Message-Id: <199601202235.RAA11753@locke.ccil.org>
Date:         Fri, 19 Jan 1996 10:40:10 LCL
Reply-To: colin@kindness.demon.co.uk
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: "C:WINSOCKKA9QSPOOLMAIL" <colin@kindness.demon.co.uk>
Organization: None
Subject:      Re: TECH: Nested relative clauses
To: John Cowan <cowan@LOCKE.CCIL.ORG>
Status: OR
Content-Length: 2748

I've at last found time to have a look at this (for the first time)
and get my mind round it.

I understand the problem Veijo is tackling, and I agree that the
centre-embedded construction is unwieldy (and therefore unlikely
to be used).

But I'm not at all happy about the proposed solution. I can see that
it might well work in practice, but it strikes me as the same category
as 're xirma' - a construction that is in (has been in from the
beginning) for no better reason than that people want keep using it,
but that betrays the structure of the language with a surface kludge.

The level of structure I'm talking about is that of (grammatical) category. In
 English
(and Chinese) a word or phrase can often function as several different
 grammatical
categories. In Finnish it usually cannot, because its form defines its category.
 In
Lojban we have invariable words, but clearly defined categories: 'xirma' is a
 brivla; it
can therefore function as a selbri, or a bridi, or indeed a jufra. It cannot be
 a sumti -
it needs an explicit converter, normally a gadri.
Now my objection to 're xirma' is that a laivla (quantifier word) is being used
 as this
converter. Clearly it can be made to work because it has been; but in my view
 it's a
kludge, in large part because it means 're lo xirma' and that 'lo' is part of
 the
skeleton of the phrase. (The presence of 'le re xirma' complicates the
issue further)

I think the same problem arises in Veijo's proposal:

>   *le xoi le tcadu cu se klama ku'o nanmu cu se viska xu'o verba
>
What is the converter that makes 'nanmu' a sumti rather than a selbri? It must
 be the
whole phrase 'xoi le tcadu cu se klama ku'o'. Clearly this can be made to work
 (once
Veijo and John have finished hammering it out), and at one level there's no
 reason why it
shouldn't (any more than a laivla); but I belive that the gadri is really in
 there still,
and I don't like seeing it disappear.

I don't have a very strong objection, and if we were in a time of
flux, I might be prepared to let it in; but really my response is that
I don't like it, (though I see Veijo's point) and it's certainly not
sufficiently major to change things at this point. [In practice I don't
think that deep embedding is likely to get used whatever the
construction].


--
---------------------------------------------------------------------------
| Colin Fine    33 Pemberton Drive, Bradford BD7 1RA  01274 733680        |
|                          colin@kindness.demon.co.uk                     |
| "There are no extraordinary people: There are only ordinary people doing|
|       extraordinary things with what they have been given" - K.B.Brown  |
---------------------------------------------------------------------------