From LOJBAN@CUVMB.BITNET Sat Mar 6 22:58:28 2010 Reply-To: jimc@MATH.UCLA.EDU Sender: Lojban list Date: Tue Dec 17 10:07:42 1996 From: Jim Carter Subject: Re: BEGINNER'S QUESTION: internal sumti X-To: nsummers@hutch.com.au X-cc: lojban@cuvmb.columbia.edu To: John Cowan In-Reply-To: Your message of "Fri, 13 Dec 96 01:17:47 +1000." <199612121418.GAA16878@lambda.ben2.ucla.edu> X-UIDL: 7f0804900a96ba77feca9c419459b964 X-Mozilla-Status: 0011 Content-Length: 4037 X-From-Space-Date: Tue Dec 17 10:07:42 1996 X-From-Space-Address: - Message-ID: Nick Summers writes: > ... ta cu tavla be do bei le melbi ku be'o vecnu > ... > The translation offered for this bridi in the grammar is: > > "That is a talker-to-you-about-the-beautiful-thing(s) salesperson. > (or, more simply) > That's a salesperson who talks to you about beautiful things." I think your diagram (not reproduced above) is just fine, plus the inferences you have drawn from it. My only disagreement is that I would tend to interpret "Y vecnu" as a salesperson who sells Ys, because "vecnu" is transitive in an essential way. (That's not a Lojban concept.) Thus: "X is a salesman of sweet-talking dolls". You are correct that "le vecnu" means salesperson (or salesdroid or whatever) by itself. (Also, "le se vecnu" = the goods for sale; "le te vecnu" = the buyer; "le ve vecnu" = the low, low price.) You use tanru to jazz up the predicate (for example to give a teaching example of an over-complicated tanru where the modifying element has arguments stuck on). There are only about 1400 gismu, and they cover semantic space too coarsely for the nuances of meaning often desired. The tanru relation is deliberately unspecified (more unspecified than I feel comfortable with, by the way.) Speakers are expected to invent tanru; there is not. and probably never will be, any official list. The modifying element is identified as having explicit arguments because "be" links the first one to it alone (no ambiguity what item "be" is linking the arg to.) The end of the unit is marked by "be'o", which often can be elided, but in this case it can't. The element isn't "almost clausal", it IS clausal (bridi), and would have been clausal even without the arguments. "blari'o" is a lujvo, which is related to a tanru, but the rules for interpreting it semantically are a bit more structured. There is a chapter in the grammar about that. As for "who...", I like to think of it like this: a bridi expresses a relation between all of its arguments, even if some are not explicitly designated by words. Thus there must be one or more specific persons who did get, or will get, sweet-talking dolls sold to them, at one or more specific prices, even though the sentence does not identify them. There is the philosophical argument: a self-proclaimed poet never writes any poems; does he truly qualify for that predicate? I'm not going to try to answer that one. Also note, with modal cases ^H^H^H^H^H arguments included, every predicate potentially has infinitely many arguments, which complicates any analysis. Let the philosophers rage! A "bridi" is a "selbri" (bare predicate) with its arguments (sumti). A "sumti" is (omitting one or two specializations) a special kind of bridi with determiners ("le" etc.) and other stuff stuck on. A selbri can have a variety of forms, the most common of which consists of a gismu (primitive predicate word) or a lujvo and zero or more tanru modifying elements, which themselves are bridi (suitably endmarked in the complicated cases). In computer science terminology, a selbri is like a function. X1, X2... are its formal parameters. Sumti are its actual parameters. In a sumti, X1 is left open, and becomes the functions' value; whereas a bridi is a boolean expression (true or false depending on the specific sumti). Generally a bridi or sumti represents an ensemble of referents. It is important for the beginner to pay attention to self-creating tanru and lujvo, because the gismu are intended to form a "basis" of semantic space, but only combinations can cover it completely (so we hope). ---- But you certainly can say a lot just with the gismu. As a beginner you have shown a surprising grasp of what's going on. I hope you will continue in Lojban. James F. Carter Voice 310 825 2897 FAX 310 206 6673 UCLA-Mathnet; 6115 MSA; 405 Hilgard Ave.; Los Angeles, CA, USA 90095-1555 Internet: jimc@math.ucla.edu (finger for PGP key) UUCP:...!{ucsd,ames,ncar,gatech,purdue,rutgers,decvax,uunet}!math.ucla.edu!jimc