Date:         Thu, 16 Jul 1998 08:56:12 +0200
Reply-To:     Lionel Bonnetier <leo@EASYNET.FR>
Sender:       Lojban list <LOJBAN@CUVMB.BITNET>
From:         Lionel Bonnetier <leo@EASYNET.FR>
Subject:      [lojban] Thread starting on comp.ai.nat-lang
To:           Multiple recipients of list LOJBAN <LOJBAN@CUVMB.BITNET>
Message-ID: <900572279.1113334.0@listserv.cuny.edu>
Content-Length: 1372
Lines: 34


-- quotation --

From: "Russell Harper" <rharper@intouchsurvey.com>
Newsgroups: comp.ai.nat-lang
Subject: Spoken Language: Different Approach?
Date: 13 Jul 1998 14:40:58 GMT

I was wondering, instead of trying to work with existing languages, with
all of their computer unfriendly characteristics, why not invent a spoken
language that is ideal for research? E.g. only the most easily recognizable
sounds are used, each word is readily identifiable, as is every sentence.
Using either a fixed word order, or inflections, the subject, object and
verb can be determined, the same for emotive and temporal aspects.

>From this framework, it might be possible to research the higher level
problems, like how to determine context, and when to use tricky things like
humour, deception, sarcasm etc. Perhaps later, the research can be applied
to natural languages.

Anyway, I was wondering if any work has been done using this approach?

Russell

-- end quotation --

   Lionel Bonnetier <leo@easynet.fr>   Ph: +33 478 601 862