From jay.kominek@colorado.edu Tue Aug 07 11:10:21 2001
Return-Path: <kominek@ucsub.colorado.edu>
X-Sender: kominek@ucsub.colorado.edu
X-Apparently-To: lojban@yahoogroups.com
Received: (EGP: mail-7_2_0); 7 Aug 2001 18:10:21 -0000
Received: (qmail 86440 invoked from network); 7 Aug 2001 18:09:42 -0000
Received: from unknown (10.1.10.26) by l7.egroups.com with QMQP; 7 Aug 2001 18:09:42 -0000
Received: from unknown (HELO ucsub.colorado.edu) (128.138.129.12) by mta1 with SMTP; 7 Aug 2001 18:09:42 -0000
Received: from ucsub.colorado.edu (kominek@ucsub.colorado.edu [128.138.129.12]) by ucsub.colorado.edu (8.11.2/8.11.2/ITS-5.0/student) with ESMTP id f77I9gf27047 for <lojban@yahoogroups.com>; Tue, 7 Aug 2001 12:09:42 -0600 (MDT)
Date: Tue, 7 Aug 2001 12:09:41 -0600 (MDT)
To: <lojban@yahoogroups.com>
Subject: Re: [lojban] Re: Transliterations survey
In-Reply-To: <01080713343508.01223@neofelis>
Message-ID: <Pine.GSO.4.33.0108071140410.12191-100000@ucsub.colorado.edu>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
From: Jay Kominek <jay.kominek@colorado.edu>
X-Yahoo-Message-Num: 9298


On Tue, 7 Aug 2001, Pierre Abbat wrote:
> On Tue, 07 Aug 2001, Evgueni Sklyanin wrote:
> >> Fukushima, Japan
> >>
> >> fukucima.
> >> fikicima.
> >> fykycima.
> >>
> >I wonder why no one considered the variant {fukusima.}
> >A few points in favor of {si} versus {ci}:
>
> The 'f' is also an allophone; it is a variant of 'h' which is pronounced AIU as
> 'wh' before 'u' (compare English "who" and "hoot"; the first sound is a
> different phoneme but is pronounced identically). Lojban, however, has lost 'h'
> as a consonant (it merged with 'x' and the sound is now used only between
> vowels as a voiceless glide). So what do we do with Japanese words beginning
> with 'h'?

I'd drop the 'h'.

Possibly meaningless statistics:

In a Japanese dictionary of ~58k words, 3629 start with 'h'. (6.18%)

Of those 3629 words, there are 1059 words where, if you remove the 'h',
the newly created word is in the dictionary.

So dropping the 'h' leaves you with a ~33% chance of namespace collision.

For fu'ivla, the prefixed rafsi should eliminate any further confusion.
For names... well. I didn't claim the idea was perfect. :)

- Jay Kominek <jay.kominek@colorado.edu>
  Waiting Is.