Linguistic Miscellany Thread

[Moose-tache]

Both!

In Lakota, for example, reduplication is fairly regular: sha "it is green," shasha "they are green."
Creek has hatkis "it is white" and hathatkis "they are white."
Creek also uses infixes. hayyis "it is cold" becomes hayhoyis "they are cold." These infixes are a more usual way to derive plural verbs in Muskogean.
Some of these markers do double duty. Lakota reduplication can be distributive or repetitive. The most common pluralizer in Muskogean is ci, which is also a causative marker.

But as far as I can tell, every language that uses plural verbs has at least some pairs that simply have to be memorized. In Choctaw, iyalitok is "I went," but "we went" is iliɬkolitok.

[bradrn]

Eastern North American Verbs, Part 1: Pronominal Prefix Sets

Correct me if I’m wrong, but could this post be summarised as ‘most Eastern North American languages have fluid-S or split-S alignment’?

(Also, if you’re going to write more posts, could this get split off into its own thread please?)

[Moose-tache]

I thought about that, but I didn't want to commit, and I'm still learning since my knowledge of some of the midwestern Algonquian languages is embarrassingly poor. If I have two or three more of these in me I'll separate them out.

Anyway, split-S is definitely how a lot of linguists have described some of the languages in this region. I wanted to describe it here in the most pragmatic way possible so it would be clear what's going on. And as I hinted earlier, other aspects of the grammar like how noun phrases agree with the verb complicate things.

Huh. I just double checked the wikipedia page on active-stative alignment, and it lists all the Iroquoian languages. Either that's a mistake, or the author has an extremely loose definition of active-stative. As far as I can tell, Mohawk forms like wak (3>1) and sa (3>2) are never used for 1st or 2nd person subjects, unless we define subject as "whatever argument is highest in the animacy hierarchy."

[bradrn]

Huh. I just double checked the wikipedia page on active-stative alignment, and it lists all the Iroquoian languages. Either that's a mistake, or the author has an extremely loose definition of active-stative. As far as I can tell, Mohawk forms like wak (3>1) and sa (3>2) are never used for 1st or 2nd person subjects, unless we define subject as "whatever argument is highest in the animacy hierarchy."

Just checked Hopkins’s Topics in Mohawk Grammar, and it says the following:

Mohawk has an "active-stative" system of pronominal marking … In Mohawk … semantic agents … of both transitive and intransitive verbs are similarly marked. Semantic patients of both intransitive and transitive verbs are also similarly marked. Transitive and intransitive verbs in the stative aspect take patient prefixes

That being said, she doesn’t actually list the pronominal prefixes, and the example sentences are glossed unclearly. However, Bonvillain’s A Grammar of Akwesasne Mohawk does have a list, and talks about how some intransitive verbs take ‘subjective pronominal prefixes’ and others take ‘objective pronominal prefixes’.

[Moose-tache]

I have that book, and read just the opposite. Let me go back and check, and I'll respond later today.

EDIT: OK, so I’ve consulted the various documents I have on Mohawk, and I cannot find any evidence that the object series is used for subjects. This is a more difficult thing to confirm than you might think, because in Mohawk the subject and object markers have almost completely merged, and there are only a few words where you can definitively say “Aha! That is an object marker!” For example, Bonvillain gives the example ukwatv:anahlo: “I’ve been witched.” At first glance this appears to include just a generic first person singular prefix k, except the root verb does not have a w in it. That w tells us that this is not Proto-Iroquoian first person singular subject marker *k, but Proto-Iroquoian first person singular object marker *kʷ! You see what I mean about this being difficult to do in Mohawk. And since inanimate third person is usually unmarked, “something witches me” isn’t going to give us a smoking gun about its transitivity. So I think what’s happening here is that when third person subjects are deprioritized, Mohawk sentences drop into a Schrodinger’s Uncertainty Metaphor, in which the distinction between the subject and object markers is difficult to determine objectively. I guess it's reasonable to call this split-S if you want, but you could also say it's just a normal subject-object distinction that's been worn down so severely that additional features mapped over top of it have become more visible.

What Mohawk lacks is a situation in which an argument can be a) unambiguously marked with object agreement as opposed to subject agreement, and b) unambiguously a syntactic subject as opposed to syntactic object by the language's own rules. That might sound like a high bar to clear, but every Muskogean language, most Siouan languages, and Cherokee, all clear it easily.

[bradrn]

What Mohawk lacks is a situation in which an argument can be a) unambiguously marked with object agreement as opposed to subject agreement, and b) unambiguously a syntactic subject as opposed to syntactic object by the language's own rules. That might sound like a high bar to clear, but every Muskogean language, most Siouan languages, and Cherokee, all clear it easily.

So, does this mean that verbs cannot unambiguously be called ‘transitive’ or ‘intransitive’? Because from this, it sounds like it’s impossible to tell in general whether a verb takes one or two arguments.

[Moose-tache]

Well, in most cases there are two marked participants. So wak- tells you it's a third person subject and a first person object.

Here's the thing. Proto-Iroquoian had two sets, with subject coming first and object coming second. So hsi+kw would be 2>1, and ki+tsa would be 1>2. If you saw tsa by itself, you knew there was an inanimate third person subject. This is basically how Cherokee works, except there are cases where tsa might appear on a verb where the second person is treated as a syntactic subject, so we know it's not 3>2.

In Mohawk, and in Northern Iroquoian generally, the pronominal prefixes are a) mostly fused into opaque units, b) rely on the traditional positioning system described above to distinguish subjects from objects, and c) for third person inanimates apply something that looks kind of like direct-inverse.

In Mohawk, when we see k- we can assume it's a 1st sing subject with or without (because it wouldn't be marked either way) third person inanimate object. For most combinations like 1>2 or 2>1, we have a combination of prefixes, albeit fused to shit and back, but what if we want to show that our 1st sing is an object and we don't have a good prefix to represent our subject? We can't rely on k vs kw anymore, so we repurpose a third person marker wa, and get wak. You can think of this as an extention of the prefix ordering system, but Bonvillain calls it a prefix that indicates the direction of animacy (i.e. inanimate > animate, as opposed to the normally assumed animate > inanimate).

So in short Mohawk works like this:
1) if no inanimates are involved, just use one or two prefixes, subject first and object second, and then put them in a blender.
2) if your inanimate participant is an object, just skip it.
3) if you inanimate is a subject, then it isn't inanimate; use wa and try again.

The thing is, if we accept Bonvillain's perspective that wa- is an animacy hierarchy inverter, and if the subject and object markers aren't distinct, then we could see some of these words as having k as an object marker even when it represents the subject, because whether k is a subject or object marker at that point is a matter for the philosophers.

[Man in Space]

I was today years old when I found out about prophetic perfect tense. Anybody do anything like this in your conlangs? It’d be interesting to derive a future tense from a perfect.

[xxx]

by no longer having a predicate,
default tense is a stative out of time,
that of definitions,
not so far from prophetic tense...

[zompist]

I was today years old when I found out about prophetic perfect tense. Anybody do anything like this in your conlangs? It’d be interesting to derive a future tense from a perfect.

I would take the whole concept with a large dose of salt. For one thing, the major citation is Gesenius, and with all due respect to the dude, linguistics has advanced since 1813. For another... hoo boy, is there a lot of dissension on exactly what the two paradigms of Hebrew verbs mean.

Here's an exhaustive summary. (There's a download link at the bottom of the page.) The distinction was once thought to be tense, then aspect, and then all sorts of strange combinations. About the one thing that's certain is that the SC (suffixing conjugation) is neither "perfect" nor a "tense".

Now, the SC in Biblical Hebrew is for the most part not that complicated: its prototypical use is for completed past events. And the meaning of the PC (prefixing conjugation) is best thought of as not SC: so, non-complete, non-past, or modal.

What to do with future uses of the SC depends on how you analyze its core meaning. If it's tense then it's anomalous and has to be given a weird label like "prophetic tense." If it's completive then it's not terribly surprising that it can be used for future events viewed as complete. There's also the point that if you have a vision, you might well describe it using the past.

Finally, languages using the "wrong tense" for effect is common as dirt. You can use the past tense for future events in English and French, for instance, without even being a prophet. Is there a "science fiction future past tense" in English because most sf is written in the past tense?

[Ares Land]

I was today years old when I found out about prophetic perfect tense. Anybody do anything like this in your conlangs? It’d be interesting to derive a future tense from a perfect.

Yes, I have something very much like that (except it's a remote past tense, not a perfect. I don't remember basing on Hebrew, but maybe that was where I got the idea?)

I ran into the issues zompist describes, more or less; figurative use of the past tense is pretty common.

[Otto Kretschmer]

A question about Ukrainian.

What's the origin of -o in Ukrainian words like Dnipro, Petro, Dmytro, Mykhailo etc. ? All other Slavic languages have cognates of these words that end in a consonant (i.e. Dniepr, Pyotr, Dimitriy, Mikhail)

[jal]

What's the origin of -o in Ukrainian words like Dnipro, Petro, Dmytro, Mykhailo etc. ? All other Slavic languages have cognates of these words that end in a consonant (i.e. Dniepr, Pyotr, Dimitriy, Mikhail)

Though I don't know the answer to the question, "Дняпро" is, according to Wikipedia, one of the possible names for the River in Belarussian. It also states that older forms are Dniper/Dnipr. All Slavic names derive from Old East Slavic "Дънѣпръ" (Dŭněprŭ).

EDIT: Wikipedia mentions the -o, but unfortunately doesn't state it's origins.

JAL

[Otto Kretschmer]

What are some potential phonological processes by which a language with a small syllabe inventory (like Hawaiian or Japanese) could gain more syllabes?

[Man in Space]

What are some potential phonological processes by which a language with a small syllabe inventory (like Hawaiian or Japanese) could gain more syllabes?

Lag assimilation plus umlaut.

*puti > pyti > pytsi > pyts
*ta?i > t’a?i > t’E?i > t’E

Vowel contraction in hiatus.

This.

[zyxw59]

What are some potential phonological processes by which a language with a small syllabe inventory (like Hawaiian or Japanese) could gain more syllabes?

the most obvious thing that comes to mind is losing some vowels, causing new consonant clusters and coda consonants to form. or losing consonants, generating new diphthongs from vowels in hiatus.

borrowing is another option, perhaps particularly borrowing sounds that already exist in the language but in positions or combinations that don't already occur.

[bradrn]

This.

Ah, one of my favourite papers!

(The author has many other good ones, by the way, all open-access on his site: http://alex.francois.free.fr/AFpub_articles_e.htm)

[Darren]

(The author has many other good ones, by the way, all open-access on his site: http://alex.francois.free.fr/AFpub_articles_e.htm)

Those are some really interesting papers. It's nuts that Vanuatu at one point had 138+ languages within a population of under 80,000.

[hwhatting]

What's the origin of -o in Ukrainian words like Dnipro, Petro, Dmytro, Mykhailo etc. ? All other Slavic languages have cognates of these words that end in a consonant (i.e. Dniepr, Pyotr, Dimitriy, Mikhail)

Though I don't know the answer to the question, "Дняпро" is, according to Wikipedia, one of the possible names for the River in Belarussian. It also states that older forms are Dniper/Dnipr. All Slavic names derive from Old East Slavic "Дънѣпръ" (Dŭněprŭ).

EDIT: Wikipedia mentions the -o, but unfortunately doesn't state it's origins.

It has been interpreted as one of the traces of an old o-stem nominative singular in -o(s) that also pops up in other Slavic langugages (SCr. names like Branko, patronymics in -oševič). It goes back to the question of what is the outcome of final PIE *-os and *-om in Balto-Slavic and Proto-Slavic. The communis opinio now seems to be that for Proto-Slavic, it's *-os > -*um > -ъ ( > Zero in the modern languages), and *-om splitting into -*um > -ъ and *-a > -o depending on the accent type, but there are other theories. The one I favour is that *-os > *-as and *-om > *-um /*-am in Balto-Slavic, with Slavic then importing the *u to the Nominative and (East) Baltic generalizing the *-a in the Accusative; the Personal names endings in -o would then be holdovers of old *-as (personal names resisting an analogical alignment of NOM and ACC the longest makes sense). I wrote a term paper on this back at university; if I can find it, I may post references (but they'd be from the 1980s or older).

[Otto Kretschmer]

About Ossetian cases

Are Ossetian genitive and dative descended from PIE cases or a separate creation?