(L&N) Q&A Thread - Quick questions go here [2010-2019]

A forum for discussing linguistics or just languages in general.
User avatar
Frislander
mayan
mayan
Posts: 2088
Joined: 14 May 2016 18:47
Location: The North

Re: (L&N) Q&A Thread - Quick questions go here

Post by Frislander »

All4Ɇn wrote:
Frislander wrote:That's just suffixing the possessed noun with a pronominal possessive suffix, it's not its own case at all.
Even if the noun cannot change cases in the situation? Thanks for the info also
Even in that circumstance yes, but then neither nor any other natlang I know with both cases and pronominal possessive suffixes neutralises the case distinction when the noun is possessed (I mean why should it be? I can't think of any semantic justification for it.).
User avatar
MrKrov
banned
Posts: 1929
Joined: 12 Aug 2010 02:47
Location: /ai/ > /a:/
Contact:

Re: (L&N) Q&A Thread - Quick questions go here

Post by MrKrov »

Frislander wrote:I don't know of a language which relies just of the possessed marker and apposing the possessor.
You know I mentioned Fijian right there.
(2) Fijian (Austronesian; Fiji; Dixon 1988: 36)
a.
a mata-i Jone
art eye-poss John
‘John’s eye’

b.
a liga-i ‘eirau
art hand-poss 1du.excl
‘our hand(s)’
User avatar
Frislander
mayan
mayan
Posts: 2088
Joined: 14 May 2016 18:47
Location: The North

Re: (L&N) Q&A Thread - Quick questions go here

Post by Frislander »

MrKrov wrote:
Frislander wrote:I don't know of a language which relies just of the possessed marker and apposing the possessor.
You know I mentioned Fijian right there.
(2) Fijian (Austronesian; Fiji; Dixon 1988: 36)
a.
a mata-i Jone
art eye-poss John
‘John’s eye’

b.
a liga-i ‘eirau
art hand-poss 1du.excl
‘our hand(s)’
But Fijian also has possessive suffixes, so I don't think it counts.
User avatar
MrKrov
banned
Posts: 1929
Joined: 12 Aug 2010 02:47
Location: /ai/ > /a:/
Contact:

Re: (L&N) Q&A Thread - Quick questions go here

Post by MrKrov »

But not at the same time, so the pattern does exist.
Yoruba also has a variant (if limited in expression) and so it exists again. It's logically possible and attested, even if mariginally. It should suffice for their purposes.
Last edited by MrKrov on 13 Jul 2017 14:11, edited 1 time in total.
Trebor
cuneiform
cuneiform
Posts: 164
Joined: 24 Nov 2014 18:53

Re: (L&N) Q&A Thread - Quick questions go here

Post by Trebor »

Aside from Celtic and Semitic, are there additional languages or language families with regular series of inflected/conjugated adpositions?

Are there Mande languages of West Africa that form relative clauses in a fashion more familiar to native English speakers than the internally-headed or correlative type found in, e.g., Bambara?
User avatar
MrKrov
banned
Posts: 1929
Joined: 12 Aug 2010 02:47
Location: /ai/ > /a:/
Contact:

Re: (L&N) Q&A Thread - Quick questions go here

Post by MrKrov »

Yes to the first. It occurs in miscellaneous languages.
(1) Maybrat (Dol 1999: 88)
T-ai m-kah ara.
1sg-hit 3sg.n-with stick
‘I hit with a stick.’
(11) Abkhaz (Hewitt 1979: 103)
a. a-jə̀yas a-q’nə̀
def-river 3sg-at
‘at the river’

b. sarà s-q’ənt˚’
I 1sg-from
‘from me’
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: (L&N) Q&A Thread - Quick questions go here

Post by Salmoneus »

Frislander wrote:
All4Ɇn wrote:
Frislander wrote:That's just suffixing the possessed noun with a pronominal possessive suffix, it's not its own case at all.
Even if the noun cannot change cases in the situation? Thanks for the info also
Even in that circumstance yes, but then neither nor any other natlang I know with both cases and pronominal possessive suffixes neutralises the case distinction when the noun is possessed (I mean why should it be? I can't think of any semantic justification for it.).
I can see four justifications (some related)...

First, many language have morphological rules limiting the number of 'slots' that a word can have. This sometimes means that adding one affix means losing another important affix like case. [Whether that means that case is unmarked for possessed items, or that case is impossible for them, probably varies with the language]

Second, many languages have syntactic rules limiting what can be done with things like possessors and possessions (relativisation, etc). It may simply be that possessions are restricted to a single role, and hence don't require case. It's easy to imagine possessions being syntactically-semantically barred from being subjects, for instance, as they inherently lack some degree of agency. Likewise they could be barred from being recipients quite easily. So they could be only direct objects (or else in a prepositional phrase).

Third, many languages only use possessive inflexion for certain types of relationship in which the possessor is tightly bound to the possessor semantically. Body parts, actions, intimate items, emotions, etc - its quite reasonable in these cases to see the possessor as the real argument of the verb. So you wouldn't need to encode "the man's arm hit me" vs "I hit the man's arm" - instead, you'd just say "the man (his arm) hit me" and "I hit the man (his arm)", with 'man' carrying all the case marking. This is unlikely for more remote possession ("the man's axe" - where other people might hit you with his axe!), but there a different construction could be used ("the axe for the man", for instance).

Fourth, and I'm not sure if you'd call this semantic, syntactic, morphological or all of the above: some languages treat possessions almost like verbs - possessive affixes and subject agreement affixes are often the same, for instance. If verbs can't take nominal affixes like case marking, then perhaps possessions can't either.
User avatar
gach
MVP
MVP
Posts: 513
Joined: 07 Aug 2013 01:26
Location: displaced from Helsinki

Re: (L&N) Q&A Thread - Quick questions go here

Post by gach »

Trebor wrote:Aside from Celtic and Semitic, are there additional languages or language families with regular series of inflected/conjugated adpositions?
Having these seems natural to me whenever a language has possessive suffixes and its adpositions are structurally relational nouns. They are commonly found across Uralic and northern Samoyedic goes as far as not even having separate forms for the personal pronouns beyond the core cases nominative, accusative, and genitive. Instead they provide the pronominal local cases through the possessive case forms of the adposition ńa- in Nenets, no- in Enets, and na- in Nganasan. As an example, here's the singular paradigm in Nganasan (which in addition merges the core cases into single forms),

Code: Select all

        SG1       SG2       SG3
N/A/G   mənə      tənə      sïtï
LAT     nanə      nantə     nantu
LOC     nanunə    nanuntə   nanuntu
ELA     nagətənə  nagətətə  nagətətu
PROL    namənunə  namənuntə namənuntu
Closer to home, I should also mention that we can conjugate our postpositions in Finnish as well. It also happens colloquially, even though the possessive suffixes are being dropped away,

Mi-tä-s sä oo-t löytä-ny pää-lle-s?
what-PART-DP SG2 be-SG2 find-PST.PART on-ALL-SG2.POSS
"What have you found on you (to wear)?"

Sama-n taki-n ku Masa-l ol-i Lapi-s pää-llä-än.
same-ACC jacket-ACC REL <name>-ADE be-SG3.PST Lapland-INE on-ADE-SG3.POSS
"The same jacket that Masa had in Lapland on him."

(DP = discourse particle)
ImageKištaLkal sikSeic
User avatar
Frislander
mayan
mayan
Posts: 2088
Joined: 14 May 2016 18:47
Location: The North

Re: (L&N) Q&A Thread - Quick questions go here

Post by Frislander »

Salmoneus wrote:
Spoiler:
Frislander wrote:
All4Ɇn wrote:
Frislander wrote:That's just suffixing the possessed noun with a pronominal possessive suffix, it's not its own case at all.
Even if the noun cannot change cases in the situation? Thanks for the info also
Even in that circumstance yes, but then neither nor any other natlang I know with both cases and pronominal possessive suffixes neutralises the case distinction when the noun is possessed (I mean why should it be? I can't think of any semantic justification for it.).
I can see four justifications (some related)...

First, many language have morphological rules limiting the number of 'slots' that a word can have. This sometimes means that adding one affix means losing another important affix like case. [Whether that means that case is unmarked for possessed items, or that case is impossible for them, probably varies with the language]

Second, many languages have syntactic rules limiting what can be done with things like possessors and possessions (relativisation, etc). It may simply be that possessions are restricted to a single role, and hence don't require case. It's easy to imagine possessions being syntactically-semantically barred from being subjects, for instance, as they inherently lack some degree of agency. Likewise they could be barred from being recipients quite easily. So they could be only direct objects (or else in a prepositional phrase).

Third, many languages only use possessive inflexion for certain types of relationship in which the possessor is tightly bound to the possessor semantically. Body parts, actions, intimate items, emotions, etc - its quite reasonable in these cases to see the possessor as the real argument of the verb. So you wouldn't need to encode "the man's arm hit me" vs "I hit the man's arm" - instead, you'd just say "the man (his arm) hit me" and "I hit the man (his arm)", with 'man' carrying all the case marking. This is unlikely for more remote possession ("the man's axe" - where other people might hit you with his axe!), but there a different construction could be used ("the axe for the man", for instance).

Fourth, and I'm not sure if you'd call this semantic, syntactic, morphological or all of the above: some languages treat possessions almost like verbs - possessive affixes and subject agreement affixes are often the same, for instance. If verbs can't take nominal affixes like case marking, then perhaps possessions can't either.
The first three (especially the second and third) definitely sound like things that could happen, though I have my doubts about the fourth. Could you show me any natlangs that do any of these things?
gach wrote:
Trebor wrote:Aside from Celtic and Semitic, are there additional languages or language families with regular series of inflected/conjugated adpositions?
Having these seems natural to me whenever a language has possessive suffixes and its adpositions are structurally relational nouns.
I'd say the same with person-marking and verb-like adpositions.

Here's the WALS map on the subject.
Trebor
cuneiform
cuneiform
Posts: 164
Joined: 24 Nov 2014 18:53

Re: (L&N) Q&A Thread - Quick questions go here

Post by Trebor »

MrKrov wrote:Yes to the first. It occurs in miscellaneous languages.
(1) Maybrat (Dol 1999: 88)
T-ai m-kah ara.
1sg-hit 3sg.n-with stick
‘I hit with a stick.’
(11) Abkhaz (Hewitt 1979: 103)
a. a-jə̀yas a-q’nə̀
def-river 3sg-at
‘at the river’

b. sarà s-q’ənt˚’
I 1sg-from
‘from me’
Thanks for these examples. It seems that both Maybrat and Abkhaz prohibit adpositions from modifying full noun phrases and thus require the use of resumptive pronouns. Maybe Celtic and Semitic are somewhat atypical in allowing all adpositions to be inflected/conjugated directly?
gach wrote:Having these seems natural to me whenever a language has possessive suffixes and its adpositions are structurally relational nouns.
Such an idea definitely makes sense. Thanks for the data and analysis. [:)]

Would the prepositions of, say, Arabic have originally been relational nouns? They don't resemble any basic vocabulary items, e.g., parts of the body. How might this feature have otherwise arisen? My supposition would be analogy with other parts of speech taking pronominal suffixes/clitics and the limited use of independent personal pronouns.
User avatar
MrKrov
banned
Posts: 1929
Joined: 12 Aug 2010 02:47
Location: /ai/ > /a:/
Contact:

Re: (L&N) Q&A Thread - Quick questions go here

Post by MrKrov »

Trebor wrote:Thanks for these examples. It seems that both Maybrat and Abkhaz prohibit adpositions from modifying full noun phrases and thus require the use of resumptive pronouns. Maybe Celtic and Semitic are somewhat atypical in allowing all adpositions to be inflected/conjugated directly?
What? Explain this to me. Those are adpositions inflected directly and there are no resumptive pronouns anywhere.
User avatar
Frislander
mayan
mayan
Posts: 2088
Joined: 14 May 2016 18:47
Location: The North

Re: (L&N) Q&A Thread - Quick questions go here

Post by Frislander »

MrKrov wrote:
Trebor wrote:Thanks for these examples. It seems that both Maybrat and Abkhaz prohibit adpositions from modifying full noun phrases and thus require the use of resumptive pronouns. Maybe Celtic and Semitic are somewhat atypical in allowing all adpositions to be inflected/conjugated directly?
What? Explain this to me. Those are adpositions inflected directly and there are no resumptive pronouns anywhere.
Perhaps they don't understand the gloss and think that hyphens are word boundaries not affix boundaries.
Trebor
cuneiform
cuneiform
Posts: 164
Joined: 24 Nov 2014 18:53

Re: (L&N) Q&A Thread - Quick questions go here

Post by Trebor »

MrKrov wrote:What? Explain this to me.
Sure.
MrKrov wrote:
...
1sg-hit 3sg.n-with stick
‘I hit with a stick.’
...
def-river 3sg-at
‘at the river’

...
I 1sg-from
‘from me’
If the adposition could be used with a full noun phrase and resumption was not needed, the glosses would look something like the following.

1sg-hit with stick
‘I hit with a stick.’
(A somewhat puzzling sentence to begin with; is there a third-person direct object missing from the English translation?)

def-river at
‘at the river’

1sg-from
‘from me’
Sumelic
greek
greek
Posts: 566
Joined: 18 Jun 2013 23:01

Re: (L&N) Q&A Thread - Quick questions go here

Post by Sumelic »

I get what you mean, but I'm not sure if "resumptive pronoun" is the right word for this, because it's not a full pronoun but merely agreement morphology on the adposition. It might qualify as clitic doubling depending on if the pronoun agreement markers can be analyzed as clitics.

I believe this actually shows up in Celtic too in some situations. It's definitely possible with coordinated pronouns in Welsh, as in

amdanom ni a nhw
about-1PL 1PL and 3PL
about us and them

(example taken from "Coordination and asymmetric agreement in Welsh", Louisa Sadler).

According to an Omniglot page I found, "Prepositional pronouns in the Brythonic languages", it's a feature of formal Welsh to have no accompanying separate pronoun word, which implies I guess that colloquial Welsh would be likely to have one.

I don't think doubling/agreement ever shows up in Celtic for prepositions followed by non-pronoun nouns, though, as in that first Abkhaz example.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: (L&N) Q&A Thread - Quick questions go here

Post by Creyeditor »

Now there is actually an interesting parallel in verbal agreement. Some languages allow verbal agreement only if there is no full noun phrase.

Language I:
*I sing-I
*the men sing-3PL.M
sing-3PL.M

Language II:
*I sing-I
the men sing-3PL.M
sing-3PL.M
Spoiler:
Of course there are also language that require pro drop if the verb agrees.

Language III:
I sing-I
*the men sing-3PL.M
sing-3PL.M

Language IV:
I sing-I
the men sing-3PL.M
sing-3PL.M
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Trebor
cuneiform
cuneiform
Posts: 164
Joined: 24 Nov 2014 18:53

Re: (L&N) Q&A Thread - Quick questions go here

Post by Trebor »

Sumelic wrote:I get what you mean, but I'm not sure if "resumptive pronoun" is the right word for this, because it's not a full pronoun but merely agreement morphology on the adposition. It might qualify as clitic doubling depending on if the pronoun agreement markers can be analyzed as clitics.
Fair point. For my less-than-accurate use of the term, we can blame being influenced by how 'resumptive pronoun' is used in connection with Arabic even though independent pronouns are never involved.
User avatar
Frislander
mayan
mayan
Posts: 2088
Joined: 14 May 2016 18:47
Location: The North

Re: (L&N) Q&A Thread - Quick questions go here

Post by Frislander »

Trebor wrote:
Sumelic wrote:I get what you mean, but I'm not sure if "resumptive pronoun" is the right word for this, because it's not a full pronoun but merely agreement morphology on the adposition. It might qualify as clitic doubling depending on if the pronoun agreement markers can be analyzed as clitics.
Fair point. For my less-than-accurate use of the term, we can blame being influenced by how 'resumptive pronoun' is used in connection with Arabic even though independent pronouns are never involved.
Again, the WALS map and associated chapter make this distinction by differentiating languages which only use such inflection to replace pronouns and those which use it with nouns as well.
User avatar
MrKrov
banned
Posts: 1929
Joined: 12 Aug 2010 02:47
Location: /ai/ > /a:/
Contact:

Re: (L&N) Q&A Thread - Quick questions go here

Post by MrKrov »

Does anyone know of a language that has diminutive agreement, but that agreement stacks with and doesn't replace regular gender agreement?
Sumelic
greek
greek
Posts: 566
Joined: 18 Jun 2013 23:01

Re: (L&N) Q&A Thread - Quick questions go here

Post by Sumelic »

MrKrov wrote:Does anyone know of a language that has diminutive agreement, but that agreement stacks with and doesn't replace regular gender agreement?
I'm not aware of any. I just looked at "Evaluative morphology and noun classification: a cross-linguistic study of Africa" (Francesca Di Garbo), which says
"Evaluative markers do not generally induce agreement. The only exception is represented by those languages where diminutives and augmentatives are noun classes or are encoded by means of gender shifts." (117)
None of the noun class/gender shift examples in the paper include stacking agreement. Stacking prefixes occurs in some Bantu languages on some nouns, but that can be analyzed as derivational, since the agreement is only with the first, diminutive prefix.

Some Bantu languages have more than one diminutive or augmentive gender/noun class. Unfortunately, the paper only gives an example of a language where multiple singular augmentive noun classes exist for forming the augmentives of nouns of different original genders (Wamey); while it seems plausible that the same phenomenon could exist for diminutives, I don't know of an example (for singular noun classes, that is: since gender and number marking are generally fusional in Bantu, there are a number of languages with one singular diminutive noun class and one plural diminutive class, corresponding to singular and plural other noun classes.) And anyway, this still isn't stacking agreement.

I also searched a bit through the Edinburgh Handbook of Evaluative Morphology and couldn't find an example of stacking agreement, although my access though Google Books is limited so it might be easier to check it if you can get a physical copy.
User avatar
Parlox
greek
greek
Posts: 495
Joined: 10 Feb 2017 20:28
Location: Ehh

Re: (L&N) Q&A Thread - Quick questions go here

Post by Parlox »

What alignment is rarest in natlangs?
:con: Gândölansch (Gondolan)Feongkrwe (Feongrkean)Tamhanddön (Tamanthon)Θανηλοξαμαψⱶ (Thanelotic)Yônjcerth (Yaponese)Ba̧supan (Basupan)Mùthoķán (Mothaucian) :con:
Locked