Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

How can anything be confirmed?

  1. Aug 31, 2004 #1
    i know quantum mechanics hints at this, but truth is, its a scary thing. how do we know that within the past 13 billion years or so, onditions have been met which distorted space-time to unrecgonicable or descriable points. much the way black holes do now, except why stop there? how can we say for certain what happened after the big bang, or if there ever was a big bang? if time and space remained constant, then our lives would be much easier. but they don't, and their distortions are unpredictable at best. if that not so, then explain to me why not.

    i know experimental results have confirmed much hypothesis, but different means can often achieve the same ends, much the way 5-3=2 and 1+1=2. that may not be a good analogy, but my point is, that there could be reasons for stuff happening that differs from popular reason. this has already been proven as theories are constantly being proven for a while, then debunked. but whos to say the cycle has an end?
     
  2. jcsd
  3. Aug 31, 2004 #2

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    1) Nothing is even proven in science. Theories may be disproven quite easily -- all you need is one experiment that does not do what is predicted -- but even 10,000 experiments aren't enough to prove a theory, because it's always possible the next experiment will not follow suit.

    2) The interaction of time and space with mass is not "unpredictable" at all. It is extremely predictable. It is so predictable, in fact, that we have a very elegant and simple theory which describes their behavior extremely well: general relativity.

    3) Science may never have an end. No one knows yet, and we'll never know 'til we get there.

    - Warren
     
  4. Sep 1, 2004 #3
    That's true about relativity, but relativity itself is just another theory and has nothing to do with the original question which concerned QM. The answer is, that according to QM there is a chance that our very existence, right here, right now, was just a random fluctuation. The entire past history of life, the universe, and everything including ourselves could be completely misleading.

    However, the odds that this is the case are so low that pigs are more likely to grow wings. Indeterminacy is what QM is all about, so if you are attempting to use it to make any kind of absolute confirmation you will always be disappointed. All you can do is make statistical predictions and, of course, there are lies, damn lies, and statistics.
     
    Last edited: Sep 1, 2004
  5. Sep 1, 2004 #4
    My only concern is that the 'Human Life Form' is probably ill-equipped to ever do so. In fact, I have seen nothing which rules out this possibility!
     
    Last edited: Sep 1, 2004
  6. Sep 2, 2004 #5
    Like what? This is a common misconception; that theories are proven and then suddenly shown to be bunk. If a theory is shown to predict results, it won't be shown to suddenly not predict results within the regime it was tested in. What can be shown is that it only predicts results within a certain domain.

    For example, I constantly hear how einstein "proved newton wrong." If that's the case, and newtonian mechanics has been proven wrong and relativity is the successor, why do newtonian mechanics still work so well for non-relativistic domains? I'm not going to use relativistic mechanics predicting the forces on a ladder or a bridge. I'm not going to use relativistic mechanics to find the distance it takes my car to stop on a road. Relativity might have a broader range of situations it can theoretically be used in, but it is so much more complicated that it is only used when necessary.

    I believe a substantial part of your post stems from your mislabeling philosophical interpretations of science as science.
     
  7. Sep 2, 2004 #6
    This can't be the case, Locrian, because then anything with predictive power can be a valid scientific theory. However, the nature of a scientific theory should be to establish an understanding of the most fundamental interactions with which it is concerned. Thus, since relativity is a generalization in some sense of Newtonian mechanics and is therefore a more fundamental theory, it is a more valid theory.

    So I'd agree with you that Newtonian mechanics isn't garbage, but one could quickly conceive of a theory with predictive powers that is complete hogwash, such as the atomic model postulated by Dalton and company.
     
  8. Sep 2, 2004 #7
    Well yea - any theory with predictive power is science. And I do believe that includes ones that sound strange to us now, or have been improved on since; if they still predict results in their regime, then they aren't debunked, though they may have been improved upon.

    I'm not sure what you could mean by Dalton's atomic model; maybe my chemistry history isn't up to par. As far as I know, Dalton never modeled the atom (though he guessed at their sizes, and was wrong), but instead built his theory of chemistry on the fact that there is such thing as an atom, and that atoms of different elements interact in integral numbers without altering each other's element. This is an extremely useful concept and, as far as I'm concerned, it is alive and well.

    Of course, Dalton's belief that heavier atoms took up more space in a gas was wrong. There was evidence against that during that period of time. Some of Dalton's assumptions concerning chemical reactions were also wrong, which were also disproven during the 19th century.

    In other words, where his theories were good predictors they still exist in modern science, and where they were bad predcitors they have been replaced. This is exactly what I said above; philosophical renderings of a theory are meaningless, the theory either predicts well or it doesn't.

    I'd love to see a list of theories the original poster feels have been proven and then debunked. I use 18th and 19th century physics all the time, so seeing people write things like that is like being told my hammer doesen't actually work because a nailgun exists, despite the fact that I use the hammer every day.
     
  9. Sep 2, 2004 #8
    First off, the hammer analogy is a misleading one in the sense that, whatever it takes to get the nail in, so long as the nail goes in everything is fine.

    Not so with science: theories have the added requirement that they must be general and fundamental. For instance, there is no theory for the law of action and reaction for nailguns - even though a perfectly useful one could be developed with acceptable predictive powers. However, common sense tells us that the same law of action and reaction should be applied to many more phenomana, and thus a more general rule is developed.

    This common sense reasoning is not philosophical, it is purely scientific. It is the motivation behind current attempts at Grand Unification Theory and it is axiomatic to all scientific endeavour. Each and every thoery deals with the general and not the particular. Because of this, theories can be developed which are quite plainly wrong even though as they are developed they are evidenced in a few or many particular examples. For instance, Dalton espoused the view that atoms were essentially nondivisible spheres. There was no reason to disbelieve him, but he was wrong. He made a general claim about atoms, and he was proven wrong. Its really that simple.
     
  10. Sep 3, 2004 #9
    Biological limits to human scientific research efforts

    • ...let us make no mistake, it is biological evolution of mental capacity, and not merely an accumulation of scientific research data that is needed. The time must come — strange as the idea may seem, even to scientists — when, unless man himself evolves greater mental capacity as such, a "principle of diminishing returns" will show itself in relation to research effort. All that is discoverable and comprehensible at our level of intelligence may soon have been discovered and comprehended.

      ...a point is reached where the message written on the cloudy face of nature will span too complex a set of relations for us to read them. Our scientific instruments may record them sensitively and our scientists may debate the riddles offered, but perhaps in vain. There will always be geniuses, one hopes; but just as a genius among homo neanderthalus, or among the primates, might not, in our culture, get beyond recognizing one-syllable words, so the genius of the next century may find himself unable to read any further in the open book of science.
    (Raymond B Cattell. A New Morality from Science. Chapter 3, The Basic Logic of Beyondism. Section 3.9. Note 5.)
     
    Last edited: Sep 3, 2004
  11. Sep 3, 2004 #10
    Merely doubting is still rational-we do not even have currently a clear,sufficient,reason,beyond reasonable doubt,that QM [the standard formalism] is more probable to be true-but claiming that science is deceived or fundamentally limited is not,from logic and all objective knowledge we have now at least.[Basically] nothing is certain (I mean enunciations about the world),not even QM,but we are still entitled rationally (more 'positive' reasons for) to define provisional,openly accepted as fallible,objective knowledge,properly justified,the best we managed at a certain moment,even in absence of certitudes :-).

    I advise you to read first Russell's 'The problems of philosophy' before attempting,potentially,to give yourself a more precise answer to your own question.Afterward try reading some serious books on philosophy of science.Certitudes,I'm afraid do not exist...The book's content is representative even now,though of course views have evolved and,as almost everything in philosophy,many serious philosophers disag[r]ee (having reasons) with some of his points.

    http://www.popular-science.net/books/russell/chapter2.html

    [edited to correct the layout]
     
    Last edited: Sep 3, 2004
  12. Sep 3, 2004 #11
    There is no such requirement on theories. To be a scientific theory, it must predict experimental results as hypothesised, and that is all. That "common sense" reasoning is philosophical, since it isn't part of the scientific method. You are adding philosophical baggage to the definition of scientific theory that may or may not be useful, but certainly doesen't define science. As for a GUTs, unless they predict new experimental results, they are nothing more than intelectual excersizes.

    So... his view that atoms were nondivisible spheres wasn't tested, and was therfore a hypothesis and never a theory. This does not fit your initial proposal that you can concieve of theories with predictive powers that are hogwash. On the contrary, this hypothesis of his had no predictive power, and therefore was tossed out when it was tested.

    Again, Dalton had no information available to speak about what an atom was made up of; what he did have available was evidence as to how they chemically interacted (which was in integral numbers). Where he followed the evidence his theory is still good, and where he didn't, he never created a a real theory, and others moved on. To put it another way, Dalton made hypothesis that were based on experiment and they are good, and he made arguments based on philosphy (unacceptably connecting the chemical interaction of an atom with it's structure) which were wrong.

    You keep saying I'm wrong, but the examples you provide support my argument.

    I'm still waiting for the initial poster - or anyone else - to provide a good list of theories that were proven and then debunked. So far all I see pointed out are parts of theories that were never proven and then debunked, which is hardly the same thing.
     
  13. Sep 3, 2004 #12
    By virtue of the fact that Dalton relied on inductive reasoning to arrive at his theory that atoms were nondivisible spheres, the theory was inherently tentative and fallible, since inductive reasoning does not lead to certainty. Tentative means that the theory was the best one to go with given the empirical evidence, and fallible means that some further studies down the road may falsify the theory. It is clear that when we look back, Dalton was not entirely accurate with the concept of the atom. However, it did provide a strong foundation for further theory development that led to a more accurate view of the atom. A hundred years in the future, scientists will most likely look back on our silly ideas and wonder what the heck we were thinking.

    Any scientific theory is tentative and fallible, regardless of the degree to which the theory corresponds with reality. Some theories may be accurate to 10+ decimal places, but this does not necessarily guarantee that the theory is perfect or complete. Some theories are extraordinarily successful at making predictions, but this cannot imply that they represent the absolute truth. The continuous perfection of scientific theories is one motivation for pursuing science.
     
  14. Sep 3, 2004 #13
    Thanks for the link, I forget sometimes how well Russell puts things.

    However, I wonder if you'll humor me for a moment. The quote above seems to me to be philosophically sound. My question is, what would you do if you called your bank and found out you had no money in it. By your calculations, you should have quite a bit of money in it. Upon asking them where it went they replied,

    "I'm sorry sir, but the knowledge of your balance you have is provisional and accepted as fallible. Keep in mind that you have not deposited any actual matter here, but instead were dealing in electonic money, which is an entirely philosophical idea and cannot be physically substantiated (which would hardly constitute proof in any case). Since you cannot provide overwhelming proof that the knowledge of which you speak is not entirely subjective, we see no reason to accept your point of view."

    Would you accept that answer, or would you say "Gimme my ******* money" ? (Or both?)
     
  15. Sep 3, 2004 #14
    You've only replied to half of my last post. Refer back to it for a clear example of a ludicrous theory.

    Early QM theory predicted a massless neutrino. With the exception of actually testing the mass of the neutrino, this theory showed remarkable success in predicting outcomes of a multitude of experiments.

    If you are too say that each and every theory is inductive by nature until such a time that all avenues of its predictive power are tested by the method of exhaustion, then you will have a tough time finding any theory that abides by your standards of scientific methodology. A theory devised to explain any one phenomena will often entail unknown (at the time) predictions concerning any number of other phenomena. This is often what shows a theory to be wrong, and not just 'a little off'.
     
  16. Sep 3, 2004 #15
    And I responded to that second half of your post in the second half of mine, which you even quoted. I do not understand what else you could be refering to.

    This is not true. Quantum Mechanics makes no statement about any particles at all without first providing data. What information was used to determine the mass of the neutrino? What assumptions and data were involved? Show these calculations and lets discuss them.

    No I won't - you misunderstand my standards of scientific methodology. I demand only that a hypothesis be experimentally tested within a regime. I consider it suspect when used outside of that domain. That a theory must be proven outside of that regime is a standard you have placed upon it, not me. You make the point that a theory can make predictions about things in which it hasn't been tested which may turn out to be wrong; I consider that the result of misusing the theory, probably due to over-philosophising its meaning.

    I want to be clear that nowhere here have I stated that a theory cannot be improved upon, or that a new theory could not encompass an old one. I contend only that once a hypothesis is shown to predict results, then that hypothesis is never debunked - well, unless the test turns out to be invalid or the scientist dishonest, in which case it was never proven to begin with. The history of physics isn't littered with the corpses of broken theories as the original poster would have us believe, but is instead a story of continual improvement, and many cases the "old" one is so robust as to be useful long after "newer" theories are developed.

    I am interested in your neutrino example, by the way, and look forward to discussing it.
     
  17. Sep 3, 2004 #16
    I was referring to the almost hilarious 'law of action and reaction for nail-guns'. I realize fully how contrived this example is, but I think it is a valid counterexample to the reasoning you have put forth so far in this forum.

    I shall show this to you next week sometime. Stay tuned (you can refer to Griffihs Particle Physics for a great expose on the subject).

    No way. This is not over-philosophising in the least. To continue with an earlier example, the fact that neutrinos have mass implies a number of other predictions within the current Standard Model of physics. These consequence have not all been tested, but the massive property of the neutrino has been shown experimentally as have a great number of other parameters within the Standard Model. Thus, even though all possible predictions have not been tested, the theory of the Standard Model is taken as valid and correct and its consequences are assumed. This is not to say that the theory might soon be thrown out the window when these consequential predictions are found to be false.

    I understand your position, and disagree with it. But that is ok. Everyone is allowed to be wrong once in awhile. :smile:

    NB. wrong is defined to be in contradiction with what I hold to be right.
     
  18. Sep 3, 2004 #17
    The key of the fallibilistic approach is that knowledge is about justification,if there is proper justification,sufficient reasons,then one is entitled to say that one knows something.There is no need for certainty (proof) though fallibilists do not claim that it is impossible to be mistaken.In the case you brought about,by applying a careful,scientific type,research it should be easy to find the sufficient reasons (with the assumption of transparency,access at all relevant documents,and of course honesty) I need to persuade all rational people (judges) that I have deposited some money in that bank and that they should still be there.The scientific method is still our best 'tool' so far there is no claim against this though,by the way,only as the norm,especially if we refer at physics where underdetermination is really a problem.In the case you put forward those who claim otherwise (namely the bank stuff) in front of such scientific results should either provide at least an equal conjecture if not better (if they still want to 'work' under the actual scientific method) or to provide an alternative method which to be at least as successful (well,fruitful).

    What you probably intended was to point that science can attain even certitudes,or at least that we can know surely that we are very close,in particular cases at macroscopical level at least,where the problems plaguing enumerative induction do not exist.Strictly speaking there is no proof that underdetermination of theories holds at all levels (once we reject all forms of idealism of course,this is one of the basic assumptions of science) but certainly it is a problem for physics (the base of chemistry,biology and so on) where non trivial changes do exist,we have examples,as history of science proves plenty,sometimes contradicting not only common sense but also enunciations accepted only after careful observations,using the best scientific methods of investigation,the most reliable in the past,involving intersubjectivity,namely Mill's methods (of agreement,differences and so on).Unfortunately scientific knowledge does not consist (as you suggest) only in a continuos grow of scientific knowledge,nontrivial changes do exist,some previoulsy accepted concepts could disappear altoghether in the new paradigms.

    Thus we can grant,in my view of course,a supporter of Feyerabend would disagree with this,only a fallible epistemological privilege to a (minimal) scientific method based essentially on the fact [that] the 'scientific method' approach is still 'progressive',as the norm,in what fecundity is concerned (in spite of the proved to exist detours,only transitory however,it is still fully in line with Lakatos' 'scientific research programmes' approach).But fallibilism is essential,there are currently,at least,no sound arguments to back both enumerative and hypothetical induction.
     
    Last edited: Sep 3, 2004
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?