Michael Foster writes:
This is a trend in the sciences in general, but I find it particularly
annoying in chemistry. You have a trend toward computerizing
everything. Hence, you get chemistry without chemicals, botany
without flowers . . .
Yes, this is deplorable.
Yes, no doubt this fine fellow is a little concerned with how ab initio
things are in his little world. First, he tells us how quantum mechanics
is the most successful theory in the history of science, a statement I
would argue with. Then we are treated to his bleatings as how it
doesn't really work, sort of.
I do not understand the technical details, but my sense is that he is
trying to pull the theory down a few notches, but as an insider he has to
watch his words. It reminds me of when Dyson tried to introduce Feynman's
theories to Oppenheimer, who was very upset, and kept heckling him
unpleasantly. Dyson had to portray this as nothing really new, just another
expression of the conventional.
By mainstream standards, this essay was breath of fresh air. It is nice to
see someone admit theory has its limits, and that electron shells may be a
mere convenience, without any actual physical reality. (They may be real
after all, but the point is, shells along with many other theoretical
constructs are justified mainly because they are convenient. And why not?)
The failure of ab initio quantum chemistry to predict such diverse and
important things as the effects of catalysts, the behavior of rare earth
elements, the existence and behavior of electrides, the quantum yield of
dye chromophores and so on really calls into question the whole damn
field.
I did not realize it had so many deficiencies. However, none of this is a
problem if only theory is considered a useful tool, and not an end in
itself or a quasi-religion. It is clear why people want ab initio theories,
and unifying theories. They are compact, and useful even when they do not
quite work right. As long as you know where the limitations and gaps are,
you are safe. But some people crave ab initio theory as an end in itself.
They hope to divorce knowledge from experiments and observations, to make
some sort of ultimate Theory of Everything. They want to play God, or be
God, it seems. Why? It does not seem like a practical goal, or even a
desirable one. Why does it upset people that we must depend partly on well
established observations that have no underlying theoretical basis? What
harm can that cause? As long as it works, who cares?
As I mentioned, one of Kuhn's main hypotheses is that there are always gaps
and leaks and unexplained problems, and every theory must eventually
crumble. Theory is a useful guide and a starting point. This notion upsets
physicists to no end, but most scientists I know personally are biologists
and electrochemists, and they take it for granted. Martin Fleischmann says
that whenever you hear people talking about some branch of chemistry or
physics being a closed book, you can bet that particular field is about to
erupt in controversy, and the theories are held together by bailing wire
and bubblegum. (Actually, bailing wire . . . is how I put it. That's an
Americanism, I believe. I am not sure how the Brits say it.)
- Jed