Theiling Online    Sitemap    Conlang Mailing List HQ   

Re: GROUPLANG: Pronouns

From:Mathias M. Lassailly <lassailly@...>
Date:Monday, October 19, 1998, 18:18
Charles wrote :

On Mon, 19 Oct 1998, Nik Taylor wrote:
> > > Well, I was going to suggest a triple system like that, but I don't > > really like it very much. I see no reason to have a seperate > > intransitive case, since it never needs to disambiguated from transitive > > cases.
There is a deeper issue than (in)transitivity here. Euro-linguists like to define an ergative language as a language where 'subject of intransitive verb has the same tag as object of transitive verb'. Can you feel the nonsense that is ? From the ergative viewpoint there is no 'intransitive subject' nor 'transitive object' whatever the change occurs when adding an ergative (like position of absolutive verbal affixed pronoun in Sumerian). So of course there is no such need you mention if you leave away MEANING in a sentence : then you can regard noun-rooted predicates as equivalent to verb-rooted predicates. But not so if you really want to allow either populations of us to use either root and still get perfectly understoiod from the other population. I mean : Carlos and I and many others will almost always find a way to derive a noun into a predicate to express things in images like English does. This implies that we find in nouns a specific meaning making them part of a process, as English does with 'hand', 'ship', 'man', 'branch', etc. To do so we need two specific 'cases' (i.e. verbal voices in a nom/acc language), namely : 1. agentive - 'to work as, to have the role of' a hammer, a mill, a chief, a home, etc. - and 2. attributive ('to have an absolute feature' = to have a head, a home, a country, a brother, strength, (to produce) smoke, etc.'. Luckily, the 4 other cases we need can be also used with verb-rooted predicates - namely : 1. ergative : 'to apply' a colour, a hammer, etc.; 2. absolutive : 'to benefit or experience' = to benefit of gift, clothes; to perceive a smell, an image; to experience a position; to use one's organs eye, lungs, etc; 3. Patientive : 'to suffer an agent or its result' = to suffer a wound, an injury, a hammer, etc.; 4. Causative : 'to make' = to make it a flower, to have someone go, etc. Computers will may never be able to analyse a noun as the potential agent of a process unless you tag on that noun both aspective (verb) and unaspective (noun) definition thereof. This is because computers will never experience in their life that a cow is at the same time an agent to bread, to eat, to watch trains, etc. But we human do. So please grant Carlos and me and other human conlangers the delight of deriving predicates from nouns AND being understood by our fellow human conlangers WITHOUT the confusion English raises with verbs like 'to man a ship'. Please allow us the pleasure of conjuring up our human experience from nouns of instrument (loom), result (fruit) and relative (brother) using agentive and attributive. :-) Actually, we all-noun promoters could claim for more than 10 cases we need to fully express ourselves in this language ;-) But we don't. We know you don't need more than 2 cases (nom/acc) + a few verbal affixes (factitive, passive, etc). Yes we need more cases. But we don't need factitive and passive either. So you'll see this system doesn't require more tags than an all-verb rooting predicate's one. The same number of tags is required. The difference is that noun-rooting predicate needs more cases, while verb-rooting predicate needs more verbal forms. So 6 cases is OK, don't you think so ? Mathias Anyways, the case system as proposed seems too "computer-like",
> > so to speak. > > There is a natlang distinction between and "active" and "passive" > intransitive, though it is somewhat rare. When the moon rises, > is it raised or does it rise? (I hope that is a good example, > but I can't remember a cited one.) > > I think computers will have enough trouble with this lang, BTW. > >
----- See the original message at -- Free e-mail group hosting at