Linguistic Miscellany Thread

[bradrn]

Are there tonal languages without a level tone? E.g. just rising and falling tone?

There’s one Sino-Tibetan language — the name of which unfortunately escapes me — which has an underlying H/L system, but where tones consistently spread by one syllable (to the right, I believe) such that all syllables have surface contour tones.

[Creyeditor]

Are there tonal languages without a level tone? E.g. just rising and falling tone?

Phonetically in isolation: yes, in most languages with a simple high vs. low contrast these tones are pronounced as a rise and a fall respectively if in isolation.
Phonologically: maybe, depends on your analysis. Some approaches in the Sinitic tradition treats all tones as contours except the level tone which is sometimes treated as the absence of tone.
Phonetically in connected speech: maybe? Phonetics of tone are often not as straightword as one would like them to be, cf. peak delay and stuff. Also, what would that mean, if any up or down pitch excursion is a rise or a fall than any language with only high and low tones would count as a contour-only language phonetically, I guess.

[bradrn]

Are there tonal languages without a level tone? E.g. just rising and falling tone?

Phonetically in isolation: yes, in most languages with a simple high vs. low contrast these tones are pronounced as a rise and a fall respectively if in isolation.

This claim I hadn’t seen before. Source please?

[Creyeditor]

Have you looked at pitch tracks from two tone languages? There are no level tones to be found in the phonetics there. Generally, speakers start from the same pitch (or a similar pitch) and then move gradually towards the phonetic target pitch.

I valued speed over accuracy here, so you can have a look at the measurements of Thai tones on page 130 here: https://citeseerx.ist.psu.edu/document? ... ff026500b1. Both high tones and low tones have pitch excursion in the citation forms, even though they contrast with contour tones. Note that a difference of -20 Hz or +38 Hz are very much audible for human ears. The authors also describe only the mid tones in citation forms as level across the whole syllable in the text on page 130. You can also have a look at the nice diagrams on page 131. The differences between level (high, low) and contour (rising, falling) tones is really not about any tone being level, but much more about the shape and size of the contour.

Of course, this does not prove the "most languages" part. I feel like this is often implicitly assumed in many descriptions, but I'll keep my eyes open to see if I find a more explicit discussion of a two-tone language.

[Richard W]

I valued speed over accuracy here, so you can have a look at the measurements of Thai tones on page 130 here: https://citeseerx.ist.psu.edu/document? ... ff026500b1.

Ah, that paper that explains why some attested Thai words don't exist! (Words with falling tones on short checked syllables, such as แร่ด, and high tones on long checked syllables, such as เค้ก.)

[WeepingElf]

What do you think of this? Does temperature have an influence on language sonority?

[zompist]

What do you think of this? Does temperature have an influence on language sonority?

Ugh. I skimmed this, but found no reference to the basic problem with doing statistical analysis on languages: language features are not randomly distributed. Therefore all the normal statistical tests are useless, and finding "correlations" is meaningless.

[WeepingElf]

What do you think of this? Does temperature have an influence on language sonority?

Ugh. I skimmed this, but found no reference to the basic problem with doing statistical analysis on languages: language features are not randomly distributed. Therefore all the normal statistical tests are useless, and finding "correlations" is meaningless.

Yes, I am very sceptical about this, too. There have been various claims of this sort, such as that languages spoken in mountainous areas were more likely to evolve ejectives. It is also such that neighbouring languages tend to have similar phoneme inventories, even if they belong to different families. Many features have geographically limited distributions (e.g. clicks are only found in Africa, front rounded vowels are common in northern Eurasia and rare elsewhere (at least, WALS says so), etc.), though one cannot really correlate them to geography or climate.

[Richard W]

What do you think of this? Does temperature have an influence on language sonority?

This observation from the paper is pretty devastating: "The correlations fluctuating around zero indicate that the correlation between sonority and temperature is largely absent within language families."

[WeepingElf]

What do you think of this? Does temperature have an influence on language sonority?

This observation from the paper is pretty devastating: "The correlations fluctuating around zero indicate that the correlation between sonority and temperature is largely absent within language families."

Yes. This is quite obviously bad science - the sort of mayhem that frequently happens when people from the "hard sciences" try to explain linguistic phenomena. They say, of course, that "bad science sometimes makes good fiction", and you may thus liberally use such correlations in your conworld (but I am not going to do that with my projects), but as a statement about the real world, it probably doesn't hold water.

Over at CONLANG, replyer Markus Klyver quoted the following from the paper:

...some other conjectures about the temperature’s effects on sonority can be proposed. Cold air is always dry because of its low water vapor capacity (55), causing water evaporation from the vocal cords’ surface, which makes phonation control difficult (5, 56) and frustrates the production of sonorants, because sonorants are commonly voiced, requiring the vibration of vocal cords. Besides, in colder climates, especially at higher altitudes, wind chill is severe and might necessitate people keeping their mouths more closed, leading to a reduction of sonorant usage (3). It has also been also suggested that colder climates discourage outdoor activities and that indoor communication at close distances would lead to better preservation of obstruent sounds (23).

Which sounds more like a preconceived idea than an empirical finding. 'Nuff said.

[bradrn]

The only really convincing correlation I’ve seen is that implosives correlate with warm climates, which is obvious enough it barely even needs statistical tooling.

[WeepingElf]

An afterthought: As human beings are warm-blooded animals and living flesh is wet, the temperature and humidity of the air in the vocal tract is probably not affected by the environment to a significant degree, especially considering that most speech sounds are made while breathing out.

[Travis B.]

An afterthought: As human beings are warm-blooded animals and living flesh is wet, the temperature and humidity of the air in the vocal tract is probably not affected by the environment to a significant degree, especially considering that most speech sounds are made while breathing out.

But but but implosives involve breathing in and obviously are tied to warm climates! /s

[Moose-tache]

Reminds me of the time a bunch of computer scientists used software designed to track the spread of a virus to determine the urheimat of PIE. Naturally, the software just returned the earliest attested location, Anatolia, as the source. Everyone involved wiped their hands and said there, we solved it.

STEM people are fricking annoying sometimes.

[WeepingElf]

Reminds me of the time a bunch of computer scientists used software designed to track the spread of a virus to determine the urheimat of PIE. Naturally, the software just returned the earliest attested location, Anatolia, as the source. Everyone involved wiped their hands and said there, we solved it.

STEM people are fricking annoying sometimes.

Yes. And they often fail the simplest plausibility checks. I have seen such a tree which placed Proto-Romance at about 1000 BC. That should have alarmed them that their clock was ticking at the wrong rate, but no, nobody noticed. I also have the feeling that the age of PIE that "resulted" from the analysis was part of the input, so the argumentation was circular. Ouch.

[Moose-tache]

I think on day one of university level higher math, the professor says "Listen, since every other field is just a specific application of ours, that means that once you get good at math, by definition you are also good at botany, sociology, meteorology, geology, neurosurgery, HVAC engineering, and Central Asian history. Go nuts."

[bradrn]

I think on day one of university level higher math, the professor says "Listen, since every other field is just a specific application of ours, that means that once you get good at math, by definition you are also good at botany, sociology, meteorology, geology, neurosurgery, HVAC engineering, and Central Asian history. Go nuts."

How strange, mine unaccountably skipped that bit.

[Raphael]

I just learned that there are Wikipedia editions in Classical (or, as they call it, "literary") Chinese, and Cantonese.

[Moose-tache]

Eastern North American Verbs, Part 1: Pronominal Prefix Sets

In the same way that we can talk about “Standard Average European,” there are sets of features that are common across North America, or at least large parts of it. The basic template for a ENA verb is something like:
1) directional, instrumental, or applicative
2) pronominal prefix
3) verb stem
4) tense/mood/evidentiality complex
Some aspects of this template can be made even more specific. Many if not most languages in the region encode number into the verb stem, and many of the same directional prefixes appear time and time again. Today, let’s talk about one specific aspect of the pronominal prefixes.

In Choctaw, to say “you run,” you would say ish-baliili. To say “I see you,” you would say chi-pisali, where the second person object chi appears instead of the second person subject ish. However, to say “you want,” the word would be chi-banna. There are two sets of prefixes (actually three but we’ll stick to two for now), called Class I and Class II in Muskogean. I’ll call them Set A and Set B to match the Cherokee terminology for convenience. Set A is for subjects and Set B is for direct objects. But Set B is also used for subjects that have low agency (such as experiencers), and is especially common on intransitive and stative verbs. This doesn’t mean we should abandon the concept of subject and just call them “high agency” and “low agency,” because other aspects of Choctaw grammar, mostly related to noun phrase marking which we won’t get into, make it clear that the second person arguments in chi-banna and ish-baliili have something in common syntactically. Similarly, we cant say that chi-hochowa "you're cold" uses a direct object and an implied subject ("It colds you"), because that too would have consequences for other aspects of the grammar that simply don't materialize. This pattern is universal among the Muskogean languages, but it doesn’t stop there.

In Lakota, wa is used in wa-u “I come,” and ma is used in o-ma-yalepi “you seek me,” but ma is also used in ma-k’uje “I am sick.” I know very little about most of the Siouan languages, but so far this alternation seems to be almost universal.

Meanwhile, the Algonquian languages have a subject set and an object set*, but the object set is not used for low-agency subjects. Algonquian does have a direct-inverse system, which is not entirely relevant here except that it will come up again in just a moment.

At last we come to the Iroquoian languages. At first glance, they appear to be doing something very different. There are subject and object prefixes, but they are thoroughly fused, and when individual morphemes can be recovered, in most languages they are simply subjects or objects. In Cherokee, however, Set B can mark low-agency subjects. At the opposite end of the family Mohawk uses the two sets simply as subjects and objects, but when and how subjects and objects are marked depends on a salience hierarchy reminiscent of the direct-inverse system of Algonquian. For Mohawk, this extra indication of argument role is very useful, since the inherited Proto-Iroquoian sets have almost merged (k and kw mostly appear as k, and hs and ts mostly appear as s). In other words, Iroquoian languages are wedged between two regional strategies for person marking, and tend to imitate whatever system for person marking is common in the languages around them.

* I know people are going to say "what object set," and that's a fair question. The overt personal prefixes don't indicate subject or object, but there are object-only markings in the peripheral slot, and these peripheral markings never describe the subject, even if it is low-agency.

[jal]

Many if not most languages in the region encode number into the verb stem

Does this mean regular alteration of a base stem, or suppletive forms?


JAL