Friday, April 18, 2008

Universal Principles, or Why Chemistry Is An Inferior Body Of Knowledge

Qualities of a Body of Knowledge

What are some relevant qualities by which we may judge the quality of a body of knowledge? We can start with the ratios of rules to facts and of facts to phenomena. We'll go on with the universality of principles, the extent of knowledge, and the simplicity & organizing power of concepts. So when we look at chemistry, what do we see?

In chemistry, there are too many rules for the facts we know. There are too many facts for the phenomena we know. Rules are so far from universal that they can never be called principles - as proved by the fact you can't name a single one that belongs entirely to chemistry. Knowledge of phenomena is very limited. And all concepts are ad hoc, baroque and with limited power.

Principles of Chemistry

But what about "Ka x Kb = Kw" and "all endothermic reactions speed up with increasing temperature"?

The problem with those principles is that they are very far from universal. Although chemicals can be acids or bases, most of them are neither. How exactly is it meaningful to talk about the pH of Uranium Oxide? Or steel? And as for reactions, the problem there is that chemical reactions consume themselves and so they have a nasty tendency of being intermittent. We're not talking then about something that persists but rather something that flickers to life and then winks out. Most chemicals are not in the middle of a reaction!

Concepts of Chemistry

What exactly are the fundamental concepts of chemistry anyways? The chemical bond is one. Unfortunately, talking about the universal laws of bonding would just be a repetition of quantum electrodynamics and electrostatic attraction.

Valence shell electron theory is a staple of chemistry. It explains the periodicity of Mendeleev's table. But Mendeleev's table itself explains nothing further. It only talks about how some chemicals are similar to other chemicals, not why they are the way they are in the first place.

Chemistry doesn't explain why there's two electrons to an orbital (physics does that) and it doesn't explain the shapes or numbers of orbitals (physics sortof does that). Hell, it's not even possible to account for the colour of gold without dragging general relativity into the mix!

Actually, if anything has any claim to being a fundamental concept of chemistry, it's orbitals. That's because orbitals are critical in chemistry yet they are far too complex to compute using physics. They're an emergent property.

Pity the orbitals don't really matter in the far reaches of chemistry. Certainly, everything is made out of them, but it doesn't seem to matter. Probably because when it comes down to it, there's only a few of them that matter at all.

Let's compare with physics.

A Brief Look At Physics

Absolutely everything in the physical universe is made out of energy. For something to even exist, it has to be made out of energy.

Do you know why photons can only travel at the speed of light? It's because if they didn't then there would be a reference frame in which they are at rest. And since their rest mass is zero, this would mean they have absolutely no energy. Thus there would be a reference frame in which a photon does not exist and another in which it does exist. And that's absurd.

Literally, Physical Existence = Energy.

Space then is a near-universal. Everything exists in space except for space itself. Space may or may not be a property that emerged from superstrings.

Information is another near-universal. Everything that exists save possibly space itself holds information in order to exist. The laws of thermodynamics are laws about information. 'entropy' can be defined as 'information we don't care about' thus neatly demonstrating why it has no role in the fundamental laws of physics. Information is a conserved quantity which follows specific laws.

From these concepts alone, it is possible to formulate a theorem about the maximum possible rate of computation in a volume of space at any non-zero temperature. Any attempt to compute more would require more energy, which would increase the energy density past the point a black hole would form which would disconnect you from the heat sink and destroy your ability to compute anything. That is a pretty fantastic thing when you think about it.

And Back To Chemistry

Chemistry has nothing like it. There are a few different orbitals and there are a few types of bonds ... but there seem to be too few types of bond to matter and the different orbitals hardly matter. So no universal laws and no universal theorems. I'm not sure why chemistry is so fragmented but it's a shame.

And that's why chemistry is an inferior body of knowledge.


Peter said...

Two technical notes:

"We're not talking then about something that persists but rather something that flickers to life and then winks out. Most chemicals are not in the middle of a reaction!"

-The process of chemicals reacting in a closed system does not end. It get to the point where the forward reaction rate is equal to the reverse reaction rate (equilibrium). In other words, although there is not a net increase in the number of products or reactants, the reaction is still going on - it's just going on at the same rate that it is going on backwards.

"'entropy' can be defined as 'information we don't care about' thus neatly demonstrating why it has no role in the fundamental laws of physics."

-Chemists care about entropy =P

And lastly:

Is there anything, in any field of study, that you would consider comparable to a universal law like conservation of energy, or did you just decide to pick on chemistry?

Richard Kulisz said...

Yeah, it's been ages since I've touched chemistry.

I meant to slag biology as well but I never got around to it. Plus, biology has evolutionary theory which does matter everywhere. That's a bunch of intertwined concepts, but they don't cover everything mostly because biology is pretty arbitrary.

Other fields? Well, OS (or any kind of complex interactive system) has graphs as a fundamental unifying concept. Everything is part of a graph. All non-graphs are special-cases (eg, trees) and are inferior to graphs.

Real economics is entirely based on the concept of 'value', although nobody bothers to formalize it as a unit of lifetime. That is, how much lifetime would you be willing to sacrifice for something. Value is the concept that underpins everything. The biggest problem is fake economists fucking it up for the real economists.

AI research is built around ideas. Unfortunately that's flawed since ideas are far from universal. The unifying concepts are 'value' and 'concept'. Emotions are emergent properties. And units of experience are percepts which can be thought of as degenerate concepts. Ideas are aggregates of concepts.

Urbanity could probably develop a unified theory around interactions, covers and a few concepts from economics.

Quantum physics has fields.

Mathematics has axiomatic systems.

Psychology has rationality and to a lesser extent empathy. Psychologists haven't discovered this. And many of them don't even agree with it. But they are there.

Political science is all about power. The only way of completely nullifying it is randomness (sortition). Round-robin rarely effectively nullifies it. For everything else, it's a question of how much power is nullified.

Astrophysics has gravitational collapse and nuclear fusion. I think.

Cosmology has not sure what.

This sucks, I know I'm familiar with more fields than these but I can't recall them at the moment.

Actually now I can narrow down what it means to have a unifying concept. It's not just a fundamental concept that underlies everything. It's a fundamental concept that varies smoothly and continuously.

For chemistry, orbitals don't do that because they're too discrete and they vary non-linearly. And that's before they interact.

For biology, ecological niches are too non-linear. And speciation is not even close to being universal.

Anonymous said...

A small nit in an otherwise interesting post:

the colour of gold may be explained with special relativity, not the general theory.