Friday, 9 September 2011

Could physics predict a giraffe?

I have a cosmological post brewing, so I thought I would touch on a slightly different topic, namely the question of "could physics predict a giraffe?" The following has the usual "buyer beware" clauses; I am a physicist, an astrophysicist at that, and not a chemist, or a biologist, and definitely no a philosopher of science, although I may end up annoying all of them. To start with, let's look at the subject, to wit, a giraffe.

The reason for the post is because of an article over at The Curious Wavefunction titled Why biology (and chemistry) is not physics. The basic argument is this;

Physics is a fundamental science, and identified the basic workings of the Universe. How do nuclei hold themselves together, how does the Universe expand, why do electrons flow through conductors etc etc. That's physics. Now, physics is "reductionist", in that all complex processes can be broken down into a the application of relatively simple underlying physical laws. An example; the thing that is the source of friction, which stops your car when you apply the breaks, and the force stopping you from falling to the centre of the Earth, both the electromagnetic force, the same force that gets electrons to flow through wires. And, of course, with just gravity and electromagnetism, we can explain the physics of everyday experiences.

However, apparently biology and chemistry are somehow different, and that if you try to reduce biological and chemical processes to the basic bits and pieces (i.e. into physics) you somehow lose something.

The issue with the article that kicked this off is that, even given all the basic rules of the Universe, the laws of quantum mechanics and general relativity, physics would be unable to "predict a giraffe". Why? Because somewhere along the line, there were random events (that mutated the genetic code of proto^n-giraffes) and physics just would not take this into account. The final paragraph of the article hammers home the point:
"This role of contingency and accident is one of the most important reasons why the reduction of chemistry and biology to physics won't work. In addition as I have described before, reductionism cannot account for variety in chemistry. Yet another reason why chemistry and biology are not physics."
The "variety in chemistry", after a little reading, suggests that chemists have broader intuition based on experience, and that is lost when you try and reduce chemistry to basic physics.

Well, I grew up in the country and often worked on farms, and I can tell you what all of this smells like (to me). Before I continue, an intermission.

Right, I'll start with something controversial. There is no such things as physics, chemistry and biology. Well, of course we have departments of Physics, Chemistry and Biology, and people who label themselves as physicists, chemists and biologists, and even journals dedicated to further, ultra-fine refinements of subsets of these areas. But where is the boundary between chemistry and physics, or chemistry and biology?

Take a look at this image

The caption for this image reads "A picture of the simplest ion channel known - antibacterial gramicidin A.  It conducts monovalent cations at near diffusion rates, causing collapse of the membrane potential and killing hapless bacteria.  Here Gramicidin A (helical dimer in red) is embedded in lipid bilayer (only the phosphate head groups in green are shown) and solvated with a KCl solution (K blue, Cl red and water is in the background).  Because of its simplicity, gramicidin offers an ideal channel structure for testing new methods and ideas."

Now, is this biology, chemistry or physics? It has a number of buzz-words from all fields, but it is actually research being done by the Computational Biophysics group at the University of Sydney. Why is this physics? Because it is being done in a department with a big sign saying Physics across the front door, by a person who labels themselves as a physicist as they did a degree in a different large building with Physics written across the door.

You don't have to look far to see that, at the edges, these subjects are blurred, and the labeling of physics, chemistry and biology can seem a little arbitrary. But people like to put things in boxes with rigid walls, even then the walls do not exist.

Back to the matter in hand. As ever, xkcd makes a deep comment on the issue;
Note, mathematics, especially pure mathematics, is not science. But this encapsulates some of the key points (although, as I pointed out, there are not rigid divides between the "fields") and even touches on physics envy.

So, could physics predict a giraffe? I think here it is important to realise that much of science is the science of complex systems, where lots of simple underlying processes interact to produce a more complex outcome. Some people think complex systems are somehow magical, but they aren't really, they produce unexpected results, yes, but the underlying processes are simple. Also, some seem to think that because a system is complex, it is somehow unpredictable. Again, this is not really the case. If we know the initial conditions well enough, we can evolve a system using the simple rules and see what happens (and yes, I do know what chaos is).

One of my fav complex systems, cosmological n-body simulations.
Lot's of little masses, following simple rules, producing complex outcomes. What's the limiting issue? Computational power. More computational power, the more things we can follow, the more detailed outcomes we can determine.

When we break it down, a giraffe really is just a bunch of fundamental (i.e. physical) processes interacting in a complicated way. With enough computational power, we could simulate all the processes going on in a giraffe, and hence an entire giraffe. With enough computational power, we could simulate the evolution of a giraffe, including all of the random possible events that happened along the way, and as well as all the giraffes that exist, we could find all the other animals that could have evolved instead of giraffes.

Now, I'm not saying this is computationally easy. We definitely don't have the computational power today to do this, and we may never really have it (it may take an actually universe to compute such things), but fundamentally it is possible.

I'll say it again, even if you hate reductionism, all processes, at the end of the day, however complex, boil down to physical interactions. You may wave your hands and go on about the variety of chemistry and complexity of biology, but this does nothing but reinforce the imaginary walls between the fields. Sure, you have intuition and ideas guided by your experience, but physics is at the basis of everything. If you don't believe this, then at some point you have to inject magic. And that just isn't science.

OK. On a lighter note, seeing we are talking about giraffes, why do they have such long necks? I'm sure you'll say something about them browsing the leaves high in the trees, but surely giraffes have long necks because their legs are so long, and without it, they would not be able to drink? 

No comments:

Post a Comment