Morphosyntactic alignment

The best topics from Languages & Linguistics, kept on a permanent basis.
User avatar
Lleu
Lebom
Lebom
Posts: 96
Joined: Sun Oct 16, 2005 10:38 am
Location: Tkaronto

Post by Lleu »

Dhana groups SA as nominative in active sentences and SP as nominative in passive sentences. PT also group together in active sentences, while R is split into two cases. So, I guess that makes Dhana active-dative.

I once had a little sketch for a conlang that had a "second passive" voice, distinguishing instrument/recipient as subject from theme as subject:

The wall was sprayed with paint by John: passive
Paint was sprayed onto the wall by John: second passive
A book was given to her by John: passive
She was given a book by John: second passive
agus tha mo chluasan eòlach air a’ mhac-talla fhathast / às dèidh dhomh dùsgadh
(mona nicleòid wagner, “fo shneachd”)

chris_notts
Avisaru
Avisaru
Posts: 275
Joined: Wed Dec 15, 2004 9:05 am
Location: Nottingham, England
Contact:

Post by chris_notts »

ngwaalq is ergative, if you only look at the case particles. However, it has a topic overlay system (ie, the topic particle replaces the case particle generally), and the topic marker tends to associate most often with the subject (A/S).
Try the online version of the HaSC sound change applier: http://chrisdb.dyndns-at-home.com/HaSC

User avatar
Nuntar
Lebom
Lebom
Posts: 128
Joined: Sat Oct 25, 2003 7:07 am
Location: [16.50.72.0]
Contact:

Post by Nuntar »

Arêndron is accusative, and I'm considering making the new version dechticaetiative.

Xephyr would tell you that dechticaetiative is properly pronounced with a /s/, but I say /dEk.tI."kaI.S@.tIv/.
[quote="Amaya"]Every now and then, the world tries to say something. I'm never quite sure whether or not to listen to it at times like that.[/quote]

Sander
Niš
Niš
Posts: 14
Joined: Thu Aug 14, 2003 8:41 am
Location: Belgium
Contact:

Post by Sander »

I've always used [dEk.tI."ki.S@.tIv] and <dechticaetiatief> [dEx.ti.kaj.sja.tif] in Dutch. but I suppose aI is more logical.
Sander/benanne
[b][url=http://benanne.net/misc/com.pdf]λemvrin eiδerial[/url][/b] (pdf)
[b][url=http://www.omniglot.com/writing/lembrin.htm]λemvrin eicrýs[/url][/b] (omniglot)

User avatar
con quesa
Lebom
Lebom
Posts: 159
Joined: Sun Apr 27, 2003 1:34 pm
Location: Fnuhpolis- The City of Fnuh

Post by con quesa »

Saimiar is Ergative-Dative, with the interesting distinction of also having Ergative syntax in many instances, including arguement omission. Juteldi is boring ol' Accusative-Dative. Dazgõ Xin, however, marks all six types of core arguements with different cases.
con quesa- firm believer in the right of Spanish cheese to be female if she so chooses

"There's nothing inherently different between knowing who Venusaur is and knowing who Lady Macbeth is" -Xephyr

User avatar
vohpenonomae
N'guny
N'guny
Posts: 91
Joined: Sat Nov 02, 2002 4:23 am

Post by vohpenonomae »

The languages of the PCM family, through Noyahtowa, work on an active system; I'm not sure how you'd label the system, but it goes like:

INSTRANSITIVE

Participant is S or O, depending on whether the active conditions are fulfilled.


TRANSITIVE

Participants are marked A and P if active conditions are fulfilled; they are S and O if not.


DITRANSITIVE

As in transitives, except R is incorporated and T marked as P (if active conditions are fulfilled; it is marked as O otherwise).
"On that island lies the flesh and bone of the Great Charging Bear, for as long as the grass grows and water runs," he said. "Where his spirit dwells, no one can say."

User avatar
Nuntar
Lebom
Lebom
Posts: 128
Joined: Sat Oct 25, 2003 7:07 am
Location: [16.50.72.0]
Contact:

Post by Nuntar »

Sander wrote:I've always used [dEk.tI."ki.S@.tIv] but I suppose aI is more logical.
Actually, in anglicised Latin and Greek <ae> is pronounced [i:], as in formulae, Caesar, Aesop etc. [aI] is the pedantic pronunciation; I guess the reason I use it is that it's one of those words I've never heard someone else say, so I don't have to conform to how other people say it.
[quote="Amaya"]Every now and then, the world tries to say something. I'm never quite sure whether or not to listen to it at times like that.[/quote]

User avatar
Radagast
Lebom
Lebom
Posts: 97
Joined: Sat May 21, 2005 9:46 am
Location: Denmark
Contact:

Post by Radagast »

On verbs Mèlw normally marks patients of transitive verbs and intransitive verbs alike. (but it has an inverse form to switch to an accusative system)Agents are marked differently. So its basically ergative.

But in dependent clauses and in main clauses this marking and the topic system are likely to cause confusion it case marks both the patient and agent NP's.
[i]D'abord on ne parla qu'en poésie ; on ne s'avisa de raisonner que long-temps après.[/i] J. J. Rousseau, Sur l'origine des langues. 1783

User avatar
BettyCross
Sanci
Sanci
Posts: 18
Joined: Mon Jun 26, 2006 8:44 pm
Location: NC Research Triangle

Post by BettyCross »

Malganic has two noun classes, Animate (based on IE Mas-Fem case endings) and Inanimate (based on IE neuter case endings).

For all subjects, Nominative.
For all direct objects, Accusative.
For all indirect objects (recipients of ditransitive verbs), Dative.

However, because early IE languages didn't have distinct endings for the nominative and accusative in the Neuter, how can we distinguish the Inanimate subject from direct object with a transitive verb?

I came up with two different solutions:

In the A dialect, word order is SOV. This is fixed for inanimate subjects, though for animate subjects more variation is allowed.

In the B dialect, there is an adposition, a separate particle placed after the inanimate Agent to distinguish it from the direct object, allowing the same flexibility of word order for all subjects. SOV is still, however, the normal word order.

B has, if I'm not mistaken, a Split-S system based on "gender" (or noun class, depending on how you want to define it).

After two Ergative languages in a row, I thought I'd give ergativity a rest.

Betty Cross
May the odds be ever in your favor.
Oi sî đât sort điri
ever be-SUBJ the odds 2S-DAT

Ezekiel
Niš
Niš
Posts: 5
Joined: Tue Jul 26, 2005 3:05 am
Location: Earth, occasionally
Contact:

Post by Ezekiel »

Holy crap.

This thread has opened my eyes to what morphosyntactic alignment means. All the wiki articles and peoples' posts here just confused the heck out of me, but these diagrams and explanations finally did it! THANK YOU SANDER!

(And, by the way, dechticaetiative is an incredibly horrible word to pronounce. It breaks my brain and my mouth every time I try.)

While I haven't really determined exactly where Xon will stand on this issue (partially because it's just a script at the moment, without even a phoneme table!) I'm very partial to the accusative-dechticaetiative system. It just seems really...clean, I guess. Very straightforward. I'm still unsure about whether I want to mark cases or not (the monstrous case systems I've seen are very frankly disturbing.) However, this has definitely helped me on my conlanging journey, so I thank you again.
And pressing of the Undefined/The definition on my mind
Held up before my eyes a glass/Through which my shrinking sight did pass
Until it seemed I must behold/Immensity made manifold
Whispered to me a word whose sound/Deafened the air for worlds around

User avatar
Rik
Lebom
Lebom
Posts: 134
Joined: Fri Jul 04, 2003 2:57 pm
Location: London, UK
Contact:

Post by Rik »

I hereby declare Ákat to be (yet another) active-dative language!

User avatar
Mwanatabu
Niš
Niš
Posts: 2
Joined: Thu Sep 14, 2006 2:46 am
Location: The Netherlands

Post by Mwanatabu »

Yen groups S and A together, and does the same with P, R and T.
Since syntactical roles are unmarked (no case system), it indeed relies completely on context to figure out which is which, although animate objects will preferably appear earlier in the verbal clause than inanimate ones.

User avatar
Ketumak
Lebom
Lebom
Posts: 231
Joined: Sun Feb 09, 2003 3:42 pm
Location: The Lost Land of Suburbia (a.k.a. Harrogate, UK)
Contact:

Post by Ketumak »

My main conlang is plain old accusative-dative. It has been tripartite-dative and will be so again. It's called Lhemburan [yeah ... I know, I know!] The changes I'm working on are great enough to warrant a name change though. So I'm thinking of going for Lhembeku to avoid confusion with Sander's Lembrinn.

Sander
Niš
Niš
Posts: 14
Joined: Thu Aug 14, 2003 8:41 am
Location: Belgium
Contact:

Post by Sander »

Mwanatabu wrote:Yen groups S and A together, and does the same with P, R and T.
Since syntactical roles are unmarked (no case system), it indeed relies completely on context to figure out which is which, although animate objects will preferably appear earlier in the verbal clause than inanimate ones.
Actually, word order can mark case just as much as morphology (affixes, mutation etc)... seeing as you say "...will preferably appear earlier..." I suppose that could count as something resembling a direct/inverse system with an (albeit limited) animacy hierarchy :)
My main conlang is plain old accusative-dative. It has been tripartite-dative and will be so again. It's called Lhemburan [yeah ... I know, I know!] The changes I'm working on are great enough to warrant a name change though. So I'm thinking of going for Lhembeku to avoid confusion with Sander's Lembrin.
Don't worry about that :) I used to have a nooblang called Areyi (cfr. guitarplayer's Ayeri). Besides, you might have been first with that name anyways. I used to call it therial or "Asofyetunnamedese" :D
Sander/benanne
[b][url=http://benanne.net/misc/com.pdf]λemvrin eiδerial[/url][/b] (pdf)
[b][url=http://www.omniglot.com/writing/lembrin.htm]λemvrin eicrýs[/url][/b] (omniglot)

User avatar
Klaivas
Sanci
Sanci
Posts: 56
Joined: Mon Jan 30, 2006 2:00 pm

Post by Klaivas »

A lot of information to take in here, but I managed to understand it :mrgreen: Can I nominate this for the L&L museum, perhaps? I know it's in C&C but I think it deserves a place there. Nevertheless, I've decided to make Ewari an ergative-dechticaetiative language... thinking this through carefully, it makes more sense in my mind to group the cases as follows:
Image
A: subject (ergative case)
T: topic (dechticaetiative case)
S P and R: objects (absolutive (?) case)

This will mean I'll need to do some editing, but I wasn't happy with the nom-acc system I had before anyway. Thanks loads Sander!

User avatar
Miekko
Avisaru
Avisaru
Posts: 364
Joined: Fri Jun 13, 2003 9:43 am
Location: the turing machine doesn't stop here any more
Contact:

Post by Miekko »

Sander wrote:
Mwanatabu wrote:Yen groups S and A together, and does the same with P, R and T.
Since syntactical roles are unmarked (no case system), it indeed relies completely on context to figure out which is which, although animate objects will preferably appear earlier in the verbal clause than inanimate ones.
Actually, word order can mark case just as much as morphology (affixes, mutation etc)... seeing as you say "...will preferably appear earlier..." I suppose that could count as something resembling a direct/inverse system with an (albeit limited) animacy hierarchy :)
I think you misunderstood what was being said. If S and A are merged, distinguishing them by word order unmerges them! It's like trying to conduct the surgical attachment of a bodypart by bringing along the tools for amputation.
< Cev> My people we use cars. I come from a very proud car culture-- every part of the car is used, nothing goes to waste. When my people first saw the car, generations ago, we called it šuŋka wakaŋ-- meaning "automated mobile".

User avatar
Mwanatabu
Niš
Niš
Posts: 2
Joined: Thu Sep 14, 2006 2:46 am
Location: The Netherlands

Post by Mwanatabu »

Sander wrote:
Mwanatabu wrote:Yen groups S and A together, and does the same with P, R and T.
Since syntactical roles are unmarked (no case system), it indeed relies completely on context to figure out which is which, although animate objects will preferably appear earlier in the verbal clause than inanimate ones.
Actually, word order can mark case just as much as morphology (affixes, mutation etc)...
That was what I was trying to say. There is just no overt case-marking in the morphological sense :)
seeing as you say "...will preferably appear earlier..." I suppose that could count as something resembling a direct/inverse system with an (albeit limited) animacy hierarchy :)
I liked that kewl word too, "animacy hierarchy" :P , but no: it mainly has to do with the preference to bind animate personal pronouns directly to the verb (object clitic) and has little to do with the P-, R- or T-role within the verbal clause.

User avatar
vohpenonomae
N'guny
N'guny
Posts: 91
Joined: Sat Nov 02, 2002 4:23 am

Post by vohpenonomae »

A clarification, if you please: when you guys speak of an "animacy hierarchy," are you referring to something like that found in Algonquian languages, where a person hierarchy determines which participant is marked with a prefix and which with a suffix? Cheyenne's hierarchy goes: 2-1-3-4-I, where 3=3rd animate, 4=obviative and I=inanimate person.
"On that island lies the flesh and bone of the Great Charging Bear, for as long as the grass grows and water runs," he said. "Where his spirit dwells, no one can say."

User avatar
Legion
Avisaru
Avisaru
Posts: 522
Joined: Sat Mar 05, 2005 9:56 pm

Post by Legion »

vohpenonomae wrote:A clarification, if you please: when you guys speak of an "animacy hierarchy," are you referring to something like that found in Algonquian languages, where a person hierarchy determines which participant is marked with a prefix and which with a suffix? Cheyenne's hierarchy goes: 2-1-3-4-I, where 3=3rd animate, 4=obviative and I=inanimate person.
Yes, I do, at least.
My language has a an animacy hierarchy working this way :

-Singular persons are higher in the hierarchy than plural ones.
-First person is higher than second person, which is higher than third persons.
-Proximate third persons are higher than obviative ones.
-Common third person is higher than Neuter one, which is higher than Group one.

Common, Neuter and Group are genders, so this gives (provided that 4=obiative) :

1SG - 2SG - 3CSG - 3NSG - 3GSG - 4CSG - 4NSG - 4GSG - 1PL - 2PL - 3CPL - 3NPL - 3GPL - 4CPL - 4NPL

User avatar
Ketumak
Lebom
Lebom
Posts: 231
Joined: Sun Feb 09, 2003 3:42 pm
Location: The Lost Land of Suburbia (a.k.a. Harrogate, UK)
Contact:

Post by Ketumak »

Sander wrote:Don't worry about that I used to have a nooblang called Areyi (cfr. guitarplayer's Ayeri). Besides, you might have been first with that name anyways. I used to call it therial or "Asofyetunnamedese"
Thanks :) I may still change it anyway though, to show that I'm mixing Lhemburan ideas with ones from my older project Lesdekan: Lhembekan or somesuch. I'll think on it. Anyone familiar with our phonologies though will know that the names Lhemburan and Lembrin are pronounced differently - my "Lh" is voiceless for e.g. and your "b" a bilabial fricative.

Coming back to the topic, I think I've identified another role, not mentioned so far - the agent-patient - that's one actor, two hats. The A-P is simultaneously the doer and done-unto of a transitive verb - so reflexive, reciprocal or middle voice generally.

The window broke
La fenetre s'est rompu (French : middle voice marked because the breaking is done by and to the window)

Sometimes these constructions are hard to spot in English, translation into a Romance language reveals them. Also:

English spoken here
Ingles se habla aqui (Spanish : "English speaks itself/gets spoken here" - medeo-passive?)

User avatar
Legion
Avisaru
Avisaru
Posts: 522
Joined: Sat Mar 05, 2005 9:56 pm

Post by Legion »

Sonib wrote:
The window broke
La fenetre s'est rompu (French : middle voice marked because the breaking is done by and to the window)
AUGH ! You can't use "rompre" with a window ! "rompre" is the idea of bent, distort something until it breaks. You can "rompre" bread or a branch, but not a window ! You must use "casser" here.

Sander
Niš
Niš
Posts: 14
Joined: Thu Aug 14, 2003 8:41 am
Location: Belgium
Contact:

Post by Sander »

Sonib wrote:
Sander wrote:Don't worry about that I used to have a nooblang called Areyi (cfr. guitarplayer's Ayeri). Besides, you might have been first with that name anyways. I used to call it therial or "Asofyetunnamedese"
Thanks :) I may still change it anyway though, to show that I'm mixing Lhemburan ideas with ones from my older project Lesdekan: Lhembekan or somesuch. I'll think on it. Anyone familiar with our phonologies though will know that the names Lhemburan and Lembrin are pronounced differently - my "Lh" is voiceless for e.g. and your "b" a bilabial fricative.

Coming back to the topic, I think I've identified another role, not mentioned so far - the agent-patient - that's one actor, two hats. The A-P is simultaneously the doer and done-unto of a transitive verb - so reflexive, reciprocal or middle voice generally.

The window broke
La fenetre s'est rompu (French : middle voice marked because the breaking is done by and to the window)

Sometimes these constructions are hard to spot in English, translation into a Romance language reveals them. Also:

English spoken here
Ingles se habla aqui (Spanish : "English speaks itself/gets spoken here" - medeo-passive?)
In this case it looks like fenêtre is the agent whereas se is the patient. Sure, they refer to the same thing... but I don't think that justifies bundling them and assigning them a new semantic role.

By the way, Lembrin is the "anglified" spelling. The language is actually called λemvrin (eiδerial) ["LEm.b4in e."TE4.jal].
Sander/benanne
[b][url=http://benanne.net/misc/com.pdf]λemvrin eiδerial[/url][/b] (pdf)
[b][url=http://www.omniglot.com/writing/lembrin.htm]λemvrin eicrýs[/url][/b] (omniglot)

User avatar
Ketumak
Lebom
Lebom
Posts: 231
Joined: Sun Feb 09, 2003 3:42 pm
Location: The Lost Land of Suburbia (a.k.a. Harrogate, UK)
Contact:

Post by Ketumak »

You can certainly view it that way and it makes for a neater theory if you don't include compound features. Thinking some more, by my analysis, you'd have to say that most languages are A=P for part of the time and not the rest of the time, which is odd and not neat.

Similar comments could apply to A=R as in: "He bought the book for himself"

I was drawing there on the essay by Rick Moreau called "The lexical semantics of a machine translation interlingua"

Its here:

http://www.eskimo.com/~ram/lexical_sema ... tml#S2_7_2


He covers middle voice in 2.7.2

User avatar
Ketumak
Lebom
Lebom
Posts: 231
Joined: Sun Feb 09, 2003 3:42 pm
Location: The Lost Land of Suburbia (a.k.a. Harrogate, UK)
Contact:

Post by Ketumak »

Oh yeah ...
legion wrote: AUGH ! You can't use "rompre" with a window ! "rompre" is the idea of bent, distort something until it breaks. You can "rompre" bread or a branch, but not a window ! You must use "casser" here.
Of course! It's been a long time since I studied French properly as opposed using trial and error when talking to our French town-twinning friends!

User avatar
Karig
Niš
Niš
Posts: 1
Joined: Sun Jan 16, 2005 11:41 pm
Location: Northern Virginia, USA
Contact:

Post by Karig »

Ummm, if I understand this properly, I'm tentatively classifying my language (Azarennya) as ergative-dative.

Nouns are "inflected" by preceding them with a particular article. (I stole this idea from Hawaiian.) This is equivalent to saying "that dog" for singular and "those dog" (no -s on "dog") for plural.

The experiencer of an event (S) and the patient of an action (P) are both marked with the same article. The agent of an action or causer of an event (A) is marked with a different article. So I guess that makes Azarennya ergative.

Meanwhile, if you're doing something to an object, that object is still the "receiver" of the action regardless of whether there is a secondary object or beneficiary. You give the book; you give the book to John; the book is being given in either case. "Book" would get the "object" article in either case. So apparently that makes Azarennya dative.
Now with [url=http://www.karig.net/]language blog[/url].

Post Reply