zompist bboard

My native Polish has a lot of them and I find them annoying, difficult to type using the AltGr key and slowing down my typing speed.
I even tried to make my own keyboard. I would use non-Polish letters like q, x, v for ó, ś, ć respectively (as the "-ść" combination is pretty common in Polish (deriving from earlier "-st") but I was still forced to use AltGr for other characters, there's just too few keys on the keyboard for all the letters.
You might wanna know that most Polish users don't use diacritics when they don't have to, typing words "miłość" as "milosc" and so on. Usually it's not difficult to understand.

So I started thinking how to get rid of diacritics altogether?
I think I was inspired by Croatian which uses "nj' for /ɲ/. So you could go further with this and change Polish [ś] into [sj], [ź] into [zj], [ń] into [nj] but it gets trickier with the others:
ł /w/ which historically developed from /ɫ/, ż /ʐ/, rz /ʐ/ (historically /rʲ/, ch /x/ (pronounced the same as [h], historically /ɦ/ I think), ó /u/ (historically /ɔ:/, now pronounced the same as ), and ę /ɛ̃/, ą /ɒ̃/.

I also like Albanian having [x] for /dz/, it would make sense to do the same in Polish since a single [c] is used for /ts/ too. So Polish [dź] could be spelt as [xj] too, by analogy. And possibly /dʐ/ would be [xz], by analogy to [sz] /ʂ/, cz /tʂ/, rz /ʐ/.

So we're left with [q] and [v] unused. Any ideas?

The difficulty here is that Polish doesn't have very restrictive phonotactics, words like "ćpun" /tɕpun/ and "rtęć" /rtɛ̃tɕ/ are allowed, for example.

http://steen.free.fr/cyrpol/
Поврут Таты, през А. Мицкевича

Пуйдьте о дятки, пуйдьте вшистке разэм
За място, под слуп на взгурэк,
Там пред цудовным клѩкнийте образэм,
Побожне змувте патëрэк.

Тато не враца ранки и вечоры
вэ лзах го чекам и трводзэ;
Розлялы реки, пэлнэ зверя боры,
И пэлно збуйцув на дродзэ;-

Слышѫц то дятки бегнѫ вшистке разэм
За място под слуп на взгурэк,
Там пред цудовным клѩкаѭ образэм,
И зачинаѭ патëрэк.

Цалуѭ земѩ, потэм в Имѩ Ойца,
Сына и Духа свѩтэго,
Бѫдь похвалëна пренайсвѩтша Труйца
Тэраз и часу вшелькего.

or

/p b t d k g/ <p b t d k g>
/ts dz ts` dz` ts\ dz\/ <c x ch xh ty dy>
/m n J/ <m n ny>
/f v s z s` z` s\ z\ h/ <f v s z sh zh sy zy h>
/l r j w/ <l r j w>
/a e o i 1 u/ <a e o i eu u>
/e~ o~/ <enh onh>

using a digraph for /1/ frees up <y> to mark palatalization

Povrut Tateu, pshez A. Mickyevicha

Pujdtie o dyatki, pujdtye vsheustkje razem
Za mjasto, pod swup na vzgurek,
Tam pshed cudowneum klenhknijtye obrazem,
Pobozhnye zmuvtye patyorek.

Tato nye vraca ranki i viechoreu
ve wzah go chekam i trvoxe;
Rozlaweu zheki, pewne zvjerza boreu,
I pewno zbujcuv na droxe;-

Sweushonhc to dyatki bjegnanh vsheustkie razem
Za mjasto pod swup na vzgurek,
Tam pshed cudowneum klenhkajanh obrazem,
I zacheunajanh patyorek.

Cawujanh zyemjenh, potem v Imjenh Ojca,
Seuna i Duha syvjenhtego,
Bonhx pohvalona pshenajsyvienhtsza Trujca
Teraz i chasu vshelkjego.

You could do what the new Kazakh orthography does and just make all the digraphs with `!

Seriously, though, it's much easier if the phonotactics are restrictive. But, in a pinch, you can always use ' or - to disambiguate clusters from digraphs. Or, as in one version of the romanisation I like for Mongolian, use dedicate characters (e.g. "h" and "z") that only appear in digraphs. also e.g. "x" in maybe the most popular alternative orthography for Esperanto.

Šọ̈́gala wrote:Seriously, though, it's much easier if the phonotactics are restrictive. But, in a pinch, you can always use ' or - to disambiguate clusters from digraphs. Or, as in one version of the romanisation I like for Mongolian, use dedicate characters (e.g. "h" and "z") that only appear in digraphs. also e.g. "x" in maybe the most popular alternative orthography for Esperanto.

Or just don't sweat it too much? Most natlang orthographies exhibit some ambiguity in this area, e.g. English pothole vs pothos. If you know the vocabulary, it's seldom a problem.

Good point. However, people tend to shy away from that for their conlang, since it starts out with literally 0 people knowing the lexicon. So if there's ambiguity in the orthography, you'll need an additional set of notes with IPA pronunciations or whatever. Not a big deal; I'm sure some people keep vocab lists for their conlangs with IPA in addition to the practical orthography anyway.

Šọ̈́gala wrote:Good point. However, people tend to shy away from that for their conlang, since it starts out with literally 0 people knowing the lexicon. So if there's ambiguity in the orthography, you'll need an additional set of notes with IPA pronunciations or whatever. Not a big deal; I'm sure some people keep vocab lists for their conlangs with IPA in addition to the practical orthography anyway.

If you don't want to go the full length to IPA, you can add a pronunciation note just showing whether the digraph stands for one or two phonemes (e.g. pothole (pot-hole).

Šọ̈́gala wrote:e.g. "x" in maybe the most popular alternative orthography for Esperanto.

I never understood how an auxlang could have so many diacritics AND leave out latin that could be used instead.
"ĉ" could be spelt as "q", "ĝ" as j, "j" as "y", "ĥ" - does anyone even hear the difference between ĥ and h? Are there any languages that have both the sounds, in non-allomorphic distribution? ĵ could be "j", ŝ could be "x", ŭ could be just "u" as in any natural language, I know what it's like in my native Polish - most teachers will tell you to pronounce "nauka" (study, science) as /na'uka/, instead /'nauka/ as most tend you pronounce it, maybe that's what Zamenhof couldn't stomach but I couldn't care less.
This would spare people the use of the clumsy digraphs with -x. Most Esperanto words hardly resemble their source vocabulary anyway.

awer wrote:Are there any languages that have both the sounds, in non-allomorphic distribution?

Arabic does

There are many examples: Kazakh, Kalmyk, Scottish dialects of English. Perhaps Proto-Indo-European.

awer wrote:"ĥ" - does anyone even hear the difference between ĥ and h? Are there any languages that have both the sounds, in non-allomorphic distribution?

In addition to those mentioned above, I'll add Arapaho, Cheyenne, Delaware, Crow, pretty much all of the languages of the Pacific Northwest, many of the Khoisan languages, many languages of the Caucasus such as Lezgian and Lak, the Tibeto-Burman language Nuosu and a number of languages in South America such as Guahibo, Amahuaca and Maka among many others. It's not really that ridiculous a distinction to make, considering what other kinds of distinctions are made in consonant inventories.

Also minor nitpick: in this case you'd want to talk about "allophony"; allomorphy is when an inflection form varies in terms of its shape for whatever reason, and operates on the phonological level, not the phonetic level.

I don't see how typing alt+s is any slower than typing s-j. It's two characters either way.

awer wrote:"ĥ" - does anyone even hear the difference between ĥ and h? Are there any languages that have both the sounds, in non-allomorphic distribution?

Me, another native Polish speaker. Actually, both of these sounds are present in my speech. The truth is I have many sounds you probably don't and my Polish could be incomprehesible to you.

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:

awer wrote:"ĥ" - does anyone even hear the difference between ĥ and h? Are there any languages that have both the sounds, in non-allomorphic distribution?

Me, another native Polish speaker. Actually, both of these sounds are present in my speech. The truth is I have many sounds you probably don't and my Polish could be incomprehesible to you.

Oo, I'm interested! Tell us more!

Ok, It'll be long but I'll try to describe it:
1. No "classical" phonemic nasal vowels: Ę and ą are /ɛN/ and /ɔN/ word-internally (N assimilates in POA to the following consonant) and /ɛ(m)/ and /ɔm/ word-finally.
2. /ɛ/ is backed, something like [ɜ] and sometimes it is identical to /ɨ/ ([ɘ]).
3. Intervocalic /g/, /gʲ/, /v/, /vʲ/, /b/, /bʲ/ get lenited to [ɰ/w/j j ʋ/w ʋʲ/ɥ β̞/w β̞ʲ/ɥ] and often disappear entirely along with /w/ and /j/ (still only between vowels).
4. Sequences like /ts tʂ dʐ/ are affricated to [t̪͡s̪ t͡ʂ d͡ʐ].
5. All the hiatuses undergo crasis, creating things like [øː œːeː yː æː] and maybe something else too (that is rather complicated and I don't wanna list all the possiblities.
6. Preconsonantal /f/ and /p/ are [h] going through an intermediate step of [ɸ]
7. Presonorantal t is (I suppose) [θ], though I can't be sure.
8. When there is VNV[-stress], the nasal becomes [w̃ ɰ̃ j̃] depending on its place of articulation and if the second vowel is a high vowel, it turns to VṼ.
9. /ɔ/ becomes [œ] unless it is next to [w].
10. /a/ is often fronted to /æ/, especially in the vicinity of palatal(ised) consonants.
If anyone would like to see specific words, I can give some transcriptions and/or audio files.

All my friends don't speak like that, they try to stick to the standard Polish except for the first change, which didn't occur only for two girls or so. It annoys me a lot as I don't believe in proper versions of languages but I try not to show that.

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:Ok, It'll be long but I'll try to describe it:
1. No "classical" phonemic nasal vowels: Ę and ą are /ɛN/ and /ɔN/ word-internally (N assimilates in POA to the following consonant) and /ɛ(m)/ and /ɔm/ word-finally.
2. /ɛ/ is backed, something like [ɜ] and sometimes it is identical to /ɨ/ ([ɘ]).
3. Intervocalic /g/, /gʲ/, /v/, /vʲ/, /b/, /bʲ/ get lenited to [ɰ/w/j j ʋ/w ʋʲ/ɥ β̞/w β̞ʲ/ɥ] and often disappear entirely along with /w/ and /j/ (still only between vowels).
4. Sequences like /ts tʂ dʐ/ are affricated to [t̪͡s̪ t͡ʂ d͡ʐ].
5. All the hiatuses undergo crasis, creating things like [øː œːeː yː æː] and maybe something else too (that is rather complicated and I don't wanna list all the possiblities.
6. Preconsonantal /f/ and /p/ are [h] going through an intermediate step of [ɸ]
7. Presonorantal t is (I suppose) [θ], though I can't be sure.
8. When there is VNV[-stress], the nasal becomes [w̃ ɰ̃ j̃] depending on its place of articulation and if the second vowel is a high vowel, it turns to VṼ.
9. /ɔ/ becomes [œ] unless it is next to [w].
10. /a/ is often fronted to /æ/, especially in the vicinity of palatal(ised) consonants.
If anyone would like to see specific words, I can give some transcriptions and/or audio files.

All my friends don't speak like that, they try to stick to the standard Polish except for the first change, which didn't occur only for two girls or so. It annoys me a lot as I don't believe in proper versions of languages but I try not to show that.

I'm also Polish and I've never heard of a dialect remotely resembling what you describe. Where are you from? My gut reaction reading this was "This can't be real. This guy's gotta be trolling us."

There

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:There

IMG_20171130_100057.png

Yeah, I've seen that but I know some people from Lodz and they speak nothing like what you're describing...

To be fair, he did say:

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:All my friends don't speak like that, they try to stick to the standard Polish except for the first change

Also Hebrew has that distinction. Diacritics can be useful in guiding pronounciation. Adding unused latin characters in their place can lead to a pretty unintuitive orthography for outsiders tryng to learn the language. Your example, Albanian, is pretty bad at this, with <x q> being used for values completely unrelated to how they are used in any other language. <x> as a filler consonant in Esparanto, on the other hand, makes sense because <x> is often associated with marks that are completely unrelated to writing but are just checks for convenience. ("x marks the spot", "mark x for yes", "x means absence", etc.)

gestaltist wrote:I don't see how typing alt+s is any slower than typing s-j. It's two characters either way.

The 'alt' key is further from the 'home' position, and one has to be careful that one figures don't slip off the bottom of the keyboard. I have a keyboard layout for the Lanna script that caters for the alphabet listed in abecedaries and I have found that using the '!' position as a dead key can work better than AltGr. It partly depends on how one positions the keyboard.

mèþru wrote:Also Hebrew has that distinction.

Isn't it just /h/ vs. /χ/ in Modern Hebrew, though?

Well, ĥ can be realised as /χ/ among some speakers and Hebrew /χ/ can be /x/ for some speakers. According to Wikipedia it's even /ʀ̝̊/ for some Hebrew speakers (didn't know that but I doubt I can hear the difference, while I know I sometimes vary between /χ/ and /x/ in my own speech). In addition, some Mizrahim have a three way distinction between /x~χ/, /ħ/ and /h/.

It should be entirely possible to use digraphs to render the Slavic languages (at least) without diacritics. See my post about using Gaelic spelling to write Russian: viewtopic.php?f=4&t=40092&p=1143994#p1143994 Polish spelling has advantages, but Polish spoils the soup with dotted consonants that represent the same sound as ones with digraphs. Actually, Magyar, the Hungarian language, has a marvelous way of writing consonants without diacritics, but it defeats the whole business with all the vowel accents and double accents. Magyar even has a 'ж', and it's written 'zs', as opposed to 'sz' which is used to write the sound of Russian 'с'. I was looking at Magyar spelling as a possible orthographical system for Rromanés (the language of the Romani, or as most of them call themselves when speaking English, Gypsies). There were exciting polemics going around a few years ago as to how Rromanés should be spelled. For a while, there were Rromanés pages on Wikipedia(c) in devanagari, the Sanskrit alphabet (since the Roma originated in India)-- it's actually a lovely and fitting orthography, but a bit difficult to learn. I have on an external disk drive a few experimental scribblings in the direction of a Roman spelling of Russian, one that more than Cyrillic (in my opinion) illuminates the etymology (true origin) of individual Russian words. It could even be stretched to spell English phonetically. Unfortunately, my operating system has lost the system address of the drive, and it will be some time before I can recover these musings of mine. Actually, the native orthography of each respective language has a charm of its own and has been carefully crafted over the centuries to fit native needs. Nevertheless, an easier phonetic spelling might be useful to many as "training wheels". I'm currently becoming enamoured with Irish Gaelic orthography; one can try for years to get used to it and finally (if their experience is like mine) "the light goes on".

Atlanta Bill wrote:I'm currently becoming enamoured with Irish Gaelic orthography; one can try for years to get used to it and finally (if their experience is like mine) "the light goes on".

QFT. This was certainly the case for me as well and over the years I've noticed that the many many complaints about how difficult it is are never from people who regular speak it or have made a serious study of the language. They basically equate to "It's not English so it's crazy!"--conveniently ignoring the fact that English orthography is far more inconsistent and confusing than Irish at its worst.

Rromanes is not really the name of the language in Romani itself but rather an adverb (literally 'Romani-ly', meaning something like 'in the Romani way').