Reductionism: Physics-Chemistry-Biology

[balrog666]

Hey guys,

Well I'm having this argument with a close friend of mine who's about to begin her 2nd year of undergraduate science. The argument is essentially about the most basic form of reductionism: I explained that physics 'comes' from mathematics, chemistry comes from physics and biology comes from chemistry.

She's someone who loves chemistry and seemed angry at the suggestion that it is reducible down to physics... I gave some examples of the way chemistry is working off the foundation of physics but they didn't seem to satisfy. I'm wonder if anyone could offer up a simple way to show her exactly what I mean - Or if there really is strong argument out there about whether chemistry is reducible to physics (I'm doubting it).

Thanks.

I disagree. Physics is not reducible to mathematics in spite of the essential nature of mathematics within physics . Ditto for chemistry/physics, the reduction is practically meaningless.

 

[cyborg]

Mathematics cannot TELL us anything about our world. It cannot be a science.

 

[TheChadd]

But everything you do in physics would be meaningless without the mathematic principles.

Ditto for chemistry/physics, the reduction is practically meaningless.

The idea isn't to say if you know physics you understand chemistry, but to point out that (as was said earlier in this thread):

I like to say that math is to physics what letters are to words. Words are to a sentence what physics is to chemistry and a sentence is to a paragraph as chemistry is to biology. Understanding grammar does not mean you will be able to grasp literature, but it is a foundational component. Grammar being essential to literature does not mean it is “better”, they are two different beasts in many ways.

 

[Slimething]

Mathematics cannot TELL us anything about our world. It cannot be a science.

I agree with you but there are several instances where anti-intuitive discoveries have been made by following mathematics alone, devoid of human observation. The most famous example is the Theory of Special Relativity.

Like I wrote earlier, I'm not going to participate in what begat what science. That's a fool's game. When I was in college, I had some nebbish physics major taunt me that chemistry was all physics so I wrote down two chemicals and told him to calculate the product of the reaction, if there was any. That worked pretty well.

We're all after the same thing, which is to understand nature and, if you're good at it, you won't limit yourself to the classical definitions of the basic sciences. Use any tool that you find useful and the names be damned.

 

[cyborg]

I agree with you but there are several instances where anti-intuitive discoveries have been made by following mathematics alone, devoid of human observation.

This is only possible when the maths relates to the real world and the math can only relate to the real world by engaging in physics.

Being able to predict unanticipated phenomena and then seek them out is one of the primary reasons maths is useful but it cannot be said the discovery is complete until the prediction is confirmed - you may be wrong after all.

 

[DeviousB]

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

I would argue that natural selection operates according to the rules of statistics.

 

[geni]

What do you mean by solvable? An analytic solution isn't even possible for helium. Even Newtonian gravitation can't be solved for the solar system, but that never seemed to be a reason to think it was wrong.

I never said it was wrong. However if you are going to describe chemisty through physics you are going to need to be able to solve the Schrödinger equation for systems or more than 1 electron. Unless you can prove it is posible to do (you don't have to do it just need to prove it is posible) this you can't prove it is posible to derive all chemistry from physics.

I'd expect that approximations to the Schrödinger equation would allow you to derive any chemical principles,

You need kinetics and relativity as well at a mininium.

although it might be practically impossible for large enough systems, like trying to predict enzyme activity. I suppose if you must satisfy the philosophy of science types it might be better to say "in principle any theory of chemistry or biology can be derived from the theories of physics".

First you need to show that it is in thoery posible to solve the Schrödinger equations for multi election systems.

 

[pgwenthold]

IFirst you need to show that it is in thoery posible to solve the Schrödinger equations for multi election systems.

To what level of accuracy?

That is the question at hand. No, the SE cannot be solved analytically for any multi-electron system. However, with numerical approaches you can come more or less as close as you want to the solution.

No, it's not sophisticated, but the reaction H + H2 is pretty well known based on QM calculations, at least the chemistry of it. Sure, there are plenty of people who are still studying it, but they are focusing on the details of the physics.

Heck, even state-of-the-art dynamics calculations (especially those that are QM based) are not all that far off the mark for complex reactions. For example, the gas phase SN2 reaction is extremely well-described computationally. Similarly, very good success has been achieved with unimolecular reactions. Reaction rates and branching ratios can be calculated.

The limiting factor is the complexity. You have to know enough to include all the relevant coordinates, and for large systems, it's tough to know which of the 3N-6 should be important. In principle, you could just include them all, but that is a resource limitation.

Yes, there is a long way in order to do the calculation on any type of system. However, that is mainly in resources. Right now, we know what we need to do in order to do it, it is just way beyond our capabilities.

The SE can be solved to within "chemical accuracy," which is the accuracy level of typical chemistry measurement. It's not easy to get there (especially with larger systems), but you can if you put enough effort into it.

 

[Ziggurat]

I would argue that natural selection operates according to the rules of statistics.

Yes, but that's an emergent property. Statistics doesn't describe mechanisms, it describes the logic of what you can conclude from incomplete or imperfect information. Which is why it pops up in so many places. The mechanisms themselves are still physical mechanisms, describable (at least in principle) by physics.

 

[Ziggurat]

You need kinetics

You aren't under the mistaken impression that the Schrödinger equation cannot handle kinetics, are you?

And yes, the Schrödinger equation can be solved numerically for arbitrary numbers of particles (if you've got the computational power).

 

[DeviousB]

Yes, but that's an emergent property. Statistics doesn't describe mechanisms, it describes the logic of what you can conclude from incomplete or imperfect information. Which is why it pops up in so many places. The mechanisms themselves are still physical mechanisms, describable (at least in principle) by physics.

Unless you are arguing for strong determinism, there is no way to get from the physical properties of an organism to an example of evolutionary complexity like an eye.

The eye is selected for by the processes of natural selection, not physics. The physics of the eye may be what confer efficiency on it but they do not result in the progressive refinement (or not) of the eye.

Physics may explain how the eye works, it cannot explain why we have them.

 

[Ziggurat]

Unless you are arguing for strong determinism, there is no way to get from the physical properties of an organism to an example of evolutionary complexity like an eye.

Well, I'm not sure what you mean by "strong determinism".

The eye is selected for by the processes of natural selection, not physics.

Natural selection is an emergent property. That doesn't mean it's not still rooted in physics. The organization of hexagonal convection cells in heated liquids is another example of an emergent property of a complex system. There's no simple way to show that such self-organization is what will result using, for example, the Schroedinger equation, but it's all still just physics. The more complex the system, the harder it is to trace emergent properties back to their physical origin (and often times, the less useful such an exercise is anyways). We label things as being something other than physics when we either can't trace the link (because of the complexity) or don't care to, not because the link doesn't exist.

 

[DanishDynamite]

And yes, the Schrödinger equation can be solved numerically for arbitrary numbers of particles (if you've got the computational power).

Is this correct, even theoretically? How powerful a computer would you need to solve the Schrödinger equation for every particle within 6,000 miles of the Earth's center? Or within 20 AUs of the center? Given that we can't solve the Schrödinger equation for a cup of pond water, I feel your view is even theoretically wrong, unless you posit close to infinite time to solve it in.

 

[pgwenthold]

Is this correct, even theoretically? How powerful a computer would you need to solve the Schrödinger equation for every particle within 6,000 miles of the Earth's center? Or within 20 AUs of the center? Given that we can't solve the Schrödinger equation for a cup of pond water, I feel your view is even theoretically wrong, unless you posit close to infinite time to solve it in.

There is no such thing as "close to infinite." Every thing else is finite, and is an infinite amount of time short of infinite.

Yes, it can be calculated in a finite amount of time. I just hope there isn't a power outage.

 

[DanishDynamite]

There is no such thing as "close to infinite." Every thing else is finite, and is an infinite amount of time short of infinite.

I know. I think my meaning is clear. Perhaps I should have put "within the expected life of the Universe assuming a closed Universe".

Yes, it can be calculated in a finite amount of time. I just hope there isn't a power outage.

Yes, but if the expected time to claculate is greater than the expected lifetime of the Universe, can it be said to be even theoretically correct?

 

[Ziggurat]

Is this correct, even theoretically?

Depends what you mean by theoretically possible. But there are no barriers other than limited computational power.

How powerful a computer would you need to solve the Schrödinger equation for every particle within 6,000 miles of the Earth's center?

Many, many, many orders of magnitude larger than we can envision ever building. The difficulty is that the numerber of calculations needed to solve this numerically scales exponentially with the number of particles you're trying to calculate for. So it doesn't take a system anywhere near this large to become impossible to calculate from a practical perspective. But that exponential scaling problem isn't even unique to quantum mechanics computations - you get the exact same problem trying to calculate Newtonian gravitational interactions of all the stars within a galaxy, for example.

 

[DanishDynamite]

Depends what you mean by theoretically possible. But there are no barriers other than limited computational power.

And time. Don't forget time. And if the maximum theoretical time remaining for this Universe is limited....is something that takes longer theoretically possible?

Many, many, many orders of magnitude larger than we can envision ever building. The difficulty is that the numerber of calculations needed to solve this numerically scales exponentially with the number of particles you're trying to calculate for. So it doesn't take a system anywhere near this large to become impossible to calculate from a practical perspective. But that exponential scaling problem isn't even unique to quantum mechanics computations - you get the exact same problem trying to calculate Newtonian gravitational interactions of all the stars within a galaxy, for example.

Exactly. The devision of chemistry and physics will last for quite a while, perhaps forever (read: until the end of the Universe)

 

[Ziggurat]

And time. Don't forget time. And if the maximum theoretical time remaining for this Universe is limited....is something that takes longer theoretically possible?

Doesn't really matter for my argument. What I'm claiming is that physics equations describe all chemical reactions. In that sense, all chemistry is physics. Whether or not WE can solve a particular set of equations is a different question from whether or not they provide an accurate description of the mechanics involved. It's not possible to solve 3-body Newtonian gravity problems analytically, and before the advent of computers not even numerically, but it was quite clear that the equations involved were still an accurate description of the mechanics. And the only limits to computational power are practical limits. It may indeed be the case that the lifetime of the universe limits computational power, but that too is only a practical (though insurmountable) limit, because there's nothing fundamental about the age of the universe.

Exactly. The devision of chemistry and physics will last for quite a while, perhaps forever (read: until the end of the Universe)

I never said otherwise. In fact, let me quote myself from my first post: "So yes, all chemistry is physics, but that statement doesn't mean much on a practical level."

 

[DanishDynamite]

Doesn't really matter for my argument. What I'm claiming is that physics equations describe all chemical reactions. In that sense, all chemistry is physics. Whether or not WE can solve a particular set of equations is a different question from whether or not they provide an accurate description of the mechanics involved. It's not possible to solve 3-body Newtonian gravity problems analytically, and before the advent of computers not even numerically, but it was quite clear that the equations involved were still an accurate description of the mechanics. And the only limits to computational power are practical limits. It may indeed be the case that the lifetime of the universe limits computational power, but that too is only a practical (though insurmountable) limit, because there's nothing fundamental about the age of the universe.



I never said otherwise. In fact, let me quote myself from my first post: "So yes, all chemistry is physics, but that statement doesn't mean much on a practical level."

I agree. I just thought that the question I posed ("If the maximum theoretical time remaining for this Universe is limited....is something that takes longer theoretically possible? ") was kind of interesting. Wanted to give it a bit more play.

 

[pgwenthold]

I know. I think my meaning is clear. Perhaps I should have put "within the expected life of the Universe assuming a closed Universe".

Why are you assuming that? I thought that at last best interpretation, the universe is essentially balanced, or perhaps even open.

Yes, but if the expected time to claculate is greater than the expected lifetime of the Universe, can it be said to be even theoretically correct?

Note that we are talking about current methodology and equipment.

There is no reason to think that computing power and methodology won't improve substantially in time, as well, such that with appropriate computing resources (say, put every processor we have right now and run them parallel) there would not be a foreseeable time limit to the answer (albeit it could be billions of years, but hey, still within a universal time constraint)