Return-Path: From: cbmvax!uunet!math.ucla.edu!jimc Return-Path: Message-Id: <9106282205.AA14523@luna.math.ucla.edu> To: lojban-list@snark.thyrsus.com Cc: "Arthur W. Protin Jr." (GC-ACCURATE) Subject: Re: le mi jurdi In-Reply-To: Your message of "Fri, 28 Jun 91 14:51:14 EDT." <9106281451.aa07246@COR4.PICA.ARMY.MIL> Date: Fri, 28 Jun 91 15:05:04 -0700 Status: RO X-From-Space-Date: Fri Jun 28 20:01:32 1991 X-From-Space-Address: cbmvax!uunet!math.ucla.edu!jimc la "Arthur W. Protin Jr." cusku di'e > Rather than say the thing of mine which is an address > "lo mi judri" > I want to say the thing which is the address of me. > something definite for > da poi da judri mi This pattern comes up fairly commonly; I see it most often with body parts and garments (defined in -gua!spi, but not Lojban, analogous to "x1 is x2's hat"). One possible convention is this: The pre-s-selbri argument ("mi" here) is usually just like any other argument of the s-selbri, and thus goes in x2 with no special linking. However, if x2 cannot accept it (because it doesn't exist, or because an abstract sumti belongs there), you fall back on "pe" (nonrestrictive) or "ne" (restrictive). By the way, the official interpretation of "lo mi judri" is "lo judri pe mi" nonrestrictive. I can't think of a circumstance when I would want the nonrestrictive "pe"; I would always want the restrictive "ne", as in pe'u ko dunda fi mi le mi mapku .enoi le tu mapku Please (imper.) give me my hat not your hat If you believe in diklujvo, the most flexible interpretation is to analyse the s-selbri as if it were going to be a type 3 (transitive) diklujvo, and then drop the pre-argument into the transitive place, if it exists. This usually gives the same answer as above, but when the transitive place is other than x2 it gives a little more flexibility. Note that the grammar allows any argument in the pre-argument site, but a long argument is atrocious style, as in le (la kristoforo kolon. ku) bloti Christopher Columbus' boat -- jimc