From: Siva Kalyan <[log in to unmask]>

>This reminds me of what are called "direct/inverse marking" systems (found, I think, in Native American languages as well as Sino-Tibetan), where the verb is unmarked if the subject is higher on the person hierarchy than the object (e.g. 1st pers. subject, 3rd pers. object), but takes an "inverse" marker if the object ranks higher than the subject (e.g. 3rd pers. subject, 1st pers. object).

Okay, that's neat! I could see a system where the higher animacy ranking is unmarked -- no need because its superiority is obvious. ;)) In Kalchian, I don't see how, for example, a dog's animacy could "outrank" a man's animacy.
If the dog sees / examines the man, then the roles are simply reversed. 

It is certainly possible for a IV-A being (a man, e.g.) to "slide" from his rank to say II-I. This is because there are two components to animacy: the innate is what you are by nature; the momentary is relative to circumstance. So, a man is by nature IV-A (the highest rank, which they call "wisdom"), but at night when he's snug in bed he becomes II-I (a rather lower, inanimate rank, which they call "instinct"). The same man could devolve further, for example, to I-A ("active") by joining in with an angry mob. If he's killed by the constabulary for participating in said mob, he will become level I-I ("stative") -- a corpse.

So, actual circumstance can trump innate quality, but this trumping happens within a being. It's not one being trumping another.

>This would mean that your system is typologically unusual, in that 1>3 is overtly marked, even though 2>3 and 3>3 are unmarked. Then again, I don't know what direct/inverse marking systems look like for verbs of perception.
>Another thing: I assume it's possible to put your second and third sentences into the middle voice, for the purpose of topicalizing the dog, or emphasizing the agency of "you" or "she". 

Actually, in these cases, I don't think so. By nature, the action of "SEEing" is a passive one and the action of "EXAMINing" is an active one. I don't know, but it may well be possible for other kinds of verbs to be rendered in the middle voice. It's a matter of perspective, as the phiosophers say: for the first person, the image of the dog comes into the windows of the eyes unbidden; but on the other hand, it is known for a fact that the

> Is there a corresponding valence-changing operation for the first sentence—i.e. a sort of "antipassive"? 

Can you rephrase that? I'm not quite sure what "valence-chaning operation" entails. I know I've heard of it here before! It's a matter of deleting the logical object (the dog) and shifting the logical agent (me) into that position and also change that agent's ending. If so, I think that would basically yield something like "ku-mue-Co sme-ssue": "I get myself seen". But that doesn't quite sound right... We just want to end up with "I see", right?

This would be used "on top of" the middle-voice marking, and would have the function of topicalizing the first person, and emphasizing the patiency of the dog.


>For that matter, now that I think of it, it looks like what you have is really a kind of (incipient?) split ergativity! Specifically, the split occurs along the person hierarchy (what you have glossed as "abl" in the first sentence would thus be [in this context] an ergative marker—indeed ergative markers do often arise out of ablatives/instrumentals).

Split ergativity is fun! It looks like the split here might actually be "certain verbal domains" rather than person or animacy or anything like that. This split happens with verbs like see, sense, touch, taste, hear, visualise, come to realise; maybe even certain emotional verbs like fear, avenge, nauseate, etc. Purely non-perception / non-emotional verbs (hit, strike, talk to, bully, fight, harvest, hunt, etc) conjugate very much like we'd expect from nom-acc languages: the subject / agent is in the nom, the object / patient is in the acc. The passive causes an inversion as one might expect.

>This may be the better way of analyzing your system—though I don't know how common it is for split ergativity to go by person (usually it goes by animacy).

Thanks for the suggestions!


>On 12 October 2013 at 23:37:47, Padraic Brown ([log in to unmask]) wrote:
>So, there is in the farthest reaches of Alaria beyond the seas a language with what
>>I think is an interesting sort of verbal conjugation. First, the examples, in this case,
>>the verb "see":
>>First person:
>>hmang-Tan-Co    sme-ssue-Ti    ku-mue-te
>>dog-A-ref    me-A-abl
>>The dog gets himself into view by me (masc).
>>Second person:
>>ũm-dang-so    saman-Ø-ni        un-g-hmang-Tan-sum
>>thou-A-nom    examine-Ø    to-lias-dog-A-acc
>>You (slave) examine to the dog.
>>Third person:
>>lĩ-Tan-so    ilt    saman-Ø-ni        un-g-hmang-Tan-sum
>>one-A-nom    that    examine-Ø    to-lias-dog-A-acc
>>She examines to the dog.
>>hmang = dog (any major species of wild, domestic or semi-domesticated)
>>sme- = "it comes into view"
>>saman- = "thoroughly examine"; "pick apart"
>>A = animacy ranking (there are seven, all together, both inherent and momentary)
>>MID = middle voice
>>Ø = active voice is unmarked default
>>abl = ablative case
>>nom = nominative case
>>acc = accusative case
>>ref = referential case
>>lias = an intrusive consonant
>>This example is in the unmarked/default "ongoing present". There is also
>>a (marked) "momentary present".
>>As you can see, and this is also the case for all other verbs of perception,
>>the first person forms are conjugated in the middle with respect to the
>>referent and with a "passive" root verb; while the second and third
>>persons are conjugated in the active with respect to the subject and with
>>an entirely different, "active" root verb.
>>It got me wondering if there is any ANADEW for this kind of conjugation:
>>one conceptual map for one person and then using an entirely different 
>>conceptual map for the other person(s), and entirely different verbs for
>>both. The distinction here being one of detached, almost impersonal experience 
>>(for the first person) vs. a more hands on, almost violent engagement with the 
>>object in question (for the second and third persons).