Sound Change Quickie Thread

[bradrn]

Interestingly enough, the Index Diachronica says that something similar can happen for vowels; that is, it seems that V > 0 / _# is attested. Do you have any ideas on why this one is natural, whereas Akangka’s change isn’t?

The tautological answer is that vowels form a natural class more easily than consonants as a group do, so that V > 0 / _ # is more analogous to [+plosive] > 0 / _# (especially if it's actually more like V > ə > 0 / _#, like [+plosive] > ʔ > 0 / _#). I'm not sure how illuminating that is, though.

That is definitely illuminating enough — thank you!

(Blevins has fairly specific ideas about the phonetic conditions that make particular changes natural, but my knowledge of such stuff is pretty basic.)

Blevins was recommended to me earlier in this thread as well… I managed to find the book in my University’s library, but I felt that I didn’t have enough time to read it thoroughly then. Maybe sometime later.

[Xwtek]

Also, while I don't know whether dental /n̪/ is realistic, why don't you palatize /n/ to /ɲ/ and simply not touching it in later sound changes?

[akam chinjir]

Also, while I don't know whether dental /n̪/ is realistic, why don't you palatize /n/ to /ɲ/ and simply not touching it in later sound changes?

Hmm. No reason that I can think of, actually. I'll do that. Thanks!

[dɮ the phoneme]

Can anyone recommend any reading on tone systems where a large number (perhaps the majority) of syllables are neutral tone? I think such systems exist. I'm particularly curious how the neutral tones interact with tone persistence, if present.

[bradrn]

Can anyone recommend any reading on tone systems where a large number (perhaps the majority) of syllables are neutral tone? I think such systems exist.

On the contrary, I would be surprised by such a system: I would expect tones to be equally distributed across syllables. (I wonder if anyone has studied these statistics? I would be fascinated to get some concrete data on this!) What you’re describing sounds closer to a pitch accent system, although I appreciate that that’s something slightly different.

I'm particularly curious how the neutral tones interact with tone persistence, if present.

What’s ‘tone persistence’? I looked it up and couldn’t find anything…

[Pabappa]

Navajo might qualify, although since it only has two real tones, the unmarked tone is called low rather than neutral. However, a four-tone analysis is possible if you consider the long vowels to be separate tones.

That isnt a very good example, but there might be another language either in the same family or in the same geographical area that might be a better model to follow.

[dhok]

Can anyone recommend any reading on tone systems where a large number (perhaps the majority) of syllables are neutral tone? I think such systems exist.

On the contrary, I would be surprised by such a system: I would expect tones to be equally distributed across syllables. (I wonder if anyone has studied these statistics? I would be fascinated to get some concrete data on this!) What you’re describing sounds closer to a pitch accent system, although I appreciate that that’s something slightly different.

Anecdotally, Tone 1 in Mandarin is significantly rarer than the other tones.

Middle Chinese probably counts as such a language, as most syllables were píng (neutral) tone, and many of the others rù ("checked", e.g. ending in a stop and therefore not really a tone at all, arguably a subset of píng) tone.

[dɮ the phoneme]

I'm particularly curious how the neutral tones interact with tone persistence, if present.

What’s ‘tone persistence’? I looked it up and couldn’t find anything…

Tone persistence is when tones exhibit the properties of independent phonological units, not bound to particular syllables. For example, you might have a (diachronic or synchronic) process of vowel deletion, after which the tone of the deleted vowel remains present by surfacing on an adjacent syllable. You can also end up with floating tones this way.

[Richard W]

On the contrary, I would be surprised by such a system: I would expect tones to be equally distributed across syllables. (I wonder if anyone has studied these statistics? I would be fascinated to get some concrete data on this!)

The Proto-Tai tone system shows an unequal distribution. Syllables ending in a plosive are called tone D ('checked' or 'dead' syllables), other syllables ('live' syllables), which to careless listening end in a vowel or resonant (/l/ or a nasal), are split between 3 tones, labelled A, B and C. This tone system lasted until recently enough that is reflected in Thai spelling. The commonest of A/B/C, dubbed A has no tone mark; the second commonest, dubbed B has a tone mark which is a single vertical stroke; the rarest, dubbed C, has a tone mark which started as a vertical stroke plus a horizontal stroke, but now looks surprisingly like a digit '2'. Looking at a line of text, most syllables have no mark. The numerous loans from Pali and Sanskrit have been assigned to tone classes A and D, increasing the bias towards tone A.

Historically, this may not be so surprising. One of B and C is supposed to reflect a final /s/, and the other a final glottal (possibly several different glottals).

Nowadays, A, B and C have become 5 tones. One of the tones derived from A is known as the 'common' tone, and remains the commonest, but its predominance is much less.

There's a corpus analysis for Thai (FREQUENCY OF OCCURRENCE OF PHONEMES AND SYLLABLES IN THAI: ANALYSIS OF SPOKEN AND WRITTEN CORPORA by Munthuli et al.) in https://www.google.com/url?sa=t&rct=j&q ... 1i8EsjkCPs (too lazy to degooglise it). The modern tones derived from the old tones as follows:

mid < most A (34%)
low < some B, some D (22%)
falling < other B, some C, some D (19%)
high < other C, some D (16%)
rising < other A (9%)

The split of A, B and C is conditioned by the Proto-Tai initial; the split of D is conditions by old initial and vowel length.

There are some interesting correlations in the co-occurrences of sounds; most are to be explained by (a) the tones of live and dead syllables being two separate systems and (b) history. I've seen some elegant explanations of the non-existence of some words that even I know. There are quite a few words that could not have derived from Proto-Tai, many being recent loans from Chinese; that is why modern Thai has to add two new tone marks.

[akam chinjir]

Can anyone recommend any reading on tone systems where a large number (perhaps the majority) of syllables are neutral tone? I [i]think[/i] such systems exist. I'm particularly curious how the neutral tones interact with tone persistence, if present.

You've already looked at Moira Yip's book Tone, I imagine, but I expect the sections about tone in African languages might be helpful.

Anyway lots of languages have been analysed as having just one or two marked tones, and a good many of those must have many syllables with no tone. Among one-tone systems there are at least two patterns you can expect to find:

• There's just one tone that's active in the phonology (by triggering OCR effects, that kind of thing), but there's a late process that inserts a default tone onto syllables that don't have the active tone. So phonetically there are still two tones. (Most often it'll be the high tone that's active, and a low tone that's inserted as a default, but you also find the opposite pattern.)
• There's only one tone that's active in the phonology, and you don't even get a default tone inserted. This means that only syllables that are marked with that tone are specified with a pitch target; the pitch of other syllables is just depends on their environment. (I guess it's this sort of system that's most naturally described in terms of a neutral tone.)

I think that by definition a neutral tone can't be phonologically active, so I don't see how you could get a floating neutral tone. By the same token, though, a syllable with neutral tone should be able to host a floating tone (though there may be OCR/sandhi effects depending on what's going on in adjacent syllables).

[Nortaneous]

Can anyone recommend any reading on tone systems where a large number (perhaps the majority) of syllables are neutral tone? I think such systems exist.

On the contrary, I would be surprised by such a system: I would expect tones to be equally distributed across syllables. (I wonder if anyone has studied these statistics? I would be fascinated to get some concrete data on this!)

I'd guess that this is generally not true. My impression is that, in Iau, tones 7, 8, and 9 are much more common than tones 2 through 6. (There's no tone 1.) Then again, the expanded tone inventory of Iau relative to the other Lakes Plain languages developed from loss of entire syllables and preservation of tonal contours. So the distribution of tones is probably less skewed in Kirikiri.

That's not to say that fewer tones necessarily implies an even tone distribution, of course, just as fewer consonants doesn't necessarily imply an even consonant distribution. Rotokas only has six consonants, but /g/ is relatively rare.

[Xwtek]

Phnom Penh dialect of Khmer has two tones, level and rising, originating from elision of /r/. You can expect the tone distribution to be very uneven.

[Xwtek]

Is it realistic to have sound change from /b/ to /d/ syllable finally? This is because I want symmetry with the shift of its nasal allophone /m/ > /n/ in the same context, so the coda can't be labial.

[dɮ the phoneme]

How plausible is debuccalization of voiceless coda stops to [ʔ] only in post-vocalic position? So [pot] > [poʔ] but [post] > [post].

[Zaarin]

How plausible is debaucalization of voiceless coda stops to [ʔ] only in post-vocalic position? So [pot] > [poʔ] but [post] > [post].

Many dialects of English, my own for instance, do exactly that. (Well, with the caveat that /t/ is usually debuccalized after a resonant, too.)

[Xwtek]

1. Is it realistic to have voiced nasal fricatives as an allophone of voiced fricatives next to nasal vowel?
2. If it's not, is it realistic to have a nasality spread through voiced fricatives?

[Pabappa]

Ive argued before that I dont believe nasal fricatives exist, and that Wikipedia's examples of languages claiming to have nasal fricatives always depend on ambiguous definitions, and some may even just be incorrect, such as labeling an <s̃> as a nasal fricative when it is in fact just an orthographic convention for labialization. However, Im not as confident about ruling out nasal fricatives as mere allophones.

[Xwtek]

Ive argued before that I dont believe nasal fricatives exist, and that Wikipedia's examples of languages claiming to have nasal fricatives always depend on ambiguous definitions, and some may even just be incorrect, such as labeling an <s̃> as a nasal fricative when it is in fact just an orthographic convention for labialization. However, Im not as confident about ruling out nasal fricatives as mere allophones.

Still Is it possible for nasal vowel to spread through voiced stop, liquid and voiced fricatives (and turning the first two into the nasal version), while voiceless fricatives and stop block them? Like:

auzã > ãũzã, but eixẽ > eixẽ.

[Pabappa]

Sorry, I dont know. I answered only the first question because I dont really know enough to answer the second one. Ive only created two languages with nasal vowels and in both languages they quickly disappear. Instinctively I'd say you could do it, and I've done something similar where /zã/ > /nã/ > /na/ while /sã/ just shifts to /sa/, but that was wholly based on my personal preferences and not on researching similar changes in natlangs.

[Nortaneous]

AFAIK, nasal fricatives are unattested except at the labial and glottal POAs, except for the bizarre /ħ̃ʲ/ in Tofa. Nasalized glottal fricatives are usually glottal fricatives that condition nasality on the following vowel, or something like that.