File spoon-archives/bhaskar.archive/bhaskar_2002/bhaskar.0205, message 169

Subject: RE: BHA: Mainstream Philosophy of Science
Date: Fri, 17 May 2002 10:57:20 -0400


I've interspersed replies below.

> We seem to be inhabiting different paradigms. Holding up empiricism as
> some kind of model or template for locating deficiencies in CR doesn't
> cut any ice with me because I accept the CR critique of empiricism. Why?
> The same reason as I accept any other critique - I make an assessment of
> the arguments in favour of both outlooks and choose.

I'm not holding up empiricism as a model. What I am doing is saying that
these philosophers are addressing issues that I feel are not adequately
addressed in CR. Some of the philosophers are empiricist, some are not. I'm
not particularly wedded to the answers they give or even the criteria they
use for what counts as a good answer. Instead, I'm simply pointing to the
questions they raise and some of the arguments they make about those
questions. I think questions like "Besides empirical experience, what else
can science know?" are important.

> >what's important is
> >the mathematical properties (structure) of light's behavior.
> Isn't this precisely the CR view, expressed in examples by Bhaskar over
> an again? I.e. what we have is an ontology of causal powers (plus
> fundamental structuring principles - the categories); i.e. in the terms
> used in the later philosophy, dispositional and categorial realism. When
> Bhaskar gives examples from the natural sciences of the former he cites
> precisely mathematical formulas for structures.

No, I don't think it is, at least in some versions of CR. As I think I
mentioned in my original post, this comes close to the CR view but is not
identical with it. The empiricists claim the only thing we can know is what
we experience. The author who proposed structural realism (I don't have the
book with me and I don't recall which of the authors I cited wrote this)
said the structure of mathematical functions describing light's causal
properties is real, in addition to the empirical experience of light. In
other words, he extended the "real" beyond the empirical to include the
structure of causal powers as they are exercised. CR, on the other hand,
might accept this but goes further. It would say, for instance, that at some
point science can identify photons and similar things as being real.
"Structural realism" deliberately avoids such claims about the real
existence of photons because of the philosophical difficulties involved. CR
also uses "structure" differently. Rather than designating the patterning
and organization of exercised causal powers, CR uses "structure" to
designate the internal relations of things, from which these things derive
their causal powers as emergent properties.

I must admit (confess?) that I've not kept up with Bhaskar's more recent
work, having turned my attention to other CR authors  whose work is more
directly relevant to my own (Sayer, Archer, etc.). You may be right that
dispositional realism is the same as "structural realism." Still, if you
look for instance at Danermark, et al., there is no mention of dispositional
realism or of the different kinds of theories one finds in science. (Again,
I don't have the philosophy of science book with me, so I can't cite the
author -- but I did find the discussion of different kinds of theories,
including such things as Brownian motion and ideal gas laws, quite useful.
It allows us, for instance, to distinguish between theories intended to be
literal depictions of the world from those that are intended as heuristics.)

Might we not, for example, think of Marx's simple commodity production as
just such an heuristic? Many Marxists have taken it literally and thought of
it as a sort of pre-capitalism. Bhaskar himself is probably better on this
(the heuristic nature of some scientific theories) than many of his
followers. Still, insofar as I know, even RB has not given such a clear
designation of the different kinds of theories one finds in science.
Furthermore, it would be very very helpful for practicing critical realists
to understand these different roles for theory. I think too often we see the
 only role for theory as being to identify real entities whose causal powers
arise as emergent properties for structures of internal relations and which
are manifested contingently as actual events.

> Nor does (in you previous email) waving asside the theory of alethic
> truth - *the* (D)CR theory of truth - on the spurious grounds that it is
> >truth in terms of the world rather than in terms of human knowledge
> cut any ice. It is a theory about the human achievement of ontological
> truth. Ultimately we can know the world because we are an emergent part
> of it and epistemology is contained within ontology - we are not set off
> dualistically over against it. We do achieve knowledge of the world and
> could not have evolved successfully if that weren't so.

I don't necessarily disagree. But at the very least CR's case would be
helped by linking it to what are apparently widespread themes in
contemporary philosophy of science. The "human achievement of ontological
truth" seems to me to be another way of phrasing the "no miracles" argument
for realism. The latter takes its name from an article by Hillary Putnam in
which he says something like, "The best argument for realism is that it's
the only theory that doesn't make the success of science a miracle." Compare
this, for instance, with RB's critique of Rorty in Reclaiming Reality. If I
remember correctly, RB starts by saying Rorty is puzzled by why science is
doing so well of late. Wouldn't it have been stronger to say that Rorty is
ignoring and dismissing the entire "no miracles" argument for realism? (Much
as you're criticizing me for overlooking crucial arguments in CR.) Then RB
could launch into a critique of Rorty and show why his philosophy leads him
to regard science's success as a miracle.

I also wasn't dismissing alethic truth. I was simply saying that the theory
of alethic truth does not help us in concrete situations in which we're
trying to explain the truth-value of terms like photons or the capitalist
mode of production. Yes, we can say that there are real unobservables and
that we think that photons and the CMP are such entities. But this just
establishes their candidacy as real entities. It does not elect them. In
these specific cases, I think the human species could have evolved perfectly
well to its current state without knowing about photons or the CMP.
Furthermore, insofar as the concept of photons helps us make laser devices
or the concept of the CMP helps us make socialism, I think the pragmatists
are  right. All we know for sure is that these are concepts that help our
practice. They may be literally true or they may just be  convenient devices
that help us cope. To establish their truth as concepts literally
designating real entities in the world requires an altogether different
theory of truth than the more general claims about alethic truth. In other
words, I'm not dismissing alethic truth, but I am saying we could use
something more.

> Nor does invoking scientific revolutions as if they demonstrate that the
> truth-claims of the predecessor theory were all false cut any ice. CR
> has a powerful theory sustaining the rationality of scientific
> revolutions - see esp. the Hegel-derived epistemological dialectic of
> science in the early part of DPF.

Not having read all of DPF, I'll have to look this up. The role that
scientific revolutions play in the literature I brought up is this. If we
accept that scientific revolutions happen, then what science says today may
be considered wrong tomorrow. Therefore, we have no strong grounds for
accepting today's science as literally true. The literature goes on to say
that some realists attempt to get around this problem by claiming today's
science is "approximately" true. The question then becomes what does
"approximately" mean. It also raises interesting questions about historical
episodes in science. Was Priestley's "dephlogistacated air" "approximately"
Lavoisier's oxygen? Or, were they entirely two different things such that
were we to believe in phlogiston we would be totally wrong about air? I'll
have to look again at DPF to see how it addresses such issues.

> How do we choose between theories of late capitalism and of post-
> industrial society? We've been over this territory before. We make an
> assessment of their relative explanatory power. (Which theory better
> explains the imbecile hucksterism of the moronic Empire? And its
> expansionary dynamic? Do we see an end of ideology? Etc etc etc.) There
> are no ready made formulae or algorithms that we can apply mechanically
> - we make an assessment of the overall arguments in favour of each. On
> the Bhaskarian arguments on facts and values etc, all we need at the
> outset is commitment to truth (a necessary condition of all discourse,
> so we're not importing it into the scientific endeavour - it is an
> aspect of our core universal human nature) and everything else follows,
> i.e. politics needn't determine our science, it can be the other way
> around.

Yes, but then we need to elaborate on what we mean by explanatory power. I
also think this is not enough. Milton Friedman, in _An Essay on Positive
Economics_, argues that theories are like black-box calculating devices.
Their value is in helping us make good predictions about the world (not
necessarily, or not even primarily, predictions about the future state of
the world, but rather predictions about the patterns of data in our
scientific studies). Thus, he argues, the fantastic constructs in
neoclassical economics are perfectly legitimate. We do not care if people
really behave as "rational economic man" or if there is perfect information,
etc. All we should care about, according to Milton, is that these constructs
help us make good predictions.

In some cases, I am inclined to agree with uncle Milty. If we're trying to
project a town's population for next year, then a good projective model is
all we need. Of course, we also need a good realist theory of population so
that we have cause to believe the demi-regs under which the model operates
will still be in effect next year. However, Milty wants to go beyond this.
He wants to license science as being strictly concerned about predictions
which, in the Hempel-Popper tradition, is the same thing as an explanation.
I believe the plausibility of the theory's terms -- in other words the
theory's internal realism about the world -- is also important. Late
capitalism is a better theory than post-industrialism because (1) capitalist
relations still govern production (of not only industrial goods but also
services, cultural goods like music, etc.) and because (2) the bulk of
economic production (and a good deal of value) still involves industrial
goods, although now their production is more global so that countries within
the over-developed world may in fact have more "knowledge workers" than
industrial workers. Nonetheless, for some purposes, we might have an easier
time using the post-industrial theory to explain things. For instance, if we
consider changes in Chicago's political machine, we might explain it more
easily by pointing to the shift from blue-collar industrial workers to
white-collar post-industrial workers, so that the machine's patronage was no
longer an effective mechanism for political governance. A theory based on
late-capitalism may be more difficult to apply precisely because it is based
on a deeper level of reality so that the connections to empirical events are
more complex and contingent. Still, I would throw my hat in with the
late-capitalism theory precisely because it penetrates reality more deeply.

I think CR points us in one of two directions, which are almost equivalent.
(1) We can define a good explanation to be one that's based on real things
and causal powers necessarily arising as emergent properties from their
internal relations. Or, (2) we can accept a more open and imprecisely
defined notion of "explanation" and add a second criterion for accepting a
theory besides its explanatory power. This second criterion would be the
theory's realism.

However we resolve this, we're straying a bit from the original discussion
of scientific truth. As I mentioned in an earlier posting, I'm attracted to
a criteria for the true existence of a scientific entity as requiring it to
have two independent causal properties which are verified through scientific
research. This would, for example, uphold the existence of the CMP because
it explains both business cycles and the global spread of commodity
production, two observable phenomena. Nonetheless, someone could come along
with a theory explaining business cycles in terms of the money supply and
the spread of commodity production by psychology. In other words, the idea
that we postulate some thing to explain one phenomenon and then take the
fact that the thing, so postulated, explains an entirely independent
phenomenon as proof of the thing's existence makes intuitive sense and seems
consistent with CR. Nonetheless, even this argument may have logical flaws.
It seems CR can only benefit by engaging in debates about such logic with
those who are not sympathetic to CR. This way we can work out the strengths
and weaknesses of our arguments. Otherwise, we're just preaching to the


	Marsh Feldman

     --- from list ---


Driftline Main Page


Display software: ArchTracker © Malgosia Askanas, 2000-2005