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

A forum for discussing linguistics or just languages in general.
shimobaatar
darkness
darkness
Posts: 8716
Joined: Fri 12 Jul 2013, 22:09
Location: PA → IN

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

Post by shimobaatar » Mon 17 Jul 2017, 19:40

Tripartite's pretty rare. I don't know for certain if it's the rarest, though.
User avatar
Parlox
sinic
sinic
Posts: 279
Joined: Fri 10 Feb 2017, 20:28
Location: North-East Favorc shoreline, Suot village, the Saon.

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

Post by Parlox » Mon 17 Jul 2017, 19:41

Ok, thank you.
My main conlangs are Podmåri, and Gwynwth.
i have several other smaller conlangs, such as Tänggorosepero , Makǔdorın, La Patagonê, and Nòvgraďà.
Along with a unnamed sign language, and multiple conlangs that are on hold.
User avatar
MrKrov
cleardarkness
cleardarkness
Posts: 2414
Joined: Thu 12 Aug 2010, 01:47
Location: /ai/ > /a:/
Contact:

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

Post by MrKrov » Mon 17 Jul 2017, 20:27

Aww bummer, Sumelic.
Adarain
greek
greek
Posts: 642
Joined: Fri 03 Jul 2015, 14:36
Location: Switzerland, usually

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

Post by Adarain » Mon 17 Jul 2017, 21:03

Parlox wrote:What alignment is rarest in natlangs?
This question has the false premise of assuming there to be a finite number of alignments. Basically all languages are either purely nom/acc or show some kind of split system (bear in mind however that many languages we consider to be purely accusative are actually split systems between accusative and direct/nonmarking. For example all western IE languages I’m aware of do not distinguish nominative and accusative in the neuters. Syntactically these languages will be accusative though). I’m not aware of a language that is strictly ergative at all (even some of the most ergative languages of Australia, featuring even deeply ingrained syntactic ergativity, have splits in the morphology).

Tripartite, arising only when accusative and ergative systems overlap, is usually restricted to a small set of words; very few languages show it in a significant fashion.

Finally, transitive alignment is just actually dumb, and we only know of it in one corner of one language (Rushan, restricted to I think past tense or sth like that), and it seems to have arisen recently and is already fading out of use again.
At kveldi skal dag lęyfa,
Konu es bręnnd es,
Mæki es ręyndr es,
Męy es gefin es,
Ís es yfir kømr,
Ǫl es drukkit es.
shimobaatar
darkness
darkness
Posts: 8716
Joined: Fri 12 Jul 2013, 22:09
Location: PA → IN

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

Post by shimobaatar » Mon 17 Jul 2017, 21:07

Adarain wrote:
Parlox wrote:What alignment is rarest in natlangs?
This question has the false premise of assuming there to be a finite number of alignments.
Yes, thank you! I couldn't figure out how to articulate this.
User avatar
LinguistCat
hieroglyphic
hieroglyphic
Posts: 73
Joined: Sat 06 May 2017, 06:48

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

Post by LinguistCat » Mon 17 Jul 2017, 21:11

Moved this from the wrong thread over to here.

When/how did Classical Japanese 好く, a yodan verb according to this, become Modern Japanese 好き(な), a na-adjective? Was it about the same as 嫌ふ/嫌う becoming 嫌い(な), which was also a yodan verb and also became a na-adjective? I suppose since I can see a few ways it could happen, I'm more interested in when the change started and how long (approx) it took. Thanks.
User avatar
eldin raigmore
fire
fire
Posts: 5636
Joined: Sat 14 Aug 2010, 18:38
Location: SouthEast Michigan

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

Post by eldin raigmore » Mon 17 Jul 2017, 21:36

Parlox wrote:What alignment is rarest in natlangs?
I'm assuming, as shimobaatar assumes, that you mean "morphosyntactic alignment of intranstive Subjects with transitive Agents and/or transitive Patients".
shimobaatar wrote:Tripartite's pretty rare. I don't know for certain if it's the rarest, though.
According to http://wals.info/feature/98A#2/25.5/148.9, if we're talking about case-marking of full noun-phrases, the rarest are Tripartite and Active/Stative (Split-S or Split-Intransitive), at about 2% each.
Accusative/Nominative with a Marked Nominative is also pretty rare, at about 3%.
Then Ergative/Absolutive at about 17%;
then Standard Accusative/Nominative (i.e. with unmarked Nominative) at about 24%;
then Neutral (that is, these three roles don't differ by case-marking morphology on the noun when they're full NPs) at about 52%.

According to http://wals.info/feature/99A#2/25.5/148.5, if we're talking about case-marking of independent pronouns, Tripartite, Active/Stative, Accusative/Marked-Nominative, and None, are all equally rare at about 2%.
("None" is for languages in which pronouns are not permitted in (at least) one of the three roles; either the intransitive Subject can't be a pronoun, or the transitive Agent can't be a pronoun, or the transitive Patient can't be a pronoun.)
Ergative/Absolutive is next at about 12%;
then Standard Accusative/Nominative at about 35%;
then Neutral at about 46%.

According to http://wals.info/feature/100A#2/16.6/148.5, if we're talking about agreement-marking on the verb,
the least common is Hierarchical Alignment, at about 3%.
(In Hierarchical Alignment, nouns and pronouns are arranged in a memorized hierarchy of person and definiteness and animacy and maybe other things as well, such as empathy or topicality or salience. The verb probably agrees only with the highest participant. A transitive verb is probably Voice-marked with "Inverse Voice" when the lower participant is the Agent and the higher participant is the Patient. It may be unmarked, or it may be Voice-marked, to show it's in "Direct Voice" when the higher participant is the Agent and the lower participant is the Patient.)

Next rarest is Ergative Alignment at 5%. (In this context, Ergative marking on the verb would be because transitive verbs agree with their Patient the same way as intransitive verbs agree with their Subjects. Transitive verbs may not agree with their Agents at all; or, they may agree with them, but not the same way intransitive verbs agree with their Subjects.)

Next is Active/Stative Alignment at around 7%; intransitive verbs sometimes agree with their Subjects the way transitive verbs agree with their Agents, and sometimes intransitive verbs agree with their Subjects the way transitive verbs agree with their Patients. I can't tell whether the author means this to be co-extensive with Fluid Alignment (where even with the same noun for Subject and the same intransitive Verb the alignment might be Agent-like or Patient-like depending on semantic factors), or merely to include Fluid Alignment. I think they mean it to include not only Fluid Alignment, but also Dynamic/Stative or Active/Inactive Alignment in which, as long as the same noun or pronoun is the Subject of the same intransitive verb, the alignment is fixed to either be Agent-like or Patient-like.

Close behind is Split Alignment at around 7%. This is not really an alignment type. Rather, this occurs when two or more of the other alignment types both occur in the language. For instance, maybe
  • the language is Accusative/Nominative when the verb is Past Tense, but Ergative/Absolutive when the verb is Future Tense, and Tripartite when the verb is Present Tense.
  • Or Accusative/Nominative when the verb is Perfective Aspect, but Ergative/Absolutive when the verb is Imperfective Aspect.
  • Or Accusative/Nominative when the verb is Realis Mood, but Ergative/Absolutive when the verb is Irrealis Mood.
  • Or the nouns and pronouns are in hierarchy; and the language is Accusative/Nominative for participants at the top of the hierarchy, Ergative/Absolutive for participants at the bottom of the hierarchy, and Tripartite for participants in the middle of the hierarchy
And there could be other splits besides or instead.

Next more common is Neutral at about 22%. In this context, "neutral" means the verb doesn't agree with its participants.

Commonest at about 56% is Accusative/Nominative Alignment. Transitive verbs agree with their Agents the same way intransitive verbs agree with their Subjects. Transitive verbs may not agree with their Patients at all; but if they do, it won't be the same way intransitive verbs agree with their Subjects.

[hr][/hr]

One would expect that it would be unusual to find both that the full noun-phrases in these three roles are not case-marked differently, and also that the verb does not agree with them differently.
But http://wals.info/combinations/98A_100A#2/25.5/148.9 shows 21 languages out of the 188 languages for which both features 98A and 100A are recorded in their database are Neutral/Neutral. That's about 11%.

One would expect that it would be unusual to find both that the pronouns in these three roles are not case-marked differently, and also that the verb does not agree with them differently.
But http://wals.info/combinations/99A_100A#2/25.5/148.9 shows 19 languages out of the 179 languages for which both features 99A and 100A are recorded in their database are Neutral/Neutral. That's about 11%.
(There are no None/Neutral nor None/Split nor None/Hierarchical. IIANM by definition there couldn't be.)

It might be worthwhile to compare how languages align the case-marking of their pronouns vs how they align the case-marking of their full noun-phrases.
See http://wals.info/combinations/99A_98A#2/25.5/148.9
Note that all three of the languages for which 99A is "None" have 98A as "Neutral".
Notice that, for 142 of the languages for which both 98A and 99A are recorded in their database, 98A and 99A have the same value. Out of all 175 of the languages in their database for which they've recorded both features, around 81% have the same answer for the full noun-phrases (98A) as for the pronouns (99A).

[hr][/hr]

If you look at http://wals.info/combinations/98A_99A_100A#2/25.5/148.9 you will see several combinations that have either 0 languages or only 1 language.

[hr][/hr]

I hope that helps!
Last edited by eldin raigmore on Tue 18 Jul 2017, 19:07, edited 2 times in total.
Iyionaku
roman
roman
Posts: 1422
Joined: Sun 25 May 2014, 13:17

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

Post by Iyionaku » Tue 18 Jul 2017, 05:36

Adarain wrote:
Parlox wrote:What alignment is rarest in natlangs?
Finally, transitive alignment is just actually dumb, and we only know of it in one corner of one language (Rushan, restricted to I think past tense or sth like that), and it seems to have arisen recently and is already fading out of use again.
What is this? I assume subject and object of a transitive verb both in the same case and the subject of an intransitive verb in another?
Heaven and Earth, but I feel the color of the cake when you keep the Victoria.
I had a mantra on the moss and I had to go to bed.


Oh, and there is a [ɕ] in my name!
Trebor
cuneiform
cuneiform
Posts: 167
Joined: Mon 24 Nov 2014, 18:53

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

Post by Trebor » Tue 18 Jul 2017, 07:27

Adarain wrote:For example all western IE languages I’m aware of do not distinguish nominative and accusative in the neuters.
Which (if any) languages possessing masculine, feminine, and neuter genders also morphologically distinguish--usually or always--accusative from nominative?
Iyionaku
roman
roman
Posts: 1422
Joined: Sun 25 May 2014, 13:17

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

Post by Iyionaku » Tue 18 Jul 2017, 08:17

Trebor wrote:
Adarain wrote:For example all western IE languages I’m aware of do not distinguish nominative and accusative in the neuters.
Which (if any) languages possessing masculine, feminine, and neuter genders also morphologically distinguish--usually or always--accusative from nominative?
Apparently not, at least definitely no Indo-European language.
Heaven and Earth, but I feel the color of the cake when you keep the Victoria.
I had a mantra on the moss and I had to go to bed.


Oh, and there is a [ɕ] in my name!
shimobaatar
darkness
darkness
Posts: 8716
Joined: Fri 12 Jul 2013, 22:09
Location: PA → IN

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

Post by shimobaatar » Tue 18 Jul 2017, 11:46

Iyionaku wrote:
Adarain wrote:
Parlox wrote:What alignment is rarest in natlangs?
Finally, transitive alignment is just actually dumb, and we only know of it in one corner of one language (Rushan, restricted to I think past tense or sth like that), and it seems to have arisen recently and is already fading out of use again.
What is this? I assume subject and object of a transitive verb both in the same case and the subject of an intransitive verb in another?
As far as I know, that's exactly right.
Adarain
greek
greek
Posts: 642
Joined: Fri 03 Jul 2015, 14:36
Location: Switzerland, usually

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

Post by Adarain » Tue 18 Jul 2017, 13:07

Trebor wrote:
Adarain wrote:For example all western IE languages I’m aware of do not distinguish nominative and accusative in the neuters.
Which (if any) languages possessing masculine, feminine, and neuter genders also morphologically distinguish--usually or always--accusative from nominative?
Germanic languages. Except in the neuter, as said. Slavic languages, same story (based on information I just got from wikipedia). Romance, Baltic and Celtic seem to have collapsed the gender system into two (I know that this is not stricly true of Romance at least, some remnants of the neuter remain, but Romance languages also don’t distinguish a whole lot of cases anymore so it’s moot), anything further east I know absolutely nothing about, but I did check and Sanskrit retains the three-gender system and does not distinguish nom/acc in neuter but does everywhere else.
At kveldi skal dag lęyfa,
Konu es bręnnd es,
Mæki es ręyndr es,
Męy es gefin es,
Ís es yfir kømr,
Ǫl es drukkit es.
Iyionaku
roman
roman
Posts: 1422
Joined: Sun 25 May 2014, 13:17

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

Post by Iyionaku » Tue 18 Jul 2017, 13:14

Adarain wrote:
Trebor wrote:
Adarain wrote:For example all western IE languages I’m aware of do not distinguish nominative and accusative in the neuters.
Which (if any) languages possessing masculine, feminine, and neuter genders also morphologically distinguish--usually or always--accusative from nominative?
Germanic languages. Except in the neuter, as said. Slavic languages, same story (based on information I just got from wikipedia). Romance, Baltic and Celtic seem to have collapsed the gender system into two (I know that this is not stricly true of Romance at least, some remnants of the neuter remain, but Romance languages also don’t distinguish a whole lot of cases anymore so it’s moot), anything further east I know absolutely nothing about, but I did check and Sanskrit retains the three-gender system and does not distinguish nom/acc in neuter but does everywhere else.
Hooops, I got the question completely wrong. I thought the question was "Are there any languages wherein nom-acc is distinguished in neuter gender?" This is indeed, as Adarain has confirmed, apparently not existent.
Heaven and Earth, but I feel the color of the cake when you keep the Victoria.
I had a mantra on the moss and I had to go to bed.


Oh, and there is a [ɕ] in my name!
User avatar
Omzinesý
mayan
mayan
Posts: 2220
Joined: Fri 27 Aug 2010, 07:17
Location: nowhere [naʊhɪɚ]

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

Post by Omzinesý » Tue 18 Jul 2017, 14:25

You mean in IE languages?
Adarain
greek
greek
Posts: 642
Joined: Fri 03 Jul 2015, 14:36
Location: Switzerland, usually

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

Post by Adarain » Tue 18 Jul 2017, 15:58

Iyionaku wrote:
Adarain wrote:
Trebor wrote:
Adarain wrote:For example all western IE languages I’m aware of do not distinguish nominative and accusative in the neuters.
Which (if any) languages possessing masculine, feminine, and neuter genders also morphologically distinguish--usually or always--accusative from nominative?
Germanic languages. Except in the neuter, as said. Slavic languages, same story (based on information I just got from wikipedia). Romance, Baltic and Celtic seem to have collapsed the gender system into two (I know that this is not stricly true of Romance at least, some remnants of the neuter remain, but Romance languages also don’t distinguish a whole lot of cases anymore so it’s moot), anything further east I know absolutely nothing about, but I did check and Sanskrit retains the three-gender system and does not distinguish nom/acc in neuter but does everywhere else.
Hooops, I got the question completely wrong. I thought the question was "Are there any languages wherein nom-acc is distinguished in neuter gender?" This is indeed, as Adarain has confirmed, apparently not existent.
I wouldn’t say it’s nonexistant. I know absolutely nothing about the Indo part of Indo-European. It would not surprise me if there was at least one language in which analogy kicked in and neuter took on a distinction between nom and acc. I am just not aware of any. Iirc in anatolian neuters took on ergative alignment instead.
At kveldi skal dag lęyfa,
Konu es bręnnd es,
Mæki es ręyndr es,
Męy es gefin es,
Ís es yfir kømr,
Ǫl es drukkit es.
User avatar
KaiTheHomoSapien
sinic
sinic
Posts: 400
Joined: Mon 15 Feb 2016, 06:10
Location: Napa Valley, California

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

Post by KaiTheHomoSapien » Tue 18 Jul 2017, 17:32

I'd be curious to know the answer to that question too. Outside of IE, I mean. No IE languages distinguish nom/acc in the neuter that I know of (Adarain is right about Hittite, where neuters take on ergative alignment--not sure about the other Anatolian languages). But what about outside of IE? Is the masculine/feminine/neuter division largely limited to IE languages to begin with?
Don't live to conlang; conlang to live.

My conlang: Image Lihmelinyan
Adarain
greek
greek
Posts: 642
Joined: Fri 03 Jul 2015, 14:36
Location: Switzerland, usually

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

Post by Adarain » Tue 18 Jul 2017, 18:31

KaiTheHomoSapien wrote:I'd be curious to know the answer to that question too. Outside of IE, I mean. No IE languages distinguish nom/acc in the neuter that I know of (Adarain is right about Hittite, where neuters take on ergative alignment--not sure about the other Anatolian languages). But what about outside of IE? Is the masculine/feminine/neuter division largely limited to IE languages to begin with?
Similar divisions do exist in other corners of the world for sure. Division into masculine/feminine is very common, so is division into animate/inanimate; the classical IE system is simply what you get when you overlay those. If we take neuter=inanimate then it becomes quite easy to find more examples of course. From all I can tell, Tamil distinguishes nominative (unmarked) and accusative (marked with a suffix) equally in all nouns, and distinguishes the classes masc.sg, fem.sg, human.pl, nonhuman.sg, nonhuman.pl
At kveldi skal dag lęyfa,
Konu es bręnnd es,
Mæki es ręyndr es,
Męy es gefin es,
Ís es yfir kømr,
Ǫl es drukkit es.
User avatar
eldin raigmore
fire
fire
Posts: 5636
Joined: Sat 14 Aug 2010, 18:38
Location: SouthEast Michigan

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

Post by eldin raigmore » Tue 18 Jul 2017, 19:13

According to WALS.info (again you can look it up),
which defines "gender" as "concordial noun-class" (that is, other words in the clause must agree with the noun's gender):

Most of the world's languages don't have gender.

Most languages that do have gender, have sex-based gender.
(They define "sex-based gender" to mean "either at least one gender contains all adult female humans and no male creatures, or at least one gender contains all adult male humans and no female creatures".)

Nearly every language (indeed every language in their database) that has gender, but doesn't have sex-based gender, has animacy-based gender.


[hr][/hr]


If "Agent" and "Patient" are both inanimate, it might not make much difference which is which.
As Sancho Panza says in "Man of La Mancha": "Whether the rock hits the pitcher or the pitcher hits the rock, either way it's going to be bad for the pitcher".

Languages with hierarchical morphosyntactic alignment, and direct/inverse voice system, and an obviative vs proximative distinction in third persons, usually don't distinguish which of two inanimate third-person participants is the agent and which the patient.

Since, in many languages with sex-based gender, "Neuter" is nearly equivalent to "Inanimate", it makes sense that in the neuter declensions there need be little difference between nominative and accusative, or between ergative and absolutive (if the languages' morphosyntactic alignment and case system make such a remark meaningful and applicable to the language in question).

Is Hittite an exception?
It would be nice to have a decent sample of exceptions.
Adarain
greek
greek
Posts: 642
Joined: Fri 03 Jul 2015, 14:36
Location: Switzerland, usually

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

Post by Adarain » Tue 18 Jul 2017, 19:37

In Ergativity, Dixon makes a little summary list of all split systems known to exist which split based on animacy/topic-worthiness. There are example languages given to each system, but no example sentences sadly.

All systems follow the rule that for more animate NPs you find accusative marking, and for less animate ones ergative marking. There may be gaps in-between (no distinctions) or an overlap (tripartite). I will list them as he did, with left = more animate. For example AS≠P > ASP would mean that there’s accusative marking (P distinct) in more animate, and no marking (all equal) in less animate NPs.
  1. Part accusative: AS≠P > ASP (Latin)
  2. Part ergative: ASP > A≠SP (Burushaski)
  3. Part accusative, remainder ergative: AS≠P > A≠SP (Kuku-Yalanji, Ngiyambaa; this is also where Hittite falls)
  4. Part accusative, part ergative, overlapping: AS≠P > A≠S≠P > A≠SP (Cashinawa, Yidinʸ)
  5. All accusative, part ergative: AS≠P > A≠S≠P (no example known)
  6. All ergative, part accusative: A≠S≠P > A≠SP (Waga-Waga)
  7. Part accusative, part ergative, with a gap: AS≠P > ASP > A≠SP (no example known)
At kveldi skal dag lęyfa,
Konu es bręnnd es,
Mæki es ręyndr es,
Męy es gefin es,
Ís es yfir kømr,
Ǫl es drukkit es.
User avatar
Lambuzhao
earth
earth
Posts: 7143
Joined: Sun 13 May 2012, 01:57

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

Post by Lambuzhao » Tue 18 Jul 2017, 20:11

Completely off-topic.

But I [<3] your infra-quote from the Hávamal, Adarain, so I translated it into Sadrås:

uolle þe Tage duon þë Antnochte glutramegght
uolle þe Fvran duon þë Norgebrente,
uolle der Knobv duon þë Fumeghte
uolle þe Joungeße duon þë Geisvreghte
uolle ye Floðefïll geihinne þrœgeghte,
Uolle ye Bjroü duon yë Drägghte


Some notes
Spoiler:
Norgebrent - Lit. 'corpse-burning'. Nowadays, The word Nor is used in Sadrås more for 'meat' or 'flesh' (Cf. Norbællen 'meatball') than 'corpse'. Also, the more contemporary word Ünsrljaight, betokens more of a Doctrinist custom of laying to rest rather than kindling. but would make sense in this older Pagan context as 'pyre' or 'funeral bier'.

Knobv - 'knife', means anything from an easily-concealed dirk or stiletto, to well-nigh a machete or machæra-like weapon, and does duty for most Sadrasåler Dingmenn.

þë Geisvreghte - The normal word for wedding is Svreghtniß, but the collective noun 'vows' makes as much sense here.

Floðefïll geihinne þrœgeghte - 'Snowfield when crossed'
:mrgreen:
Post Reply