Things Decided for Stupid Reasons

[TurkeySloth]

I like <ň> for /ŋ/. <ñ> is also fine. <ƞ> looks nice. <ņ> would be cool if the cedilla connected to either one of the "legs", but it doesn't (at least in any font that I know, and if it does in some stylistic font, it's basically irrelevant), so it's generally not what I'd use even if every other diacritic in the orthography was a cedilla. <ṋ> or <ṉ> work, too. You could even use <ᵰ> or <ʼn>, or maybe <ǹ> if <ń> was used for the palatal nasal or something. The common romanisation of Sanskrit and other Indian languages uses <ṅ>, which is also alright. Actually, <ng> is one of the few digraphs I really like as well.

...basically, in my opinion it doesn't matter how it's romanised as long as it's some kind of "N-letter". Sometimes even letters that look nothing like N could work. Anything but <q>, really.

Except that <ƞ> stands for a vowel, [i]—specifically, in Greek. Speaking of, while my need to have letters fit into the same phonological category as their parent language may not be stupid, it prevents me from using some more inventive initial transcriptions. "Initial transcriptions" because I refuse to call use of non-Latinate letters "Romanization."

N with a long right leg <ƞ> is a different letter than eta <η>.

Very true. However, that's an understandable mistake because the letters look exactly the same. Also, note that n with a long right leg comes up as a selling error, whereas eta doesn't. For the record, the same is true of other pairs of Greek letters and Latinate variants thereof.

[Zaarin]

Very true. However, that's an understandable mistake because the letters look exactly the same. Also, note that n with a long right leg comes up as a selling error, whereas eta doesn't. For the record, the same is true of other pairs of Greek letters and Latinate variants thereof.

They look different in Times New Roman, for what it's worth. This board's font unhelpfully conflates a lot of symbols, including lowercase eks and lowercase chi. At least it registers combined diacritics now...

[Pabappa]

Not mine, but David J. Peterson has remarked that if he's working on a language where, for whatever reason, he has decided the orthographic representation of long vowels will be a doubled vowel letter, then that language can never have long /e/ and long /o/ - he always gets rid of them by dipthongizing them or something else. This, he says, is because no matter what, no matter how many explanations are given, the actors will always, always read orthographic <ee> and <oo> as /i/ and /u/.

I know I answered this once already, but I've been working on paleo-Pabappa again and it came to me that I think there is a perfect balance point somewhere when you can have just enough use of doubled vowels but not too much. Perhaps its my English speaking background , but I think English pretty much gets it right. German is another good example. Dutch uses double vowels too much ... I've heard the same said of Finnish, but as I said earlier in this thread, I
prefer double vowels whenever the long vowels pattern the same way as diphthongs and are comparable in frequency to diphthongs. Thus Finnish has not just <oo>, but also <uo>, <iö>, <yö>, etc. Whereas Dutch's diphthongs seem less common than the simple vowels and I think may even contrast in length themselves. So if I were designing the orthography for Dutch, I would want to use macrons.

[Dē Graut Bʉr]

Diphthongs in Dutch do not contrast in length. Perhaps you're confused by the existence of "short ei" and "long ij", but those are pronounced identically, with the words "short" and "long" here serving only to distinguish between the two different spellings.

[Pabappa]

Diphthongs in Dutch do not contrast in length. Perhaps you're confused by the existence of "short ei" and "long ij", but those are pronounced identically, with the words "short" and "long" here serving only to distinguish between the two different spellings.

Okay, thanks, I wrote that from my phone and without looking anything up. But I was really thinking of words like kraai and leeuw, and Im sure there are others, where we have a long vowel followed by a glide. it seems this is not considered a diphthong in Dutch, but in the raw phonetic sense, I think it still is .... it seems there still is never a pure length contrast, but if I was to redesign an orthography for Dutch that replaced all of the double vowels with macrons, I think I'd use macrons over these double vowels too. Is there ever an <eei> at the end of a word?

[Alargule]

I (native Dutch speaker) have never seen it. Probably because it would be too similar to the ee sound, like in nee, mee, ree, slee, gedwee etc.

[jal]

I think y'all are confusing stuff here. It's better to use the terms tense and lax instead of long and short when talking about Dutch phonology, as a number of traditional "long" vowels are actually just as short as the traditional "short" vowels. Then there's r-lengthening, but that's allophonic, and there's words of foreign origin that may receive length (even where the donor language doesn't have it, like <beige> [bE:Z@]). Wikipedia has a pretty good overview. As for <kraai> and <leeuw>, these are considered diphthongues, but they aren't any longer than diphthongues that have a lax starting point.

As for Dutch orthography, it's a bit of a mess, though macrons wouldn't help it :).


JAL

[din]

As for Dutch orthography, it's a bit of a mess, though macrons wouldn't help it .

Really? I think it makes a hell of a lot more sense than just about any other orthography used for a language spoken in western Europe. There are a few arbitrary rules and a few exceptions, but for the most part, it's pretty straightforward

[jal]

Really? I think it makes a hell of a lot more sense than just about any other orthography used for a language spoken in western Europe. There are a few arbitrary rules and a few exceptions, but for the most part, it's pretty straightforward

Ok Din, really, how much other ortographies do you know? Of course, English and French are bad, and so is Danish, but I find, say, Spanish, Italian, Polish, Swedish, Finnish, and any flavour of former-Yugoslavian a lot easier to pronounce based on spelling than Dutch is. And German isn't worse than Dutch, imho. And the ridiculous stupidity of representing tense vowels by a double vowel, but only in closed syllables, otherwise we use a single vowel but hey, that's also used for lax vowels in closed syllables, so we just double the consonant then, but only intervocally, otherwise we don't.

EDIT: Oh, and I forgot to fulminate against the "d" and "b" are written at the end of a word but pronounced [t] and [p], but hey, let's write "v" and "z" as "f" and "s" instead, to fuck up consistancy.


JAL

[Alargule]

I'd say Dutch orthography is 'consistent within its inconsistency': the rules per se might not make a whole lot of sense, but at least it does so in a regular and predictable way. Apart from six year olds and cases of severe dyslexia, I never see any mistakes made against short or long vowels ('wij maaken' instead of 'wij maken' or 'wij stopen' instead of 'wij stoppen'). The real difficulty seems to lie in the d/t distinction at the end of past participles, or rather: whether that distinction should only be applied to pp's (yes) or to 2nd and 3rd person singular present tense as well (no). So it's 'hij gelooft' and 'hij heeft geloofd', but due to the latter you see 'hij geloofd' a lot - 'hij heeft gelooft' seems to be a lot less frequent.

[din]

Really? I think it makes a hell of a lot more sense than just about any other orthography used for a language spoken in western Europe. There are a few arbitrary rules and a few exceptions, but for the most part, it's pretty straightforward

Ok Din, really, how much other ortographies do you know? Of course, English and French are bad, and so is Danish, but I find, say, Spanish, Italian, Polish, Swedish, Finnish, and any flavour of former-Yugoslavian a lot easier to pronounce based on spelling than Dutch is. And German isn't worse than Dutch, imho. And the ridiculous stupidity of representing tense vowels by a double vowel, but only in closed syllables, otherwise we use a single vowel but hey, that's also used for lax vowels in closed syllables, so we just double the consonant then, but only intervocally, otherwise we don't.

EDIT: Oh, and I forgot to fulminate against the "d" and "b" are written at the end of a word but pronounced [t] and [p], but hey, let's write "v" and "z" as "f" and "s" instead, to fuck up consistancy.


JAL

I'm pretty familiar with the orthographies of all languages spoken in Europe (unless they're really tiny). Do note that I said 'for a language spoken in western Europe'. Sure, Spanish, Italian and Finnish are more regular, but German sure as hell isn't. Just think of how they deal with long vowels... I think Alargule describes it well when he says that the inconsistency is applied consistently. I definitely wouldn't have made the same decisions if it were up to me to come up with an orthography, but when I taught the language I was pretty much always able to provide my students with a spelling rule for every situation, and we actually never spent too much time on the subject, because most of them mastered it pretty quickly.

[statelessnation]

redacted

[HourouMusuko]

I dropped /θ/ from my main conlang (and replaced all instances of it with /f/) because it would've been the only sound that I could not conventionally represent with a single character (diacritic or not) of the Latin alphabet. I was writing it <th> and I didn't want any digraphs.

[Pabappa]

I actually used <f> for /θ/ for a while for the same reason ... though I wasnt against digraphs so much as against odd-looking characters. I ended up changing it to ṣ (s-underdot if it doesnt show). i usually reserve stop + <h> for aspiration.

[Pabappa]

The syllable order for the parent script of all of my conlangs is derived from an article I read in a parenting magazine saying that the baby's first consonant sounds are "/h/, /l/, and the glottal stop". This was 1997 or so, and even back then, I was trying to make a language based around baby talk. I remember being surprised by this article, since I had assumed that the baby's first sounds would be the familiar labials /p b m w/ as I had read in a different article earlier on. I believe that the /h l ʔ/ thing was talking about the infant vocalization stage that comes even before babbling, but even so, I'm not sure that the article was correct. If it is correct, it means that babies develop flexible tongues and then lose that ability, and dont regain it for in some cases years (witness some 5 yr olds still having trouble with /l/).

I flipped the order to /l h ʔ/ (Im not sure the original article was intending them to be in any particular order anyway), but its taken me a long time to admit that I dont really have a good internal explanation for why the speakers would order their alphabet like that. There's also the problem that the syllabary reflects sound changes that don't belong to a language I still have .... I just want the order that way because I've memorized it and would find it hard to change. The order is /a la i ha ka u ma ca na li sa pa ta pi hi ti si ci mi ni pu hu tu su lu ku mu cu nu ki/. (Yes there are only 30 syllables in the basic inventory, but *thats* not a problem for me, because the languages that need more simply use digraphs or add new letters just like Earth languages do.)

Furthermore, some of the sound changes in that lost language were ones i came up with during a transitional stage where i was just learning about how sound changes actually work. previously, i had not really thought much about it,and i created script-based sound changes where a consonant would shift to a consonant adjacent to it in the alphabet, ...e.g. with one language, /r/ shifted to /s/, and /p/ shifted to /r/. Unconditionally of course. And often i would do sound changes individually for each word, meaning that each word underwent random changes and did not follow ordered lists of rules. Thus /enodab/ > /ende/, but /alakob/ retains its final /-ob/. The language with the babytalk syllabary didnt have anything as obviously wrong in its diachronic history as these examples, but i am not sure that everything really makes sense.

edit: reordered two sentences and added a new one

[Pedant]

Simple one in my case: back when I was starting out in the late 2000s (and give me a break, I was all of ten years old) I wanted to try and represent sounds that weren't normally included in the alphabet, namely [ð] and [ʒ]. So this fellow, instead of asking whether or not there was a way to represent these, just stuck a line over the digraphs <th> and <sh> to represent that they were voiced. I've got at least three scripts (two alphabets and an abjad) with <c> as well as <k> and <TH> and <SH> with big fat lines over them. I think it even carried over into the orthography of the early Barkaloanan script (ironically the only character I can't remember from its daughter script, Barkal Eya, is for [ð])...

[Qwynegold]

My "favorite stupid" transcription for /ŋ/ is ɳ, which is for some reason common in Hungary.

I believe this is derived as a mis-Latinization of η, which was often used as a substitute for the proper thing. Today Unicode also has Latin Ƞ ƞ though, probably for some recently-written language that managed to establish eta for /ŋ/…

Isn't ƞ usually used for marking nasalization?

[Vijay]

[ƞ] in IPA is a retroflex nasal. I've never seen it used anywhere else (i.e. I've only see it in IPA as far as I remember, at least if we exclude what Tropylium said).

[Kuchigakatai]

[ƞ] in IPA is a retroflex nasal. I've never seen it used anywhere else (i.e. I've only see it in IPA as far as I remember, at least if we exclude what Tropylium said).

No, Qwynegold is right. ƞ (n with a long right leg) is not the same as ɳ (n with a retroflex hook).

According to Unicode, Inc. (informally the Unicode Consortium), ƞ is used to indicate nasalization in Lakota (the Wikipedia article prefers the regular velar ŋ instead) and was also formerly used to represent Japanese ん n'.

[Qwynegold]

and was also formerly used to represent Japanese ん n'.

Nani?! This was new to me.