Reductionism: Physics-Chemistry-Biology

[Ziggurat]

No-one (in biology) is arguing that biological creatures do not follow the principles of physics and chemistry. And if you want to say NS is 'emergent', fine. But it is strongly emergent; it is not directly explainable by the mechanisms of the underlying system. Ergo, does not reduce.

We cannot reduce it now (and might not ever). But we also have limited intelligence to figure these things out. Our particular limitations don't mean that it is impossible to reduce. It's just like irreducible complexity in biology: without proof that something is irreducible, the inability to reduce it should not be taken as a sign of anything more profound than our own inadequacy.

 

[DeviousB]

We cannot reduce it now (and might not ever). But we also have limited intelligence to figure these things out. Our particular limitations don't mean that it is impossible to reduce. It's just like irreducible complexity in biology: without proof that something is irreducible, the inability to reduce it should not be taken as a sign of anything more profound than our own inadequacy.

In otherwords you have no proof that NS can be reduced, and introduce a strawman as justification. Natural selection isn't an 'irreducibly complex' bit of physics, it isn't physics.

What element of physics is essential to the operation of NS? How do we then create computer models with no physics in them at all and still have them evolve through NS?

Can you demonstrate any link between NS and physics other than they can both apply to physical objects?

 

[Ziggurat]

In otherwords you have no proof that NS can be reduced,

Proof? No, there can be no proof that it's reducible unless you actually reduce it. Rather, I have reason to believe it is reducible and NO reason to think it isn't. More below.

and introduce a strawman as justification.

It's not a strawman unless I attribute the position to you. I did not.

Natural selection isn't an 'irreducibly complex' bit of physics, it isn't physics.

This is turning into a semantic debate about what it means to "reduce" something from biology to physics, which is really not very interesting. More below.

What element of physics is essential to the operation of NS? How do we then create computer models with no physics in them at all and still have them evolve through NS?

How good a model is depends upon what you demand of that model. That you can throw away details from the model and still have it work doesn't mean those details aren't critical to the real-world operation of the system. For example, your natural selection model probably doesn't take into account details like sperm motility, womb pH levels, etc. even though those are all biological factors which the real-world phenomenon most certainly depends upon. Why? Because the model works even without that level of detail, because the demands on the model are very limited. A model which included those details wouldn't be worse, it would probably be better, but the added complexity probably does not produce enough improvement to justify the extra effort. The fact that something can be reduced does not mean from a practical standpoint that it should be. As a general rule, one should use the simplest model that is accurate enough, not the most accurate model available.

If you could model the entire physics of an ecosystem (something we cannot do and quite possibly never will be able to do for purely practical reasons), I expect we would indeed find that it behaves according to natural selection, WITHOUT putting natural selection into the model explicitly. That is what it means to be an emmergent property, and if a property is emmergent, then it is at least in principle reducible. If it is NOT reducible, then that means that no matter how complex we make our physical simulation, it would not match reality. And that cannot happen unless there is some process going on which does NOT obey physics as we understand it, but actually violates it. While that might be possible (I can't prove that natural selection is reducible, and we cannot construct such massively complex models to test them against reality), there is no evidence that any such processes actually occur (ie, there's no reason to think it isn't reducible).

 

[Slimething]

Okay, point taken. I phrased things poorly. What I am trying to say is that all the forces involved are known. At least the important ones.
From this we can conclude that if we could analyse these forces well enough we could model the interactions. Niether I nor anyone else is saying that we actually can do so, just that the forces involved are known.

Roboramma, I apologize for the tone of my response. It was generally unfriendly and I was more peeved at shneibster than anything else. I can only conclude the man has not the faintest idea of what chemistry involves or he would not have written something so ignorant. I believed you were parroting him.

Let's get to it. All the major forces known, yes. All the forces known, no. Even then, we can say the same about the more popular sports as the gaming world has gotten sophisticated enough that the leading software includes player's tendencies and individual athletic skills. So, why not just model the games? Because, as human foibles, there are also certain aspects of chemistry that it would take too long to include in a model or are not yet known. Hell, if we could model to that degree, it would be a cinch to program cancer cell receptors and design the magic bullet to obliterate these cells.

If you aren't suggesting that there is some other force at play, then I don't see how chemistry isn't reducible to physics. Will any particle be affected by any other in a way that contradicts known physics?
Or are you suggesting that this is some other force involved.
This is a serious question, I don't know that much about either chemistry or physics and am hoping to have my own errors shown to me.

And it's a very good question. Philosophically, it's easy to say that, as all materials are made of atoms and physics can characterize a simple atom, that every science which studies materials made of atoms must be reducible to physics. Theoretically, I would agree that, if physics were capable of defining higher complexities, that it would rule all the known scientific universe. But, it can't. Don't get me wrong, I love physics and am spellbound what has been achieved but it can't yet, nor every will, replace chemistry let alone biology which is even more complex. I believe that Godel's Theorem probably comes into play here. Yes, there must be forces at play that neither physics nor chemistry know yet.

Are you suggesting that the characterists of a compound don't depend on the characterists of the atoms that it's made of and the way that those atoms interact, or that the way that those atoms interact is not itself dependent upon physics?

I'll declare it outright. Yes, there are characteristics of compounds that do not depend on the interaction of the atoms it consists of. That's not surprising to someone who knows chemistry. Not all characteristics can be deduced from the components therein. One of the major aspects that physics can't address is toxicity. Ordinary table salt comprises two deadly elements but it's relatively non-toxic to us. Odor cannot be modeled from atomic interaction. Strangely, the color of conjugated organic molecules can be modeled but that's more from empiricism than theory.

If physics doesn't describe the chemically relevent characteristics of atoms (I think it does, but whatever) are they not dependent upon the characteristics of the particles that make up those atoms and the way that they interact?

I think it's fairly established that chemically relevant characteristics are not wholly predictable from current physical knowledge. Why? Dunno. Maybe you'll find out.

Well, as I don't pretend to be an educated person, just one who's trying to understand, I won't pretend to even know what you mean by organic synthesis.

You are an educated person. It was I acting like a boor. Organic synthesis is the preparation of more complex organic molecules from simpler organic or inorganic molecules. Those are the guys in the white coats with all the fancy glassware. These are the guys who can make pharmaceuticals from basic building blocks. They are quite talented and, no, they cannot be replaced by a computer program. Admittedly, a lot of the theory behind what is done there has been successfully modeled but billions are still spent "playing the game".

So, there's my point. Be well.

 

[pgwenthold]

I'll declare it outright. Yes, there are characteristics of compounds that do not depend on the interaction of the atoms it consists of. That's not surprising to someone who knows chemistry.

I guess I am the exception, then. yes, I know chemistry.

Not all characteristics can be deduced from the components therein. One of the major aspects that physics can't address is toxicity. Ordinary table salt comprises two deadly elements but it's relatively non-toxic to us.

Actually, someone who knows chemistry will know that table salt is not made of atomic elements, but is made of ions.

Their are huge differences between elemental sodium and sodium ion, and they originate in and are reflected in the physics.

Odor cannot be modeled from atomic interaction.

What? Odor is all about molecular shape and intermolecular interactions!

Strangely, the color of conjugated organic molecules can be modeled but that's more from empiricism than theory.

Only to the extent that the concept of "color" is empirical. To the extent that color results from the absorption of light, it is completely calculable! In fact, that is one thing that is done pretty well nowadays (although it requires adequate dynamic correlation in the excited state). Not quite as accurate as structures and vibrational absorption frequencies, but still pretty good. In fact, many people use "color" calculations to help identify transient species.

 

[Slimething]

I guess I am the exception, then. yes, I know chemistry.

OK, pg, explain stereochemical phenomena in light of strictly intramolecular interaction between atoms.

Actually, someone who knows chemistry will know that table salt is not made of atomic elements, but is made of ions.

You're just quibbling here. How am I going to give you vial of sodium ion? What you are saying is that elemental sodium and elemental chlorine will NOT form table salt on combination?

Their are huge differences between elemental sodium and sodium ion, and they originate in and are reflected in the physics.

Please do go on about the physics of elemental sodium and ionic sodium, pg. This will be fascinating. Please do explain. Explain why and how they are different using only physics. This will be rather delightful.

What? Odor is all about molecular shape and intermolecular interactions!

Ahem, the point in question is to whether or not physics can account for all chemical phenomena, pg. Even so, can you put up the physics as to why some chemicals smell sweet and some not? Start with carbonyl chloride. That's an easy one that smells like hay. I'll just stand over here so as not to be in your way.

Only to the extent that the concept of "color" is empirical. To the extent that color results from the absorption of light, it is completely calculable!

Thanks, that's what I wrote. But it's only true in some cases, pg. Again, please, do carbonyl chloride. Please calculate its color using only physics.

 

[Cuddles]

I think Dawkins discussed this whole thing rather well.

Look at an aeroplane. You can see the whole plane as one thing and study how it flies. You can look at it as made of several large structures like wings and body and see how they interact with each other. You can look at individual components and how they connect, how the engines work and so on. You can look at the materials used to make components and see how they react under different conditions and with each other. You can look at the atoms in the materials and see how they effect their properties. You can look at the particles that make up atoms and see why the atoms behave the way they do. You can look at the fundamental things that make up the particles in atoms and see why it is even possible for atoms to exist. No doubt one day we will be able to go even further than this.

All these are perfectly valid ways of looking at a plane, and they are all reducible to a more fundamental view, but this does not mean that one is the correct way, or that any way is better than any other. A plane is made of atoms, you cannot deny this, but this does not mean that designing an engine based on quantum physics is sensible, practicle and certainly not useful. In the same way, chemistry is entirely made up of physics, because it studies things that are entirely governed by the laws of physics. This does not mean that it is in any way useful to try to derive all of chemiistry through quantum mechanics, but it does mean that it is just plain stupidity to try to claim that it is not based on physics. In the same way, to study the function of the liver you use biology because you simply don't care what individual quarks are doing, even though everything that happens only happens because of them.

And as others have pointed out, maths is completely different. Biology, chemistry and physics are all sciences that focus on different scales. Maths is the language that science speaks.

 

[pgwenthold]

Thanks, that's what I wrote. But it's only true in some cases, pg. Again, please, do carbonyl chloride. Please calculate its color using only physics.

Do you think the absorption bands of carbonyl chloride cannot be calculated?

You give me a good computer with code for EOM- or MR-CCSD(T) that I can extrapolate to an infinite basis set, and I'll be able to calculate the long wavelength absorption of carbonyl chloride.

Can it be reduced to physics? Absolutely! Actually, I am surprised that you even bring this up, because it is such a trivial example of where physics _can_ be used to calculate "chemical" properties.

In terms of odor, yes, it is complicated. But it is doable, which is the question (show that the substrate will activate the same receptors as does those in freshly cut hay - that's an issue of structure and intermolecular forces, which are subject to physics principles)

Please do go on about the physics of elemental sodium and ionic sodium, pg. This will be fascinating. Please do explain. Explain why and how they are different using only physics. This will be rather delightful.

Oh, where to start...how about "total angular momentum of the wave function"? Sodium ion S = 0. Sodium atom: S = 1/2

How about "total electronic charge"? Sodium atom = 0, sodium ion = +1

Charge and angular momentum are basic physics concepts.

How about stability? Can physics tell me that Na is more reactive than Na+? Absolutely. Calculate the first and second ionization energies of Na, and they are extremely different (I mean, 40 eV different)

Man, the physics of elemental sodium and sodium ion are so completely different to be laughable.

How am I going to give you vial of sodium ion?

I can give you a beam of it, if you want (you have to be careful when talking to people who might be a mass spectrometrist)

What you are saying is that elemental sodium and elemental chlorine will NOT form table salt on combination?

Of course they will, but then again, butadiene reacts with ethylene to give a substance that is neither butadiene nor ethylene. It has properties that are distinct from its reagents. Similarly, sodium and chlorine react to give a substance that contains neither atomic sodium nor atomic chlorine. Properties change when electrons move. The differences in the properties of Na and Na+ illustrate that clearly.

 

[DeviousB]

It's not a strawman unless I attribute the position to you. I did not.

What you actually said is, "It's just like irreducible complexity in biology: without proof that something is irreducible, the inability to reduce it should not be taken as a sign of anything more profound than our own inadequacy."

Sorry, that it is like irreducible complexity in biology is an unproven assertion on your part. Please demonstrate that this is so. Without such proof you have created a proposition that is simpler to refute instead of attacking the original argument. So it is either a strawman argument, if you are implying that your proposition is analagous to mine, or a meaningless digression, if you are not.

How good a model is depends upon what you demand of that model. That you can throw away details from the model and still have it work doesn't mean those details aren't critical to the real-world operation of the system [...]

You are confusing natural selection with the biology of the organism. Nowhere did I say that I was modelling biological organisms and selecting upon them. In fact most computer models of selection do no such thing, Dawkins, for example, had a population of letters evolve into the sentence "METHINKS IT IS LIKE A WEASEL".

Natural selection requires (1) that a population of objects reproduce with a degree of inheritence, (2) that there is also a degree of variation between the parent and child object, and (3) that the variation affects the ability of the object to reproduce. That's it.

Inheritence, variation, fitness.

In biological organisms, the inheritence is provided by DNA, the variation by mutation, the fitness by (for argument's sake) differential survival. These are physical properties of the organism and can be potentially explained as physics (as I have already accepted).

They are, however, just postulates of natural selection (to use John Maynard-Smith's terminology). Other postulates in other systems will serve just as well. Natural selection reduces directly to mathematics, do not pass physics, do not collect 7 extra dimensions.

If you construct a model of the real world accurate enough to demonstrate NS, it will be because you have directly modelled the mathematics from which NS arises.

 

[Slimething]

Do you think the absorption bands of carbonyl chloride cannot be calculated?

I know they can. I asked you to do it.

You give me a good computer with code for EOM- or MR-CCSD(T) that I can extrapolate to an infinite basis set, and I'll be able to calculate the long wavelength absorption of carbonyl chloride.

You made the claim, pg. You arrange for your own tools. You are in a skeptic's forum so I'm calling you on your claim.

Can it be reduced to physics? Absolutely! Actually, I am surprised that you even bring this up, because it is such a trivial example of where physics _can_ be used to calculate "chemical" properties.

For it being so trivial, you are not doing very well. All I asked you to do was to back up your claim. You chose the easiest of the various challenges, completely disregarding the stereoisomerism question, and all I see is words. No physics, just words.

In terms of odor, yes, it is complicated. But it is doable, which is the question (show that the substrate will activate the same receptors as does those in freshly cut hay - that's an issue of structure and intermolecular forces, which are subject to physics principles)

All I asked you to do was to back up your claims, not give me a song and dance about how it would be possible if... You cannot do it. Fess up.

NOTE: You may now want to return to what I wrote and read it more carefully. What I wrote was that, if you use facile thinking and philosophically define physics as the science that defines atoms, then anything reduces down to physics. However, physics does not define atoms yet.

Oh, where to start...how about "total angular momentum of the wave function"? Sodium ion S = 0. Sodium atom: S = 1/2

How about "total electronic charge"? Sodium atom = 0, sodium ion = +1

Charge and angular momentum are basic physics concepts.

How about stability? Can physics tell me that Na is more reactive than Na+? Absolutely. Calculate the first and second ionization energies of Na, and they are extremely different (I mean, 40 eV different)

Man, the physics of elemental sodium and sodium ion are so completely different to be laughable.

I'm not laughing. This is getting rather tiresome. I know all this stuff. You told me you could explain the difference between the ion and the element. I see nothing here regarding the types of interactions that either form would have with anything else. You have to do better than this. Again, describe the differences in such relevant things as solubility, color, odor, toxicity, etc of both sodium ion and metallic sodium using only physics. Time's a-wasting.

I can give you a beam of it, if you want (you have to be careful when talking to people who might be a mass spectrometrist)

One of these days, you'll truly get to know just how much I enjoy smart ass answers.

Of course they will, but then again, butadiene reacts with ethylene to give a substance that is neither butadiene nor ethylene. It has properties that are distinct from its reagents. Similarly, sodium and chlorine react to give a substance that contains neither atomic sodium nor atomic chlorine. Properties change when electrons move. The differences in the properties of Na and Na+ illustrate that clearly.

The statement above could have been ripped out of any elementary chemistry book, pg. What happened to the physics you were so loudly proclaiming?

 

[DeviousB]

They are, however, just postulates of natural selection (to use John Maynard-Smith's terminology). Other postulates in other systems will serve just as well.

Okay, it looks like I've screwed up that sentence and the edit button has disappeared.

*Note to self: use the goddamn preview!*

 

[DeviousB]

And as others have pointed out, maths is completely different. Biology, chemistry and physics are all sciences that focus on different scales. Maths is the language that science speaks.

But it's not invariably the same math.

 

[Ziggurat]

You are confusing natural selection with the biology of the organism. Nowhere did I say that I was modelling biological organisms and selecting upon them. In fact most computer models of selection do no such thing, Dawkins, for example, had a population of letters evolve into the sentence "METHINKS IT IS LIKE A WEASEL".

If you're not selecting biological organisms, then it isn't biology. It may provide insight into biology, but that is NOT the same thing. Yes, natural selection for letters forming a sentence doesn't depend upon physics. Same with the rules of sudoku. But neither of them is biology, or even a physical science.

Natural selection requires (1) that a population of objects reproduce with a degree of inheritence, (2) that there is also a degree of variation between the parent and child object, and (3) that the variation affects the ability of the object to reproduce. That's it.

Inheritence, variation, fitness.

And if you don't account for how those values arrise for a biological population, but merely assign them, then you are brushing the actual mechanics under the rug. That may work as a practical matter, but the fact that you don't NEED to reduce the system doesn't mean it cannot be reduced.

In biological organisms, the inheritence is provided by DNA, the variation by mutation, the fitness by (for argument's sake) differential survival. These are physical properties of the organism and can be potentially explained as physics (as I have already accepted).

They are, however, just postulates of natural selection (to use John Maynard-Smith's terminology). Other postulates in other systems will serve just as well. Natural selection reduces directly to mathematics, do not pass physics, do not collect 7 extra dimensions.

As I said, this is turning into a debate about what it means to be reducible. The model itself IS NOT REALITY. That you can make a successful model independent of physical details does not mean that the phenomenon is not reducible.

It is only irreducible if it is not possible, even in principle, to model the system based only upon the physical details. Slimething gets that. He understands that there must be some NEW physical process, which we do not currently know about or which CANNOT be modeled (even in principle), in order for complex systems to be irreducible. He thinks there are such processes, I think there are not, and there's no way to prove it either way. But the logic of that requirement is quite clear. The only way to maintain the concept of irreducibility outside of that is to do what you have essentially done: include non-physical sciences. Which is, well, kind of pointless.

 

[DeviousB]

There are no known phenomena in chemistry or biology which do not operate according to the laws of physics.

Natural Selection is certainly a phenomenon of biology, it is also non-physical (as I pointed out in my original post and you disputed).

The only way to maintain the concept of irreducibility outside of that is to do what you have essentially done: include non-physical sciences. Which is, well, kind of pointless.

So in essence you are saying that only physical phenomena are reducible to physics. Which, if it is not an actual 'bait-and-switch', is hardly a profound observation.

 

[Ziggurat]

So in essence you are saying that only physical phenomena are reducible to physics.

Well, yes. That's what I thought this thread was about: physical phenomena.

 

[DeviousB]

Well, yes. That's what I thought this thread was about: physical phenomena.

My mistake, I thought the topic included Biology and Chemistry.

 

[Ziggurat]

My mistake, I thought the topic included Biology and Chemistry.

You don't think biology and chemistry are studies of physical phenomena? If not, then perhaps you have a different definition of the term than I do.

 

[Earthborn]

I explained that physics 'comes' from mathematics, chemistry comes from physics and biology comes from chemistry.

I think it is fair to say that biology is reducible to chemistry, chemistry is reducible to particle physics. That is to say, you can study these things at a smaller and smaller scale and you'll pass every all those sciences. But that of course ignores the fact that at larger scales emergent proporties arise that justify making it into different sciences. It will be very difficult to express a lion's hunting behaviour in purely chemical terms.

Physics encompasses all things, but chemistry is only reducible to a specific subset of physics: particle physics. The movements of the planets need not be considered.

The claim that physics comes from mathematics is silly, IMHO. Physicists use mathematics as a tool, but I don't think there is anyone of them who claims that physics is a subset of mathematics or that the laws of physics are somehow reducible to pure mathematical abstractions. They use mathematics for as far as it has real world applications, but mathematicians don't necessarily concern themselves with the real world (at least in their work).

 

[Ziggurat]

The claim that physics comes from mathematics is silly, IMHO. Physicists use mathematics as a tool, but I don't think there is anyone of them who claims that physics is a subset of mathematics or that the laws of physics are somehow reducible to pure mathematical abstractions.

I agree, though I approach it from a different angle. There's no limit to the number of possible mathematical models for universes, and the math can tell you whether or not these models are self-consistent. But that's not enough: the model must correspond in some way to reality in order to be physics, and the math alone can never tell you whether or not that's the case.

 

[DeviousB]

[Nevermind]