From grey.havens@earthling.net Mon Jul 10 11:21:43 2000 Return-Path: Received: (qmail 20426 invoked from network); 10 Jul 2000 18:18:07 -0000 Received: from unknown (10.1.10.26) by m2.onelist.org with QMQP; 10 Jul 2000 18:18:07 -0000 Received: from unknown (HELO postfix2.free.fr) (212.27.32.74) by mta1 with SMTP; 10 Jul 2000 18:18:06 -0000 Received: from tam.n (marseille2-1-60-180.dial.proxad.net [212.27.60.180]) by postfix2.free.fr (Postfix) with ESMTP id DEB95740B6 for ; Mon, 10 Jul 2000 20:18:01 +0200 (MEST) Date: Mon, 10 Jul 2000 20:17:38 +0200 (CEST) X-Sender: elrond@tam.n To: Lojban List Subject: Re: [lojban] Re: zi'o & otpi In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII From: Elrond > > > le se gerku be zi'o, for a dog breed > > that exists independently of actual dogs. > > 1) "gerku" relates not only actual dogs with actual dog breeds, but > allows the relation of potential dogs and/or potential dog breeds as well. > Does it even make sense to talk about a dog breed which neither is, > nor could possibly be, instantiated in any conceivable dogs? What on earth > would make it a *dog* breed, then? Well, let's imagine a fiction story in the future, where a crazy inventor is wanting to create a completely new, yet unexisting, dog breed, and this not by instanciating a sample dog from it, but by specifying on paper all the characteristics of the dog breed, more particularly by specifying the characteristics of the breed not as the characteristics of any dog of that breed, but of the breed as a "breed" (= species, somewhat). Such a specification is feasible: "I am going to create a dog breed that will be described in other books for at least one billion years". This might seem flawed, because one could argue that to create a dog breed, it is needed to instanciate a dog of that breed. However, I do not agree. When I think about a breed of any animal type, I do an abstraction (else I wouldn't understand why to use the same word, "breed", for several different animals). For me, what makes a breed a dog breed is that it is related to dogs, and is a concept that allows regrouping several different objects (here, dogs of different breeds) into one family sharing common properties when describing the whole set of them. It is then perfectly legal, IMO, to speak about properties of the breed that can be described in other books, for example. (Anyone stating that no property of a breed is not also linked with the individuals has to fight the philosophical truth that the concept of "breed" has at least the property of its very existence as a concept...) Of course, when thinking about the concept of "breed" as explained above, a separate generic predicate should logically be needed (for example, se danlu, jutsi ?), although any predicate taking a breed as first argument has most probably a set of members as another argument (and there we have the same problem all over again). I postulate that the "dog breed" is a rather simple concept that *should* be allowed not to need actual dogs to be thought about. I might be wrong, and thinking actually about "gerku girzu" instead of "se gerku". Well, a special form of "girzu", though, one that has not its (still problematic) x3 place... (Honestly, I feel that lojban here is constraining the way I'm thinking -- it seemed obvious to me to be able to think about a "girzu be fi zi'o", even a "se gerku be zi'o", and all I'm hearing for several days now just states that it "cannot"... I just am more and more agreeing with the SWH!) But let's go slightly offtopic. Actually, I am beginning to think that the whole issue is not about the ability of describing such or such concept, but about the way (western) people think and use description words to designate/name *objects*, rather than *relationships* (as does, apparently, lojban). By the way, I just noticed that this is not the first time that lojban allows constructs that are difficult to understand at first (perhaps a side-effect of being logical and providing symmetry of constructs around relationships ?) To my mind, if "le girzu be zi'o" is declared to be sensible, then symmetrically "le se girsu be zi'o" should be, no matter what we think about it. Doesn't the Book state everywhere that the order of appearance of arguments in a bridi is only a matter of emphasis? The very act of allowing "zi'o" for x2-x5/bai/fi'o places MUST allow its use for x1, too. If using "zi'o" for x1 in a construct leads to obviously paradoxical meanings, or simply to concepts that are strictly logically nonsense, then it should mean that the very place structure itself is flawed, and not that the language user "should not" build such constructs. If I were to be completely wrong and forgetting a rule of Lojban that states that the "x1" place has a special extra weight/meaning, then I would come to think that the vast potential useful features of taking the bridi as instantations of relationships (selbri) between *symmetrically* AND semantically *weight-equal* arguments are simply wasted. Cheers raph