From arosta@uclan.ac.uk Tue Sep 25 07:53:53 2001 Return-Path: X-Sender: arosta@uclan.ac.uk X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-7_3_2_2); 25 Sep 2001 14:53:53 -0000 Received: (qmail 35649 invoked from network); 25 Sep 2001 14:49:44 -0000 Received: from unknown (10.1.10.142) by l7.egroups.com with QMQP; 25 Sep 2001 14:49:44 -0000 Received: from unknown (HELO com1.uclan.ac.uk) (193.61.255.3) by mta3 with SMTP; 25 Sep 2001 14:49:39 -0000 Received: from gwise-gw1.uclan.ac.uk by com1.uclan.ac.uk with SMTP (Mailer); Tue, 25 Sep 2001 15:27:07 +0100 Received: from DI1-Message_Server by gwise-gw1.uclan.ac.uk with Novell_GroupWise; Tue, 25 Sep 2001 15:58:06 +0100 Message-Id: X-Mailer: Novell GroupWise 5.5.2 Date: Tue, 25 Sep 2001 15:57:40 +0100 To: richard , lojban Subject: zipf computations & experimental cmavo (was: Re: META : Who is everyone Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable Content-Disposition: inline From: And Rosta X-Yahoo-Message-Num: 11022 My sense is that major cmavo zipfings would not belong to existing selmaho. However, one possibility would be to try to restrict zipfer cmavo to string abbreviations i.e. a zipfer cmavo is equivalent to a determinate longer string of Lojbab words. Then the zipfer cmavo could be expanded to=20 the longer string at a preparser stage. (I'm talking about a kind of lojban equivalent of "IIRC" and "IMHO" and "OTOH", imagining that those abbreviations were pronounced "irk" "im-ho", "o-toe".) The need for zipfer cmavo of this sort could be computed automatically, if= =20 someone wrote the necessary software & all usage was fed into it. The softw= are would consider the frequency of every wordstring, relative to its syllablecount, = and calculate how many syllables in usage could be saved by replacing the wordstring by a= n experimental cmavo. Savings significant enough to justify the effort of lea= rning the experimental cmavo could then justify adding the cmavo to the language. I calculate that there are 1224 disyllabic experimental cmavo (of CVVV and CVVVV type), which is sure to exceed demand both for zipfer cmavo and for new nonzipf cmavo. Perhaps also this software could be used for making a ca= se for reassigning some monosyllabic cmavo; e.g. I would love to reassign {lau= } as an abbreviation for tetrasyllabic {tu'odu'u}. At any rate, software of t= his sort would allow us to gauge the level of 'zipfean defectiveness' in the languag= e -- the extent to which the language design imposes superfluous syllables on= usage. There still remains the question of how to zipf down long lujvo whose sylla= blecount is excessive relative to their frequency, but Type IV fuhivla might provide= a solution here. --And. >>> Richard Curnow 09/23/01 10:10pm >>> Adding new cmavo to existing selma'o is a relatively simple change in jbofi'e. Adding new selma'o would presumably involve adding new grammar rules (else why would you need them) - the complexity would vary on a case-by-case basis. Also there are parts of the 'back-end' (e.g. working out which terms are which arguments of each selbri) which have a lot of grammatical structure hard-coded in. On Thu, Sep 20, 2001 at 05:41:27PM -0400, Bob LeChevalier (lojbab) wrote: > At 12:31 PM 9/20/01 +0100, And Rosta wrote: >=20 > >Either that or > >we encourage people to experiment with using their own zipfing experimen= tal > >cmavo, possibly with concomitant grammar changes, and then see which > >experimentals catch on. >=20 > Experimental cmavo have no defined grammar, so using them is not a "gramm= ar=20 > change". However you have to be skilled enough at the language to be abl= e=20 > to communicate without having a jbofi'e to parse and translate your=20 > experiments, since jbofi'e won't likely be able to incorporate all of the= m=20 > even if Curnow wanted to try (some might not pass LALR1) --=20 Richard. ---------------------------------+------------------------------------- Richard P. Curnow | C++: n., An octopus made by Weston-super-Mare, UK | nailing extra legs on a cat. http://www.rrbcurnow.freeuk.com/ | To unsubscribe, send mail to lojban-unsubscribe@onelist.com=20 Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/=20