From arosta@uclan.ac.uk Wed Nov 14 05:38:03 2001 Return-Path: X-Sender: arosta@uclan.ac.uk X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-8_0_0_1); 14 Nov 2001 13:38:03 -0000 Received: (qmail 87851 invoked from network); 14 Nov 2001 13:38:03 -0000 Received: from unknown (216.115.97.172) by m2.grp.snv.yahoo.com with QMQP; 14 Nov 2001 13:38:03 -0000 Received: from unknown (HELO com1.uclan.ac.uk) (193.61.255.3) by mta2.grp.snv.yahoo.com with SMTP; 14 Nov 2001 13:38:03 -0000 Received: from gwise-gw1.uclan.ac.uk by com1.uclan.ac.uk with SMTP (Mailer); Wed, 14 Nov 2001 13:14:11 +0000 Received: from DI1-Message_Server by gwise-gw1.uclan.ac.uk with Novell_GroupWise; Wed, 14 Nov 2001 13:50:11 +0000 Message-Id: X-Mailer: Novell GroupWise 5.5.2 Date: Wed, 14 Nov 2001 13:49:39 +0000 To: pycyn , lojban Subject: Re: [lojban] Why is there so much irregularity in cmavo/gismu? Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable Content-Disposition: inline From: And Rosta X-Yahoo-Profile: andjamin >>> 11/13/01 09:55pm >>> #arosta@uclan.ac.uk writes: #> I think part of the problem is that Lojban has a much narrower definitio= n #> of 'grammar' and 'grammatical' than is normal in linguistics and than is #> normally included within the 'grammar' of natural languages. This is not #> necessarily a Mistake, because an invented language is a different #> sort of creature from a natural language. At any rate, the role of a #> grammar is normally taken to be the rules that define a mapping from #> phonological structures to sentence meanings; the grammar generates #> all the well-formed sentences of the language, where sentences are #> defined as pairings between meanings and phonological forms. #> Lojban 'grammar' does something totally different: it defines a set #> of phonological strings and structurings of the words therein, but says= =20 #> nothing about their meanings. Natural language simply has no #> analogue of this 'pseudogrammar'. #=20 #While I agree that there are many questions about Lojban usage and meaning= =20 #that are unresolved, I have to protest And's description -- and prescript= ion=20 #-- on grammar. While his usage is not idiosyncratic, it is not nearly so= =20 #universal as he would make it seem: still current phrases like=20 #"transformational grammar". "phrase structure grammar", "Montague grammar"= =20 #and the like refer to things in the same class and role as Lojban grammar= =20 I would grant your point but not your examples. Transformational Grammar is the sort of grammar (mapping sounds to meanings) that I was talking about. You're the expert on Montague, but all the same, it's my impression that that is what Montague grammar does to. OTOH, Phrase Structure Grammar is indeed a more abstract kind of thing, that does not necessarily involve meanings. So let me backtrack and say "There are two different senses of _grammar_ prevalent in linguistics, only one of which is equivalent to the Lojban use of the term". #(with the difference that Lojban grammar, being prescriptive, actually doe= s=20 #what it says it is to do). The most And can clearly say of Lojban is that= it=20 #has only half of what an ideal grammar would have (which is 2/3 more than = any=20 #other language has, by the way), a mapping between sound sequences and=20 #grammatical utterances. What is missing is the semantic component, from=20 #grammatical utterance to sentence meaning, where Lojban is not significant= ly=20 #better off than many ordinary langauges (except for having secure grammati= cal=20 #sentences) and may even be behind in some areas. Many grammarians would be= =20 #seriously put out if parts of the sentence to meaning mechanism were=20 #incorporated into the sound to sentence mechanism, though, again, this is = not=20 #a universal objection and several such combinations have enjoyed a measure= of=20 #success over the last half-century (and some have been disasters, but so h= ave=20 #some pure cases).=20=20 We could debate this, but it would be a debate about natlangs and natlang linguistics. I think the simplest response is that describing/defining an existing language is very different from defining a language that is in the process of being invented, and it is risky to assume that properties of the one carry over analogously to the other. #In any event, "pseudogrammar" for Lojban's quite=20 #successful syntax is misleadingly denigrating, as though Lojban's situatio= n=20 #were somehow different from -- and inferior to -- that of natural language= s.=20=20 =20=20 'Pseudogrammar' is misleadingly denigrating, but Lojban's situation is both superior and inferior to that of natural languages. It is inferior in that = the macrogrammar of natlangs are vastly more complete and elaborated, but it is superior in that what so far exists of the language is incontrove= rtibly and explicitly documented. --And.