Selected Newsgroup Message
From: "Sergio Navega" <snavega@ibm.net>
Subject: Re: Neurons and consciousness
Date: 27 Jul 1999 00:00:00 GMT
Message-ID: <379dac2e@news3.us.ibm.net>
References: <7nb0q9$rkv@ux.cs.niu.edu> <7ng8ti$ata@dfw-ixnews4.ix.netcom.com>
<7ni3e7$6ab@ux.cs.niu.edu> <7nk08r$t6a@dfw-ixnews9.ix.netcom.com>
X-Notice: should be reported to postmaster@ibm.net
X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3
X-Complaints-To: postmaster@ibm.net
X-Trace: 27 Jul 1999 12:55:10 GMT, 129.37.182.30
Organization: Intelliwise Research and Training
Newsgroups: comp.ai.philosophy,sci.physics
Steven B. Harris wrote in message <7nk08r$t6a@dfw-ixnews9.ix.netcom.com>...
>In <7ni3e7$6ab@ux.cs.niu.edu> Neil W Rickert <rickert+news@cs.niu.edu>
>writes:
>>>
>>>>As I said, your ideas about language are hopelessly simplistic.
>>
>>> It remains to be seen which of us have the simplistic view of
>>>language.
>>
>>>>> Most of our words do not have definitions.
>>
>>> I suppose dictionaries only cover a fraction of the language,
>then?
>>
>>Perhaps you have never noticed that most dictionary "definitions"
>>refer to other words, and their "definitions" in turn refer to still
>>other words. In effect the dictionary is a system of circular
>>"definitions" which therefore fails to define anything.
>
>
> Therefore?? People like you should be barred from use of the
word
>"therefore." <sigh>. What you say follows logically _only_
if no
>other sources of knowledge of definitions of words BUT the dictionary
>are available. However, since that is rarely the case your remarks are
>irrelevent. An English dictionary doesn't define anything only if you
>have no English at all. However, given a definition source other than
>the dictionary for only a few words, a dictionary can usefully defining
>the _rest_ of the words in a language. Which may be most of them
>(though they will be a different set of words for different people,
>which is why you can't leave any out). All of which being why
>dictionaries exist. I suppose you are going to argue that dictionaries
>don't really give any definitions at all, but most people, excluding
>you of course, are too dumb to have figured that out, and pay out money
>in a giant scam, year after year? Hmmm?
>
> It would not take many extra-dictionary definitions to make English
>quite accessable though a dictionary, to a civilization which knew it
>other than a dead language. After all, ancient Egyptian writing and a
>number of other formerly mysterious ancient written languages have been
>deciphered starting with a single inscription which essentially defines
>a few words in terms of known words in another language (by repeating
>an inscription in several languages). After that, the rest of the
>language can be worked out by induction, from use of unknown words in
>context along with known words. Which is (by example) also what many
>dictionaries do (though they also do much more). It's not quite
>circular, since it proceeds outward from a kernal or core. None of
>this supports your statement that most words don't have definitions.
>
Steven, as much as I want to agree with you in other aspects (I also
regard induction as important), dictionary definitions are far from
what one can mean by "definition" in a more deep way. It's all in the
"core" that you cite.
I guess that this idea can be more easily grasped if one thinks about
what are the preconditions in order to use a dictionary. Put aside
the more lexical aspects such as alphabetical knowledge. We take a
definition of a word and then what do we do?
There are two cases I would like to highlight. First, take the word
"avalanche", for instance. American Heritage Dict. goes this way:
"A fall or slide of a large mass, as of snow or rock, down a
mountainside". It is obvious that you'll have to know what "slide"
or "fall" means, also what is a large mass, snow, rock, mountain,
sides of mountains. If we go on and take the dictionary definitions
of each of these words we will find ourselves trapped in a large
circular and self-referential definition with only one way out: to
resort to sensory concepts, which *can't* be defined. That's the
limit of the dictionary 'definitions'. Nobody can define these
concepts. I can't say what is "red" or what is "movement" or
what is "rough" or what is "salty", or "loud", or
"painful".
Although this seems to be enough trouble, I wouldn't say it is the
worse. The worse is the second case, that of the *abstract words*
which can't be defined *even in terms of single sensory concepts*.
One of my preferred cases is the word "justice". American Heritage
take on this word is almost laughable: "The quality of being
just; fairness". It is not trivial to reduce a case like this
to basic sensory concepts. What is at stake here is not the
"definition" of the terms, but our ability to *recognize* its
appearance in certain cases, the structure of sub-concepts that
make the concept emerge.
Thus, I'd be more confortable to substitute our capacity of
"definining" things with two other abilities. Our ability to
*recognize* and our efficacy in *generating* situations where
the concepts appear clearly, with the added condition that
the ability to recognize precedes that of generation.
This is, in my opinion, very important to AI. There are a lot
of projects out there that insist on using "definitions" as the
support of its "knowledge base". They try, hopelessly, to put
as many definitions as possible, in a desperate attempt to find
that "critical mass" where the whole crap will act intelligently.
It is easy to perceive that in my opinion this will never work.
Without recognition (which means, without *perceptual* abilities)
no system will be able to *generate* useful knowledge, which is
one of the most important aspects of any AI system. Notice that
I've said "perceptual", not sensory. I have a strong impression
that the former is more important than the latter.
Regards,
Sergio Navega.
From: "Sergio Navega" <snavega@ibm.net>
Subject: Re: Neurons and consciousness
Date: 31 Jul 1999 00:00:00 GMT
Message-ID: <37a30df3@news3.us.ibm.net>
References: <7nb0q9$rkv@ux.cs.niu.edu> <7ng8ti$ata@dfw-ixnews4.ix.netcom.com>
<7ni3e7$6ab@ux.cs.niu.edu> <37A2DFC4.593463ED@travellab.com>
<37A2EFC6.3E8D@online.no> <37A2FE70.6BFB4785@travellab.com>
X-Notice: should be reported to postmaster@ibm.net
X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3
X-Complaints-To: postmaster@ibm.net
X-Trace: 31 Jul 1999 14:53:39 GMT, 129.37.183.227
Organization: Intelliwise Research and Training
Newsgroups: comp.ai.philosophy,sci.physics
Henning Strandin wrote in message <37A2FE70.6BFB4785@travellab.com>...
>Tore Lund wrote:
>>
>> Henning Strandin wrote:
>> >
>> > Neil W Rickert wrote:
>> > >
>> > > Perhaps you have never noticed that most dictionary
"definitions"
>> > > refer to other words, and their "definitions" in turn refer
to still
>> > > other words. In effect the dictionary is a system of circular
>> > > "definitions" which therefore fails to define anything.
>> >
>> > This assertion is simply wrong. Most definitions in a dictionary are not
>> > circular at all, but hierarchical. The word bee is defined as an insect,
>> > insect is defined as an arthropod, an arthropod is defined as an
>> > invertebrate having jointed limbs and an invertebrate is defined as an
>> > animal without backbone.
>>
>> Of these four definitions the latter three seem OK (to nonexperts like
>> myself at least). But the first one - bee is an insect - is merely a
>> classification. We should have to add lots of other information to get
>> an idea of what a bee is. And even so, if we had never seen a bee, this
>> would just give us an *idea*.
>
>The point in defining a word is to give you an idea of what it means.
>
That is right, but the idea of what it means is (one way or another)
already "inside" the reader. It could be through an analogy or example
(where you already have fundamental notions) or through terms that
can't be further defined, such as color atributes, shapes, smells, etc.
Dictionaries may be circular in regard to some words and hierarchical
in relation to others. But they always stop at the level of syntactic
relationships. "Blue" can only be experienced, not defined.
Regards,
Sergio Navega.
From: "Sergio Navega" <snavega@ibm.net>
Subject: Re: Neurons and consciousness
Date: 02 Aug 1999 00:00:00 GMT
Message-ID: <37a59841@news3.us.ibm.net>
References: <37A3302B.3733F782@travellab.com>
<37A3474E.16D3BC2F@clickshop.com> <7nvqsd$396@ux.cs.niu.edu>
X-Notice: should be reported to postmaster@ibm.net
X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3
X-Complaints-To: postmaster@ibm.net
X-Trace: 2 Aug 1999 13:08:17 GMT, 200.229.243.58
Organization: Intelliwise Research and Training
Newsgroups: comp.ai.philosophy,sci.physics
Neil W Rickert wrote in message <7nvqsd$396@ux.cs.niu.edu>...
>Seth Russell <sethruss@clickshop.com> writes:
>>Henning Strandin wrote:
>
>>> Can you define the meaning of the word bee by pointing at a bee?
>
>>Yep, imho, that's the way you do it.
>
>Agreed. However this doesn't really amount to a definition. It
>allows the learner to use the word 'bee' in ways somewhat similar to
>how you use the word. But there is no guarantee that there is any
>identity in how the word is used. This problem with ostensive
>definition is the basis for Quine's skeptical argument about
>'meaning' (in Word and Object as well as other places).
>
Can't believe you guys settled agreeing on this. While I admit
that pointing is helpful in the process of constructing meaning,
it is far from enough without a lot of accessory work.
Take a dog, for instance. I may point to it and call it 'animal'.
Everybody would agree. Now I may go to China and point to that
same dog and call it 'edible'. Everybody there would agree.
In the US, that wouldn't fit very well.
Regards,
Sergio Navega.
From: Neil W Rickert <rickert+nn@cs.niu.edu>
Subject: Re: Neurons and consciousness
Date: 02 Aug 1999 00:00:00 GMT
Message-ID: <7o4f1v$9dk@ux.cs.niu.edu>
References: <7nvqsd$396@ux.cs.niu.edu> <37a59841@news3.us.ibm.net>
Organization: Northern Illinois University
Newsgroups: comp.ai.philosophy,sci.physics
"Sergio Navega" <snavega@ibm.net> writes:
>Can't believe you guys settled agreeing on this. While I admit
>that pointing is helpful in the process of constructing meaning,
>it is far from enough without a lot of accessory work.
>Take a dog, for instance. I may point to it and call it 'animal'.
>Everybody would agree. Now I may go to China and point to that
>same dog and call it 'edible'. Everybody there would agree.
>In the US, that wouldn't fit very well.
You are right, of course. I didn't take the pointing idea as
suggesting just one instance of pointing. I took it more broadly as
including other examples of use of the term in context.
Back to Menu of Messages Sergio's Homepage
Any Comments? snavega@attglobal.net