From araizen@cs.huji.ac.il Thu Oct 24 12:04:56 2002
Received: with ECARTIS (v1.0.0; list lojban-list); Thu, 24 Oct 2002 12:04:56 -0700 (PDT)
Received: from mxout1.netvision.net.il ([194.90.9.20])
	by digitalkingdom.org with esmtp (Exim 4.05)
	id 184nHp-00014w-00
	for lojban-list@lojban.org; Thu, 24 Oct 2002 12:04:53 -0700
Received: from default ([62.0.148.37]) by mxout1.netvision.net.il
 (iPlanet Messaging Server 5.2 HotFix 0.8 (built Jul 12 2002))
 with SMTP id <0H4I00ME821REA@mxout1.netvision.net.il> for
 lojban-list@lojban.org; Thu, 24 Oct 2002 20:58:41 +0200 (IST)
Date: Thu, 24 Oct 2002 20:58:27 +0200
From: Adam Raizen <araizen@cs.huji.ac.il>
Subject: [lojban] Re: la ogYsty.
To: "lojban-list@lojban.org" <lojban-list@lojban.org>
Message-id: <0H4I00ME921SEA@mxout1.netvision.net.il>
MIME-version: 1.0
X-Mailer: Foxmail 4.1 [eg]
Content-Type: text/plain; charset=us-ascii
X-archive-position: 2311
X-ecartis-version: Ecartis v1.0.0
Sender: lojban-list-bounce@lojban.org
Errors-to: lojban-list-bounce@lojban.org
X-original-sender: araizen@cs.huji.ac.il
Precedence: bulk
Reply-to: lojban-list@lojban.org
X-list: lojban-list

de'i li 2002-10-24 ti'u li 15:45:00 la and. cusku di'e

>I am having difficulty in working out how {la .ogYsty.} can misparse
>in such a way as to make it illicit. I can't find any way of
>breaking it down into smaller chunks.
>
>Other examples:
>
>   la .olbyni.
>   la cropcy.
>   la kadiu.
>   la to,ori.
>
>The same question goes for {la stylanan.} and {la stailanan}.
>
>Given that we want want the phonological structure of cmevla to
>be as unconstrained as possible, was there any reason why the
>rule wasn't something like "break the string up into pause-
>delimited chunks, and then parse those chunks into the 
>smallest possible licit words"?

There's probably nothing wrong with 'la stylanan'; 'la' embedded within a name after a 'y' is in the same class as embedded after a consonant: you can't have a word break there without a pause, though I don't think that CLL explicitly allows it. 'la stailanan' would probably break up as 'lastai la nan.' ("It plastic-shapes South"?) As for the rest, I think the motivation is that it gives simple and easy to check-for rules, while at the same time allowing names to stay relatively close to their original shape. If the rule were "anything which is not a brivla or a cmavo and cannot be broken down into one is a valid cmene", it would be significantly more difficult to check whether a given cmene is valid, and you probably couldn't do it on-the-fly in most cases.

mu'o mi'e .adam.