • november: frumajiuleis "before yule" → þrumajais /ðrúmajɛs/ → ðrumjes /ðrúmjɛs/ → ðrumjes /ðrúmjɨs/
  • december: jiuleis "yule" → iujais /jájɛs/ → jejmalz /jéjmælz/ (month element "-malz" added) → jimus /jéməs/
  • january: aftumajiuleis "after yule" → aftaumjais /áftɔmjɛs/ → aftojes /æftójɛs/ → aftojes /æftɒ́jɨs/
• word + "mēnōþs" (month) compounds:
  • february: haírutmēnōþs "deer month, stag month" → hairaudmenþs /hɛ́rɔdmenðz/ → erunbalz /ɛrúnbælz/ → erundus /ɛrúndəs/
  • march: grasmēnōþs "grass month" → grasmenþs /grázmenðz/ → granbalz /gránbælz/ → grandus /grándəs/
  • april: áustrōmēnōþs "Easter month"** → ausdromenþs /ɔ́ʊzdromenðz/ → ozdrumalz /ɔzdrúmælz/ → ozdrumus /ɔzdróməs/
  • may: áikmēnōþs "oak month" → aigmenþs /ɛ́ʊgmenðz/ → engbalz /éŋbælz/ → engus /jéŋgəs/
  • june: falgōmēnōþs "fallow month" → falgomenþs /fálgomenðz/ → falgumalz /fælgúmælz/ → fogumus /góməs/ (in careful speech /vgóməs/)
  • july: hawimēnōþs "hay month" → hawaumenþs /hávɔmenðz/ → awmalz /ávmælz/ → awmus /wávməs/
  • august: asanmēnōþs "harvest month" → asanmenþs /ásanmenðz/ → azomalz /æzómælz/ → azomus /æzɒ́məs/
  • september: triumēnōþs "wood month" → ddjiumenþs /ʤjámenðz/ → zjamalz /ʒámælz/ → djamus /ʒáuməs/
  • october: weinamēnōþs "wine month" → wonamenþs /vwánamenðz/ → wonmalz /vónmælz/ → womus /vɒ́məs/

**(if there was an actual gothic cognate to english "easter" it didn't survive into any existing texts, and at least in the christian sense had been replaced by greek "paskha")

coming up with this list necessitated figuring out the words for deer/stag, oak, fallow, and easter, which were all derived from proto-germanic. the element /menoþ/ had been reduced by phase 2 to /mal/, which was further mangled to /bal/ after some consonants; the one-two punch of phase 3's loss of /l/ in the syllable coda and unstressed vowel reduction in final syllables led to the reinterpretation of everything after that element's initial consonant (/m/, in some cases changed to /b/ or /d/ or /g/!) as a case ending. however, the dative singular ("in July") and genitive singular ("the 14th of March") are possibly the most commonly used forms for any month name, and their endings begin with vowels, meaning the /l/ deletion wasn't triggered. so in these two positions only, i've retained the -al- ending (which is rendered /-áulɨ/ <-ale> in dat sg and /-áʒɨs/ <-eljes> in gen sg)

conversely, as noted above, the word for december added the -mal element, as it was shorter than all the other month names and would have otherwise been the phonetically rather indistinced /jéjɛs/. but of course this element went through the same reduction as described in the last paragraph

i want to link to the dictionary entries i have on anthologi.ca but i'm having permissions issues so as a consolation prize i'll include a sound change development chart explaining the singular forms for "march" (numbers refer to specific sound changes; see the relevant post):

found some scans of (non-calligraphic/non-uncial) handwritten gothic from a couple of property deeds (one here, and a portion of another one here), so i used the opportunity to develop lowercase versions of the letters. the capital versions are just standardized versions of the calligraphic forms (shown below in the form that shows up when i paste the letters into ms paint), but i worked out the lowercase versions by hand, and then (since i don't have my font creator program on this computer) edited together the best approximation based on times new roman. (i have two versions of the lowercase "j" in the image, neither of which i'm completely satisfied with, but really it's a font issue; the descenders in times new roman are too small and it makes the letter look weird either way.) the latin letters above are the transliteration (the gothic letters <𐍁> and <𐍊> had no phonological value and were included in the alphabet solely for use as numerals; in modern gothic i've pressed the former into use to represent /ð/). first draft obviously, curious to hear people's thoughts

I really like it!

My first thought is that a few of those lowercase letters look overly Latinate in their details, especially ⟨r a⟩ /ɡ d/. Meanwhile ⟨ρ⟩ /þ/ looks overly Greek (and is misaligned). Also, some letter pairs look rather close: most notably /u p/ and /r f/.

More broadly, the whole thing is mildly inconsistent: e.g. I can see two separate styles of serifs. [EDIT: Interestingly, looks like that’s a property of the original Times too.] Furthermore the lowercase is not yet harmonised with the uppercase. I think a lot of problems should work themselves out as you continue thinking about this situation. I’d suggest taking the lowercase /a j/ as a starting point for refining the style of the other letters: I really like their current design, and out of all these letters they feel the most characteristically ‘Gothic’ to me. Trying to actually handwrite the letters would help a lot too. (I recommend doing so with a broad-edged nib, if you have one.)

Also, regarding /j/: I think I like the rightmost variant a little more. But perhaps try flattening out that acute angle a bit? You could look at the design of glyphs like ⟨ꞔ 𝼝 ᶗ ʂ ᶎ ʐ⟩ for inspiration. (Indeed the first of these is almost identical to yours, though not quite the same: its hook is a little smaller.)

bradrn wrote: ↑Fri May 16, 2025 3:33 am I really like it!

My first thought is that a few of those lowercase letters look overly Latinate in their details, especially ⟨r a⟩ /ɡ d/. Meanwhile ⟨ρ⟩ /þ/ looks overly Greek (and is misaligned). Also, some letter pairs look rather close: most notably /u p/ and /r f/.

More broadly, the whole thing is mildly inconsistent: e.g. I can see two separate styles of serifs. [EDIT: Interestingly, looks like that’s a property of the original Times too.] Furthermore the lowercase is not yet harmonised with the uppercase. I think a lot of problems should work themselves out as you continue thinking about this situation. I’d suggest taking the lowercase /a j/ as a starting point for refining the style of the other letters: I really like their current design, and out of all these letters they feel the most characteristically ‘Gothic’ to me. Trying to actually handwrite the letters would help a lot too. (I recommend doing so with a broad-edged nib, if you have one.)

Also, regarding /j/: I think I like the rightmost variant a little more. But perhaps try flattening out that acute angle a bit? You could look at the design of glyphs like ⟨ꞔ 𝼝 ᶗ ʂ ᶎ ʐ⟩ for inspiration. (Indeed the first of these is almost identical to yours, though not quite the same: its hook is a little smaller.)

so, the question of the styling, serifs, etc. — everything relating to the font rather than the letterforms themselves — is not really what i'm asking about, since as mentioned i don't have a working font program at the moment so these were cobbled together from times new roman characters in an old photoshop knockoff. if i'm understanding you correctly irt harmonizing with the capitals that might be the same issue, since TNR doesn't have gothic letters so a different font had to be used for the capitals in the image

i did do the initial development by hand, though it was with a pencil as i don't have any calligraphy pens. i'm curious what you mean by certain letters looking "overly latinate" or "overly greek". is it the shapes of the letters that looks this way, or is this another issue of the font (e.g. the distribution of the thick vs thin parts of curves, etc)? also unclear what you mean by the <þ> character being "misaligned"

i think you certainly do raise valid points about certain characters looking similar (although i will point out that the difference between my <r> and <f> is the same as the difference between latin <n> and <h>!). it's interesting that you bring up both <u> and <p> as i originally had different forms for both of these. the earlier version of <u> was a mirror image of the <n> form, which i rejected as being too similar to that letter. the <p> originally looked like a latin <n>, but then that would look like a small version of the capital <U>. so i then went with a form that looks similar to the current version but with the top bar sticking out to the left; it looked great handwritten, but bizarre in the font version. i might give it another go. also, just as an aside, the descender on <ð> can either go straight down as depicted, or curve to the right

it would probably be helpful if i post the handwritten versions, i can do that later tonight

Emily wrote: ↑Fri May 16, 2025 3:55 pm it would probably be helpful if i post the handwritten versions, i can do that later tonight

I think this would be extremely helpful, yes!

so, the question of the styling, serifs, etc. — everything relating to the font rather than the letterforms themselves — is not really what i'm asking about, since as mentioned i don't have a working font program at the moment so these were cobbled together from times new roman characters in an old photoshop knockoff.

I do understand this. But I also think that stylistic harmony is more important than commonly thought — it helps to make the whole thing look like a single writing system.

(Also, might I suggest using a vector editor such as Inkscape? Generally font design is done in vectors, not rasters.)

if i'm understanding you correctly irt harmonizing with the capitals that might be the same issue, since TNR doesn't have gothic letters so a different font had to be used for the capitals in the image

I believe so, yes.

i did do the initial development by hand, though it was with a pencil as i don't have any calligraphy pens. i'm curious what you mean by certain letters looking "overly latinate" or "overly greek". is it the shapes of the letters that looks this way, or is this another issue of the font (e.g. the distribution of the thick vs thin parts of curves, etc)?

It’s sort of… a general feeling about how those scripts usually look, I guess? Mostly involving the overall structure of the letterforms. Details like serifs and thicks and thins certainly contribute, I think, but they’re secondary. Like, to take a few examples, trying to explain as best I can:

• ⟨þ⟩ is very obviously from Greek ⟨ρ⟩. To me the smooth transition between the loop and the descender feels very characteristic of Greek — Latin and Cyrillic tend to make them two separate strokes, as in ⟨p⟩ which has the straight line continuing slightly above the intersection. (But the letter also has a Latin serif at the bottom.)
• ⟨u⟩ is very obviously from Latin ⟨n⟩. The shape of the rightmost stroke, curving back down from an oblique angle, feels very distinctively Latinate to me. (Armenian is the only other script I can think of which uses this shape regularly.) Cyrillic tends to prefer straight lines at right angles, as in ⟨п⟩; Greek also likes right angles, though more often they’re not quite exactly right angles, and they use circular rather than straight lines.
• As a counter-example, the most distinctive of these letters is ⟨a⟩. Neither Greek nor Latin nor Cyrillic make much use of oblique descenders like that in this letter. Neither do they make much use of triangles, at least in the lowercase. (To a first approximation, Latin tends to round them off; Cyrillic tends to square them off; Greek has mostly kept them in the uppercase, but made them more rounded in the lowercase.)

also unclear what you mean by the <þ> character being "misaligned"

To me it looks off-centre compared to the other letters — the bowl feels like it ‘should’ be centred rather than an ascender.

all right, here are the handwritten forms (sorry for the bad cell phone pictures). this is the lowercase forms, followed by the capital and lowercase versions of each letter next to each other: and here are some words written out: the top three panels are the declensions for, respectively, ods "dog", zgwohs "shoe", and adus "hand". the rest of the image is words from the lexember thread. the backwards tilde appearing over some of the consonants (or consonant clusters) indicates palatalization and derives from a gothic <j>

and here are some images from the development tests. first is the letter forms i found in the aforementioned deeds: these (particularly the forms from the arezzo deed) formed the basis of my experiments

here are some of the tests themselves (second image is annotated):

Thanks!

Those handwritten forms indeed look a lot more coherent than the Times-based rendering. Perhaps I’ll have a go myself at turning it into something more polished.

bradrn wrote: ↑Sat May 17, 2025 2:58 am Those handwritten forms indeed look a lot more coherent than the Times-based rendering. Perhaps I’ll have a go myself at turning it into something more polished.

i would prefer to do it myself, to be honest. but i'm curious to hear your thoughts (or anyone else's of course) about the forms having now seen the written versions

Emily wrote: ↑Sat May 17, 2025 3:13 am

bradrn wrote: ↑Sat May 17, 2025 2:58 am Those handwritten forms indeed look a lot more coherent than the Times-based rendering. Perhaps I’ll have a go myself at turning it into something more polished.

i would prefer to do it myself, to be honest. but i'm curious to hear your thoughts (or anyone else's of course) about the forms having now seen the written versions

To be honest, it’s quite hard to say without having actually attempted it myself…

For now, I think my main thought is that letterforms which originally seemed very Latinate look distinctly less so when handwritten. Compared to Latin, it looks to me like ⟨a⟩ has a relatively smaller bowl, ⟨u⟩ is missing a stroke, ⟨ʀ⟩ /r/ has a stroke sticking out the top and (again) a smaller bowl, and ⟨y⟩ has a stem curving the other way. Similarly, ⟨п⟩ quite consistently has a stroke sticking out the left which is absent in Cyrillic. And /o/, which looked like an upside-down ⟨ɤ⟩, here looks more like an ⟨ℓ⟩.

This is a very nice, well-designed script - I like it! Rock on!

Seconded. If I had to describe Modern Gothic with one phrase, it would be "fanciful realism," and this script fits that vibe perfectly.

I have to say that I very much like the handwritten forms myself.

I think they're nice, speaking as someone who tried and failed to do the same thing many years ago.

I'm a huge fan of Germanic langs and I'm really liking what I'm seeing here; can't wait to see more!

alice wrote: ↑Tue May 20, 2025 2:32 pm I think they're nice, speaking as someone who tried and failed to do the same thing many years ago.

So you'd say Dekavurian was a failure?