The 'Is this attested?' Thread

[Yalensky]

Is the T-V distinction only a European thing? An areal feature? Or are there languages outside of Europe that use the 2nd person plural pronoun as a formal singular?

[bbbosborne]

from a quick google search and look at wikipedia, it seems tagalog definitely has it.

[Xwtek]

from a quick google search and look at wikipedia, it seems tagalog definitely has it.

It seems like Tagalog is heavily influenced from Spanish, though.

[mèþru]

It isn't only European, but it definitely is far more common in Europe.
https://en.wikipedia.org/wiki/T–V_distinction

[Tropylium]

Nouns in my current proto-language inflect for class and number, quite a lot of classes as in Bantu languages. I also quite like the prefixal shape of them but I had an idea that the root word that gets inflected has sort of "slots" for lack of a better term, where placement of the class/number marker dictates whether it is singular or plural.

Ex.) bas /bɑs̱/ "person, thing, entity, stuff, whatchamacallit etc."
Singular
C1: Sentients : a- : abas "person, individual"
C2: Dang. Sent. : ge- : gebas "criminal, bad person, enemy"
C3: Beasts : ki- : kibas "beast"
C4: Animals : i- : ibas "animal"

Plural
C1: Sentients : <(h)a> : bās "people"
C2: Dang. Sent. : <ge> : bages "criminals, bad people, enemies"
C3: Beasts: <ki> : bakis "beasts"
C4: Animals: <(h)i> : bais "animals"

Sounds like former morphology — maybe -s was also a class marker of some sort that got fossilized behind the productive class markers in some cases. I don't know if you can find "thematic consonants" like this in nominal inflection elsewhere, but they appear in verbal inflection in e.g. Ket, Burushaski and Athabaskan.

Also, I like your "Dang Sentients" class

That reminds me also: does any language have two orthogonal class systems? So e.g. maybe with one slot for shape markers (flat/tall/round/etc.) and another unrelated one for size markers (big/medium/small) or animateness or color or what have you.

[Vilike]

That reminds me also: does any language have two orthogonal class systems? So e.g. maybe with one slot for shape markers (flat/tall/round/etc.) and another unrelated one for size markers (big/medium/small) or animateness or color or what have you.

Michif nominals (mostly of French origin) have two genders assigned to them, one French (m/f, based on etymology), the other Cree (an/in, based on the gender of the Cree traduction of the word); the former triggers agreement on preposed adjectives and articles, the latter triggers agreement on demonstratives and verbs. But the two systems never surfaces both on the same word.

• Iı æ:rple:n awa a:še:-pajI-w
• ART.M.SG airplane this.AN back-move-3.AN
• 'This airplane goes back.'

(BAKKER Peter, A Language of Our Own: The Genesis of Michif, the Mixed Cree-French Language of the Canadian Metis, p.99-109)

[2+3 Clusivity]

That reminds me also: does any language have two orthogonal class systems? So e.g. maybe with one slot for shape markers (flat/tall/round/etc.) and another unrelated one for size markers (big/medium/small) or animateness or color or what have you.

Check out chapter 2 of Classifiers by Aikhenvald. TL; DR: yes.

Per that source, some languages have two noun class (or gender) systems: "pronominal noun class" and "nominal noun class". These class types differ in a number of features including semantics (animacy/sex versus animacy/sex/shape/size/etc.), number of classes, variability of class assignment (pronominal is (typically?) fixed and nominal can be variable), location of marking (personal pronouns/demonstratives/verbal cross-referencing versus adjectives/numerals), etc.

Many language in South America, apparently, will have a small sex/animacy based pronominal noun class system marked on personal pronouns, articles, and in verbal agreement and larger shape/sex/animacy based nominal noun class marked on adjectival and numeral modifiers. For example, Baniwa (North Arawak) has feminine/non-feminine marked on demonstratives, personal pronouns, and verbal cross-referencing AND a set of 44 noun classes marked on adjectives based on shape, form, structure, sex, and animacy.

Better yet, two langauges in the study extend marking of both types onto the same word: Paumarí (Arawá) and Mba (Niger-Congo). In Paumari, non-demonstrative modifiers and verbs can be marked with agreement markers for both pronominal gender (feminine v. non-feminine) and a shape/structure based class marked by the presence or absence of a prefix <ka->. The pronominal noun class system is marked on demonstratives, suffixes on inalienably possessed nouns, and on three verbal suffixes; the nominal (Non-)KA-class markers are prefixed to verbs (and numerals which are stative verbs) and appear as prefixes on inalienably possessed nouns.

A few non-co-occurring examples:

<<ka-voroni-'a-hi> ida ojoro> KA.CLASS-fall-ASP-TH:FEM this.FEM turtle(KA CLASS, FEM) "The turtle fell"

<Ø-varoni-'a-ha ada kahami> NON.KA.CLASS-fall-ASP-TH:MASC this.MASC palm.tree(NON-KA CLASS, MASC) "the palm tree fell"

And two co-occuring examples (recall that numerals are stative verbs which can take both types of class marking):

<ada kasi'i ka-hoara-na> dem.MASC crocodile(KA-CLASS, MASC) KA.CLASS-one-MASC "this one crocodile"

<ida kanawa ka-hoara-ni> dem.FEM canoe(KA-CLASS, FEM) NON.KA.CLASS-one-FEM "this one canoe"

Certain areas seem to be more prone to leaning one way or the other. Demonstratives are noted by Aikhenvald to hold a middle ground between pronominal noun class and nominal noun class marking in Amazonian languages. It is noted that the occurrence of the two class types appears to be based on an animacy scale ranging from personal pronouns, to demonstratives, to common nouns. Starting from the common noun side, shape-based noun (i.e. nominal noun) classes are more common; conversely, starting from personal pronoun side, animacy based (i.e. pronominal) noun classes are more common.

Mba, the other language allowing co-occurrence, nearly exemplifies this pattern and shows that the co-occurrence occurs in the middle of the two poles. It allows nominal gender on nouns, adjectives, demonstratives, possessive constructions (optional), certain interrogatives, numerals (optional) but not on personal pronouns. On the other hand, pronominal gender occurs on personal pronouns, numerals, certain interrogatives, demonstratives (optional) but not on possessive constructions, adjectives, or nouns.

[Zju]

What is the most extreme case of regularly ocurring metathesis documented? I know that some language (Rotokas?) has a synchronic VC → CV / _# in some of its morphology and that Proto-Slavic underwent {e o}R > R{ē ō} / C., but these are relatively tame. Is there something regular to the likes of, say, wV(C)bV → bV(C)wV?

[Pabappa]

Bump, because it's been a month and this is a good question. My own instinct tells me that those pacific langs are the winner, with Slavic close behind , but I'd be curious to know more too.

[Nortaneous]

What is the most extreme case of regularly ocurring metathesis documented? I know that some language (Rotokas?) has a synchronic VC → CV / _# in some of its morphology and that Proto-Slavic underwent {e o}R > R{ē ō} / C., but these are relatively tame. Is there something regular to the likes of, say, wV(C)bV → bV(C)wV?

Rotuman, not Rotokas.

There's derivational metathesis in the verbal system of Saanich, but I'm not sure how widespread / productive it is. The Basque aspiration did some wild stuff (the details of which I can't remember off the top of my head) that I don't think is metathesis strictly speaking but is at least in the same ballpark. I've heard of some Chinese dialects where erhua becomes a medial retroflex lateral, but I have no idea how I'd go about finding this again. Greek has both quantitative metathesis and yod metathesis. There are a few more examples on Wikipedia - Hebrew, dialectal Finnish, etc. - and more here.

There's also Armenian *Pr > rP, which might have been regular: (and apparently also *sr > *rs?)
- *swidro- > kʰirtn
- *ḱubhros > surb
- *daḱru > *draḱu > artaws-r
- *gʷreh2won > erkan
- *bhreh1wr > aɫbewr (dissimilation)
- *bhreh2ter > eɫbayr (more dissimilation)

[Karch]

Kwara'ae underwent a deliberate metathesis of the word-final vowel and the preceding consonant, making most words end in consonants.

[chris_notts]

I've been thinking about differences in the scope of negation for modal verbs in Sint, and how that interacts with raising of negation. I had the idea of the mood of the complement affecting the scope: with an optative complement, negating a deontic modal has scope over the modal itself (e.g. must -> don't have to), whereas with a potential complement it doesn't (e.g. must -> must not). This compares to positive polarity contexts, where the complement of a deontic modal is always in the optative.

The idea was that since the optative is used in main clauses to describe what should occur, and the potential is used for possibility (future, habitual, ability, ...) then negating "must" with an optative complement should negate the obligation, whereas negating it with the potential should negate the possibility. To gloss very roughly, it would be the difference between "it mustn't be that he should go" (-> he doesn't have to go, but can), and "it mustn't be that he could go" (-> it shouldn't be possible that he goes).

Does anyone know any language where changes in the mood of a modal complement affect the scope of negation marking on the matrix verb?

The behaviour of the two modal classes differs under negation. Firstly, the complement clause can be negated, in which case the mood of the complement is the same as in the positive polarity examples above:

lees isınaal
lees is-n-naa-l
it.must.be NEG-1EXC-go-OPT
`I musn't go'

It is also possible, though, to raise the negation to the matrix clause. Epistemic modals still take the potential in this case, but the mood of deontic modal complements is more complex. They can take either the optative or potential with a difference in meaning. The choice alters the scope of negation: with an optative complement negation has scope over the modal, whereas with a potential complement the modal has scope over the negation. Compare the following:

vilees nınaal
vi-lees n-naa-l
NEG-must.be 1EXC-go-OPT
`I don't have to go'

vilees nınaaje
vi-lees n-naa-e
NEG-must.be 1EXC-go-POT
`I musn't go'

vimalın nınaaje
vi-malın n-naa-e
NEG-might.be 1EXC-go-POT
`I might not go'