Newsgroups: comp.ai,comp.speech,comp.ai.nat-lang,sci.lang,comp.ai.philosophy,comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!nntp.club.cc.cmu.edu!miner.usbm.gov!news.er.usgs.gov!stc06.ctd.ornl.gov!fnnews.fnal.gov!cbgw1.lucent.com!cbgw3.lucent.com!news.pbi.net!news.mathworks.com!howland.erols.net!usc!news.isi.edu!gremlin!shomase!jbarnett
From: jbarnett@charming.nrtc.northrop.com (Jeff Barnett)
Subject: Re: two agendas of artificial intelligence
Message-ID: <E4HH5M.4CK@gremlin.nrtc.northrop.com>
Sender: news@gremlin.nrtc.northrop.com (Usenet News Manager)
Reply-To: jbarnett@charming.nrtc.northrop.com
Organization: Northrop Automation Sciences Laboratory
References: <53t89q$irs@bignews.shef.ac.uk> <5406r5$dtm@nntp.seflin.lib.fl.us> <elnaDzqpsA.KuG@netcom.com> <32E5559C.606@ohsu.edu> <32e8652b.11176921@news.mindspring.com> <32E6968D.496F@ohsu.edu>
Date: Thu, 23 Jan 1997 22:58:33 GMT
Lines: 31
Xref: glinda.oz.cs.cmu.edu comp.ai:43707 comp.speech:12122 comp.ai.nat-lang:5872 sci.lang:68960 comp.ai.philosophy:51402 comp.ai.alife:7342

In article <32E6968D.496F@ohsu.edu>, Stuart Robinson <robinstu@ohsu.edu> writes:
|>   .............  moby deletions .........................
|> I spoke of the utility of invented languages, it was in the context of
|> an invented langauge for the purpose of computer speech recognition.
|> 
|> I am not aware of much "empirical" work on invented languages, such as
|> Esperanto.  If there exists any, would someone supply references?

I'm not sure if the following is the sort of invention that you had
in mind, but was a kind frequently used during the 1970's for
the ARPA-sponsored Speech Understanding Research Project.  We all
used what we called "pragmatic" grammars in our systems.  A pragmatic
grammar would be defined in terms of fitting constiuents such as
   Was <person> <rank-name> of the <organization-name>?    or
   Was <officiers-name> <military-rank-name> of the <cadre>?
Where the first matches "Was Joe CEO of IBM?" and the first and
second both match "Was Jimmy Carter Captain of the Seawolf?".
Essentially, grammars were writen with extentional names for some
of the meta variables in addition to the usual parts of speech
and other-syntax references.  Also note that the case (Filmore
stuff) was buried in the pragmatic syntax.  The reson for all of
this was that it made parsing infinetly easier because the sytax
was also the sematic constraints of an albeit more constrained
language.  If your point in worying about invented language is
to find something which is manageable on one hand, fairly natural
on the other, AND, at the same time easy for the client to know
the boundaries, then think along the lines indicated above.

Jeff Barnett


