Sound Change Quickie Thread

[Herr Dunkel]

Uh, probably not? Breathy and creaky voice are on opposite sides of modal voice.

Thats kind of what I was concerned about, but I thought it might be workable anyways. What's the best way for ejectives to develop?

The glottalic theory of (Pre-)PIE had Ejective > Voiced, and since sound changes generally can work in reverse, too, you can have Voiced > Ejective
Or Geminate > ejective, C+h > C+ʔ > C', C+ʔ > C', t͡s > t' etc.

[8Deer]

Uh, probably not? Breathy and creaky voice are on opposite sides of modal voice.

Thats kind of what I was concerned about, but I thought it might be workable anyways. What's the best way for ejectives to develop?

The glottalic theory of (Pre-)PIE had Ejective > Voiced, and since sound changes generally can work in reverse, too, you can have Voiced > Ejective
Or Geminate > ejective, C+h > C+ʔ > C', C+ʔ > C', t͡s > t' etc.

Hmm, I think I'll have the voiced plosives go ejective and then have the breathy voiced become regular voiced. Thanks.

[Whimemsz]

No, stop. Darkgamma doesn't know what the fuck he's talking about.

(1) The glottalic theory of PIE isn't necessarily correct, so it's dangerous to draw conclusions about sound changes from the implications of positing the glottalic theory

(2) SOME sound changes can go both ways with ease, but there are many that are either much rarer or practically impossible in reverse. Voicing stops between vowels is extremely common; devoicing them much less so. /k/ can debuccalize to /ʔ/ easily; /ʔ/ > /k/ is, if not unattested, at least phenomenally rare. I'm not aware of any case where voiced stops developed into voiceless ejectives (although a change plain voiced > implosive > voiceless ejective might be more plausible).

In other words, Darkgamma has made two enormous and unfounded assumptions, and his suggestion is worthless (however, his other suggestions, e.g. (C+h >) C+ʔ > C' are much more plausible)

[sirdanilot]

Uh, probably not? Breathy and creaky voice are on opposite sides of modal voice.

Thats kind of what I was concerned about, but I thought it might be workable anyways. What's the best way for ejectives to develop?

The glottalic theory of (Pre-)PIE had Ejective > Voiced, and since sound changes generally can work in reverse, too,

No. You're probably getting at fortition here, but fortittion is not a 'reverse sound change'. You cannot just add a random feature to a phoneme and be done with it; it has to come from somewhere. For example, devoicing the coda consonant in many languages comes because of the 'silence' at the end of a word (which is [-voice]).

you can have Voiced > Ejective

I don't think so. Any attestation?

Or Geminate > ejective,

Maybe, I'd love to see attestion of this because the reverse is much more likely.

C+h > C+ʔ

C+h clusters aren't very likely, because clusters with /h/ are disfavored crosslinguistically. If they happen, they will most likely be on morpheme barriers (like English withhold).
It probably happens in some languages, but at least in Zeelandic Dutch /h/ simply disappeared rather than becoming glottal stop. Perhaps a glottal stop is pronouncing in words like 'dog' */hɔnd/ [(ʔ)ɔ̃nt], but that has other reasons; actually epenthetic [n] is preferred above [ʔ] 'the dog' */də hɔnd/ [dnɔ̃nt]. There is no difference between 'the lord' /də hɪːr/ [dnɪːr] and 'the honour' */də ʔɪːr/ [dnɪːr] ; indeed, in careful speech one would pronounce a glottal stop in the second one rather than in the first.

C+ʔ > C',

This sounds the most plausible to me, though there are many, many languages that do perfectly well with glottal stops directly after consonants.

t͡s > t' etc.

Affricates are fairly stable segments and I can't really see this happening...

Ejectives can turn into pharyngeals, as happened in Nootka; however, I could hardly imagine the opposite happening. I would think that Semitic pharyngeals might have emerged from ejectives as well, though the pronounciation of, for example, emphatic consonants in Biblical Hebrew is uncertain.

Generating marked phonemes like ejectives is often hard to explain, and most languages that have them have had them for a very long time. Some more suggestions I could make is t -> ʔ before nasals and word-finally (a bit like in Cockney), then dropping the vowel preceding ʔ, or so. *katna - kaʔna - kʼana. /q/ also tends to become glottal stop (as in some Arab dialects), and stops generally tend to become glottal stops before nasals (if not doing other fun things like devoicing the nasal etc.)

EDIT: And yes, aspiration is obviously common, but that is not the same as consonant + /h/. I couldn't see aspiration turning ejective either; again, the other way round is much more likely.

[Herr Dunkel]

No, stop. Darkgamma doesn't know what the fuck he's talking about.

(1) The glottalic theory of PIE isn't necessarily correct, so it's dangerous to draw conclusions about sound changes from the implications of positing the glottalic theory

(2) SOME sound changes can go both ways with ease, but there are many that are either much rarer or practically impossible in reverse. Voicing stops between vowels is extremely common; devoicing them much less so. /k/ can debuccalize to /ʔ/ easily; /ʔ/ > /k/ is, if not unattested, at least phenomenally rare. I'm not aware of any case where voiced stops developed into voiceless ejectives (although a change plain voiced > implosive > voiceless ejective might be more plausible).

In other words, Darkgamma has made two enormous and unfounded assumptions, and his suggestion is worthless (however, his other suggestions, e.g. (C+h >) C+ʔ > C' are much more plausible)

You can essentially draw up a sound change that changes /ħ/ into /ʙ/ or /ʟ/ or /ǂ/
And I didn't say all sound changes can go both ways, just that they generally can (unless they're implausible like 0 > q!ʰ).
You can turn voiced plosives into ejectives like this:
/b/ > /ɓ/ > /ɓ̥/ > /p'/

Affricates are fairly stable segments and I can't really see this happening...

If you allow Aspirate > Ejective, you can have:

/t͡s/ > /tʰ/ > /t'/

[finlay]

And I didn't say all sound changes can go both ways, just that they generally can (unless they're implausible like 0 > q!ʰ).

This is still false. You're too vague about "implausibility", too.

[Herr Dunkel]

This is still false. You're too vague about "implausibility", too.

What about this:

Implausible sound change - sound change that is unlikely to happen due to the conditions and enviroment before the change occured and the general likelyhood of such a change occuring

'S that better?

[sirdanilot]

/b/ > /ɓ/ > /ɓ̥/ > /p'/

You haven't got a clue what you're talking about. You cannot pull sound changes out of your ass and call them realistic.

Stop + glottal stop turning into an ejective is plausible, but other than that, it's going to be very, very hard to get marked phonemes such as ejectives into your language. I find it more fun anyway to start with such marked phonemes and then break them down.

[WeepingElf]

I can only second what Whimemsz and finlay have said. Darkgamma is talking about things he hasn't understood.

[Bob Johnson]

Darkgamma is talking about things he hasn't understood.

and people are only just noticing?

[Herr Dunkel]

/b/ > /ɓ/ > /ɓ̥/ > /p'/

You haven't got a clue what you're talking about. You cannot pull sound changes out of your ass and call them realistic.

Stop + glottal stop turning into an ejective is plausible, but other than that, it's going to be very, very hard to get marked phonemes such as ejectives into your language. I find it more fun anyway to start with such marked phonemes and then break them down.

Heck, "/b/ > /ɓ/ > /ɓ̥/ > /p'/" was proposed by Whimemsz

[----]

Where? I don't see that in any of his(?) posts.

[Herr Dunkel]

(although a change plain voiced > implosive > voiceless ejective might be more plausible).

[sirdanilot]

Well, I don't agree with that either. Perhaps if there are already ejectives, that voiceless implosives could turn ejective by analogy (they do sound similar despite being totally different articulatory) but nothing like this is attested nor very plausible. If you want ejectives so bad, just put them in your language from the beginning. That's what I did, and it is perfectly fine and realistic.

[finlay]

psst... ejectives aren't actually voiceless.

I can certainly see a connection between ejectives (egressive glottalic phonation) and implosives (ingressive glottalic phonation), though. I think in Mayan /ɓ tʼ kʼ/ form a series, where pʼ was presumably replaced by ɓ because ɓ is more distinct and easier to produce than pʼ.

[finlay]

This is still false. You're too vague about "implausibility", too.

What about this:

Implausible sound change - sound change that is unlikely to happen due to the conditions and enviroment before the change occured and the general likelyhood of such a change occuring

'S that better?

Well, no. The fact of the matter is that's just a woolly definition, and doesn't tell us anything about what is implausible and what isn't. You may have to go about it on a case-by-case basis, in the sense of "this is plausible because" and "this is implausible because". When you make generalisations such as "most sound changes are plausible both ways", be prepared to prove it. That may require you to show us in painstaking detail that every possible sound change is plausible both ways, with good reasoning, and that's a) far too much of a task and b) you don't have the necessary skills or vocabulary. You might be able to get away with showing it using statistics, but there is no way to ensure that you have taken a fair sample. So it's a pointless exercise overall.

[sirdanilot]

psst... ejectives aren't actually voiceless.

I can certainly see a connection between ejectives (egressive glottalic phonation) and implosives (ingressive glottalic phonation), though. I think in Mayan /ɓ tʼ kʼ/ form a series, where pʼ was presumably replaced by ɓ because ɓ is more distinct and easier to produce than pʼ.

Oh, that's interesting. And you are right, ejectives are just not specified for voice.

[Herr Dunkel]

When you make generalisations such as "most sound changes are plausible both ways"...

Don't put words into my mouth.

[Mr. Z]

When you make generalisations such as "most sound changes are plausible both ways"...

Don't put words into my mouth.

O RLY?

And I didn't say all sound changes can go both ways, just that they generally can

If that doesn't mean that most sound changes are plausible both ways, then what does it mean?

[chris_notts]

I had ejectives develop from geminate voiced stops in one of my conlanging projects, Mɛdíʈʂai. See:

http://zbb.spinnwebe.com/viewtopic.php?f=4&t=39158

In my sound changes, a distinction of voiceless / voiced singleton / geminate changed from:

----- t t: -------- d d: -------
spread glottis constricted glottis

to:

tʰ ----- t -------- d ------- d̰
spread glottis constricted glottis

via aspiration of the geminate voiceless series, and creakiness of the geminate voiced series. Changes in phonation served to make the geminate series more distinct, which was important since they became very common initially where there are fewer cues as to length.

Finally, further constriction and loss of voicing as a distinctive feature of the creaky series occurred, giving ejectives. While not relevant to this discuss, the voiced series then fricativised and the unvoiced series became voiced, giving a three way ejective / voiced / aspirated distinction in the stops.

Whether you believe that this is plausible or not is up to you. The reason I think this is reasonable is actually described in the PDF grammar I've been writing. I'll quote it:

.. Development of Geminate Contrast into Laryngeal Contrast

[] provides several examples of cases where geminates have been made more distinctive by concomitant
variation in laryngeal seing. is seems to be especially common for initial geminates which are not
merged with their singleton counterparts, as the duration of the closure for obstruents can be hard to
percieve word initially. For example:

In Taba (Bowden and Hajek ), which also shows word-initial geminate/singleton contrasts,
phrase-initial geminates show variation across the hyper-to-hypoarticulation continuum. In
careful speech, phrase-initial geminates are realised ”with a greater degree of tension and
more articulatory force”; while in casual speech ”they can be realised with the same reduced
tension and articulatory force as singletons. is sort of variation in phrase-initial position is
expected, since it is precisely in initial position that the closure duration of a voiceless stop is
difficult to perceive due to the avsence of a cue for the onset of stop closure.

e author continues later on:

In careful speech, hyper-articulation of initial geminates can result in shis from geminate
consonants to fortis or aspirated consonants. In normal or casual speech, the geminate closure
duration is produced, as in phrase-medial contexts, but insufficient perceptual cues result in a
singleton precept, giving rise to the evolution of word initial degemination. Many instances of
each development are found in the historical record. In Tuvaluan, a Polynesian language (Mil-
ner ), unstressed vowel deletion gave rise to initial geminates which were subsequently
reinterpreted as aspirated stops: *ke-kémo > kkémo > khémo ‘blink’, *pu-púni > ppúni > phúni
‘(be) shut, blocked’, *ta-táki > áki > tháki ‘lead’ etc. in line with hyperarticulation.

e author also points out that:

e tendency for geminate oral stops to be voiceless as opposed to voiced is noted by Klingen-
heben (), and has been argued to follow from the aerodynamic properties of stop produc-
tion (Jaeger ; Ohala a). e general argument is that, the longer the stop closure, the
more difficult it is to sustain voicing. Supragloal air pressure build-up will inhibit vocal-fold
vibration unless some active step is taken to increase the volume of the supra-gloal cavity.

e author argues that these modifications, such as larynx lowering, to maintain voicing during the
geminate voiced stops are the motivation for the Sindhi sound change *bː > ɓ.

A similar step presumably took place in the development of the voiced geminate stops reconstructed
for Mɛdíʈʂai. As a result of these changes, creaky voicing on the following vowel developed, aer which
the voicing of the segment was secondary and the development to ejective stops occurred.

A similar change may also have taken place in the Bantu language Southern Sotho, which according
to [] underwent sound changes such as *mp -> ph and *mb -> pʼ.

I think the Southern Sotho example is most interesting, because it shows a combination of voiced sounds, none of which is glottal, developing into an ejective, although the intermediate steps aren't clear.

[Herr Dunkel]

When you make generalisations such as "most sound changes are plausible both ways"...

Don't put words into my mouth.

O RLY?

And I didn't say all sound changes can go both ways, just that they generally can

If that doesn't mean that most sound changes are plausible both ways, then what does it mean?

Many can, but not all, heck.
Syncope is irreversible, as are plenty other changes.

[finlay]

For fuck's sake, you are so fucking tiresome.

[Herr Dunkel]

For fuck's sake, you are so fucking tiresome.

Do you know how many times I want to scream that into your face and spam your inbox with flames?
Tolerance is bliss.

[Mr. Z]

When you make generalisations such as "most sound changes are plausible both ways"...

Don't put words into my mouth.

O RLY?

And I didn't say all sound changes can go both ways, just that they generally can

If that doesn't mean that most sound changes are plausible both ways, then what does it mean?

Many can, but not all, heck.
Syncope is irreversible, as are plenty other changes.

We're not saying you said all can. We're saying you said most can. That's what you said, isn't it?

[Herr Dunkel]

We're not saying you said all can. We're saying you said most can. That's what you said, isn't it?

NO