How do physicists deal with emergent properties?

[nightflyer]

I recently read a couple of books by Paul Davies, in which one of the main themes was that biology is characterized by emergent properties that can not be reduced to mere physics (self organization, consciousness and so on). I would like to hear what physicists have to say about this important topic? Why is it that we can not understand biology based on physical models today? Is it because biology is ultimately too complex to be described by mathematics (and will therefore never be understood in mathematical terms) or is it a matter of developing mathematical models that are complex enough? What I mean by the second alternative is that biology will ultimately be reduced to the more fundamental sciences of mathematics and physics once we know enough about how it works, by means of better technology to observe it.

 

[ZapperZ]

This is an example of how physicists deal with emergent properties:

http://physicsandphysicists.blogspot.com/2006/10/theory-of-everything.html

Zz.

 

[vanesch]

This is a touchy subject, and several physicists have different viewpoints on it. Broadly, there are two categories: reductionists, and anti-reductionists.

Reductionists claim that the abstract mathematical structure that describes correctly the microphysics should contain also all observable "emergent" phenomena.

Anti-reductionists claim that the laws of physics change when several entities come together, and that there is no general link between the microphysics and the macrophysics, except in those cases where one can explicitly derive a link.

Both categories of people agree however upon the methodological need to study macrolaws, in that even reductionist recognize that, even though according to them, the macrolaws can in principle be derived from the microlaws, this is not the most practical way to proceed (and maybe even in principle impossible because of mathematical-technical difficulties).

So although the philosophical views are different, the operational approach of both is the same.

I hope that I've been fair (I'm a reductionist).

 

[HallsofIvy]

Actually, I think the best answer is in what you said initially, "emergent properties that can not be reduced to mere physics". Since it the cannot be reduced to physics (I'm not sure any physicists would like the word "mere" here!), it is outside physics and physicists do NOT deal with it.

Of course, like anyone, physicist are interested in things outside their own field and have, as others have shown dealt with it in a fairly philosophical way.

 

[complexPHILOSOPHY]

I would assert that emergent properties are simply assumed as not very well understood, currently. As Vanesche pointed out, many reductionists believe the 'emergent properties' are simply the product of undiscovered or unproved variables and will eventually be understood. If the macroscopic, physical reality that we experience is an emergent aspect of the underlying, microphysics (with no visible transitions or interactions for us to observe) -- then we will never be certain as to how the universe fully operates and this poses a serious problem for physicists.

It seems to me that when we develop a field which appears in some ways, counter-intuitive to the previous approach (QM in relation to CM), we are initially confronted with 'emergent phenomena' which we must then find precedence for as we work through the problems.

Reductionists tend to presuppose the existence of causality, which operates from top-to-bottom as well as from bottom-to-top, so assuming there is ultimately some cause or variable, connecting the parts and the whole, it will eventually emerge.

I too struggle with the notion of emergent entities, properties and phenomena because I like to assume the existence of deterministic, physical laws, however when confronted with systems in which the whole is greater than the sum of the parts -- it is hard to maintain a rigid, reductionist position. Perhaps we need to start discussing the ontological status of how we classify and describe the constituents of reality to discover the root of the problem?

Have you read anything by Roger Penrose? He is (as far as I know), a staunch proponent of emergance (atleast in relation to quantum theories of mind) but I might be mistaken.

Do a search for 'holism' and you might find some interesting perspectives.

I also think that when you attempt to deconstruct complex, organized systems down into quantum or sub-atomic entities, that there will inevitably be massive holes, where we cannot find causation between the microscopic and macrocosmic. We are trying to reduce the whole of reality, into constituents that we have no direct interaction with -- it's difficult.

We are defining the universe on so many different scales, its a complex and intricate process that will take us a long time to develop an intuition for.

 

[ZapperZ]

I would assert that emergent properties are simply assumed as not very well understood, currently. As Vanesche pointed out, many reductionists believe the 'emergent properties' are simply the product of undiscovered or unproved variables and will eventually be understood. If the macroscopic, physical reality that we experience is an emergent aspect of the underlying, microphysics (with no visible transitions or interactions for us to observe) -- then we will never be certain as to how the universe fully operates and this poses a serious problem for physicists.

Whoa! Back off a bit. I don't think even vanesch would claim that.

Emergent properties such as superconductivity, fractional quantum hall effect, collosal magnetoresistance, etc.. etc.. are ALL well-understood! They are all described by quantum mechanics very well. It is just that the starting point for the description is the many-body ground state, not the microscopic interaction of the individual particles! There is nothing that says that QM must start at the individual particle interaction only.

I strongly suggest you read up on the references of emergent phenomena in physics. You'll find that some of the most well-known and well-described phenomena, with the highest degree of certainty, are these emergent phenomena. These are where we got the standard accepted values of "e" and "h".

Zz.

 

[complexPHILOSOPHY]

Whoa! Back off a bit. I don't think even vanesch would claim that.

Emergent properties such as superconductivity, fractional quantum hall effect, collosal magnetoresistance, etc.. etc.. are ALL well-understood! They are all described by quantum mechanics very well. It is just that the starting point for the description is the many-body ground state, not the microscopic interaction of the individual particles! There is nothing that says that QM must start at the individual particle interaction only.

I don't think I expicitly stated that, or atleast that wasn't my intentions. I stated that causality operates from top-to-bottom as well as from bottom-to-top.

I strongly suggest you read up on the references of emergent phenomena in physics. You'll find that some of the most well-known and well-described phenomena, with the highest degree of certainty, are these emergent phenomena. These are where we got the standard accepted values of "e" and "h".

Zz.

I am aware of this fact (and I wasn't trying to dispute it, I apologize if I constructed my language in that way), however, I guess my concern is not knowing precisely how these properties emerge and simply accepting the results.

I have not worked up into higher level maths or QM so my overall conception is still vague, however, my impression of emergent systems is one in which the whole is greater than the sum of the parts -- so perhaps, it is my assumption or presupposition that is creating the distorted perception?

 

[nightflyer]

Thank you for your replies!


ZapperZ:

If I understand you correctly, you are saying that emergent properties such as superconductivity might very well be understood in terms of quantum mechanics. However, what do you have to say about biological phenomena such as protein folding, DNA decoding and encoding and so on? Are these processes ultimately understandable in mathematical terms, or do they require "higher", emergent laws? I very much appreciate the link you sent me. The reason I am asking about these things is that I am trying to decide whether to study biology or physics; I am very much interested in medicine, but being a reductionist at heart I have a strong feeling that the "life sciences" of today will become more and more fundamental in the future, and move into the fields of engineering. Thoughts?

 

[ZapperZ]

I am aware of this fact (and I wasn't trying to dispute it, I apologize if I constructed my language in that way), however, I guess my concern is not knowing precisely how these properties emerge and simply accepting the results.

But you had no problem in "simply accepting" the value of "e" or "h".

We simply didn't just accept the results. We know the mechanism that caused each of these phenomena. The fact that there isn't a "bridge" currently that can work its way from the individual particle, to me, makes no difference in how accurate we can describe the system. And in many systems, it make no sense to try to go to the individual particle interaction. A "phonon" is a collective excitation. It has no definition when you go down to the individual particle scale. That's like trying to find the temperature of an individual particle.

Zz

 

[Q_Goest]

Hi Night'

If I understand you correctly, you are saying that emergent properties such as superconductivity might very well be understood in terms of quantum mechanics. However, what do you have to say about biological phenomena such as protein folding, DNA decoding and encoding and so on? Are these processes ultimately understandable in mathematical terms, or do they require "higher", emergent laws?

Regarding protein folding, Laughlin is a condensed matter physicist who has published a paper on this topic. He specifically includes protein folding as an "emergent" phenomenon. You can find his paper here:
http://www.pnas.org/cgi/reprint/97/1/32.pdf
I suspect DNA decoding and encoding could also be considered "emergent" but he doesn't specifically say that in this paper.

As you say, Paul Davies also writes quite a bit about this topic. Don't know if you've read this one yet, but you may also find it interesting:
http://arxiv.org/abs/astro-ph/0408014

The reason I am asking about these things is that I am trying to decide whether to study biology or physics; I am very much interested in medicine, but being a reductionist at heart I have a strong feeling that the "life sciences" of today will become more and more fundamental in the future, and move into the fields of engineering. Thoughts?

Hey Night'. If you are "very much interested in medicine" then don't worry about what's reductionist, emergent or whatever. It makes no difference. What's important is that you enjoy what you do. Don't go into another subject because you're concerned it may change in the future, it won't. Or at least how you feel about the work won't change.

"People rarely succeed unless they have fun in what they are doing."
~ Dale Carnegie
(so do what you enjoy!)

Regarding emergence in general, I think Vanesch put the right spin on it. There's a lot of discussion in philosophical circles regarding emergence which is largely BS IMO. The problem of emergence regards whether or not some phenomena is reducible. Unfortunately, I believe the problem regards how we can know if something is reducible or not. Laughlin for example has pointed out many examples, and I believe he's right, but I believe the problem with proving he's right is a lack of a solid method to determine how we can know if something is not reducible. What method can you use to prove a phenomena such as protein folding is irreducible? I don't think there's a good answer to that.

 

[ZapperZ]

Thank you for your replies!


ZapperZ:

If I understand you correctly, you are saying that emergent properties such as superconductivity might very well be understood in terms of quantum mechanics. However, what do you have to say about biological phenomena such as protein folding, DNA decoding and encoding and so on? Are these processes ultimately understandable in mathematical terms, or do they require "higher", emergent laws? I very much appreciate the link you sent me. The reason I am asking about these things is that I am trying to decide whether to study biology or physics; I am very much interested in medicine, but being a reductionist at heart I have a strong feeling that the "life sciences" of today will become more and more fundamental in the future, and move into the fields of engineering. Thoughts?

I think the problem here is that you're putting possible philosphical implications ahead of the "substance".

The link that I gave you to my blog entry had two references from Laughlin's PNAS papers. You might want to seriously consider reading those first and see if your question might have been answered there.

As far as the biological aspect of it, I don't want to make anything speculation based on something that I don't have an intimately knowledge of. I'm willing to use examples from condensed matter physics because that is what my background is in and I know what is involved in those phenomena that I listed.

Zz.

 

[Chaos' lil bro Order]

I think you are all talking about unifying QM with Classical physics. This is probably one of the biggest challenges in physics and there is no point in theorizing when, or if, it will ever be 'bridged', as Zapper says. What's important is that both accurately describe their respective orders of magnitude well enough to make predictions that are fallsifiable.

This also reminds me of the classic analogy, where a person does not need to know how an engine works, to properly drive a car. (But if the car breaks down, they may need to learn some QM to get it working again

 

[ZapperZ]

I think you are all talking about unifying QM with Classical physics. This is probably one of the biggest challenges in physics and there's no point in theorizing when or if it will ever be 'bridged', to use Zz's analogy. What's important is that both accurately describe their respective orders of magnitude well enough to make predictions that are fallsifiable.

Nope, that's not right. Superconductivity, which is an emergent property, is the clearest manifestation of quantum phenomena at the "macroscopic" scale. It isn't classical, nor are magnetoresistance and fractional quantum hall effect.

None of what has been talked about here has nothing to do with "bridging" classical and quantum description.

Zz.

 

[neurocomp2003]

if your looking to study biology or physics...look for a programme that does both (biophysics). Studies the whole cellular/molecular/protein level. Also there are some physicists who study DNA via strings...funny concept to me but they do.

From Discovery or Science: Marvin Minsky was quoted as saying something like if we knew allt eh fundamental laws of whatever system we are studying, we should be able to simulate it on a computer (of course to numerical accuracy...then again you can always increase it)...

pick up a book by Gary Flake.

and go to google scholar and search for articles on high performance computing in the subject of interest(ie. protein folding or molecular dynamics) and see what laws they use.

 

[vanesch]

To the outsiders: Zz and I have long-standing "debates" over these reductionist/non-reductionist issues

Emergent properties such as superconductivity, fractional quantum hall effect, collosal magnetoresistance, etc.. etc.. are ALL well-understood! They are all described by quantum mechanics very well. It is just that the starting point for the description is the many-body ground state, not the microscopic interaction of the individual particles! There is nothing that says that QM must start at the individual particle interaction only.

I would like to point out that even reductionists do not claim that a many-body system cannot have properties that are meaningless for individual particles, and that many-body states (especially in quantum theory) have properties which do not even make sense for individual particles. For instance, "temperature" is something which is quite meaningless for a system of only a few particles, but becomes meaningful for large ensembles of particles.

What reductionists claim, is simply that the emergent property can in principle be derived (or better, "mathematically exists, in a Platonic sense") by writing out the microlaws for the constituents, without any added input.
That is: you give me the laws by which electrons, nucleae etc... are governed, and - in principle - this statement contains all there can be said about any conglomerate. Anti-reductionists claim that you need new information beyond these fundamental laws without which the emergent properties cannot arrise.

However, reductionists are well aware of the formidable (and maybe impossible) mathematical task it might be in most cases to derive any of these properties - let even apart knowing what to look for ! As such, pragmatically, they agree with anti-reductionists in that the best way to make some progress in a field which is confronted to many-degree-of-freedom systems (like condensed matter physics), that it often is to get hints from experiment, and even to construct phenomenological models just starting from these observations, without trying to make directly a link with the underlying microphysics. The difference resides only in that reductionists believe that such a link exists, in principle, and anti-reductionists think that in many cases, no such link exists.

So in a certain way, reductionists always remain unsatisfied when a new "macrolaw" is found, as long as they don't have at least a toy model, based on microlaws only, from which they can derive a similar behaviour for a similar macroscopic quantity - while anti-reductionists don't believe this has anything to do with nature.

Anti-reductionists can point to the fact that many macroscopic phenomena exist for which no detailed derivation "from first principles" has been found yet. ZapperZ gave some examples.

Reductionists can point to the fact that many macroscopic phenomena DID find finally an explanation on the basis of derivations from first principles (at least in toy models, and sometimes in realistic models). This goes from many simple thermodynamic properties in statistical mechanics (starting with the kinetic theory of a perfect gas) to certain results in solid state physics.

Finally, there is sometimes a remark that has not much to do with this debate, but which is sometimes mentioned: it is very well possible (even rather probable) that what is now, today, considered as "fundamental physics" are in fact nothing else but approximate "macrolaws" of an underlying "microphysics", and so all we think of right now as "fundamental" are nothing else but "emergent properties" of yet a deeper layer of nature.

But this doesn't relate in any way to the discussion "reductionists" vs "anti-reductionists", because the same set of arguments from both sides would then simply apply to this next layer.

 

[Chaos' lil bro Order]

Nope, that's not right. Superconductivity, which is an emergent property, is the clearest manifestation of quantum phenomena at the "macroscopic" scale. It isn't classical, nor are magnetoresistance and fractional quantum hall effect.

None of what has been talked about here has nothing to do with "bridging" classical and quantum description.

Zz.

Nope. You are wrong. Superconductivity is a quantum phenomena discovered by Onnes in 1911. It does not manifest at the macroscopic scale. You should provide references or a clear example if you want to make such claims, this is a physics forum.

 

[nightflyer]

vanesch:

If I understand you correctly you are saying that even reductionists acknowledge macrophysical laws that might not be possible to derive from the microphysical ones, but the macrophysical laws can be understood in mathematical terms? Or am I missing something here? (I have a feeling I am...) In other words - the way I understand your argument is that different mathematical laws govern micro- and macrophysical laws, and that the macrophysical ones can not be derived from the microphysical ones?

ZapperZ:

I will have a closer look at the links, thanks!

 

[ZapperZ]

Nope. You are wrong. Superconductivity is a quantum phenomena discovered by Onnes in 1911. It does not manifest at the macroscopic scale. You should provide references or a clear example if you want to make such claims, this is a physics forum.

.. and if you have followed my previous argument about superconductivity, you would have SEEN the references I've given. For example, try Carver Mead's PNAS paper[1]:

Although superconductivity was discovered in 1911, the recognition that superconductors manifest quantum phenomena on a macroscopic scale came too late to play a role in the formulation of quantum mechanics. Through modern experimental methods, however, superconducting structures give us direct access to the quantum nature of matter. The superconducting state is a coherent state formed by the collective interaction of a large fraction of the free electrons in a material. Its properties are dominated by known and controllable interactions within the collective ensemble. The dominant interaction is collective because the properties of each electron depend on the st ate of the entire ensemble, and it is electromagnetic because it couples to the charges of the electrons. Nowhere in natural phenomena do the basic laws of physics man ifest themselves with more crystalline clarity.

So take that! But even without such a reference, we KNOW about superconductivity (at least *I* do), and that we can OBSERVE its effects macroscopically! Just look at those levitated trains, for heaven's sake! Those, by themselves, are clear indication of a quantum effect at the macroscopic level already! I didn't realize that I had to be THIS explicit in pointing it out.

Now it is MY turn to ask you for references to back your claim that "emergent" properties are classical, and that this whole thing is simply a classical-quantum transition.

Zz.

[1] C. Mead, PNAS v.94, p.6013 (1997).

 

[vanesch]

vanesch:
If I understand you correctly you are saying that even reductionists acknowledge macrophysical laws that might not be possible to derive from the microphysical ones, but the macrophysical laws can be understood in mathematical terms? Or am I missing something here? (I have a feeling I am...)

No, of course not, that is the anti-reductionist viewpoint (unless the subtlety lies in the word "derive", see further)

Well, reductionists think that the "Platonic mathematical solution" which is supposed to exist to the microphysics problem (that is, in condensed matter physics, the entire set of solutions to the unsimplified Schroedinger equation with something like 10^25 degrees of freedom or so) contains all the emergent properties that one can think of. Only:

- they are not necessarily sure that one day we will find a feasible derivation, simply because of the monstrosity of the mathematical problem (it is not because the solution "exists" that we can (even in principle) construct it explicitly: this is the "Platonic" part of it).
- they also acknowledge that even if one were given this entire solution, that it might not even be evident to recognize any macrolaws in it, if not hinted by experiment. (In other words, what quantity one should distill from this big mess ?).

They even recognize that "new physics" might be needed, but this would then be a fundamental change on the microscale: in other words, quantum theory, or the specific hamiltonian, would then have to be inadequate (we missed an interaction, or quantum mechanics is wrong, or whatever).

THIS is the fundamental difference with the anti-reductionist viewpoint.

What I don't know, is whether anti-reductionists consider:
- that the entire solution to the mathematical problem of the unsimplified microproblem simply doesn't exist (they might take on a constructivist mathematician's viewpoint (and not a Platonic one), and say that as long as there is no explicit derivation of a construction of the solution, well, then this solution doesn't exist)
- that this solution might exist, but will not reveal certain properties, even when looked for. In other words, for some or other reason, the observables corresponding to the emergent property cannot be defined over the microsolution.
-something else.

 

[Q_Goest]

Per Carver Mead's PNAS paper: Although superconductivity was discovered in 1911, the recognition that superconductors manifest quantum phenomena on a macroscopic scale ...

What is the philisophical difference between the phenomenon of superconductivity and conventional magnatism? I deal with MRI manufacturers regularly, and even have a patent in the field, but I fail to see the philisophical difference here.

Example: I have a magnet on my desk that hold papers up. The force is derived from the motion of the electrons in the iron. I'd consider this a manifestation of quantum phenomena at a macroscopic scale. Similarly, light reflecting off a surface and many other phenomena require interactions at a quantum scale.

So how is superconductivity fundamentally different? It is a phenomena which requires an explanation at the quantum level, like so many other things.

Note however, forces and momentum of objects, stresses in material, gravity, fluid dynamics, heat transfer and many other things also create phenomena, but they don't require explanation at the quantum level. Classical mechanics in general doesn't require any explanation of phenomena at such a level, and classical mechanics is obviously reducible to smaller 'chunks' of material, right the way down to the mesoscopic level.

One last thought. If we define "emergence" as Vanesch did (in terms of reductionism) then I think the explanation of why superconductivity is "emergent" requires one to address the issue of why it is irreducible even in principal to the interactions of electrons with a material.

 

[ZapperZ]

What is the philisophical difference between the phenomenon of superconductivity and conventional magnatism? I deal with MRI manufacturers regularly, and even have a patent in the field, but I fail to see the philisophical difference here.

Example: I have a magnet on my desk that hold papers up. The force is derived from the motion of the electrons in the iron. I'd consider this a manifestation of quantum phenomena at a macroscopic scale. Similarly, light reflecting off a surface and many other phenomena require interactions at a quantum scale.

So how is superconductivity fundamentally different? It is a phenomena which requires an explanation at the quantum level, like so many other things.

Actually, the study of "magnetism" (not magnetic fields) IS, in fact, quantum mechanical and an emergent phenomena. Quantum magnetism, a field of study in condensed matter physics, study why certain things are ferromagnetic, diamagnetic, antiferromagnetic, etc... etc. So in essence, the origin of magnetism in materials is as profound as superconductivity.

The only difference here being that there is already an "phenomenological" classical model of magnetism and magnetic field that deal with the fields being created - the Maxwell Equations. Superconductivity has no such counterpart in classical physics. And superconductivity has one very obvious aspect to it that is not obvious in magnetism - the manifestation of various phenomena due to the onset of long-range phase coherence. The easily-achieved destruction of such phase coherence (or decoherence) is what is currently thought to be the source of the "transition" between quantum mechanical world, and classical world.

One last thought. If we define "emergence" as Vanesch did (in terms of reductionism) then I think the explanation of why superconductivity is "emergent" requires one to address the issue of why it is irreducible even in principal to the interactions of electrons with a material.

This is because you need to consider what KIND of interaction is necessary, and to what degree. In many-body theory, the simplest approach is to use the mean-field theory approximation, in which the interaction of one electron with ALL other electrons in the fermi sea is approximated by some average field. Now this is already starting with a "many-body" approximation. Now only that, you now have to deal with the rest of the ions in the solid. In superconductivity, these ions are the "source" of lattice vibrations, or phonons, which themselves are only defined as a "collective" excitation.

So now, already you have a gazillion degree of freedom to deal with here if you decide to not care about applying these many-body approximation. Many of these emergent entities such as phonons and mean-field approximations simply aren't obvious and not present when you work out each individual interactions ad nauseum. There are no obvious insight that make you look at the physics and then say "ah hah! I see a pattern emerging that points to superconductivity at the larger scale". No one has seen that.

Zz.

 

[Chaos' lil bro Order]

.. and if you have followed my previous argument about superconductivity, you would have SEEN the references I've given. For example, try Carver Mead's PNAS paper[1]:



So take that! But even without such a reference, we KNOW about superconductivity (at least *I* do), and that we can OBSERVE its effects macroscopically! Just look at those levitated trains, for heaven's sake! Those, by themselves, are clear indication of a quantum effect at the macroscopic level already! I didn't realize that I had to be THIS explicit in pointing it out.

Now it is MY turn to ask you for references to back your claim that "emergent" properties are classical, and
Zz.

[1] C. Mead, PNAS v.94, p.6013 (1997).

I never claimed emergent properties are classical and that this whole thing is simply a classical-quantum transition. You should read my posts more clearly, I said 'I think', there is a big difference between claiming something and offering one's take on the situation. I thought you would know this without me explicitly pointing it out to you. Oh well.

As for superconducting maglevs... that is a good example, its also prudent to note that many maglevs do not use superconducting magnets (yet!).

On the flip side, Zapper, can you give me a good example of a non-emergent phenomenon?

 

[nightflyer]

Vanesch:

I kind of like Feynman's metaphor about how we find ourselves in a game of chess that is played by mother nature. We see how the pieces move, and from these observations we make up rules that are more or less successful at predicting what will happen next. However, the game of chess is much more complex than just following some basic rules, and we are not yet able to play the game (in other words, reduce chemistry and biology to physics).

However, knowing the rules must be a good start. And we did actually manage to build a chess computer that beat the world champion based on algorithms, so maybe there is hope that we will be able to understand nature's game of chess based on the physical laws we are able to observe... But then again, what do I know.

 

[complexPHILOSOPHY]

Vanesche, that was the most clear and precise description regarding 'emergent phenomena/properties' in physics that I have heard.

When you describe 'platonic mathematics' you are referring to the notion that mathematics exist objectively in nature and independent of human mind, correct? I wrote an in-depth critique to Plato's metaphysics, opposing his arguments in one of my philosophy classes.

Does your use of 'platonic mathematics' relate to "The problem of the one and the many," and his dualism? I am familiar with the philosophy of mathematics and it's relevant schools as well, if that is more relevant to the discussion. I have a passion for philosophy, so I enjoy these discussions.