Theiling Online    Sitemap    Conlang Mailing List HQ   

Re: A concept.

From:Christophe Grandsire <christophe.grandsire@...>
Date:Wednesday, August 7, 2002, 9:47
En réponse à Nihil Sum <nihilsum@...>:

> > Could there be a language where nouns take endings to show the person > and number of the OTHER participant in an action? For example, where > a noun, which is the OBJECT of a verb, is marked to show what > person / number is the SUBJECT of that verb, and vice versa. I've got > some ideas for a language now... > It'll be based on an ergative-absolutive system, with word order > Agent-Verb-Patient (I might change my mind and make it P-V-A just to > be > difficult). Verbs will not inflect for persons and numbers, instead, > the > agent will be marked on the patient, and the patient will be marked > on the agent. > Does this sound feasible? Plausible? >
Well, that's an idea which, even if proved implausible, deserves to be part of at least one conlang!!! :))) As for its plausibility, well, there are languages which mark tense on the subject rather than the verb, so I don't think marking the person instead would be such a big step. As for the feasibility, well, anything is feasible, Maggel and Ebidesian prove it quite well ;))) .
> > sabak-um kuat utuk-an > The dog bites the man > > (dog-[with 3sg patient] bites man-[with 3sg agent]) > So sabak-um here means "a dog, as acting upon a third person singular > patient" > and utuk-an means "a man, as acted upon by a third person singular > agent" >
Now, although some redundancy is always nice, I think this is a bit much. If you have both an expressed agent and an expressed patient, I think only one would need person marking. After all, the mark expresses both the agency or patiency of the marked noun and the person and number of the other noun, which is normally enough for cross-referencing. In other words, "sabak-um kuat utuk" & "sabak kuat utuk-an" are as unambiguous as the sentence you proposed. In such a system, I'd expect one marking to be more common than the other when both agent and patient are expressed. For instance, since your language is ergative, I'd expect that the most common (or even only) way to express the previous sentence would be "sabak-um kuat utuk", where the agent is overly marked, while the patient is only obliquely marked by the mark on the agent, and by itself stays unmarked (which fits quite well in an ergative paradigm). An accusative language would on the other hand favour the expression "sabak kuat utuk-an". Of course, you could allow other forms (including the overly redundant one you proposed) for topicalisation or definiteness for instance. Non-redundant forms look to me quite plausible. I don't see anything strange or implausible in marking the object for the person and number of the subject (and vice-versa) rather than marking the verb this way. I expect it to be quite rare, but not impossible.
> > Weird eh? When the AGENT takes a plural ending, the PATIENT also > takes an ending to indicate that what is acting on it is plural! And > vice versa! >
Niceness. Of course, it's plainly logical in such a scheme, and I see nothing strange about it. Agreement patterns take stranger paths :)) . With such a system (using the ergative-like proposal I gave), you could even stop bothering marking the patient even in number, since it's already marked in the agreement suffix on the agent! (I've seen languages which mark definiteness or number of the object exclusively on the verb, not on the object itself, so it's not entirely unnatural. It's weird yes, but not unnatural).
> sabak-anga > a dog, with a 1pl agent > > mim sabak-anga > we see the dog > > sabak-ana > a dog, with a 3pl agent > > mim sabak-ana > they see the dog > > sabak-ose > a dog, with a 1pl patient > > sabak-ose mim > the dog sees us > > sabak-os mim > the dog sees me > > mim utukag-eka > you (pl) see the men > > utukag-ishe mim > the men see you (pl) >
These examples are the ones that make me think such a system is extremely neat and plausible. It doesn't look to me more difficult than putting the personal marks on the verb and/or using separate pronouns.
> Let me know what y'all think. I want this to be CHALLENGED: what > possible > snags that could put an end to this concept? > Or do you need me to clarify some of this nonsense? >
Well, I don't think there is much to challenge on this system, except for the too big redundancy of the first examples you proposed. But as I said, it's easily corrected, and it makes the cross-referencing more convincing as it's more useful. Now, how would you handle intransitive and ditransitive sentences with this scheme? For intransitive sentences, I suppose you wouldn't use the affixes at all. And for intransitive sentences using a first or second person, you could use a set of independent pronouns for that. Then intransitive sentences would fit quite well in the system, especially if you use an ergative system. Patient and subject of intransitive sentences would have the same marking (nothing prevents you to make the system more active-like by allowing marking on the patient as well for topicalisation reasons or others as I explained). As for ditransitive sentences, I don't think we could extend such a scheme like that. With two arguments it stays manageable, but with three it gets a bit out of proportions, even if you use non-redundant marking (since each argument, agent, patient and oblique, should have its own set of markers to refer to the two other arguments). It's not impossible, but quite improbable. Of course, you could simplify the matter by not having real ditransitive sentences. Instead, the third participant would be marked like any other argument except agent and patient, that is adpositions, cases, or whatever you want to use for them. I can't find any case where such a system would not work. I've thought of different kinds of subclauses, argument raising and other things like that, and I cannot find a case where no natural solution (compatible with the rest) could be used. So basically I'll say: go ahead with your scheme!!! It's extremely original (although, like anything in conlanging, I'm sure someone is gonna come with an example of con- or natlang featuring it already ;))) ) and nothing implausible. Rare maybe, but not implausible. Christophe. Take your life as a movie: do not let anybody else play the leading role.


H. S. Teoh <hsteoh@...>