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

Was: Element properties.

  1. Sep 22, 2005 #1
    We can say that the macro properties are emergent properties (non existent at the previous level of complexity, that of atoms) dependent also (apart the quantum laws) on the complex arrangements of the elements at the 'lower' structural level. In the case of many such emergent properties (such as wetness, solidity, colour) 'reduction' at the quantum theories is possible (if we add some extra assumptions needed to bridge the languages used at the macro level and that used by QM). It is conceivable that the process can be extended at even higher levels (for ex. that covered now by the 'higher level' sciences such as biology, economics etc), this at least in theory, for it is becoming very quckly virtually impossible to deduce the properties of very complex systems from QM (even using approximations).

    However no one can be sure that all emergent properties are truly reducible to the laws of QM, there are scientists (see for example Paul Davies in 'The fith miracle' or his article in the book 'The next fifty years') who believe that there are truly emergent laws at very high levels of complexity which cannot be reduced (notwithstanding that they are survenient on the quantum level and quantum laws they cannot be deduced from there). The most well known case is that of consciousness who might be such an irreducible property of matter appearing due to very high complexity at very high organizational levels.

    Why is it so? More generally, why are emergent properties-reducible or not-such as we observe them? Well the answer, fully compatible with the actually accepted notion of rationality, is that we have to just accept that this is a fundamental feature of nature, this is how nature work (letting the door open to something more fundamental, such as a Creator's will, this 'cluster' of hypotheses is by no means totally discredited; though of course it has no epistemological privilege currently).
    Last edited: Sep 22, 2005
  2. jcsd
  3. Sep 22, 2005 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Ha, an old wound is ripped open :biggrin:

    Yeah, that's indeed then the consequence of letting go the "reductionist" hypothesis.

    To come back to the OP, reductionists (that is, people who think - as I do - that there is ONE universal set of laws of nature, which we may for the moment only know partly, which govern the microworld and from which ALL macroscopic properties can be, at least in principle, though not in practice) believe that IN PRINCIPLE (that means, if we had enough computing power, and if we knew well enough the laws of nature), you can derive all the properties of, say, carbon as a macroscopic body from 1) the knowledge of the carbon atom and 2) the initial conditions - what you did to them.

    In *practice* however, that's not the right approach because certain properties would involve calculations which are intractable. It is then more practical to set up a phenomenological model "as if" this were another law of nature, and there are indeed scientists who claim that this is not only a practical approach, but is fundamental.

    My "simple proof" (ok, it is kind of a joke of course) for reductionism goes as follows:
    1) we are supposed to know the laws who govern ONE atom (call it quantum mechanics).
    2) if we know that N atoms are governed by the microscopic laws of nature, then we can add ONE EXTRA atom, having now N+1 atoms, and we also know how they are governed by the microscopic laws of nature (in QM, we have the hilbert space H_N x H_1 and the hamiltonian gets extra terms).
    3) by induction follows now that we know that a set of M atoms is governed by the microscopic laws of nature, for M any natural number.
    4) any macroscopic system contains a finite, natural number of atoms M, and hence is governed by the microscopic laws of nature.

    :smile: :smile:

    But apparently, there are serious people holding other opinions on the subject. In any case, they are right *in practice* for certain properties, because when correlations become long, the computations become often totally untractable.
  4. Sep 23, 2005 #3
    'Specific physics answers' can be done only in the broader philosophical context I've outlined above, the problem of emergence is not so easy as some might believe.
  5. Sep 23, 2005 #4
    Yes the actual scientific program is reductionist in nature, many scientists hold to this view (actually only a 'weak' kind of reductionism is possible, we always need extra assumptions to link the languages used by theories dealing with different levels of complexity). However many of them are skeptical that we will ever have the 'computational power' to reduce 'in fact' biology or economics, for example, to some basic laws. So we should better accept a compartimentalization of the domains of science upon the level of complexity (Feynman for example talked of such a division), each with its specific laws without bothering too much of reduction (though holding that in principle it is possible, as you did). Thus it is fully acceptable to have, for example, a theory of good observations which is not expressed in terms of QM.

    Can we sustain reductionism based on previous experience? Well we hit here the full might of the problem of induction, personally I'd argue that there is no good reason to accept full reductionism only based on reliabilism (it 'worked' well thus far), at most we can grant it currently the status of first research program deserving to be pursued further (your argument is not valid, anyway it is inductive and cannot count as a sufficient reason).

    Besides as I've stressed before enough many scientists (and way many philosophers) believe that there are laws of complexity which cannot be reduced (not even in theory) to some fundamental laws at an ultimate (previous) level of complexity (these laws do not contradict those basic laws, being in fact survenient on them). Now it is very difficult to make the difference between true emergent (irreducible) properties and emergent properties that cannot be reduced due to some computational limits...remain to be seen if new facts could give us new insights.

    I stop here, as I see philosophy is not welcome here, be it only philosophy of science...but things are not at all so simple how some believe, without philosophical / epistemological investigation I doubt we can make any progress (why do so many scientists lack a 'healthy' training in logic and philosophy of science? I think this is the result of the educational process, I myself discovered these facts only after university...when I realized that we can talk even of 'indoctrination', true science does not need this to establish itself as the best 'tool' we have at our disposal, [currently at least], to make sense of observed realities...).
    Last edited: Sep 23, 2005
  6. Sep 23, 2005 #5


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2018 Award

    Open a solid-state physics or many-body physics text.. and these are not "reductionist" approach either!

    But oh wait, such physics studies are irrelevant since these things can only be done in the broader philosophical context that only you have outlined.

    Oh brother!

  7. Sep 23, 2005 #6


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    But that is exactly the status in practice, and it will probably always remain so. In a way, the "reductionist in principle" approach is a moot point for more practically inclined people, because it might be that certain computations "ab initio" would require a computer about 10 times the size of the observable universe and calculate several billion years to find its answer. As Zapper pointed out, in condensed matter, although SOME things are done almost "ab initio", many things are phenomenological models inspired by more fundamental theories and with extra experimental input.

    The problem I have with "holism" (if that's the negation of reductionism) is that "anything goes". If at any level of complexity, just any "new" law of nature (not contained in the "underlying layer", so not an emergent property of the underlying laws, and so totally UNPREDICTABLE from the underlying layer) can pop up, then this would undermine a lot of experimental results. Just to give you some silly examples: we know about the inner structure of the earth mainly by seismic wave propagation, and by using a "reductionist" approach that 1) the materials down under under pressure and high temperature behave as smaller samples in similar conditions do in the lab, hence giving us the constitutive equations which determine the wave propagation and 2) the agreement between the observed propagations and these models.
    But in a holistic approach, there's no reason why a molten ball of iron and nickel of a certain size should not simply start to behave completely differently than what the "reductionist" approaches of continuum mechanics would indicate. There's no reason why suddenly no new law of wave propagation would pop up out of nothing. Hey, maybe a big ball of marshmallows behaves about the same as what we expect (erroneously) of an iron/nickel mixture when it is big enough. You never know if there's no new law of nature that only works for marshmallow spheres larger than 76 km in diameter. You cannot (holism!) deduce that from smaller samples.

    When I said "by induction" I meant MATHEMATICAL induction, namely that:

    1) if a property is valid for 1
    2) if the statement: "if the property holds for n THEN it holds for n+1" is true


    3) the property is true for all natural numbers.

    Except for the "loose grounds". That's why I have a hard time understanding these scientists (although, as you say, there are many serious people believing that). Here's the problem: suppose that for certain cases, you CAN do the computations, and derive macroscopic properties and even estimate the errors due to the approximations you make, ab initio. Suddenly, for these cases it is FORBIDDEN for new laws to arrise - that would contradict the results of the calculations of the reductionist laws ? They can only arrise when we CANNOT do the calculations ? So the laws of nature depend on our ability to do calculations ?

    This is what I wanted to illustrate with my marshmallows example (which is of course a bit ridiculous). We think we master the propagation of elastic waves in "marshmallow". We never (I think :smile:) tested it on a ball of 76 km diameter but we assume that the "reductionist" laws are valid because we have no indication that they are not working (no long range correlations etc). So could there nevertheless be a NEW law of physics only working for wave propagations in marshmallow balls bigger than 76 km IN CONTRADICTION with the reductionist calculation ?

    Other silly example: maybe the law of Archimedes is only valid for objects smaller than 512.34 m. Anything larger doesn't follow that law anymore and melts immediately when immersed. New law. Not contained in any way in previous laws. Could happen.

    If that is true, physics is nothing else but "stamp collecting" because at just any scale, just any new law can pop up and invalidate the most secure calculations using previously well established "reductionist" laws.

    But again, I think that *in practice* there will always be a partly non-reductionist approach using a mixture of phenomenological models and "ab initio" calculations, depending on what works best.
  8. Sep 24, 2005 #7

    Nothing impedes the possibility that aberrant theories (slightly different versions than very successful otherwise theories) could be more accurate, I do not know of any sound argument strongly justifying the belief that the universe is uniform and isotropic at all levels. As a matter of fact philosophers of science have a very difficult time to justify the preference for simpler theories, see for the example 'Goodman's new riddle of induction'.

    But not of this possibility talk those who believe that there are irreducible laws of complexity. Those new laws of complexity are merely supervenient on the laws at 'lower' levels, in fact they do not contradict those laws, they are only irreducible logically to those laws (as I said consciousness is one of the most important candidates at this, that is its properties might depend on some fundamental laws of physics, being nonetheless supervenient on them, irreducible logically to them). In other words consciousness might be caused by the properties of matter at 'lower' levels being in the same time irreducible logically to them. This implies that such laws are purely phenomenological laws and should be considered fundamental features of nature, the way nature works stop; our only way to realize that they exist being by direct observation at those levels of complexity.
    Last edited: Sep 24, 2005
  9. Sep 24, 2005 #8
    If I understood well the question was way broader, so it is at least misguided from the part of a 'serious' scientist like you to fail to mention that there might exist irreducible emergent properties for which we could only say that 'this is how nature work, they are fundamental properties of nature' (or that there are higher level laws for which we haven't found yet any reductionist approach-which really gives us 'explanations'). I'd rather suggest you to take a good course in the philosophy of science before taking such action of moving the threads (you did the same things in the past with some other threads, otherwise totally legitimate though the scientific community do not prefer the ideas expressed there). I appreciate your passion and dedication you put in moderating your forums but I'd appreciate you even more if you wouldn't do that in the Torquemada style (he had shown the same passion and unshaken belief in his 'truth' as you do). No we cannot make a sharp distinction between science and pseudo sicence neither have we the right to judge alternatives from a 'god's eye' standpoint as you dogmatically do, so often unfortunately...

    [edit to add]I have to add few clarifications, this is the last thing I will add here (and as much as you'll be moderator on the QM forum I'll never post there for your relief). If you don't explain also the limitations of our theories you are doing a poor job. That's simply poor science, you are deluding people to think that our models are complete and that at least from a set of phenomenological laws + some quantum considerations we can explain virtually all macro properties. This might be right but for the moment we just don't have sufficient reasons to think so. Finally where is the honest recognition that science (especially physics) is always provisional? It's easy to say that you are doing only 'science' unfortunately doing this uncritically (meaning much more than merely working inside a certain methodology or paradigm) is equal with zero, personally I expect way important new insights from maverick scientists than from rigid thinkers like you. Don't take as something personal it's a mere constatation, unfortunately actual science is not the unassailable monolith which so many believe. I really believe that a real science enabling us to understand even the meaning of life need to go way above the actual methodology...

    Basically there is no reason currently to believe that logic apply wioth necessity to natural facts. The hidden premises made by science must always be made evident, we must clearly say that they are, provisionally accepted, basic assumption of science (not truths set in stone forever). I'm afraid some part of the old dogmatism existing before Copernicus, Galileo, Newton and so on still live with us today...in my view we should clearly explain people that we take as basic assumptions that we can understand nature (indeed we have a problem here, dogs for example cannot understand QM why should we be so sure that we can understand the basic laws of nature?), that there is a single path toward truth, that finally nature itself make the difference between truth (at least approximate truth) and falsity (new data can make the clear difference between alternatives)...By the way better change your motto, the greatest scientists were philosophers too (though with their blunders) that would certainly raise your credibility.
    Last edited: Sep 25, 2005
  10. Sep 30, 2005 #9

    That's a shame. Still, I expect you'll be plenty entertained posting here with all the 'serious' philosophers.
  11. Sep 30, 2005 #10


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    This is what I don't understand: the "lower laws" predict - at least in principle - completely the *behaviour* on "higher" levels. Sometimes this is mathematically untractable, but at least there is in principle a single and unique solution to ANY behavioural question that could be asked, at just any level. So in what way are those "higher laws" supervenient but not in contradiction with the lower laws ? Because OR the higher laws give exactly the same predictions as the lower laws, in which case I don't see the utility of them, OR the higher laws give DIFFERENT predictions than the lower ones, in which case they are certainly in contradiction with them.
    The flaw is probably to say that "the lower laws do not predict every aspect" of the complex system. Sorry, they do - in principle. You can write down the mathematical problem that would give you the solution. In many cases, we don't know how to SOLVE the mathematical problem but the question is well-posed, so a solution exists (in the Platonic sense). It is this existence, in principle, of a solution which makes the claim of "higher, irreducible but non-contradictory" laws impossible. Or they are compatible but superfluous, or they are incompatible, in which case we simply learned that the lower laws are not valid.

    The example of consciousness is interesting of course - especially because I'm a kind of believer that consciousness is a needed concept to explain quantum theory (at least in its current form). But there's a fundamental difficulty with consciousness as a "supervenient law" because consciousness in no way has any behavioural implication. A stone can be conscious or not, you'd never find out and all its behavioural properties (hardness, melting, ...) would be independent from the fact whether it was conscious or not.
  12. Sep 30, 2005 #11
    That makes no sense.

    Even if emergent properties are independent of the systems which make up a "high-complexity object", they still affect that object or they would not be detectible properties. If they affect the object, they affect the systems that make it up-- unless you want to postulate some "emergent systems" made of "emergent matter" which are acted upon by "emergent forces".

    If they affect those systems, we have physical systems being affected by properties and physical laws that do not apply to them. This is a contradiction.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook