Spelling in conlangs...

[decem]

I was wondering, do any of your conlangs have deep orthographies with irregular spellings like English? I'd expect most would be fairly phonetic, but perhaps someone was daring to deal with all the IPA or X-SAMPA transcriptions they'd have to do?

My conlang is not fully phonetic but the spelling is very regular and corresponds with pronunciation rules.

[xinda]

The conlang that I'm working on right now has a semi-conservative orthography that I'm planning on developing into something worse in potential daughterlangs. Right now though, the main non-phonemic orthographic features include the historical pronunciation of final glottal stops, representation of /ʋ/ by three different letters, and unedited coronal harmony. I really want to do something interesting with vowels, but that'll have to come later.

<c'odrubthık> "that he is very strong"
/ɕa oʈups̪ək̚(ʔ)/
[ɕotupɕəʔ]

[mbrsart]

I don't have any irregular spellings that come to mind (unless you count the wholesale slaughter of vowels in some dialects), but there are several loan words from a dead language that introduce irregular phonology, most notably geminated consonants. I'll have to throw some irregular spellings in!

[Lambuzhao]

I know this sort of thing would make my 's more realistic. Perhaps, in time, some will develop. However, I kept this sort of thing to a minimum all around. I detested the American English school subject of "phonics" as a kid, and conscientiously avoided trying to replicate any trace of that sort of wild divergence between the spoken and the written in my 's.

...Except in my Gavik. Vowels and diphthongs can have pretty different pronunciations if a tilde/macron is placed over them. Likewise, certain consonants may become fricated or turned into a glide if the following vowel/diphthong received the tilde/macron.

...and somewhat in my Çetara. /c/ and /g/ are realized differently before front\mid\back vowels, similar to Italian & French. And most words that end in -ore or
-ure are really pronounced with just a final accented [u:]; no [re] at all in colloquial speech.

[Orion113]

Due to a weird logogram fetish I have, most of my languages thus far have been semanto-phonetic. However, that doesn't mean there's not room for weird orthography.

There's a lot I could do to make my scripts more natural, introducing different readings, as in Japanese, would be a good start. But so far, I suffer a problem common to many conlangers I know, I can't muster the will-power to gunk up a perfectly good writing system.

[Xing]

I know this sort of thing would make my 's more realistic. Perhaps, in time, some will develop. However, I kept this sort of thing to a minimum all around. I detested the American English school subject of "phonics" as a kid, and conscientiously avoided trying to replicate any trace of that sort of wild divergence between the spoken and the written in my 's.

...Except in my Gavik. Vowels and diphthongs can have pretty different pronunciations if a tilde/macron is placed over them. Likewise, certain consonants may become fricated or turned into a glide if the following vowel/diphthong received the tilde/macron.

...and somewhat in my Çetara. /c/ and /g/ are realized differently before front\mid\back vowels, similar to Italian & French.

Are these regular sound changes? If so, I'd hardly regard them as "wild divergence between the spoken and the written".

[Solarius]

Juvelan, my Scandinavian isolate conlang, has a fairly archaic orthography. For instance, <cyygewut> is [qɑ˧ˤɛ˦my˨] and <ëraeso> is a mere [ɛ˧ˤo˦].

[Click]

My Kàipói uses simple phonemic spelling.

In my artlang, the spelling is a bit more complex, but it is still predictable.
For example, /ʃ ʒ ɕ ʑ ʧ ʤ ʨ ʥ/ are spelled as ‹s ́ z ́ h ́ ġ ́ t ́ d ́ k ́ g ́› (the acute is placed on the following vowel).

[Tsiasuk-Pron]

The main divergence between Itlani spelling and pronunciation is due to the palatalization (I hope that's the right term) of certain consonants when they are followed by the consonant "y."

ovatya - to catch, pronounced: o-'va-cha

doladya - to buy - do-'la-ja

Sometimes it is not sounded at all:

mishya - to go - 'mi-sha

fechya - to mix - 'fe-cha

This can occasionally lead to confusion. For example, fechya (to mix) and fetya (to joke) sound identical in Itlani (in most dialects).

[Click]

Tsiasuk-Pron wrote: The main divergence between Itlani spelling and pronunciation is due to the palatalization (I hope that's the right term) of certain consonants when they are followed by the consonant "y."

Yes, it is the right term.

Tsiasuk-Pron wrote: o-'va-cha do-'la-ja 'mi-sha 'fe-cha

Is that /o.ˈva.tʃa/, /do.ˈla.dʒa/, /ˈmi.ʃa/ and /ˈfe.tʃa/ in the IPA?

[Tsiasuk-Pron]

Tsiasuk-Pron wrote: The main divergence between Itlani spelling and pronunciation is due to the palatalization (I hope that's the right term) of certain consonants when they are followed by the consonant "y."

Yes, it is the right term.

Tsiasuk-Pron wrote: o-'va-cha do-'la-ja 'mi-sha 'fe-cha

Is that /o.ˈva.tʃa/, /do.ˈla.dʒa/, /ˈmi.ʃa/ and /ˈfe.tʃa/ in the IPA?

2-4, khay! Vey tsorni yibya makayare u sholnirdiór onyaru zhoy ra! Secha kozhá isharit ta IPA-teynarun onyaru zhoy vutova iíz fereriya ra-ishyaru. Varvari bashit klanateynafosey sonyaru.

2-4, yes, And now you can perceive that I am in no way a linguist! Although I am somewhat familiar with the IPA notation I don't know how to reproduce it here. Thanks for your transcription.

[Click]

I'm also not a linguist.
As for the IPA notation, you can paste it from the Wikipedia or make a custom keyboard layout.
You can use the X-SAMPA if you can't be bothered to do that because the X-SAMPA can be typed wih a standard keyboard.

[Omzinesý]

Gorrim has some consonants that can be written different ways.
The system is, however, quite regular.
Yet, the lax-tense distinction of vowels is marked in the following consonant (a voiced mark after the tense ones and an unvoiced mark after the lax ones, voicedness is distinctive only word-initially); and palatalisation of consonants is marked in the vowels around it (the Irish way).

[Ànradh]

This was the goal with Iriex but I've not put much thought into it since the writing system would never have appeared in the book. I had have sketched ideas about letting vowel harmony be spelled optionally since the rules were so regular, and, of course, nothing I left unmarked in the romanisation would be marked in the native script either. Not much beyond that though.

[vampireshark]

My conlangs:
·Telèmor is generally shallow orthographically. The consonants are one-to-one correspondences (as an example, <c> is /k/; never /t͡ʃ/ or /s/) Some of the vowels are digraphs with multiple ways of writing them, but some of it arises from stress patterns and plural endings. For example, <èa>, <ea>, and <ae> are all /æ/, but the first is when the syllable cannot be stressed and the third is for certain plural endings. (<aè> can also be /æ/, but rarely and only in certain words.)

·Ilian is perfectly shallow. Each letter corresponds to a single sound, though some palatalization can occur in certain dialects and lead to digraphs. Thus, while <gj> is normally /gj/, it can be realized as [ɟ] in the standard register. (Colloquially, it's just [gj].)

·Oshaháru uses an abjad; this means that, when not vocalized, the orthography isn't very shallow.

·Mís, one of my other on-off projects, uses a very deep orthography, though I'd need to find my notes to be able to give some good examples. (I should computerize them.)

[ian9113]

In one of my langs, <th> can be pronounced /tʰ s θ ʃ ʒ/. Needless to say, this language has a fairly deep orthography, especially when it comes to consonants. Typically, the pronunciation of a word is never set in stone until I put it in the lexicon. For instance, one word, <thet> meaning "road" at least went through stages of being /sɛtʰ/, /θɛt/, /ʃe/, and its current iteration, /sətᵡ/.

[Xing]

In one of my langs, <th> can be pronounced /tʰ s θ ʃ ʒ/.

How did this situation come about? Were /tʰ s θ ʃ ʒ/ originally allophones of the same phoneme, that evolved into distinct phonemes?

[ian9113]

How did this situation come about? Were /tʰ s θ ʃ ʒ/ originally allophones of the same phoneme, that evolved into distinct phonemes?

Yes and no.
<th> represents /tʰ/ (or /tᵡ/) when the <h> is supposed to represent /χ/, which usually transforms into /ʰ/.
<th> representing /θ/ is more historical, because at one point /t/ was dental, and the combination of /t̪χ/ became /t͡θ/ became /θ/.
<th> representing /ʃ/ or /ʒ/ also occurred because of that change: when /t̪/ became /t/, /tχ/ became /t͡ʃ/ became /ʃ/. Voicing occurs between vowels.

I'm not entirely sure, in retrospect, how plausible those changes are, especially to have both a fronting (?) and backing (?) of the same phoneme.

[Ànradh]

My girlfriend's dialect of Gaelic kind of does this, so that /x/ no longer appears: /x/ is palatalised to /ç/ when adjacent to slender vowels as normal, but adjacent to broad vowels it becomes /k/ when syllable final and becomes /h/ when syllable initial.

[Xing]

I'm not entirely sure, in retrospect, how plausible those changes are, especially to have both a fronting (?) and backing (?) of the same phoneme.

What's the triggering condition? I suppose that whatever caused backing anc fronting in various environments, it must have disappeared (since the split seems to have become phonetic). Also, I suppose that the distinction between /ʃ/ and /ʒ/ must have become phonetic (since you write them in slashes).