From LOJBAN%CUVMB.bitnet@YaleVM.YCC.YALE.EDU Sat Mar 6 22:59:50 2010 Received: from YALEVM.YCC.YALE.EDU by MINERVA.CIS.YALE.EDU via SMTP; Wed, 27 Jan 1993 13:06:54 -0500 Received: from CUVMB.CC.COLUMBIA.EDU by YaleVM.YCC.Yale.Edu (IBM VM SMTP V2R2) with BSMTP id 3185; Wed, 27 Jan 93 13:05:31 EST Received: from CUVMB.BITNET by CUVMB.CC.COLUMBIA.EDU (Mailer R2.07) with BSMTP id 7472; Wed, 27 Jan 93 13:05:24 EST Date: Wed, 27 Jan 1993 17:36:11 GMT Reply-To: C.J.Fine@BRADFORD.AC.UK Sender: Lojban list From: C.J.Fine@BRADFORD.AC.UK Subject: TECH.ADV - non-specific SE To: Erik Rauch Status: RO X-Status: X-From-Space-Date: Wed Jan 27 17:36:11 1993 X-From-Space-Address: @YaleVM.YCC.YALE.EDU:LOJBAN@CUVMB.BITNET Message-ID: I have noticed a small problem, and have come up with a solution which is much more general than the problem, but I think might be an interesting feature to have nonetheless. The problem is how to refer to more than one (or a non-specific) tergismu of a selbri. This is not something we often want to do; but there are certain brivla where two of the tergismu do form a natural collection and we might want to. For example: kanxe is a conjunction, stating: x2 (du'u) and x3 (du'u) are both true tanru phrase compound is a binary metaphor formed with words/concepts x2 and x3, giving meaning x4 in language x5 mintu is the same/identical thing as x2 by standard x3 gletu copulates/fucks/has coitus/sexual intercourse with x2 How, without being long-winded, can we refer in lojban to 'the components of a tanru'? The straightforward way is "le se tanru be ko'a jo'u le te tanru be ko'a" but this is both long-winded, and loses a lot of commonality. It can be shortened to "le ke se tanru jo'u te tanru ke'e be ko'a" but you need the "ke ... ke'e" to get the grouping right, so it isn't much shorter, and in any case the semantics of internally connected tanru are not quite clear. I might mean "ja" rather than "jo'u". Of course there are work-rounds like "le tanrypau be ko'a" but these are annoyingly obtuse. My suggestion is a new SE word (I will use xe'e) that leaves the assignment of sumti to tergismu specifically undefined. Thus in this use le xe'e tanru means the thing(s) taking part in a tanru relationship. Of course this can include the tanru, the meaning, and the language as well: I am relying on the intentionality of "le" (and the 'signalled ambiguity' of "xe'e") to save this. Similarly we can have le xe'e kanxe for the arms of the conjunction However, we can go further, (and make it more useful, I believe). I suggest that when using "xe'e" all untagged sumti are interpreted freely, but any explicitly FA-tagged sumti not only apply to the un-converted selbri, but also take that particular tergismu out of consideration. So le xe'e tanru befa lu zdani blanu li'u means (in full) (some of) the things which are in a tanru relationship with "xdani blanu" as the tanru - ie the modifier, the modificand, the meaning, the language. and again, we can use "le" and context to conclude that the two parts of the tanru are what is wanted. Then again le xe'e kitsa means the lovers but in a specific sense that if we are talking about more than one person they are in that relationship to each other, as opposed to le kitsa which has no such implication. The further extension of this idea is to use it predicatively: mi la mrvin. xe'e patfu I and Mervyn are in a father-relationship without specifying which way round. Again this leaves open the possibility that one of us is the te patfu (the mother). This could be handled by mi la mrvyn. xe'e patfu fizo'e where the "fizo'e" plays the function of removing the x3 from the tergismu soup. Note that this proposal does not require any changes to the syntax. It just needs a word adding to an existing selma'o, and it also needs a new interpretive rule; but we do not have these expressed in any formal way anyway. lujvo and tanru =============== tanrypau tanru pagbu tergismu te gismu but I propose to define it "x1 is the x3'th argument-role of (selbri) x2" i.e. the x2 of tergismu is not restricted to a gismu, but is any selbri. The alternative would be to define the place structure of selbri, and then pick out the appropriate place (role) as "selselbri" or "terselbri", which would a) be confusing, and b) depend on the argument structure of "selbri". This is incidentally an example of the distinction between "se broda" and "selbroda" that was discussed recently: "te gismu" unequivocally means "is an argument role of gismu ...", but we can define "tergismu" to apply to any selbri. "selma'o" is widened in a similar way. Note also the distinction between "tergismu" and "te bridi". "Te bridi" are actual arguments in a predication. "tergismu" are argument roles in a selbri (suggesting why "selbri" was once rendered as "*kunbri" (empty predicate))