Knowledge's Discretion
Jeff Rubard
discrete: the inferential status of knowledge is determined, not only
by the inferential relations a piece of knowledge stands in, but by
epistemic warrants awarded to individual pieces of knowledge on the
basis of their relation to the world. Much as it is possible to know
something without knowing all its consequences, it is possible to know
something without knowing all its prerequisites. The limits on the
level of consequences drawn from knowledge indicate that useful
knowledge is clearly not individuated by its object, in which case
knowledge would have to be as faithful to the completely given object
as possible, in order to qualify as correctly related to it. Given
this, what must a knower be?
Firstly, knowledge requires a knower to have such a structure as to
unobtrusively register epistemic facts: if the subject was such as to
psychologically constitute all its information about the world, no
piece of that information could constitute a piece of knowledge for the
reason that all which was contained was information about a subject
continuous in all respects relevant to epistemic assessment (serving
not only as in a certain way the object of the information but also as
its epistemically undifferentiated matter). In other words, sense
cannot constitute knowledge for the reason that there is no
individuation of known contents in such a way as to make their
epistemic assessment possible: a piece of knowledge requires not only
an object, but criteria indicating what piece of knowledge it is, and
these cannot be supplied by the subject.
So empirical idealism is a nonstarter for the reason that it is unclear
what we are talking about when we attribute knowledge of perception to
the idealist subject. What would count as knowledge on this score?
Pieces of information differentiated, not by the object or subject, but
by their inferential relations internal to a body of knowledge: a piece
of knowledge has not only an origin but an epistemic *standing* as part
of a more-or-less-systematic body of interconnected propositions. And
as such, it is clear that what a knower must do is not only to
accurately capture the contours of the object (this much can be done by
sense) but to maintain knowledge's interrelations: that is, to let
knowledge be knowledge as an independent realm of thought relating
itself to itself.
Knowledge's objectivity derives from this, that it is formally checked
only by other pieces of knowledge, forming an independent unity
separate from psychological states: and any entity worthy of the name
"knower" needs to duplicate, not psychological states, but the
independence of the epistemic qua information: a body of knowledge
composed of neither objectively nor subjectively determined information
could exist separate from the consciousness of the individual subject.
What it could not exist without is the recognition of knowledge as
inferential discreteness: that is to say, there can be no theory of
knowledge suitable to the task which is not, in the end, atomistic.
--
Jeff Rubard
http://opensentence.tripod.com/
Essays on theory, culture and politics
Wolf Kirchmeir
[...]>
> Firstly, knowledge requires a knower to have such a structure as to
Why should there be a knower? There is no reason to suppose that knowing
requires a knower - anymore than raining requires a rainer.
The problem with verbs is that they refer to actions, processes, states
of being, changes in states of being, logical functions, and so on. Just
which is the case with "know" is a matter of investigation, not assumption.
Attempt to describe the issue in terms of knowing as a process rather
than a knower that acts. Then the problem is both simpler and more
difficult. It's simpler because you don't have to worry about the
relation between knower, knowing, and knowledge - they're all merely
grammatically different terms for the same thing. It's more difficult
because our language makes it difficult to talk about any process
without implying a processor. But when you consider what we expect from
Jim when we ask questions about his "knowledge", it's clear that what
Jim does, and only what Jim does, is what will satisfy us that he
actually knows something or other. Note that although Jim may claim that
he knows something, this is insufficient to satisfy - we want Jim to do
a lot more than utter a claim.
So I would start with "What is the evidence that some system knows
something or other?" Work back from that. Eg, what makes me conclude
that Jim knows math? By how he solves equations, constructs proofs,
applies formulas, analyses problems numerically or geometrically or etc,
and so on. An AI system that can do the sorts of things Jim does to
prove his claim that he knows math would be quite useful IMO.
Would such a system need "knowledge representation"? Maybe. It's not
obvious one way or the other IMO. And I know an immense amount of work
has been done knowledge representation, especially on common sense and
contextual knowledge. But it doesn't seem to be getting anywhere. There
is something else at work in a human that "knows" what the sentence "I
thought we could go out" means.
To put it another way: we humans are pretty good at sussing each other
intensions -- ordinary dialogue consists to a very large extent of
incomplete utterances that require the partners to grasp each other's
intensions. How do we do it? Not by "knowledge representation", IMO.
To put it still another way: A search engine will be "smart" when it's
able to provide me not only with docs that contain the search string,
but also with docs that relate semantically to the search string. And
just generic (dictionary based?) semantic relationships, but specific to
me and my interests, ie, the context in which I am searching for
information. (BTW, I dislike that Google drops "and" in my search
strings - sometimes "and" is the crucial word distinguishing two
otherwise nearly identical compound names. By including the "and" there
would be far fewer hits for me to wade through. The writers of Google
need some serious training in language and literature. NB how important
the "and" is in that final noun phrase! Oh, I know I can use Intext: to
force the inclusion of the "and", but inclusion should be the default,
not the work-around.)
patty
>
> Why should there be a knower? There is no reason to suppose that knowing
> requires a knower - anymore than raining requires a rainer.
>
Can you give an example of knowing without a knower ?
Well there are the marks of the sentence, there is some agent that
*interprets* those marks to stand for (the process of) us going out
(indexically adjusted). I think that those of us who analyze knowledge
in semiotic terms, don't end up with the kind of problems you are
highlighting. In semiotic terms, knowledge (knowing or whatever
grammatic form you wish to analyze) sans the interpreter is a category
error; as, imho, it should be. Knowing without a knower is just the
noumenal fizz of uninterpreted marks.
>
> To put it another way: we humans are pretty good at sussing each other
> intensions -- ordinary dialogue consists to a very large extent of
> incomplete utterances that require the partners to grasp each other's
> intensions. How do we do it? Not by "knowledge representation", IMO.
>
> To put it still another way: A search engine will be "smart" when it's
> able to provide me not only with docs that contain the search string,
> but also with docs that relate semantically to the search string. And
> just generic (dictionary based?) semantic relationships, but specific to
> me and my interests, ie, the context in which I am searching for
> information.
I agree, we're anxiously awaiting that :)
> (BTW, I dislike that Google drops "and" in my search
> strings - sometimes "and" is the crucial word distinguishing two
> otherwise nearly identical compound names. By including the "and" there
> would be far fewer hits for me to wade through. The writers of Google
> need some serious training in language and literature. NB how important
> the "and" is in that final noun phrase! Oh, I know I can use Intext: to
> force the inclusion of the "and", but inclusion should be the default,
> not the work-around.)
Actually Google doesn't drop "and" inside a string. To search for
"Bonnie and Clyde", put the quotes around your string like this
<http://www.google.com/search?q=%22Bonnie+and+Clyde%22>. To search for
the string "knowledge representation", put the string in quotes too
<http://www.google.com/search?q=%22knowledge+representation%22>.
Outside of a string the "and" is implied and not necessary. It works
perfectly and i can't think of how it could be improved.
patty
JGCASEY
Wolf Kirchmeir wrote:
> Jeff Rubard wrote:
> [...]>
> > Firstly, knowledge requires a knower to have such a structure as to
> > unobtrusively register epistemic facts: [...]
>
> Why should there be a knower? There is no reason to suppose that
knowing
> requires a knower - anymore than raining requires a rainer.
>
> The problem with verbs is that they refer to actions, processes,
states
> of being, changes in states of being, logical functions, and so on.
Just
> which is the case with "know" is a matter of investigation, not
assumption.
>
> Attempt to describe the issue in terms of knowing as a process rather
> than a knower that acts. Then the problem is both simpler and more
> difficult. It's simpler because you don't have to worry about the
> relation between knower, knowing, and knowledge - they're all merely
> grammatically different terms for the same thing. It's more difficult
> because our language makes it difficult to talk about any process
> without implying a processor.
But are there not physical -requirements- for knowing to take
place just as there are, to use your analogy above, physical
requirements for raining to take place? In order for say two
binary patterns (numbers) to be transformed into say the pattern
representing the addition of those numbers isn't the physical
existence of the central processing unit required for this to
happen. This physical requirement being the "processor". It is
the physical details that someone wanting to build intelligent
machines is looking for. Just as the physical details of air
masses with different humidity, temperature, ground conditions
may be required for "raining" to take place? Moist air is
"processed" into water droplets. We can make a rain processor
that inputs moist warm air and outputs water.
Now why would a system utter I (the knower) knows something?
What is the physical requirement for a system to behave as
if it is a "knower"?
> ... But when you consider what we expect from
May I suggest that the physical requirements for doing maths
is Jim and that is what we mean objectively by the "knower"?
Neil W Rickert
Hash: SHA1
"Jeff Rubard" <jeffr...@online.ie> writes:
>It seems an important point to me that the structure of knowledge is
>discrete:
Whereas it seems obvious to me that knowledge is not discrete.
>Firstly, knowledge requires a knower to have such a structure as to
>unobtrusively register epistemic facts:
What is a epistemic fact, and what distinguishes that from an
ordinary fact?
> if the subject was such as to
>psychologically constitute all its information about the world, no
>piece of that information could constitute a piece of knowledge for the
>reason that all which was contained was information about a subject
>continuous in all respects relevant to epistemic assessment (serving
>not only as in a certain way the object of the information but also as
>its epistemically undifferentiated matter).
Hmm. Can somebody translate that into English?
> In other words, sense
>cannot constitute knowledge for the reason that there is no
>individuation of known contents in such a way as to make their
>epistemic assessment possible: a piece of knowledge requires not only
>an object, but criteria indicating what piece of knowledge it is, and
>these cannot be supplied by the subject.
I could use a translation of that, too.
>So empirical idealism is a nonstarter for the reason that it is unclear
>what we are talking about when we attribute knowledge of perception to
>the idealist subject. What would count as knowledge on this score?
I think you are just expressing your opinion as opposed to empirical
idealism, and you are spewing a pile of sophistry as an alleged
argument in support of that opinion.
>Pieces of information differentiated, not by the object or subject, but
>by their inferential relations internal to a body of knowledge:
Maybe you could start by defining "piece of information".
> a piece
>of knowledge has not only an origin but an epistemic *standing* as part
>of a more-or-less-systematic body of interconnected propositions.
A definition of "piece of knowledge" would also be useful at this
point.
>Knowledge's objectivity derives from this, that it is formally checked
>only by other pieces of knowledge, forming an independent unity
>separate from psychological states:
Why is that a basis for objectivity, rather than a basis for
solipsism?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.3.91 (SunOS)
iD8DBQFByfqCvmGe70vHPUMRArA0AKDrEUvnYtqALorAJ97bCa8NEs5HkACg5uEG
byqwudZjcPH4LWJp8tfAnAk=
=3LY/
-----END PGP SIGNATURE-----
patty
... that it (the system) emits interpretive behavior.
Interpretive behavior is acting as if marks (or precesses) in one
context stand for marks (or processes) in another context. For example
the mark "rock" might stand in this message for that rock over there.
Of course without some agent making that connection, then there is no
interpretive behavior; and that is the point. Now the marks that are
interpreted need not be external; they may be private behaviors that
only one agent can access. For example there are private processes in
my head now that i am interpreting as "knowing that someone has broken
my window with a rock". You, however, can take the quoted marks in my
last sentence as evidence that i know that my window was broken by a
rock. The private processes stand in my head for the dastardly act that
happened to my window, and i interpret those processes as knowing this
event actually happened. You however will need to use your
interpretation of my quoted string to represent this vandalism.
Unfortunately i have the broken glass itself to deal with.
PS: Note that there are 855,000 results from using the search string
"knowledge representation" with the quote marks and that each of them
has those two words concatenated exactly as in the string; but there are
8,590,000 when you don't enclose the search in quotes. And for
"knowledge representation" anded to Quine anded Longley we have only 8.
<http://www.google.com/search?q=%22knowledge+representation%22+Quine+Longley>
Sorry, i just wanted to be painfully clear about this.
patty
Skup
> Jeff Rubard wrote:
> [...]>
>
>> Firstly, knowledge requires a knower to have such a structure as to
>> unobtrusively register epistemic facts: [...]
>
>
> Why should there be a knower? There is no reason to suppose that knowing
> requires a knower - anymore than raining requires a rainer.
>
> The problem with verbs is that they refer to actions, processes, states
> of being, changes in states of being, logical functions, and so on. Just
> which is the case with "know" is a matter of investigation, not assumption.
>
Personally, I would have to agree with you. Knowledge is simply a
collection of facts. However... For something to be known, then there
needs to be a knower. There really is a spin you need to put on the
term to come across with the meaning you intend. Humans as a whole have
a vast amount of knowledge we've gained over the years, centuries,
millenia, etc. No one person can claim to posess all of the knowlege.
At the same time, it's all still knowlege. It may be stored "off-line",
by ways of books, video, audio tape, backup media?, etc.
> Attempt to describe the issue in terms of knowing as a process rather
> than a knower that acts. Then the problem is both simpler and more
> difficult. It's simpler because you don't have to worry about the
> relation between knower, knowing, and knowledge - they're all merely
> grammatically different terms for the same thing. It's more difficult
> because our language makes it difficult to talk about any process
> without implying a processor. But when you consider what we expect from
> Jim when we ask questions about his "knowledge", it's clear that what
> Jim does, and only what Jim does, is what will satisfy us that he
> actually knows something or other. Note that although Jim may claim that
> he knows something, this is insufficient to satisfy - we want Jim to do
> a lot more than utter a claim.
>
> So I would start with "What is the evidence that some system knows
> something or other?" Work back from that. Eg, what makes me conclude
> that Jim knows math? By how he solves equations, constructs proofs,
> applies formulas, analyses problems numerically or geometrically or etc,
> and so on. An AI system that can do the sorts of things Jim does to
> prove his claim that he knows math would be quite useful IMO.
>
Though... I think if we had an effective AI to solve math problems...
Mathmatics scores in public schools would definately take a hit. :)
> Would such a system need "knowledge representation"? Maybe. It's not
> obvious one way or the other IMO. And I know an immense amount of work
> has been done knowledge representation, especially on common sense and
> contextual knowledge. But it doesn't seem to be getting anywhere. There
> is something else at work in a human that "knows" what the sentence "I
> thought we could go out" means.
>
I think "knowledge representation" would not be required, but to stay
inline with my own thought process, it would. Maybe it's just me, but
when I think of an apple, I can visualize the apple. I can tell you the
most common color, I can recall the way it tastes, the way it smells,
and how good a hot cup of cider is in the fall. All of these are
triggered by the mental representation of my knowledge of an apple.
I suppose this is the one of the reasons this group exists. Open
discussion of this type of problem. How does one apply thier concept of
*thought* to the concept of AI? ...or, better yet, *should* you apply
your own thought process to it? Generally... What any one person can
think up, a group of people can tear apart. :)
> To put it another way: we humans are pretty good at sussing each other
> intensions -- ordinary dialogue consists to a very large extent of
> incomplete utterances that require the partners to grasp each other's
> intensions. How do we do it? Not by "knowledge representation", IMO.
>
An excellant example of this is any of the over-used catch-all slang
terms. i.e. dude, buddy, etc.
> To put it still another way: A search engine will be "smart" when it's
> able to provide me not only with docs that contain the search string,
> but also with docs that relate semantically to the search string. And
> just generic (dictionary based?) semantic relationships, but specific to
> me and my interests, ie, the context in which I am searching for
> information. (BTW, I dislike that Google drops "and" in my search
> strings - sometimes "and" is the crucial word distinguishing two
> otherwise nearly identical compound names. By including the "and" there
> would be far fewer hits for me to wade through. The writers of Google
> need some serious training in language and literature. NB how important
> the "and" is in that final noun phrase! Oh, I know I can use Intext: to
> force the inclusion of the "and", but inclusion should be the default,
> not the work-around.)
Oddly enough, internet search is the main thing I do in the 9-5 world.
If you include the AND as a required part of the search, if it's not a
literal string, you will get a bunch of noise. If you make the literal
term a default action, you won't be permitte the same amount of laziness
when searching for *rambo movie review*. Pages with a hit like "Joe
Blow's Review of the Movie Rambo" would totally miss. Generally
speaking, the easier software is to use, the less power the user is given.
--
--Skup
Good judgement comes from bad experience. Alot of that comes from bad
judgement...
patty
> Wolf Kirchmeir wrote:
>
>> Jeff Rubard wrote:
>> [...]>
>>
>>> Firstly, knowledge requires a knower to have such a structure as to
>>> unobtrusively register epistemic facts: [...]
>>
>>
>>
>> Why should there be a knower? There is no reason to suppose that
>> knowing requires a knower - anymore than raining requires a rainer.
>>
>> The problem with verbs is that they refer to actions, processes,
>> states of being, changes in states of being, logical functions, and so
>> on. Just which is the case with "know" is a matter of investigation,
>> not assumption.
>>
>
> Personally, I would have to agree with you. Knowledge is simply a
> collection of facts. However... For something to be known, then there
> needs to be a knower. There really is a spin you need to put on the
> term to come across with the meaning you intend. Humans as a whole have
> a vast amount of knowledge we've gained over the years, centuries,
> millenia, etc. No one person can claim to posess all of the knowlege.
> At the same time, it's all still knowlege. It may be stored "off-line",
> by ways of books, video, audio tape, backup media?, etc.
>
My point would be that the marks in the books are not a collection of
facts, nor are they knowledge. If every human alive were wiped off the
face of the earth, these marks would be nothing more than noumenal
scratches. To know that "X" is to act in relationship to X such that
"X" stands for you as an interpreter in casual relationship to X. In
other words: (1) "X" represents X to you, (2) "X" is casually related to
X, and (3) your actions (publicly and privately) are consistent with (1)
and (2). Sorry, im working on the wording of that, perhaps someone can
help.
Happily Google had enough serious training in language, literature, and
information retrieval to give us the best or both worlds: the default
anding of terms combined with explicit quoting of literal strings ...
power *and* ease of use :) You can even do not(s), the only thing you
can't do it or(s) ... but just make multiple retrievals for that.
patty
Jeff Rubard
> Jeff Rubard wrote:
> [...]>
>
>> Firstly, knowledge requires a knower to have such a structure as to
>> unobtrusively register epistemic facts: [...]
>
>
> Why should there be a knower? There is no reason to suppose that knowing
> requires a knower - anymore than raining requires a rainer.
The argument is actually that a "knower" is not what makes knowledge
knowledge: knowledge is independent of subjectivity in the strong
(epistemological) sense of being independent of psychological states as
opposed to weak (metaphysical) sense of being observer-independent.
> The problem with verbs is that they refer to actions, processes, states
> of being, changes in states of being, logical functions, and so on. Just
> which is the case with "know" is a matter of investigation, not assumption.
For grammatical completeness, it seems convenient to say that the result
of a successful "implementation" of knowledge is a knower.
> Attempt to describe the issue in terms of knowing as a process rather
> than a knower that acts. Then the problem is both simpler and more
> difficult. It's simpler because you don't have to worry about the
> relation between knower, knowing, and knowledge - they're all merely
> grammatically different terms for the same thing. It's more difficult
> because our language makes it difficult to talk about any process
> without implying a processor. But when you consider what we expect from
> Jim when we ask questions about his "knowledge", it's clear that what
> Jim does, and only what Jim does, is what will satisfy us that he
> actually knows something or other. Note that although Jim may claim that
> he knows something, this is insufficient to satisfy - we want Jim to do
> a lot more than utter a claim.
I agree with the Lichtenbergian ("It thinks") character of this account,
but to describe knowledge as processual seems to introduce adventitious
metaphysics. Since most things have some processual characteristics, it
seems anything good enough to count as knowing something will too -- but
as for knowledge itself, I am not sure which characteristics those would
be.
> So I would start with "What is the evidence that some system knows
> something or other?" Work back from that. Eg, what makes me conclude
> that Jim knows math? By how he solves equations, constructs proofs,
> applies formulas, analyses problems numerically or geometrically or etc,
> and so on. An AI system that can do the sorts of things Jim does to
> prove his claim that he knows math would be quite useful IMO.
>
> Would such a system need "knowledge representation"? Maybe. It's not
> obvious one way or the other IMO. And I know an immense amount of work
> has been done knowledge representation, especially on common sense and
> contextual knowledge. But it doesn't seem to be getting anywhere. There
> is something else at work in a human that "knows" what the sentence "I
> thought we could go out" means.
It seems unfortunate, though, to link epistemology to semantics quite so
strongly: much as whatever is good enough to count as knowledge will fit
into some process ontology, anything close enough to the knowledge
possessed by humans as to have strong similarities to an conscious state
possessing intentionality will have some semantic elements. The attempt,
however, is to avoid the "unified school district" approach of focusing
on intentionality and to highlight some of knowledge's distinctive
characteristics. "Knowledge representation" seems to have the right
idea, in that in knowledge (if not in mental states generally as per the
language of thought hypothesis) we are dealing with a sui generis mode
of organizing the mind, one which can afford to be taken singly for
analytical purposes.
--
Jeff Rubard
http://opensentence.tripod.com/
Essays on theory, culture, and politics
Wolf Kirchmeir
> Wolf Kirchmeir wrote:
>
>>
>> Why should there be a knower? There is no reason to suppose that
>> knowing requires a knower - anymore than raining requires a rainer.
>>
>
> Can you give an example of knowing without a knower ?
Depends on how you define "knowing", doesn't it? That was my point. If
you define knowing in such a way that you have knowing only with
knowers, you've begged the question.
The issue is not whether I can give an example of "knowing without
knowers", but whether "knowing requires knowers" is a true statement. I
just don't know. Do you?
Actually, it's a version of Descartes' fallacy - that thinking requires
thinkers. Ie, he defined "thinking" in such a way that it was a property
of a conscious self, wqhich means his "Cogito, ergo sum" begs the
question. BTW, Descartes went on to use this "proof" to argue that
animals were mere machines with no "soul," by which he presumably meant
whatever it is that thinks.
By saying knowing requires a knower, you are doing the same thing: you
are defining knowing as the property of a knower. Why?
I'm with Yeats: "Who can tell the dancer from the dance?"
An' if you don't dig it, 'splaining ain't no use to you.
Wolf Kirchmeir
Dunno, but I do know it's not the same as the physical requirements for
being able to say "I am a knower."
[snip "Jim doing math" as an example]>
>
> May I suggest that the physical requirements for doing maths
> is Jim and that is what we mean objectively by the "knower"?
IOW, for any X that Jim can do, the "physical requirements for doing X"
must be Jim? H'm. In that case, when Jim is doing math, he is a
"mather", I suppose.
I have a camera that is much better than I am in figuring the correct
exposure for a phtograph. It "knows" how to do that because it's built
to acquire information and calculate outputs which control the camera
settings. I presume you would be willing to say it's what we mean
objectively by the "knower."
[...]
"Know" (and its grammatical variants)is too subjective a term to be of
much use, I think.
Jeff Rubard
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> "Jeff Rubard" <jeffr...@online.ie> writes:
>
>
>>It seems an important point to me that the structure of knowledge is
>>discrete:
>
>
> Whereas it seems obvious to me that knowledge is not discrete.
>
Knowledge is continuous? There are no points of separation between
things known? Facts don't come singly?
>
>>Firstly, knowledge requires a knower to have such a structure as to
>>unobtrusively register epistemic facts:
>
>
> What is a epistemic fact, and what distinguishes that from an
> ordinary fact?
>
A loose usage, intended to focus attention on the epistemological status
of knowledge rather than "nuts and bolts" considerations of "knowledge
representation". It is an epistemic fact that I am sitting in front of
the computer, as this is both true and relevant for consideration of my
epistemic status.
>> if the subject was such as to
>>psychologically constitute all its information about the world, no
>>piece of that information could constitute a piece of knowledge for the
>>reason that all which was contained was information about a subject
>>continuous in all respects relevant to epistemic assessment (serving
>>not only as in a certain way the object of the information but also as
>>its epistemically undifferentiated matter).
>
>
> Hmm. Can somebody translate that into English?
>
If it really was just "all in your head", there would be no knowledge.
Knowledge is of things "outside your head", that are identified using a
schematism of the world contributing the relevant standards of
objectivity via restrictions on what counts as a piece of knowledge.
>
>> In other words, sense
>>cannot constitute knowledge for the reason that there is no
>>individuation of known contents in such a way as to make their
>>epistemic assessment possible: a piece of knowledge requires not only
>>an object, but criteria indicating what piece of knowledge it is, and
>>these cannot be supplied by the subject.
>
>
> I could use a translation of that, too.
>
Knowledge isn't just what you perceive, it places several relevant
restrictions on perception such that the percepts come to resemble
objects meeting certain epistemic criteria of knowability. You don't
"know" your pains.
>
>>So empirical idealism is a nonstarter for the reason that it is unclear
>>what we are talking about when we attribute knowledge of perception to
>>the idealist subject. What would count as knowledge on this score?
>
>
> I think you are just expressing your opinion as opposed to empirical
> idealism, and you are spewing a pile of sophistry as an alleged
> argument in support of that opinion.
>
It's an exercise in self-clarification.
>
>>Pieces of information differentiated, not by the object or subject, but
>>by their inferential relations internal to a body of knowledge:
>
>
> Maybe you could start by defining "piece of information".
>
A transaction with the external world meaningful to a knower.
"Information" in this sense is defined, not by a mathematical theory of
communication, but by its polymorphous characteristics: all manner of
things are ranked as "information" in the colloquial sense.
>> a piece
>>of knowledge has not only an origin but an epistemic *standing* as part
>>of a more-or-less-systematic body of interconnected propositions.
>
>
> A definition of "piece of knowledge" would also be useful at this
> point.
>
A thing known. On this interpretation, the term does not permit of a
stricter definition since knowledge is not usefully reducible to, say,
physical things.
>
>>Knowledge's objectivity derives from this, that it is formally checked
>>only by other pieces of knowledge, forming an independent unity
>>separate from psychological states:
>
>
> Why is that a basis for objectivity, rather than a basis for
> solipsism?
Because of the restrictions placed on subjectivity by knowledge, see above.
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.3.91 (SunOS)
>
> iD8DBQFByfqCvmGe70vHPUMRArA0AKDrEUvnYtqALorAJ97bCa8NEs5HkACg5uEG
> byqwudZjcPH4LWJp8tfAnAk=
> =3LY/
> -----END PGP SIGNATURE-----
>
--
Jeff Rubard
http://opensentence.tripod.com/
Essays on theory, culture, and politics
Wolf Kirchmeir
> Wolf Kirchmeir wrote:
>
>> Jeff Rubard wrote:
>> [...]>
>>
>>> Firstly, knowledge requires a knower to have such a structure as to
>>> unobtrusively register epistemic facts: [...]
>>
>>
>>
>> Why should there be a knower? There is no reason to suppose that
>> knowing requires a knower - anymore than raining requires a rainer.
>>
>> The problem with verbs is that they refer to actions, processes,
>> states of being, changes in states of being, logical functions, and so
>> on. Just which is the case with "know" is a matter of investigation,
>> not assumption.
>>
>
> Personally, I would have to agree with you. Knowledge is simply a
> collection of facts. However... For something to be known, then there
> needs to be a knower. There really is a spin you need to put on the
> term to come across with the meaning you intend. Humans as a whole have
> a vast amount of knowledge we've gained over the years, centuries,
> millenia, etc. No one person can claim to posess all of the knowlege.
> At the same time, it's all still knowlege. It may be stored "off-line",
> by ways of books, video, audio tape, backup media?, etc.
>
[...]
No, what's stored are artefacts. They aren't knowledge unless someone
can interpret them. IOW, _knowing_ is what counts, not the stored
artefacts. Look at Etruscan - no on knows how to read it, hence the
"knowledge stored in Etruscan inscriptions is forever lost," as more
than one commentator has opined in the usual cliches.
Just as the song exists only while singing is going on, knowledge exists
only while knowing is going on. Damn, but it's hard to talk about this
without implying what one doesn't want to imply! I'll quote Yeats again:
Wolf Kirchmeir
[...]>
>
> I think "knowledge representation" would not be required, but to stay
> inline with my own thought process, it would. Maybe it's just me, but
> when I think of an apple, I can visualize the apple. I can tell you the
> most common color, I can recall the way it tastes, the way it smells,
> and how good a hot cup of cider is in the fall. All of these are
> triggered by the mental representation of my knowledge of an apple.
Well, I would say that the "mental representation" is all that behaviour
you describe "I" to be doing. And all that behaviour is triggered by
something else - in this case, by your reading about "knowledge
representation", which you must have discussed in just those trems of
recalling an apple that you just used. (I could give a summnary accouint
of self-conditioning, and chained behaviours, etc, but I'll spare you.)
> I suppose this is the one of the reasons this group exists. Open
> discussion of this type of problem. How does one apply thier concept of
> *thought* to the concept of AI? ...or, better yet, *should* you apply
> your own thought process to it? Generally... What any one person can
> think up, a group of people can tear apart. :)
Um, yes. But some ideas are made of rip-stop nylon... :-)
Neil W Rickert
Hash: SHA1
Jeff Rubard <jeffr...@online.ie> writes:
>Neil W Rickert wrote:
>> "Jeff Rubard" <jeffr...@online.ie> writes:
>>>It seems an important point to me that the structure of knowledge is
>>>discrete:
>> Whereas it seems obvious to me that knowledge is not discrete.
>Knowledge is continuous? There are no points of separation between
>things known? Facts don't come singly?
Discrete and continous are not the only possibilities.
I don't agree that knowledge is a collection of facts. The term
"things known" is hopelessly vague.
Knowledge is what you acquire through learning. You don't learn in
discrete chunks. Learning often seems to be more of an accretive
process.
>>>Firstly, knowledge requires a knower to have such a structure as to
>>>unobtrusively register epistemic facts:
>> What is a epistemic fact, and what distinguishes that from an
>> ordinary fact?
>A loose usage, intended to focus attention on the epistemological status
>of knowledge rather than "nuts and bolts" considerations of "knowledge
>representation". It is an epistemic fact that I am sitting in front of
>the computer, as this is both true and relevant for consideration of my
>epistemic status.
Okay. So by "loose usage", you mean that you tossed in a lot of
superfluous verbiage so that it might impress. The down side is that
others will recognize it as sophistry.
>>> if the subject was such as to
>>>psychologically constitute all its information about the world, no
>>>piece of that information could constitute a piece of knowledge for the
>>>reason that all which was contained was information about a subject
>>>continuous in all respects relevant to epistemic assessment (serving
>>>not only as in a certain way the object of the information but also as
>>>its epistemically undifferentiated matter).
>> Hmm. Can somebody translate that into English?
>If it really was just "all in your head", there would be no knowledge.
>Knowledge is of things "outside your head", that are identified using a
>schematism of the world contributing the relevant standards of
>objectivity via restrictions on what counts as a piece of knowledge.
I'm not sure that is the correct translation.
In any case, here you insist that it is of things outside the head.
But later, you say:
>>>Knowledge's objectivity derives from this, that it is formally checked
>>>only by other pieces of knowledge, forming an independent unity
>>>separate from psychological states:
That seems to be only an internal criterion, and thus subject to my
earlier criticism of solipsism.
>> Maybe you could start by defining "piece of information".
>A transaction with the external world meaningful to a knower.
This appears to be circular.
>"Information" in this sense is defined, not by a mathematical theory of
>communication, but by its polymorphous characteristics: all manner of
>things are ranked as "information" in the colloquial sense.
You are playing pointless word games.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.3.91 (SunOS)
iD8DBQFByjgzvmGe70vHPUMRArOaAKC6ZqCv0Epp1xBKanm/SkxjkAJA7wCeJQiV
4fJfXfph+gss+Db6AHaYPb8=
=Suzs
-----END PGP SIGNATURE-----
Wolf Kirchmeir
> Wolf Kirchmeir wrote:
[...]
>
> For grammatical completeness, it seems convenient to say that the result
> of a successful "implementation" of knowledge is a knower.
Grammatical completeness is not a very convincing reason. If we spoke a
different language, we might not need this particular grammatical
completeness.... For a sense of just how grammar can interfere with
clear conceptualisation, go to sci.lang, and observe the arguments about
whether or not some language is more or less logical than another -
based on the differeing kinds of "grammatical completeness" forced on
the speakers of the different languages, of course. :-)
>> Attempt to describe the issue in terms of knowing as a process rather
>> than a knower that acts. Then the problem is both simpler and more
>> difficult. It's simpler because you don't have to worry about the
>> relation between knower, knowing, and knowledge - they're all merely
>> grammatically different terms for the same thing. It's more difficult
>> because our language makes it difficult to talk about any process
>> without implying a processor. But when you consider what we expect
>> from Jim when we ask questions about his "knowledge", it's clear that
>> what Jim does, and only what Jim does, is what will satisfy us that he
>> actually knows something or other. Note that although Jim may claim
>> that he knows something, this is insufficient to satisfy - we want Jim
>> to do a lot more than utter a claim.
>
>
> I agree with the Lichtenbergian ("It thinks") character of this account,
> but to describe knowledge as processual seems to introduce adventitious
> metaphysics. Since most things have some processual characteristics, it
> seems anything good enough to count as knowing something will too -- but
> as for knowledge itself, I am not sure which characteristics those would
> be.
Well, that unceratinty may arise form the fact that "knowledge" is
another otem of "grammatical completness." In English, we can usually
build at least three of four types of words on a given base:
i) the verb referring to some action/activity/process
ii) a noun denoting the actor/processor
iii) a noun denoting the product/result of the action/process
iv) a noun denoting the action/activity/process
Examples (note that the process-noun and gerundial endings are used
ambiguously - that should be enough of a warning sign, IMO.)
educate, educator, education, education
conspire, conspirator, conspiracy, conspiracy
magnify, magnifier, magnification, magnification
know, knower, knowledge, knowing
grow, grower, growth, growing
paint, painter, painting, painting
teach, teacher, teaching, teaching
shape, shaper, shape, shaping
view, viewer, view, viewing
act, actor, action, acting/action
eat, eater, ---, eating
sail, sailor, ---, sailing
rain, ---, rain, raining
glow, ---, glow, glowing
Etc.
A comprehensive list would display some interesting relationships
between semantics and the words built on the base. It should also raise
a number of questions.
>> So I would start with "What is the evidence that some system knows
>> something or other?" Work back from that. Eg, what makes me conclude
>> that Jim knows math? By how he solves equations, constructs proofs,
>> applies formulas, analyses problems numerically or geometrically or
>> etc, and so on. An AI system that can do the sorts of things Jim does
>> to prove his claim that he knows math would be quite useful IMO.
>>
>> Would such a system need "knowledge representation"? Maybe. It's not
>> obvious one way or the other IMO. And I know an immense amount of work
>> has been done knowledge representation, especially on common sense and
>> contextual knowledge. But it doesn't seem to be getting anywhere.
>> There is something else at work in a human that "knows" what the
>> sentence "I thought we could go out" means.
>
>
> It seems unfortunate, though, to link epistemology to semantics quite so
> strongly: much as whatever is good enough to count as knowledge will fit
> into some process ontology, anything close enough to the knowledge
> possessed by humans as to have strong similarities to an conscious state
> possessing intentionality will have some semantic elements. The attempt,
> however, is to avoid the "unified school district" approach of focusing
> on intentionality and to highlight some of knowledge's distinctive
> characteristics. "Knowledge representation" seems to have the right
> idea, in that in knowledge (if not in mental states generally as per the
> language of thought hypothesis) we are dealing with a sui generis mode
> of organizing the mind, one which can afford to be taken singly for
> analytical purposes.
The problem is that if you assume "knowledge representation", you assume
that knowledge is something apart from the knower, in which case you
also need some process of decoding the representation. But the decoded
entity is also a representation, so you need a decoder for that; and so
on. I don't like regresses of this kind - they tend to go off into the
wild blue never-never.
JGCASEY
Surely they both count? You can't have one without the other.
What is the point of all this nit-picking?
> Look at Etruscan - no on knows how to read it, hence the
> "knowledge stored in Etruscan inscriptions is forever lost," as more
> than one commentator has opined in the usual cliches.
So knowledge (patterns of some kind) have no meaning without
the context of a physical translator. 6502 code requires a
6502 cpu to give it any kind of meaning. Etruscan inscriptions
require an Etruscanian translator. It seems a pointless
observation for anyone wanting to make a physical machine
capable of processing patterns in some useful (goal orientated)
manner? Of course knowledge in its static form has no meaning
without something being wired up to make use of it. It seems
a statement of the obvious.
> Just as the song exists only while singing is going on, knowledge
exists
> only while knowing is going on. Damn, but it's hard to talk about
this
> without implying what one doesn't want to imply! I'll quote Yeats
again:
> "Who can tell the dancer from the dance?"
Trip the dancer up and there will be the dancer
without the dance.
I have no trouble distinguishing between the person
doing the dance (dancer) and the dance itself. The
dance can be done by any person who has the physical
requirements to do the dance. The dance itself is not
dependent on any particular person only that that
person has the physical requirements to dance.
A dancer can exist without the dance providing that
the potential to dance exists. Just as a 6502 processor
can exist without running a program providing it has
the potential to run 6502 code, that is, there is a
physical requirement in the cpu chip for it to be called
a 6502 processor. There is a physical requirement for
a person to be called a dancer.
There is a physical requirement to be someone who
understands Etrucscan inscriptions.
You can bounce your verbs and nouns around as much
as you like at the end of the day we have a physical
system that needs to be described and understood.
What are the physical requirements for someone to
dance? What are the physical requirements for a
machine to carry out some intelligent action?
I guess your reply, if you have read this far, will be,
"An' if you don't dig it, 'splaining ain't no use to you."
:-)
John Casey
JPL Verhey
"Wolf Kirchmeir" <wwol...@sympatico.ca> wrote in message
news:Horyd.28237$GK5.1...@news20.bellglobal.com...
But maybe this explains why we can go on discussing things forever :-)
"The only constant is change". When the knower knows this.. for a moment
s/he feels eternal, and it changed him/her forever.. Change will go on
and on and on...until we regressed into our graves. I assume that life,
organic and inorganic, will go on though.
"Life is not a creation, but a creativity"
(-)
Lester Zick
<wwol...@sympatico.ca> in comp.ai.philosophy wrote:
>Jeff Rubard wrote:
>[...]>
>> Firstly, knowledge requires a knower to have such a structure as to
>> unobtrusively register epistemic facts: [...]
>
>Why should there be a knower? There is no reason to suppose that knowing
>requires a knower - anymore than raining requires a rainer.
Technically, clouds are rainers; low pressure systems are clouders;
high pressure systems are low pressure systemers; and low pressure
systemers are high pressure systemers.
>The problem with verbs is that they refer to actions, processes, states
>of being, changes in states of being, logical functions, and so on. Just
>which is the case with "know" is a matter of investigation, not assumption.
Investigation always begins with assumptions to be investigated.
Regards - Lester
Lester Zick
<wwol...@sympatico.ca> in comp.ai.philosophy wrote:
>patty wrote:
>> Wolf Kirchmeir wrote:
>>
>>>
>>> Why should there be a knower? There is no reason to suppose that
>>> knowing requires a knower - anymore than raining requires a rainer.
>>>
>>
>> Can you give an example of knowing without a knower ?
>
>Depends on how you define "knowing", doesn't it? That was my point. If
>you define knowing in such a way that you have knowing only with
>knowers, you've begged the question.
Not unless you contend that knowers and knowledge are the same. The
only issue is whether knowing has a cause and if so what the nature of
that cause is. If knowing has no cause, it almost certainly has no
knowers. If knowing has a cause, it has knowers of whatever form.
>The issue is not whether I can give an example of "knowing without
>knowers", but whether "knowing requires knowers" is a true statement. I
>just don't know. Do you?
Depends on what you mean by knowers. If you don't know what you
mean by knowers, then you'll have some difficulty knowing the answer
to your own question. Of course, if you regress the problem of knowing
and knowers to processes and processors, you'll find the same problem
awaiting you there.
Regards - Lester
Lester Zick
comp.ai.philosophy wrote:
[. . .]
>Firstly, knowledge requires a knower . . .
Sounds like the beginning of a Socratic exercise in futility. I can't
think of a better illustration of analogical reasoning where in this
Xmas season (to paraphrase Bart Simpson) men of all faiths come
together to worship Jesus H. Christ, they do so by telling endless
series of apocryphal stories of what is what according to other
endless series of stories of what is what, world without end, amen.
Regards - Lester
JPL Verhey
"Skup" <tarz...@remove.yahoo.com> wrote in message
news:wYCdnbpP3IF...@wideopenwest.com...
...
> I think "knowledge representation" would not be required, but to stay
> inline with my own thought process, it would. Maybe it's just me, but
> when I think of an apple, I can visualize the apple. I can tell you
> the most common color, I can recall the way it tastes, the way it
> smells, and how good a hot cup of cider is in the fall. All of these
> are triggered by the mental representation of my knowledge of an
> apple.
" Consciousness is knowing that you think,
science is thinking that you know. "
I think..:)
Jeff Rubard
> Jeff Rubard wrote:
[...]
>>
>> so strongly: much as whatever is good enough to count as knowledge
>> will fit into some process ontology, anything close enough to the
>> knowledge possessed by humans as to have strong similarities to an
>> conscious state possessing intentionality will have some semantic
>> elements. The attempt, however, is to avoid the "unified school
>> district" approach of focusing on intentionality and to highlight some
>> of knowledge's distinctive characteristics. "Knowledge representation"
>> seems to have the right idea, in that in knowledge (if not in mental
>> states generally as per the language of thought hypothesis) we are
>> dealing with a sui generis mode of organizing the mind, one which can
>> afford to be taken singly for analytical purposes.
>
>
> The problem is that if you assume "knowledge representation", you assume
> that knowledge is something apart from the knower, in which case you
> also need some process of decoding the representation. But the decoded
> entity is also a representation, so you need a decoder for that; and so
> on. I don't like regresses of this kind - they tend to go off into the
> wild blue never-never.
The idea of the original post was that the knowing subject is
"discreet", that is to say not very much subjectivity is involved in
knowledge; I think in retrospect I detect some regress-blocking in this
concept, for the reason that it does not require "encoding" per se. That
is to say, the level which is eliminated is "decoding" of knowledge for
an inner realm of subjectivity. To explain, if knowledge is independent
of pure subjectivity, sense, then there's no reason to "translate"
knowledge into the idiom of subjectivity: knowledge is held by the
subject "in public", as it were, using shared idioms.
Jeff Rubard
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Jeff Rubard <jeffr...@online.ie> writes:
>
>>Neil W Rickert wrote:
>>
>>>"Jeff Rubard" <jeffr...@online.ie> writes:
>
>
>>>>It seems an important point to me that the structure of knowledge is
>>>>discrete:
>
>
>>>Whereas it seems obvious to me that knowledge is not discrete.
>
>
>>Knowledge is continuous? There are no points of separation between
>>things known? Facts don't come singly?
>
>
> Discrete and continous are not the only possibilities.
What are the other ones?
> I don't agree that knowledge is a collection of facts. The term
> "things known" is hopelessly vague.
>
> Knowledge is what you acquire through learning. You don't learn in
> discrete chunks. Learning often seems to be more of an accretive
> process.
>
Interesting point, but there's no reason there has to be a homology
between the process by which a thing comes to be known and resulting
knowledge.
>>>>Firstly, knowledge requires a knower to have such a structure as to
>>>>unobtrusively register epistemic facts:
>
>
>>>What is a epistemic fact, and what distinguishes that from an
>>>ordinary fact?
>
>
>>A loose usage, intended to focus attention on the epistemological status
>>of knowledge rather than "nuts and bolts" considerations of "knowledge
>>representation". It is an epistemic fact that I am sitting in front of
>>the computer, as this is both true and relevant for consideration of my
>>epistemic status.
>
>
> Okay. So by "loose usage", you mean that you tossed in a lot of
> superfluous verbiage so that it might impress. The down side is that
> others will recognize it as sophistry.
>
Perhaps on some measure one word ("epistemic") counts as "a lot of
verbiage", and perhaps on some level epistemology is an attempt to
impress upon someone the non-negotiability of certain claims. But other
than flagging the discussion as epistemological, I'm not sure exactly
what I did, even if it would be impressive enough for sophistry.
>>>> if the subject was such as to
>>>>psychologically constitute all its information about the world, no
>>>>piece of that information could constitute a piece of knowledge for the
>>>>reason that all which was contained was information about a subject
>>>>continuous in all respects relevant to epistemic assessment (serving
>>>>not only as in a certain way the object of the information but also as
>>>>its epistemically undifferentiated matter).
>
>
>>>Hmm. Can somebody translate that into English?
>
>
>>If it really was just "all in your head", there would be no knowledge.
>>Knowledge is of things "outside your head", that are identified using a
>>schematism of the world contributing the relevant standards of
>>objectivity via restrictions on what counts as a piece of knowledge.
>
>
> I'm not sure that is the correct translation.
>
> In any case, here you insist that it is of things outside the head.
> But later, you say:
>
>
>>>>Knowledge's objectivity derives from this, that it is formally checked
>>>>only by other pieces of knowledge, forming an independent unity
>>>>separate from psychological states:
>
>
> That seems to be only an internal criterion, and thus subject to my
> earlier criticism of solipsism.
>
The criteria are intended to be the same across the two definitions: the
"schematism of the world" would be what defines something as knowledge,
and the argument is that such a definition is internal to epistemology
(that is, does not involve defining knowledge in terms of other entities).
>>>Maybe you could start by defining "piece of information".
>
>
>>A transaction with the external world meaningful to a knower.
>
>
> This appears to be circular.
>
Here I'm defining information in terms of meaning, not the other way
around (as is more common).
>>"Information" in this sense is defined, not by a mathematical theory of
>>communication, but by its polymorphous characteristics: all manner of
>>things are ranked as "information" in the colloquial sense.
>
>
> You are playing pointless word games.
Take these comments as what they are: epistemological clarification,
which is not such a very advanced art.
Neil W Rickert
Hash: SHA1
Jeff Rubard <jeffr...@online.ie> writes:
>Neil W Rickert wrote:
>>>>Whereas it seems obvious to me that knowledge is not discrete.
>>>Knowledge is continuous? There are no points of separation between
>>>things known? Facts don't come singly?
>> Discrete and continous are not the only possibilities.
>What are the other ones?
It could consider of several distinct parts, each of which is
continuous.
To give a different example, consider oatmeal porridge. It isn't
discrete. But it isn't exactly continuous. It is lumpy.
I won't continue the discussion on other topics. If it wasn't
already apparent, I'll just note that I am skeptical of that branch
of philosophy known as "epistemology". Knowledge and knowledge
acquisition have been the subject of empirical studies, such as those
of Piaget. We should pay attention to such empirical studies in
preference to the writings of philosophers.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.3.91 (SunOS)
iD8DBQFBy4fVvmGe70vHPUMRAiAjAKCLqkTKNdrTf3K+LfQKb/fzRwmg4gCfXX3h
VD//lk9NRkZNJbaHka91kJE=
=psCD
-----END PGP SIGNATURE-----
patty
> patty wrote:
>
>> Wolf Kirchmeir wrote:
>>
>>>
>>> Why should there be a knower? There is no reason to suppose that
>>> knowing requires a knower - anymore than raining requires a rainer.
>>>
>>
>> Can you give an example of knowing without a knower ?
>
>
> Depends on how you define "knowing", doesn't it? That was my point. If
> you define knowing in such a way that you have knowing only with
> knowers, you've begged the question.
>
Well i though of an example. Suppose you have a species where all
know-how is accessible to all instances; in other words, private
behavior was impossible for the species. There would be no difference
between what one "individual knew" and what any other "individual" knew.
To say "patty knows X" would be redundant, this animal needs only say
"knows X".
> The issue is not whether I can give an example of "knowing without
> knowers", but whether "knowing requires knowers" is a true statement. I
> just don't know. Do you?
>
Yes, knowing requires a knower. Make statements like "knows X" in any
venue and nobody will be able to use your information; except that maybe
they will assume some implied agent, like for example whatever group of
people that are present at the moment.
> Actually, it's a version of Descartes' fallacy - that thinking requires
> thinkers. Ie, he defined "thinking" in such a way that it was a property
> of a conscious self, wqhich means his "Cogito, ergo sum" begs the
> question. BTW, Descartes went on to use this "proof" to argue that
> animals were mere machines with no "soul," by which he presumably meant
> whatever it is that thinks.
I think that is just a change of subject.
> By saying knowing requires a knower, you are doing the same thing: you
> are defining knowing as the property of a knower. Why?
... not, not as a "property" of a knower. We are defining "A knows X"
as a relationship between the mental activity (private behavior) of an
individual and the extension of X. Sans A we have just the
relationship between "X" and X hanging in mid air. It would be a misuse
of the concept called "know".
>
> I'm with Yeats: "Who can tell the dancer from the dance?"
Actually it's "How can we *know* the dancer from the dance?"
<http://www.emule.com/poetry/?page=poem&poem=2595>
> An' if you don't dig it, 'splaining ain't no use to you.
Any analysis ... any coding in language ... must needs partition
natural events in some manner according to the convenience of the
author. Those partitions are *not* in the events themselves (a priori);
no, they are in the interpretations of the author ... and were she
understood, those of her audience. So to look for a knowing which
"requires" knowers, a priori, is futile. Our linguistic practice has
already settled that matter for us.
But sure, i can blur my apprehension to this distinction ... i can Zen
out ignoring the natural partition between my inner activity and the
outer dance. The question is, can a behaviorist suspend his assumption
that the inner mental activity is to be ignored, long enough for him to
study the relationship between *it* and the extension of a truthful
utterance. Or is he doomed to deny that there may *be* a relationship
between something that he can measure and something that he cannot
measure.
patty
Joseph Legris
Of course raining requires a rainer. The "it" in "it is raining" is a
general reference to the status of the local weather. Knowledge and
knowing require a knower just as behaviour and behaving need a behaver.
Is that begging the question too?
--
Joe Legris
Wolf Kirchmeir
[...]
> Of course raining requires a rainer. The "it" in "it is raining" is a
> general reference to the status of the local weather. Knowledge and
> knowing require a knower just as behaviour and behaving need a behaver.
> Is that begging the question too?
>
> --
> Joe Legris
The reason we use "it" in "it is raining" is because English sentences
require subjects. Indo-european languages generally have this rule,
although there are exceptions. Eg, in Spanish, you may omit the subject
pronoun if you want - you can't do this in English. There is similar
freedom in Russian IIRC. In Russian, you can omit the copula ("be") --
you can say "I sick", while in English you must say "I am sick" -- which
causes a lot of blather about the "real" meaning of "be."
In some other languages, you don't need a subject for certain verbs. You
just say what literally translated would be "raining" or "rained", etc.
I used to know what they were, sorry, I've forgotten. Go to sci.lang and
ask for information.
Languages differ in what they must express, whether it's germane or not.
Eg, in most IE languages, you must denote the number (singular/plural)
of every noun, and every adjective modifying a noun must express the
same number. In many languages of the world you must express whether or
not a noun in the complement is an object, and what kind of object
(case); and so on. In many languages, you must denote whether or not the
referent of a noun is animate or not. And so on and so on and so on.
Much of what passes for philosophy might better be called "the higher
grammar."
IOW, the claim "of course there is a rainer" arises from grammar. Your
explanation that "it" refers to the general state of the weather may
make sense (at least to anyone with an incomplete understanding of the
vagaries of grammars), but it's not the general state of the weather
that rains. Ie, an agent noun refers to an agent, not the state of a
system. As a matter of wet fact, "raining" _is_ the state of the weather
system (more precisely, "raining" is one part of element of the system).
There is no agent.
Mind you, when I was a child, I was told a charming story about Frau
Holle, who was the snower - when she shook out the duvets of the people
who lived above the sky, the feathers that flew came to earth as snow.
ernobe
> inferential discreteness: that is to say, there can be no theory of
> knowledge suitable to the task which is not, in the end, atomistic.
It is still not clear, however, whether the epistemic standing of a piece
of knowledge is derived from it being formally checked by other pieces of
knowledge, or from its suitability to the task.
JGCASEY
Isn't the system, of which raining is one part, the agent?
Regardless of the rules of any language it is not the rules
but the meaning that counts. Include or exclude whatever you
like the meaning is either conveyed or it is not.
John Casey
Wolf Kirchmeir
> Wolf Kirchmeir wrote:
>
[...]
>
> Regardless of the rules of any language it is not the rules
> but the meaning that counts. Include or exclude whatever you
> like the meaning is either conveyed or it is not.
Well, you can say so if you like, you can say anytnbing you like, in
fact, but it's not the usual meaning of agent, which differentiates
between the agent, the action, and the result. In a weather system, they
are one and the same thing.
JGCASEY
Wolf Kirchmeir wrote:
> JGCASEY wrote:
> > Wolf Kirchmeir wrote:
> >
> [...]
> > Isn't the system, of which raining is one part, the agent?
> >
> > Regardless of the rules of any language it is not the rules
> > but the meaning that counts. Include or exclude whatever you
> > like the meaning is either conveyed or it is not.
>
> fact, but it's not the usual meaning of agent, which differentiates
> between the agent, the action, and the result. In a weather system,
> they are one and the same thing.
I don't agree. If you said "raining" I would imply a weather
system causing the rain and I think others would too. A weather
system may be "snowing" and that implies the same agent but a
different action to produce a different result.
Agent = weather system
action = mixing of warm moist air with cool air (I think)
result = rain
Agent = Jim
action = moves legs in a particular way
result = tap dance
It is the action (the inner mechanics) that is not always
known but assumed, be it for intelligent behaviour or rain.
Agent = computer
action = ?
result = winning game of chess
Agent = Einstein
action = ?
result = theory of relativity
John Casey
Joseph Legris
> Joseph Legris wrote:
> [...]
>
>> Of course raining requires a rainer. The "it" in "it is raining" is a
>> general reference to the status of the local weather. Knowledge and
>> knowing require a knower just as behaviour and behaving need a
>> behaver. Is that begging the question too?
>>
>> --
>> Joe Legris
>
Is it begging the question or not?
--
Joe Legris
patty
Sometimes i think that the restraints of one dimensional strings are too
great and that a better language would be at least 2 dimensional and
allow spirals in time ... like:
..... context ..... agent ..... action -----> result ...
.
change
.
..... result <----- action ...... agent ..... context ...
See it makes a spiral.
And there are other triples that are best coded as spirals like
context subject verb object ...
context condition action consequence ...
context patty knows x ....
context agent action result ...
One dimensional language is not very good at representing change.
patty
JPL Verhey
"JGCASEY" <jgkj...@yahoo.com.au> wrote in message
news:1104030157....@z14g2000cwz.googlegroups.com...
It appears that in this form it denotes linear causal chains: A causes
B. The line A-->B is then the action or transformation path.
But I agree with Wolf that in a weather system (any system?) there can
hardly be said to be an agent. Indeed in grammer onew can say things
that suggest agents such as "rainer", buit then there will be so many
agents possible, like "clouder", "temperature raiser" etc etc..that
grammar and reality become rather incompatible.
But when I'm pissing I do feel to be the agent. So maybe the only agent
sensible is the one with a point of view.
JGCASEY
It wasn't meant so much as a causal chain just the observation
that if something happens like raining or running there is a
physical collection of elements (agent) required for that action
to occur. There is no "running" without some multicellar organism
being required for it to happen.
The agent is the physical requirement. It is the physical details
and the physical requirement that are of interest to me, be it
raining, running or thinking. AI is about the physical requirements
to build a machine that behaves intelligently.
> But I agree with Wolf that in a weather system (any system?) there
can
> hardly be said to be an agent. Indeed in grammer onew can say things
> that suggest agents such as "rainer", buit then there will be so many
> agents possible, like "clouder", "temperature raiser" etc etc..that
> grammar and reality become rather incompatible.
Cannot an agent itself be a society of agents?
Can I say that a human cannot be an agent because they have
multiple functions and I would need an agent for each function?
> But when I'm pissing I do feel to be the agent. So maybe the only
agent
> sensible is the one with a point of view.
I am not attributing a conscious Self to a weather system.
When you "feel to be the agent" you are using the word differently
to the way I have. You have introduced the conscious Mind. And I
feel that is your rejection of a weather system being the agent of
"raining", "snowing"... as it doesn't have a Self?
So forget the definitions, can you perceive what I mean?
This comes back to what is important is not the technical structures
of different languages and their rules of use but rather what they
mean. It is this need for meaning that reveals itself in trying to
translate between languages or understand anothers point of view.
Japanese sentence --> meaning of sentence --> English sentence/s
John Casey
JPL Verhey
...
>> It appears that in this form it denotes linear causal chains: A
> causes
>> B. The line A-->B is then the action or transformation path.
>
> It wasn't meant so much as a causal chain just the observation
> that if something happens like raining or running there is a
> physical collection of elements (agent) required for that action
> to occur. There is no "running" without some multicellar organism
> being required for it to happen.
Ok, I see your point now. But I find Wolf's line of thinking still
compelling also in this case. We could say that some multicellar
organism (agent) is indeed required to exhibit, or "be in a state of"
running or any of the other identifiable activities/behaviors specific
to that (type of) agent. Also weather systems are specific agents in
that sense. But still the agent as we abstracted it conceptually, is not
different from the specific behaviors or dynamic state the system is in.
The "agent" in fact refers to those behaviors in the sense of being
identical to them. Hence the idea that for instance the agent "the
knower" refers to knowing. In most cases all this might be trivial, but
in the case of knower and knowing, or "observer observing the observed",
or "consciousness and content" ie mind-body duality it is perhaps of
special interest.
>
> The agent is the physical requirement. It is the physical details
> and the physical requirement that are of interest to me, be it
> raining, running or thinking. AI is about the physical requirements
> to build a machine that behaves intelligently.
I still don't understand the problem many have with AI, to be honest. AI
stands for "man-made intelligence" as i understand it. Man-made is clear
what it means.. and intelligence refers to certain types of behaviors
that we call intelligent behaviors. It is a man-made evaluation and
decision of what we call intelligent (machine) behavior. And all this is
not about mind or consciousness aqa strong-AI (where btw "artificial
consciousness" or AC is a better term to be used) but weak-AI. So our
world is already filled with weak-AI machines and the physical
requirements are well known and no secret. Of course there are very
simply weak-AI machines such as a thermostat but also very complicated
ones that to a degree can self-organise and adapt their behavior.
Or is your interest strong-AI/AC?
>
>> But I agree with Wolf that in a weather system (any system?) there
> can
>> hardly be said to be an agent. Indeed in grammer onew can say things
>> that suggest agents such as "rainer", buit then there will be so many
>
>> agents possible, like "clouder", "temperature raiser" etc etc..that
>> grammar and reality become rather incompatible.
>
> Cannot an agent itself be a society of agents?
Yes, in the sense of one system of activity that can be seen as a
compound of interactive subsystems.
>
> Can I say that a human cannot be an agent because they have
> multiple functions and I would need an agent for each function?
A human being consists of many identifiable subsystems many of which are
interdependant.
>
>> But when I'm pissing I do feel to be the agent. So maybe the only
> agent
>> sensible is the one with a point of view.
>
> I am not attributing a conscious Self to a weather system.
Me neither.
>
> When you "feel to be the agent" you are using the word differently
> to the way I have. You have introduced the conscious Mind. And I
> feel that is your rejection of a weather system being the agent of
> "raining", "snowing"... as it doesn't have a Self?
Not really. To me what is an agent or not merely depends on the
definition. I dropped in this discussion reading about a knower that is
knowing something. Some argued that knowing needs the agent of a knower.
Wolf pointed out, correctly imo, that they are the same.
When a human system called Jim is knowing something, we mean that as
non-Jims we see Jim exhibit certain behaviors from which we conclude
that he is indeed knowing certain somethings. When we happen to *be*
that Jim, added will be an internal / mental behavior which is called
conscious thinking which is "known-to Jim" only.
Knowing usually refers to this internal, 1st person knowing, conscious
mind. In weak-AI this is not relevant. In strong-AI/AC it is crucial. An
entirely different ball game. To *be* Jim, or Mary.. and be in a state
of conscious knowing/thinking I feel an "agent" of new sorts is needed.
>
> So forget the definitions, can you perceive what I mean?
>
> This comes back to what is important is not the technical structures
> of different languages and their rules of use but rather what they
> mean. It is this need for meaning that reveals itself in trying to
> translate between languages or understand anothers point of view.
>
> Japanese sentence --> meaning of sentence --> English sentence/s
You mean that here the agent comes in, ie as the translator/intepreter?
JGCASEY
JPL Verhey wrote:
[...]
>I still don't understand the problem many have with AI, to be honest.
>AI stands for "man-made intelligence" as i understand it. Man-made is
>clear what it means.. and intelligence refers to certain types of
>behaviors that we call intelligent behaviors. It is a man-made
>evaluation and decision of what we call intelligent (machine)
behavior.
>And all this is not about mind or consciousness aqa strong-AI (where
>btw "artificial consciousness" or AC is a better term to be used) but
>weak-AI. So our world is already filled with weak-AI machines and the
>physical requirements are well known and no secret. Of course there
>are very simply weak-AI machines such as a thermostat but also very
>complicated ones that to a degree can self-organise and adapt their
>behavior.
>
>Or is your interest strong-AI/AC?
My interest is in computer programs that carry out tasks that
if carried out by human beings would require intelligence.
[...]
>> This comes back to what is important is not the technical structures
>> of different languages and their rules of use but rather what they
>> mean. It is this need for meaning that reveals itself in trying to
>> translate between languages or understand anothers point of view.
>>
>> Japanese sentence --> meaning of sentence --> English sentence/s
>
>
>You mean that here the agent comes in, ie as the
translator/intepreter?
Meanings do have a physical requirement, an agent if you like.
I was responding to talk about the syntax differences between
languages and am suggesting that they are *equivalent* in their
power to convey meaning. And it is the meanings that are important
not the language.
In other words we need to consider that branch of linguistics
concerned with meaning, semantics, when analysing verbal or
written communications.
John Casey
JPL Verhey
>
> JPL Verhey wrote:
>
> [...]
>
>
>>I still don't understand the problem many have with AI, to be honest.
>>AI stands for "man-made intelligence" as i understand it. Man-made is
>>clear what it means.. and intelligence refers to certain types of
>>behaviors that we call intelligent behaviors. It is a man-made
>>evaluation and decision of what we call intelligent (machine)
> behavior.
>>And all this is not about mind or consciousness aqa strong-AI (where
>>btw "artificial consciousness" or AC is a better term to be used) but
>>weak-AI. So our world is already filled with weak-AI machines and the
>>physical requirements are well known and no secret. Of course there
>>are very simply weak-AI machines such as a thermostat but also very
>>complicated ones that to a degree can self-organise and adapt their
>>behavior.
>>
>>Or is your interest strong-AI/AC?
>
> My interest is in computer programs that carry out tasks that
> if carried out by human beings would require intelligence.
I like that approach :-)
>
> [...]
>
>
>>> This comes back to what is important is not the technical structures
>>> of different languages and their rules of use but rather what they
>>> mean. It is this need for meaning that reveals itself in trying to
>>> translate between languages or understand anothers point of view.
>>>
>>> Japanese sentence --> meaning of sentence --> English sentence/s
>>
>>
>>You mean that here the agent comes in, ie as the
> translator/intepreter?
>
> Meanings do have a physical requirement, an agent if you like.
I think our thinking about meaning and in as far as it refers to
something real, IS physical. Would you agree with that?
>
> I was responding to talk about the syntax differences between
> languages and am suggesting that they are *equivalent* in their
> power to convey meaning. And it is the meanings that are important
> not the language.
>
> In other words we need to consider that branch of linguistics
> concerned with meaning, semantics, when analysing verbal or
> written communications.
Yes, but I guess it is not always easy to *experience* the meaning of a
certain language that is (very) different from our own. The meaning is
also formed by cultural content etc. A scientist may relatively
succesfully describe similar meaning generated via different languages..
but the chance that he missed a lot because he lacks the life-experience
appears pretty big!
JGCASEY
JPL Verhey wrote:
[...]
>> Meanings do have a physical requirement, an agent if you like.
>
>I think our thinking about meaning and in as far as it refers to
>something real, IS physical. Would you agree with that?
Meaning is a relationship between physical things. For example
a word has a physical existence of some form but the meaning
is in its relationship with another system that uses that word.
My mother has a different meaning to me than she would for you.
The meaning of something is in the things it triggers in some
thing else.
The meaning that one thing has for the other can be described
by the effect it has on the other.
>> I was responding to talk about the syntax differences between
>> languages and am suggesting that they are *equivalent* in their
>> power to convey meaning. And it is the meanings that are important
>> not the language.
>>
>> In other words we need to consider that branch of linguistics
>> concerned with meaning, semantics, when analysing verbal or
>> written communications.
>
>Yes, but I guess it is not always easy to *experience* the meaning of
a
>certain language that is (very) different from our own. The meaning is
>also formed by cultural content etc. A scientist may relatively
>succesfully describe similar meaning generated via different
languages..
>but the chance that he missed a lot because he lacks the
life-experience
>appears pretty big!
I would suggest that we are all essentially the same regardless
of our cultural background. That is, we all have the same needs
just different ways (due to culture, personality) of expressing
those needs.
So although I may not be able to *experience* the meaning of
a certain language I suspect I never the less have the same
experiences and can express them in my own language.
--
John Casey
Stephen Harris
"Jeff Rubard" <jeffr...@online.ie> wrote in message
news:TVCyd.7983$RH4....@newsread1.news.pas.earthlink.net...
I haven't read the whole thread, but this portion didn't seem quite
precise because public subjectivity is the collection of individual
subjectivities so subjectivity itself is still required in "knowing" where
subjectivity itself is a function of conscious awareness.
I will clarify my point by providing some background which excludes
a God or a universe aware of itself Platonic realms, etc.
There is the big bang theory which proceeds to our planetary evolution.
This describes our 'universe' or reality. It is itself, and has independent
existence prior to any human knowledge of reality. This is assumed in
how we evolved into existence later in the time frame.
If you say that the universe/reality/existence is synonamous with
knowledge, a semantic distinction/association of definition is performed.
If you say that the universe/reality/existence pre-existed knowledge
so that universe/reality/existence is not synonamous with knowledge
it is still a matter of definition. And since this is the commonly agreed
(public) consensus defintion it seems to provide a clear conclusion.
As in most cases, people know the definitions of words and use them
correctly, or they do not. And it is the people who are ignorant of the
definitions that tend to convert definitional statements into philosophical
propositions.
Knowledge is a human abstraction which is never congruent with
reality, it describes some subset of observed or inferred human
observation of reality which is finitely limited by human potential even
aided by mechanical extension.
Knowledge is a human invention to describe some apect of reality
and mathematics or Physics is a subset of that knowledge as well
as a bird in the hand is worth two in the bush. Knowing is the
relationship formed between knower and the known. The knower
requires having a measurement potential. It would be anthropomorphic
to say rocks know things, that rocks experience things, just because
reality impacts the rock causally. The known originated or had its
roots in reality which unfolded in a sequentially timely manner. Reality
is a concept that does not require human awareness in order to exist.
Knowledge is a term that (as we use the term) describes what humans
know about some portion of reality and it requires intelligence.
We don't require the term intelligence to describe the universe prior
to the creation of our planet, that is a religious tact. Now that humans
exist in the universe, describing reality itself as knowledge, rather than
the source which is used to create knowledge seems to be a category
mistake. I don't think it helps to change the relationship from that of
'knower to that known' to that of 'subject to object' where the subject
is required to have a perception, while not bending definitions of
words (equivocation) so that a camera taking a picture is a perception
or leaving a message on somebody's phone is called a conversation
because one has responded in words to words inviting one to leave
a message by mechanical means similar in physical principle to an echo.
(A speaker generates a time delayed audio signal which is reflected so
that another entity may hear it.) To me, some of this debate has revolved
around considering knowledge to be reality; rather than about reality in a
more abstract or removed to another level sense, and then confused.
Knowlege as being part of reality has used the meaning of reality
differently, an undifferentiated reality compared to a human differentiated
reality which now includes the human presence in the description.
Regards,
Stephen
Stephen Harris
> Neil W Rickert wrote:
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> "Jeff Rubard" <jeffr...@online.ie> writes:
>>
>> idealism, and you are spewing a pile of sophistry as an alleged
>> argument in support of that opinion.
>>
>
> It's an exercise in self-clarification.
>
They have been arguing over a good definition of knowledge for
quite some time. I haven't followed your more specific arguments
which relate to aspects of current human knowledge.
If there is no God, the universe doesn't know itself, and there are
no other intelligent species, the universe existed in an undifferentiated
state (no intelligence to make comparisons or take measurements)
until humans arrived and carved an area of their exploration of that
reality.
Reality before humans arrived (with prior stipulations) happened in some
particular time frame. Knowledge is a creation which coincided with
the evolution of humans, so is part of their time frame in universal
history,
and is an actual physical (brain) process in current, now time. When
humans ponder a time frame prior to their existence, they are building
a representation of an earlier time, but the knowledge is not occuring
in that time frame. It occurs now in the human time frame. So that is
why one can say that epoch continues to be undifferentiated in that
time frame of the universe's existence. However, it is differentiated now.
I am not sure the words continuous or discrete describe undifferentiated
in that earlier historical period of the universe. I am not sure that before
the big bang that terms like cotinuous or discrete are apt descriptions.
I think the problem of defining knowledge accurately is that it has a
foundation in causality and time which are not properly understood.
I think it is safe to say that instances of knowledge are causality reports,
now that we have that concept, which would include subjective delusions.
One does not say I know pain, but one feels pain, but once can certainly
know that one feels pain so your remark does not seem to be on target.
even if the odd causality recognition is random,
Stephen
ernobe
It would be no use if the system could prove that it knows math, but could not
test others for their knowledge of math. Knowledge representation of some form
would be required here, because it would have to take into account the
differences in the ability to use math, set at appropriate limits beyond which
no testing is done. For example, it wouldn't test Jim on whether he can
multiply 6 digit numbers in under a fraction of a second, and it would have to
test him for abilities that are peculiar to human beings, but only such as
pertain to their use of computers, not such as they evince among themselves.
Perhaps 'knowledge representation' is not the best term to use, since the
abilities it tests in human beings or in anything else are defined by humans
themselves, so in effect it would be more of an identification process rather
than its "knowing" something. Likewise, it could not really test a human for
its knowledge of math, but could for example "know" that Jim is working on a
certain math formula and keep track of his progress. Now it doesn't seem
possible that if friend of Jim is working on a similar formula that it could
investigate areas of application for the formulas to work together. The symbols
used in algebraic formulas are specific to the formulas, and the purpose of the
formula is to represent a change in that which the symbols represent. Change
is inherent in the nature of things and computers calculate the nature of the
change, but changes cause things to transform themselves into each other so
that nothing remains of their previous state, therefore how can a computer use
certain formulas to derive new ones that bear no resemblance to them?
