Thanks to all who've replied to my question on this. Partly, that was
the intention of the question: to stimulate thought, rather than to
get a specific answer. And in fact, the replies certainly stimulated
my thoughts! (Aside: my apologies to those who were left wondering
what I was up to with my question -- I intended to respond earlier
than this.)
The discussion also prompted me to go back and re-read Winograd and
Flores' book "Understanding Computers and Cognition". The first part
of the book is an interesting synthesis of Heidegger's philosophical
work with Maturana and Varela's interpretation of living systems,
based on their biological investigations. Mark, I recommend it as a
good introduction to an alternative viewpoint on statements, facts,
truth, fallibility, etc. By the way, I think it might be
complementary rather than contradictory to Popper's ideas of
fallibilism.)
The replies very roughly fall into two categories:
- A fact is a statement of a particular kind;
- A fact is something else than a statement (a "thing", a thing or
occurence with actual existence, a piece of information, an article of
truth).
I especially liked Ray Harrell's "Generally fact refers to mutual
agreements while truths are pillars of cultural reality."
For my take on this, I'll use Mark's characterization in LO30880 as a
point of departure:
> A "fact" is something that has existence, or some aspect of reality. A
> "statement" is a claim about a fact. So the question is, can the latter
> ever "correspond" with the former?
I believe it's important to recognize that statements are not isolated
things, but are embedded in the language we use constantly to think in
and to communicate among ourselves. We are languaging creatures (cf.
Maturana), and our evolution as a species and development as
individuals (and as organizations) happens in that context.
Every language, by its nature, entails certain commitments to the way
the world should be considered to be organized. Also, one shouldn't
consider, say, the English language to be a monolithic thing; it's
constantly being revised, extended, transmogrified, etc., in many
different places and in many different ways.
To assert a correspondence between a claim and a fact is itself an act
of language; the correspondence, then is not just a relationship
between claim and fact, but includes the speaker and the hearer(s);
it's also usually situated in the context of a conversation, which
conditions the kind of statement that will be made and the kind of
correspondence that will be claimed. The "tests and evaluations" that
are used to determine what, if any facts correspond to a statement
will also be acts of language.
With this in mind, I recommend a study of DP's three interviews in
LO30919 for some nice examples (and I think they provide a good answer
to John Sibley's question).
Here's an simple dialog from Winograd and Flores' book, exploring the
meaning of "water":
A: Is there any water in the refrigerator?
B: Yes.
A: Where? I don't see it.
B: In the cells of the eggplant.
What role do truth and fallibility play in this dialog? What important
aspects do they fail to cover?
Some other possible responses of B:
1. Yes, condensed on the bottom of the cooling coils.
2. No, there's no water, but there's some lemonade.
3. Yes, there's a bottle of water with some lemon and sugar in it to hide
the mineral taste from the pipes.
1 may actually be a reasonable response from A's point of view, for example
if A were looking for sources of unexpected humidity. The distinction
between 2 and 3 isn't necessarily clear. Is lemonade water?
[Hosst's Note: Winograd and Flores -- This is a VERY interesting book!
..Rick]
What we've been discussing so far, on the "knowledge and information"
thread, has mostly been about explicit knowledge -- statements in
languages. I'd like to hear what folks think about Polyani's "tacit
dimension" in this regard. (You may remember that there was a long
discussion of Polyani's works a while ago on this list.) If "we know
more than we can tell", do we need "tacit knowledge management"?
All that said, I think it's very much worth introducing fallibility,
and the notion of what we consider to be true, into knowledge
management. I think it should be understood as a healthy step in the
right direction, though, rather than a complete characterization of
"manageable knowledge". Perhaps a good next step would be to
introduce the concepts of autopoietic systems, languaging behavior,
and consensual domains. (And perhaps a good introduction would be
Maturana and Varela's book "The Tree of Knowledge" -- I say "perhaps"
because I haven't read it yet, although the Amazon reviews are quite
positive.)
Cheers,
--Don Dwiggins "Umntu Ngumntu Ngabantu" d.l.dwiggins@computer.org (a person is a person through other people) -- South African saying
[Host's Note: In assoc w/Amazon.com, this link
Understanding Computers and Cognition : A New Foundation for Design by Terry Winograd (Author), Fernando Flores (Author) (Paperback ) http://www.amazon.com/exec/obidos/ASIN/0201112973/learningorg
.. Rick]
Learning-org -- Hosted by Rick Karash <Richard@Karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>
"Learning-org" and the format of our message identifiers (LO1234, etc.) are trademarks of Richard Karash.