From colin@KINDNESS.DEMON.CO.UK Sun Jul 02 01:42:48 2000 Return-Path: Received: (qmail 4020 invoked from network); 2 Jul 2000 08:42:48 -0000 Received: from unknown (10.1.10.27) by m4.onelist.org with QMQP; 2 Jul 2000 08:42:48 -0000 Received: from unknown (HELO finch-post-12.mail.demon.net) (194.217.242.41) by mta1 with SMTP; 2 Jul 2000 08:42:47 -0000 Received: from kindness.demon.co.uk ([158.152.216.198] helo=arac) by finch-post-12.mail.demon.net with smtp (Exim 2.12 #1) id 138fKt-000KpQ-0C for lojban@egroups.com; Sun, 2 Jul 2000 08:42:43 +0000 To: Subject: RE: [lojban] Complements and adjuncts Date: Sun, 2 Jul 2000 09:49:24 +0100 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0) In-Reply-To: <20000629041906.39746.qmail@hotmail.com> X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Importance: Normal From: "Colin Fine" X-Yahoo-Message-Num: 3367 I've been thinking more about complements and adjuncts, and I want to lay out some meanderings. First, a bit of theoretical background: in a prominent theory of grammar, X-bar theory, there is a commonality to the structure of phrases across different types of phrase (noun phrase, verb phrase etc) and across (all?) natural languages. It is claimed that there is a two level structure, with an X-phrase (noun phrase, for example) containing an X-bar, which in turn contains an X. The significance of the two-level structure is that the X-bar level contains items which are said to be crucial to the meaning of the X (complements) whereas the X-phrase may contain optional modifiers (adjuncts). Thus Pinker (in The Language Instinct) discusses the phrase The Senator from California from Illinois which he analyses as NP:[Nbar:[The Senator from California] from Illinois] 'The Senator from California' is clearly a unit - part of the meaning of a senator is the constituency. This is an idea that Lojbanists are well acquainted with. What happens if we put this into Lojban? In a sumti, we can formally mirror the layers clearly: le xelflaka'i be la kalifornias be'o pe (ra'i) la ilinOIs I suggest that this is not just a formal resemblance, but that a PE phrase really is comparable with an adjunct, as opposed to the complement within the selbri. In a bridi, we do not at first sight have this option: we may not say *ta xelflaka'i la kalifornias pe la ilinOIs because PE phrases are not permitted in bridi. We are supposed to use some construction such as ta xelflaka'i la kalifornias gi'e se krati la ilinOIs and that's find, except that we have to commit ourselves on the relationship to la ilinOIs. (I suppose if we really don't want to we can always say gi'e co'e la ilinOIs, but that is much more vague than the English. Perhaps he xebna la ilinOIs). But there is a formal parallel we can use to the sumti case: ta xelflaka'i be la kalifornias ra'i la ilinOIs This has the property (like the English and the sumti case), that it cannot be permuted - xelflaka'i be la kalifornias is a constituent and cannot be interrupted. The question is whether there is any semantic significance to the grouping. I guess what I am asking is whether ta xelflaka'i be la kalifornias ra'i la ilinOIs and ta xelflaka'i la kalifornias ra'i la ilinOIs are synonymous, or does the structural difference have a semantic correlate? Since we have already had the opinion that any expressed tagged sumti join the predefined terbridi in being complements, it is hard to see what semantic difference there could be. But I would like to hear people's thoughts on the matter. **************************************************************************** **** Colin Fine "Don't just do something! Stand There" - from 'Behold the Spirit' workshop colin@kindness.demon.co.uk **************************************************************************** ****