Some phonotactics “universals”?

A forum for discussing linguistics or just languages in general.
Post Reply
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Some phonotactics “universals”?

Post by eldin raigmore »

I recently saw in the Universals Archive that:
If C1C2- is an onset in a language, that makes it likelier that -C2C1 will be a coda in that language, than that -C1C2 will.

The Archivist chose to interpret this as:
If C1C2- is an onset in a language, and -C1C2 is a coda in the same language,
then -C2C1 will also be a coda in that language.

(I somewhat doubt those two statements are necessarily logically equivalent!)

That got me wondering.

For a given consonant-pair C1 and C2, there are sixteen possibilities:
C1C2- occurs as an onset, or it doesn’t;
C2C1- occurs as an onset, or it doesn’t;
-C1C2 occurs as a coda, or it doesn’t;
-C2C1 occurs as a coda, or it doesn’t.

Which combinations actually occur in natlangs?

For instance, in English, /st/ and /ts/ both occur as codas, and /st/ occurs both as a coda and as an onset, but /ts/ does not occur as an onset in English.
Similarly, for English /sk/ is both an onset and a coda, and /sk/ and /ks/ are both codas, but /ks/ isn’t an onset.

....

I’m sure it’s easy to think of pairs of consonants each of which is allowed in either syllable-margin in that language, and allowed to be part of two-consonant clusters in either onsets or codas in that language, but aren’t allowed to occur consecutively tautosyllabically in either syllable-margin in any syllable of that language. I think /b/ and /f/ might be such a pair in English.

I think in English /b/ and /d/ never co-occur consecutively in an onset, and /db/ never occurs as a coda, but /bd/ does occur as a coda.

And so on; I probably haven’t run out of examples, but I’m out of time for now.

I’m looking for a couple of examples of each possibility, in a couple of languages, preferably languages that either aren’t in the same family, or aren’t from the same linguistic area (sprachbund), or aren’t the same “type” (whatever that means; are there any two languages that are exactly the same type in every typology?)

....

I also have some more “possible phonotactics universals” to ask about. I’ll come back to this when I have more time!
User avatar
Pabappa
greek
greek
Posts: 577
Joined: 18 Nov 2017 02:41

Re: Some phonotactics “universals”?

Post by Pabappa »

greek allows onset /sf-/ ("sphere" etc) and coda /-fs/ (in names like Zeus) without allowing onset /fs-/ or coda /sf-/ as far as I know. So there's one positive data point. But yet at the same time Greek also allows onset /ps- ts- ks-/ without allowing coda /-sp -st -sk/; and it allows coda /-ps -ks/ and possibly /-ts/. So there's a point against it. Also it's possible that all of the words with coda /-fs/ are merely spelling pronunciations of ancient Greek words ending in -aus and -eus, which might suggest that they are not really part of the demotic lexicon any more than English "/ts/unami" etc.

If you analyze /ts/, /tš/, etc as clusters you can break this pretty easily since there are definitely languages that allow /tš/ as an onset and a coda but do not allow /št/ in either position. They simply consider /tš/ to be a single consonant. Even English only has coda /-št/ in verbs, for example "whooshed", which are largely onomatopoeic anyway.
Kavunupupis, šiŋuputata.
When I see you pointing at me, I know I'm in trouble. (Play)
Khemehekis
mongolian
mongolian
Posts: 3884
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: Some phonotactics “universals”?

Post by Khemehekis »

Pabappa wrote: 02 Sep 2020 00:51 Even English only has coda /-št/ in verbs, for example "whooshed", which are largely onomatopoeic anyway.
And the borrowed word "borscht", don't forget.
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 86,336 words and counting

31,416: The number of the conlanging beast!
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: Some phonotactics “universals”?

Post by Salmoneus »

Khemehekis wrote: 02 Sep 2020 01:25
Pabappa wrote: 02 Sep 2020 00:51 Even English only has coda /-št/ in verbs, for example "whooshed", which are largely onomatopoeic anyway.
And the borrowed word "borscht", don't forget.
And the interjection "whisht". And probably at least some other loanwords from German.

And while the great majority are "verbs", we have to recognise that in some cases the cluster is found in a "participle" adjective that is vastly more common than the "verb" it's theoretically a part of: "parched" and "farfetched", for instance, are far more commonly encountered than "to parch" or "to farfetch".

In any case, appearing only in verb forms isn't really significant, is it? Given just how common the cluster is in English. Nobody hears "he lashed out" or "fortunately, there's a cached version" and thinks "my word, what a strange sound! That can't be normal English phonotactics!"


There's a difference between marginal clusters (or phonemes) that are genuinely out of place and unusual, and occur only in low-frequency onomatopoeias and learned terms (like the /fT/ in 'phthisis', for instance), and clusters that are perfectly in keeping and relatively high frequency, but simply happen for diachronic reasons not to occur widely.
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: Some phonotactics “universals”?

Post by eldin raigmore »

Getting systematic and complete about my earlier-posted question(s):

Limiting the discussions to such languages L and pairs of consonants C1 and C2 in L such that:
* both C1 and C2 occur as first consonants of two-consonant onset-clusters in syllables in L;
* both C1 and C2 occur as last consonants of two-consonant onset-clusters in syllables in L;
* both C1 and C2 occur as first consonants of two-consonant coda-clusters in syllables in L;
* both C1 and C2 occur as last consonants of two-consonant coda-clusters in syllables in L.

.....

Question A:
What are some examples of such languages L and consonants C1 and C2 such that neither C1C2 nor C2C1 occurs as an onset cluster nor a coda cluster in any syllable of L?

Question B:
What are some examples of L and C1 and C2 such that C1C2- occurs as an onset cluster in L,
but neither -C1C2 nor -C2C1 occurs as a coda-cluster in L,
and C2C1 does not occur neither as an onset-cluster nor a coda-cluster in L?

Question C:
What are some examples of L and C1 and C2 such that -C1C2 occurs as a coda-cluster in L,
but neither C1C2- nor C2C1- occurs as an onset-cluster in L,
and C2C1 does not occur neither as an onset-cluster nor a coda-cluster in L?

Question D:
What are some examples of L & C1 & C2 such that C1C2 occurs both as an onset-cluster and as a coda-cluster in L,
but C2C1 does not occur neither as an onset-cluster nor as a coda-cluster in L?

Question E:
What are some examples of L & C1 & C2 such that C1C2- and C2C1- both occur as onset-clusters in L,
but neither -C1C2 nor -C2C1 occurs as a coda-cluster in L?

Question F:
What are some examples of L & C1 and C2 such that -C1C2 and -C2C1 both occur as coda-clusters in L,
but neither C1C2- nor C2C1- occurs as an onset-cluster in L?

Question G:
What are some examples of L & C1 & C2 so that C1C2- occurs as an onset cluster but not as a coda cluster,
and -C2C1 occurs as a coda-cluster but not as an onset-cluster?

Question H:
What are some examples of L C1 C2 where C1C2 occurs both as an onset cluster and as a coda cluster,
and both C1C2- and C2C1- occur as onset clusters,
but -C2C1 never occurs as a coda cluster?

Question I:
What are some examples of L C1 C2 where C1C2 occurs both as onset-cluster and as coda-cluster,
and both -C1C2 and -C2C1 occur as coda clusters,
but C2C1- never occurs as an onset cluster?

Question J:
What are some examples where both C1C2 and C2C1 occur both as onsets and as codas?

..... ..... ..... ..... .....

Two (sets of) questions about three-consonant tautosyllabic clusters; one (set) about three-consonant onsets and one about 3-consonant codas.

3-consonant onsets.

If C1C2C3- occurs as a 3-consonant onset, how many and which of C1C2- and C2C3- and C1C3- can we safely conclude must occur as a 2-consonant onset?

1. Examples of occurring 3-consonant onsets C1C2C3- where not all of C1C2- and C1C3- and C2C3- occur as two-consonant onsets.

2. Examples of occurring 3-consonant onsets C1C2C3- where not both of C1C2- and C2C3- occur as 2-consonant onsets.

3. Examples of occurring 3-consonant onsets where no two of C1C2- and C1C3- and C2C3- also occur as 2-consonant onsets.

4. Examples of occurring 3-consonant onsets C1C2C3- where neither C1C2- nor C2C3- occur as 2-consonant onsets.

5. Examples of 3-consonant onsets C1C2C3- where none of C1C2- nor C1C3- nor C2C3- occurs as a two-consonant onset.

... ... ...

Questions 6-10 are like those save replacing “onset” with “coda” throughout.
Post Reply