Judea Pearl wrote:
> Dear Peter,
> You asked:
> For example, am I correct in suspecting that you
> would reject the validity or value of "ancillary
> inference netwworks" of the sort that I was trying
> to describe?
>
> I reply:
> Sorry, but I have no idea what "ancillary inference networks" are.
> I need to see simple compelling examples where
> those networks offer new power or greater comfort.
> Then I would like to see semi-formal
> definitions of what the nodes-links represent, what
> the input is, what questions they are designed to answer,
> and what logic stands behind the answers.
> The burden of explicating all these, I am afraid, lies
> with those who find such networks meaningful.
> Others cannot do it for you,
> =========Judea
Dear Friends,
It is sometimes better to let a dying horse or thread die. It seems
clear that Judea sees little profit in extending our exchange, and I
think he is probably right about that. Nonetheless, I do not want people
on this list to accept Judea's intimation that the notion of ancillary
networks or of networks-embedded-in-networks is a notion that I invented
entirely on my own during a summer night's delirium. David schum, for
example, has used something like the notion of netwworks embedded in
networks. See, e.g., David A. Schum, EVIDENTIAL FOUNDATIONS OF
PROBABILISTIC REASONING (Wylie, 1994). Now I know (from an earlier
exchange) that Judea does not seem to think much of the way that David
works with inference networks. But to say that an approach such as
Dave's is _wrong_ is rather different from saying that such an approach
doesn't _exist_. Moreover, most readers of Dave's work think that
Dave's approach is at least _intelligible_ -- even they think he is
ultimately wrong in one or more particulars.
I hope the members of this list won't mind if I lean so heavily on
Schum's theory. I do so because I have enormous respect for him and his
work and because I have reason to think that I have a pretty good sense
of what his theory is. (If I had no idea what he was talking about, he
probably would not have agreed to co-author several papers with me.)
Ulltimately, of course, Dave must speak for himself. I refer to Schum's
work principally because I am not sure that _I_ should be the one to
have to work an systematic alternative to Pearl's approach. (I rather
believe in a division of labor, even in matters intellectual,
theoretical, and logical. I don't know nearly as much about logic or
probability as Judea does, but I probably know a bit more than some of
you do about some attributes of evidence and inference in legal contexts
such as trials.)
I apologize if by starting this thread anyone thought that I was trying
to get "others ... to do [my theory-building] for [me]." I really
thought that I was just asking for some advice and opinions and I
thought the internet was an appropriate medium for doing that. No one is
_obliged_ to address any of my questions or comments.
Peace!
Sincerely yours,
Peter Tillers
*********************************************************
Tillers' Dynamic Evidence Site
--with "frames":
http://www.tiac.net/users/tillers/index.html
--sans "frames":
http://www.tiac.net/users/tillers/home.html
**********************************************************
Peter Tillers, Professor of Law
Benjamin N. Cardozo School of Law, Yeshiva University
55 Fifth Avenue, New York, New York 10003, U.S.A.
(212) 790-0334; FAX (212) 790-0205
--------------1930257736A12B4DC5F8B405
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit
Judea Pearl wrote:
Dear Peter,Dear Friends,
You asked:
For example, am I correct in suspecting that you
would reject the validity or value of "ancillary
inference netwworks" of the sort that I was trying
to describe?I reply:
Sorry, but I have no idea what "ancillary inference networks" are.
I need to see simple compelling examples where
those networks offer new power or greater comfort.
Then I would like to see semi-formal
definitions of what the nodes-links represent, what
the input is, what questions they are designed to answer,
and what logic stands behind the answers.
The burden of explicating all these, I am afraid, lies
with those who find such networks meaningful.
Others cannot do it for you,
=========Judea
It is sometimes better to let a dying horse or thread die. It seems clear that Judea sees little profit in extending our exchange, and I think he is probably right about that. Nonetheless, I do not want people on this list to accept Judea's intimation that the notion of ancillary networks or of networks-embedded-in-networks is a notion that I invented entirely on my own during a summer night's delirium. David schum, for example, has used something like the notion of netwworks embedded in networks. See, e.g., David A. Schum, EVIDENTIAL FOUNDATIONS OF PROBABILISTIC REASONING (Wylie, 1994). Now I know (from an earlier exchange) that Judea does not seem to think much of the way that David works with inference networks. But to say that an approach such as Dave's is _wrong_ is rather different from saying that such an approach doesn't _exist_. Moreover, most readers of Dave's work think that Dave's approach is at least _intelligible_ -- even they think he is ultimately wrong in one or more particulars.
I hope the members of this list won't mind if I lean so heavily on Schum's theory. I do so because I have enormous respect for him and his work and because I have reason to think that I have a pretty good sense of what his theory is. (If I had no idea what he was talking about, he probably would not have agreed to co-author several papers with me.) Ulltimately, of course, Dave must speak for himself. I refer to Schum's work principally because I am not sure that _I_ should be the one to have to work an systematic alternative to Pearl's approach. (I rather believe in a division of labor, even in matters intellectual, theoretical, and logical. I don't know nearly as much about logic or probability as Judea does, but I probably know a bit more than some of you do about some attributes of evidence and inference in legal contexts such as trials.)
I apologize if by starting this thread anyone thought that I was trying to get "others ... to do [my theory-building] for [me]." I really thought that I was just asking for some advice and opinions and I thought the internet was an appropriate medium for doing that. No one is _obliged_ to address any of my questions or comments.
Peace!
Sincerely yours,
Peter Tillers
*********************************************************
Tillers' Dynamic Evidence Site**********************************************************--with "frames": http://www.tiac.net/users/tillers/index.html
--sans "frames": http://www.tiac.net/users/tillers/home.html