flink@umcp-cs.UUCP (Paul V. Torek) (07/20/85)
In article <405@spar.UUCP> ellis@spar.UUCP (Michael Ellis) writes: > [...] It appears that causal explanations are possible only > if the concept of causality is fundamentally revised. John Forge > seems to evade the issue by suggesting that "a causal process is one > which is governed by scientific laws (theories)" (1982). On this > account of causality, every QM process becomes causal by definition > simply because QM is a scientific theory. > > (from _Scientific Explanation and the Causal Structure of the World_ > Wesley C. Salmon, Princton University Press, 1984.) I don't think John Forge is "evading" the issue. Quantum processes involve causation, all right; it is just that the causation is probabilistic rather than "deterministic". It is not the case that things happen entirely haphazardly. The impact of one particle on another will influence it -- will *cause* it to do something -- even though there are several possible effects. Any objections? --Paul V Torek, umcp-cs!flink
tmoody@sjuvax.UUCP (T. Moody) (08/21/85)
*** REPLACE THIS LINE WITH YOUR MESSAGE ***
In a fairly recent posting, Michael Ellis wrote:
[Ellis]
Is it possible to provide causal explanations of QM phenomena? I do
not know. Van Fraassen argues cogently, on the basis of Bell's
inequality and relevant experimental results, that "there are well
attested phenomena which cannot be embedded in any common-cause
model" (1982). It appears that causal explanations are possible only
if the concept of causality is fundamentally revised. John Forge
seems to evade the issue by suggesting that "a causal process is one
which is governed by scientific laws (theories)" (1982). On this
account of causality, every QM process becomes causal by definition
simply because QM is a scientific theory. Such problems as the EPR
paradox then disappear... For purposes of argument, we might adopt
the following definition: {the causal net == whatever structure of
relations science describes}. We could then leave "to those
interested in causation as such the problem of describing that
structure in abstract but illuminating ways, if they wish"...
(from _Scientific Explanation and the Causal Structure of the World_
Wesley C. Salmon, Princton University Press, 1984.)
-michael
... and Paul V. Torek responeded:
[Torek]
I don't think John Forge is "evading" the issue. Quantum processes
involve causation, all right; it is just that the causation is
probabilistic rather than "deterministic". It is not the case that
things happen entirely haphazardly. The impact of one particle on
another will influence it -- will *cause* it to do something -- even
though there are several possible effects.
Any objections?
--Paul V Torek, umcp-cs!flink
[]
I have been meaning for some time to comment on this. Paul Torek is
suggesting that the concept of causation is elastic enough to include
non-deterministic interactions, while Ellis (with van Fraassen) denies
this. This is a perfect example of how scientific progress calls for
conceptual revisions. It is also a good example of a _metaphysical_
problem tangled up with physics.
Since roughly the time of Newton, the concept of causation has been
bound up with that of the determination of events. That is, the cause
of an event is the exact set of circumstances that will bring about
that event, or, more concisely, _determine_ it. To put the claim
epistemically, if you have exact knowledge of the cause, you don't
need any further knowledge to predict the effect. Where words such as
"approximately" have been used, they have only been intended as a
measure of the practical limitations on human knowledge. Until
quantum mechanics, no one ever really doubted that in the world of
inanimate objects there is, underlying our imperfect knowledge, an
_exact fit_ between cause and effect.
According to quantum mechanics, there is some slack here. Quantum
events set a range of probability, but do not exactly determine,
subsequent events.
There is no question that this does violence to the classical
conception of causation. But then, to call quantum mechanical events
"uncaused" does similar violence.
It seems to me that clarity is best served if classical causation is
recognized as an emergent property of nature, one that disappears at
quantum-level scrutiny. Perhaps some new verbs are needed to describe
the kind of quantum influences that Paul Torek asserts should retain
the name of "causation". How about "channeling"?
Todd Moody {allegra|astrovax|bpa|burdvax}!sjuvax!tmoody
Philosophy Department
St. Joseph's U.
Philadelphia, PA 19131