KFL@MX.LCS.MIT.EDU (11/07/86)
From: "Keith F. Lynch" <KFL@MX.LCS.MIT.EDU>
From: hplabs!sdcrdcf!psivax!nrcvax!nrc-ut!uplherc!utah-gr!pwa-b!mmintl!franka@UCB-Vax.arpa (Frank Adams)
...
I doubt that it is possible to design an experiment whose results could
require reverse causality for their explanation.
I don't think there is any reverse causality, and if there is, I
think it is much more likely to be discovered by peering at an
oscillosope in a high energy physics lab than by peering at cards with
mystical symbols on them by candlelight, HOWEVER, I think experiments
CAN and HAVE been designed which would require reverse causality to
explain a possible outcome of. Saying that no such experiment is
possible is saying that belief in forward-only causality is non-
scientific. All scientific beliefs are in principle falsifiable by
experiment.
...Keithweemba@brahms (Matthew P Wiener) (11/08/86)
Summary: Expires: Sender: Distribution: Keywords: In article <240@sri-arpa.ARPA> KFL@MX.LCS.MIT.EDU writes: > I doubt that it is possible to design an experiment whose results could > require reverse causality for their explanation. [Frank Adams] > > I don't think there is any reverse causality, and if there is, I >think it is much more likely to be discovered by peering at an >oscillosope in a high energy physics lab than by peering at cards with >mystical symbols on them by candlelight, Here here! > HOWEVER, I think experiments >CAN and HAVE been designed which would require reverse causality to >explain a possible outcome of. Could you go into detail on this? That "require" seems a bit strong. I would say that Frank Adams' assertion is on par with Poincare's famous claim that given a choice between a physical theory using non-Euclidean geometry and its mathematically equivalent embedding in a Euclidean space, physicists would of course opt for the latter. I for one cannot think of a physical theory where reverse causality is the preferable explanation. Nor have I tried. > Saying that no such experiment is >possible is saying that belief in forward-only causality is non- >scientific. All scientific beliefs are in principle falsifiable by >experiment. Yes and no. I'd say a lot of so called "scientific beliefs" are in the category of semantic conventions. For example, does the earth go around the sun, or vice versa, or around their mutual center of gravity, etc? All those questions are answered "yes" in an appropriate frame of refer- ence. One can trivially come up with coordinate systems where the earth is "hollow", etc. The modern attitude is that while certain frames are clearly more natural, and our terminology reflects those frames, the ac- tual physical "reality" lies above particular reference frames. According to Martin Gardner, people used to get in intense flame wars over the question of whether the moon rotates or not. Willard Quine and others would go further and maintain that NO scientific beliefs are falsifiable in principle. ucbvax!brahms!weemba Matthew P Wiener/UCB Math Dept/Berkeley CA 94720 "Do not believe astrophysical observations until confirmed by theory."
harnad@mind.UUCP (Stevan Harnad) (11/09/86)
> All scientific beliefs are in principle falsifiable by experiment.
For example, I suppose, the scientific belief that all scientific
beliefs are in principle falsifiable by experiment...lambert@mcvax.uucp (Lambert Meertens) (11/09/86)
In article <240@sri-arpa.ARPA> KFL@MX.LCS.MIT.EDU (Keith Lynch) writes: [>> = franka@UCB-Vax.arpa (Frank Adams)] >> I doubt that it is possible to design an experiment whose results could >> require reverse causality for their explanation. > [...] HOWEVER, I think experiments CAN and HAVE been designed which would > require reverse causality to explain a possible outcome of. So tell us what they are/were. > [...] All scientific beliefs are in principle falsifiable by experiment. This last claim is not a scientific belief either, then, I'm afraid. To me, "scientific belief" is a contradiction in terms. The whole fuss about falsifiability can be reduced to the self-evident observation that a theory that does not produce sufficiently specific predictions (which then, of course, might turn out false) is pretty useless, at least as far as predictive power is concerned. This is just one of many reasons why a theory might be predictively worthless, others being that it consistently predicts false things, or that it predicts the same as some simpler theory, or that it is too specific. (A predictively worthless theory may still have redeeming qualities, such as providing a framework that facilitates the construction of predictively powerful theories.) These requirements on theories are basic scientific principles; they are not scientific theories themselves--nor does anyone claim they are--but follow from plain good old-fashioned common sense. To come back to the original point in question: at least in the exact sciences (and in particular physics), for which theories take the form of a mathematical model, causality is not a concept built into the theories, but a manner of speaking informally about them. At least in the way I use the word "cause", I only say that event A causes event B if A precedes B in time. If someone would come up with a "reverse causality" contraption with a knob that I can turn freely, and it appears that something in the past (B) is consistently correlated with the way I choose the knob setting (A), there is a paradox unless I can observe B only after A. There are then two possibilities: (I) For some reason it is (given the set-up) *in principle* impossible to observe B before A. (II) The correlation only holds if I refrain from observing B before A. In either case, the true event is of course B' = my observing B. Now B' follows A. Causality (my use of the word) is not violated. HAPPY ENDING: The inventor of the contraption is awarded the Nobel prize in physics. Endless discussions in sci.physics ensue whether B "really" happens or is just "observed to have happened". New Capras and Zukovs helpfully point out to us that this is what Zen was about all the time. -- Lambert Meertens, CWI, Amsterdam; lambert@mcvax.UUCP
ins_akaa@jhunix.UUCP (Ken Arromdee) (11/15/86)
>> All scientific beliefs are in principle falsifiable by experiment. >For example, I suppose, the scientific belief that all scientific >beliefs are in principle falsifiable by experiment... Actually this does make sense. This doesn't lead to a contradiction unless one such experiment produces the paradoxical result. As long as such an experiment doesn't produce such a result, there's nothing to worry about. Being falsifiable in principle isn't going to produce contradictions unless real falsification occurs. -- Never look a gift horse in the mouth, but beware of Greeks bearing gifts. Kenneth Arromdee BITNET: G46I4701 at JHUVM and INS_AKAA at JHUVMS CSNET: ins_akaa@jhunix.CSNET ARPA: ins_akaa%jhunix@hopkins.ARPA UUCP: {allegra!hopkins, seismo!umcp-cs, ihnp4!whuxcc} !jhunix!ins_akaa
harnad@mind.UUCP (Stevan Harnad) (11/19/86)
In article <3942@jhunix.UUCP>, ins_akaa@jhunix.UUCP (Ken Arromdee) writes: > >> All scientific beliefs are in principle falsifiable by experiment. > >For example, I suppose, the scientific belief that all scientific > >beliefs are in principle falsifiable by experiment... > > this does make sense... doesn't lead to a contradiction unless > one such experiment produces the paradoxical result. What is "the paradoxical result"? If it is true that "All scientific beliefs are falsifiable by experiment" (F) and F is itself a scientific belief, then either (1) F is falsifiable (and I'd be interested to hear a potential scenario for getting such a result), in which case the generalization is valid, or (2) F is not falsifiable, in which case the generalization is simply false. There's no paradox here; no self-denying statements. Just the (Popperian variant) on a problem as old as Hume: The apparent impossibility of justifying induction without already presupposing or relying on it in doing so. -- Stevan Harnad (609) - 921 7771 {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet