Date: Tue, 05 Aug 1997 16:22:09
To: bhaskar-AT-jefferson.village.Virginia.EDU
Subject: Re: BHA: Non-experimental science (was "What must the ...")


Marshall,

You wrote:
>>>Here's what I think is at stake here.  On your account we might have to
>>>start with, say, Herb Blalock's empiricist account of causal explanation in
>>>social science (which is actually a version of Humean-J.S. Mill causal
>>>explanation) and ask if it conforms to the CR account of causal explanation
>>>in experimental science.  The answer would be "no."  Does this mean CR does
>>>not apply, or is Blalock full of -AT-$#^?

To which I responded:
>>I don't know what Blalock's account is, but I do not see why we might have
>>to start with his account of the meaning of "causal explanation".  No gun
>>is at my head, and I assume you are able to expose the error of his
>>definition.  I am willing to concede that his fixed sense of that term is
>>evil while your and Doug's fixed meaning is good.  I just find the idea of
>>a fixed definition for such a term to be highly dubious.

And you replied:
>Blalock's account is similar to Mill's.  E.g., A can cause B iff A preceeds
>B, iff whenever A then B, etc.
>
>Yes, of course we can point out the errors in Mill's account.  My question
>is how can we use RTS for this?  It seems to me RTS basis its argument on
>experimental science.  If we deem this irrelevant for a critique of
>Blalock, then RTS becomes irrelevant too.
>
>Let me put the question to you.  Given Blalock's empiricist reading of
>causal explanation, how would you use RTS in a critique without relying on
>the parts of its argument that depend on experimental practice?

Are you asking how I would refute Blalock's reading of causal explanation?
If you are, why am I not allowed to rely on those parts of RTS that rely on
experimental practice?  As I said, I am unfamiliar with Blalock, but either
you think RTS as a whole is capable of refuting him, or you do not.  If you
do, then he stands refuted and your question seems to lack a point.  If you
do not, then certainly a restricted part of RTS could not be used to refute
him, if the whole does not, so again your questions seems to lack a point.

I don't see how this relates to the current issue of whether "causal
explanation" should be given a fixed definition in advance by CR (or by an
alternative theory).  In particular I don't understand why you say,
concerning basing RTS on experimental science, "If we deem this irrelevant
for a critique of Blalock, then RTS becomes irrelevant too."  This seems to
go back to your original issue concerning the extension of CR from natural
sciences to other domains.  After various exchanges, you summed up your
position by saying "that RTS attacks positivism in the area it thinks it's
strongest: experimental natural science.  From this, RB justifies
different, non-positivist meanings for "science," "cause," "explanation,"
etc.  These MEANINGS then transfer over to the domain of the
non-experimenatal sciences, where we may then have to add some other
justifications to warrant their applicability.  The latter is what I take
as the purpose of PON."  

I felt agreement with the general drift but was uncomfortable with the 
role meaning was being assigned: "Although Marshal, Doug and I are in
agreement 
on CR not being a general, a priori theory and requiring at least support in 
the various domains outside natural science, ... I disagree with [this] way 
of putting it in terms of the meanings of "science", "cause" and 
"explanation."  I don't think it is right to say that the meanings of those 
terms are fixed by a CR analysis of natural science, especially the last 
two.  On their understanding, having fixed those meanings we go on to ask
if social studies are capable of causal explanation, and if they are,
whether those explanations are scientific in character.  I would prefer to
say that first we establish what natural science is in CR terms and then
ask whether causal explanations in other domains conform to the same kind
of depth analysis.  If the answer were to turn out negative in some domain,
I don't know that we should declare that there are no causal explanations
in that domain simply because they do not conform to the CR account gleaned
from natural science." I thought we were in agreement over the legitimacy
of extending CR to other domains but disagreed over the role of meaning.




     --- from list bhaskar-AT-lists.village.virginia.edu ---