zompist bboard

I realize that "liquid" and "rhotic" are somewhat squishy categories whose membership is sometimes debatable, but...

How many languages do you guys know of that have two or more liquid consonants but either have no lateral liquids or have no rhotics? It seems like almost invariably if a language has multiple liquids it will have at least one that's lateral and at least one that isn't. I know that this isn't true, but I can't seem to think of any counterexamples at the moment. Help me out here. Even if it's, say, a dubious case where a language has a tap consonant that may or may not "count" as a rhotic, I want to hear about it.

-----

Addendum: OK, I was able to think of one example. Haida distinguishes /l/ and /lˀ/, and pretty indisputably has no rhotics. I'm guessing it's not the only language in North America with an /l/-/lˀ/ distinction and no rhotics, since while rhotics in general are kind of uncommon in indigenous North American languages for whatever reason distinguishing between laterals and rhotics specifically seems to be very uncommon. Any other ways to have two liquids without a lateral-rhotic distinction people are aware of?

-----

Further addendum: Several Salishan languages distinguish a glottalized and non-glottalized lateral liquid, although some of these also add rhotics. St'at'imcets manages to have four lateral liquids by virtue of distinguishing both glottalization and velarization, although it also has a pair of dental central fricatives that appear to be treated as sonorants; I'm not sure whether or not it's appropriate to call these "rhotics".

According to Wikipedia:

Yimas apparently has two liquids, [ɾ~l] and [ʎ~lʲ]. Wikipedia says the latter is the palatalized counterpart of the former, and that the palatalization is predictable in many, but not all words, so they seem to be phonemically contrastive.

Nauruan has /r/ /rʲ/ but no lateral liquids.

Kashaya has plain, aspirated and glottalized /l/. The same types of /r/ occur in the modern language, but only in loanwords.

Nez Perce distinguishes plain and creaky-voiced /l/ and has no /r/.

Pitjantjatjara has /r ɻ/ but no laterals
I'm an idiot.

KathTheDragon wrote:Pitjantjatjara has /r ɻ/ but no laterals

What's your source for that? Wikipedia lists /l ɭ ʎ/ alongside those rhotics.

Tlingit has five laterals, none of which are /l/, and no rhotics: /tɬ tɬʰ tɬʼ ɬ ɬʼ/.

Sumelic wrote:[snip]

All interesting. Looks like aside from phonation, palatalization is a recurring way to distinguish liquids without introducing a central/lateral distinction.

Zaarin wrote:Tlingit has five laterals, none of which are /l/, and no rhotics: /tɬ tɬʰ tɬʼ ɬ ɬʼ/.

But are any of those liquids? None of these are sonorants articulatorily, and correct me if I'm wrong, but I don't think any of them behave as sonorants in Tlingit. I'm well aware that Tlingit, most other Na-Dene languages, and various other groups of North American languages have plenty of lateral affricates and fricatives, but at least in my experience these are normally treated as part of the obstruent system.

Chengjiang wrote:

Zaarin wrote:Tlingit has five laterals, none of which are /l/, and no rhotics: /tɬ tɬʰ tɬʼ ɬ ɬʼ/.

But are any of those liquids? None of these are sonorants articulatorily, and correct me if I'm wrong, but I don't think any of them behave as sonorants in Tlingit. I'm well aware that Tlingit, most other Na-Dene languages, and various other groups of North American languages have plenty of lateral affricates and fricatives, but at least in my experience these are normally treated as part of the obstruent system.

Ah, I missed that you were specifically looking for liquids. Still, quite a few of the languages of the PNW meet your criteria--you already mentioned Haida and Salishan. To that you can add Tsimshian, Alsean, Klamath (/l l̥ lˀ/). Also Natchez has /l l̥/.

Purepecha/Tarascan (isolate, Mesoamerica) has /r ɽ/

Miunane (from small Amazonian family) has /r rʲ/

Forest Nenets (Uralic) was left with /l lʲ/ after r rʲ > ɬ ɬʲ

Xhosa (Southern Bantu) has /l l̤/ natively, plus /r r̤/ in borrowings

Nuosu/Northern Yi (Lolo-Burmese) has /l l̥/.

I believe dialectical Swedish/Norwegian, due to l>ɽ, though I don't know how thorough the replacement is

Mapuche (isolate, South America) has /l̪ l/, plus a retroflex that's in free variation [ʐ~ɻ~ɭ], though a lightly fricative [ɻ] is apparently most common.

Depending on how lenient you're willing to be, where lleismo, a guttural <r>, and a lack of r~rr distinction overlap. So dialectically in Portuguese, Occitan, Basque, Breton, French, and northern Italian.

Similarly Paiwan (Formosan) has /ʎ ɭ/ and some dialects have [ɣ] instead of a coronal trill.

OK, I'm starting to think there isn't even a particular tendency for two or more liquids > L-R distinction outside of language area effects from large, widespread families like IE and Afroasiatic that regularly make such a distinction.

vokzhen wrote:Similarly Paiwan (Formosan) has /ʎ ɭ/ and some dialects have [ɣ] instead of a trill.

Wow, Central/Southern Paiwan has /ʎ ɭ/ but no /l/. That's wild.

The Central Tai dialect of Longzhou (Lungchow in the literature) contrasts /l/ and /ɬ/ and has no rhotics. /ɬ/ is the reflex of Proto-Tai *r, *s and *z. The Proto-Tai voiceless lateral (possibly a cluster) has merged with *l to yield /l/.

More examples:

Vaeakau-Taumako apparently has /l/ and /lʰ/ and no rhotics.

Khalkha Mongolian has /r/ /rʲ/ but no lateral approximants. It does have the lateral fricatives /ɮ/ /ɮʲ/. I don't know if the lateral fricatives pattern as liquids. It looks like /l/ exists in other varieties of Mongolian.

The WALS has a chapter on lateral consonants with a map listing languages that lack laterals; that could be a place to look for more examples.

Found from the WALS:

Keresan languages have /r rˀ/.

In general, I don't think languages with palatalization, aspiration, or phonation distinctions on liquids are any more likely to have a lateral-rhotic distinction. But a lateral-rhotic distinction for liquids is more common than any of the other distinctions. So statistically, we'd expect most languages with two or more liquids to have a lateral-rhotic distinction.

Chengjiang wrote:

KathTheDragon wrote:Pitjantjatjara has /r ɻ/ but no laterals

What's your source for that? Wikipedia lists /l ɭ ʎ/ alongside those rhotics.

*facepalm*

The Chumashan family in California offers several additional examples of languages with plain /l/ vs glottalized /ˀl/ and no rhotic, as do the Yokuts languages, Salinan, Yuki and Wappo.

Cupeño, formerly spoken in part of San Diego county, had /l/ and /ʎ/, while rhotics were limited to Spanish loan words.

Klamath, formerly spoken in far northern California, had /l/, /ˀl/, and /l̥/ with no rhotics.

Washo, a language indiginous to the vicinity of lake Tahoe, has only /l/ and /l̥/.

Lao allegedly has /l/ and /lʷ/ and no rhotics. The argument that /lʷ/ is a phoneme seems to be that all onsets consist of a single consonant, and therefore apparent clusters with /w/ as second element are in fact labialised consonants. It's a very rare sound (or cluster) - one of the few words is /lʷa:/ 'donkey', but the big dictionaries tend to record that as /la:/! That argument can be extended to the neighbouring languages Northern Thai and Tai Lü.

The general pattern we seeing is, I believe, that there is a contrastive feature that is applied to a several -pairs of phonemes, and it is unsurprising that these contrasts also affect the laterals. Perhaps it would be helpful to note whether the contrasts are part of this pattern, which may, of course, be in decay. So for Lao, non-labial consonants tend to come in pairs with a contrast of labialisation. On the other hand, the Lungchow contrast /l/ v. /ɬ/ is not part of a systematic contrast.