What I Think: Gleiser's on the Theory of Everything

I read an interesting NPR Opinion Piece written by Marcelo Gleiser. Go read it. In it he argues against pursuing a so-called "Theory of Everything", or TOE for short. It makes several good points.

First off, he's right that a TOE would not actually be a theory of everything. Rather, it would be a theory of everything that is considered the most fundamental building blocks of the physical universe. This leaves out a lot of stuff, like nearly everything in day to day life.

Second, he's also right to link the desire for unity to Judeo-Christian monotheism. This may seem like an odd idea nowadays since religion and science are seen as opposing sides of a dichotomy (they aren't). Though it may seem counterintuitive, historically these ideas come from the same source. In addition, monotheism and the search for scientific unity get support in similar ways. Science is often thought of as becoming unified in the "mind of God".

Ceteris Paribus

But there is one point that he makes that I think is not quite right. He argues that the TOE doesn't make sense, but I would disagree with this. This idea comes from his argument that theories will always be incomplete.

"A physical theory can only be proved wrong, never right, at least in any permanent sense. This is because every theory is necessarily incomplete, always ready for updates as we learn more about the physical world. What we can say about Nature depends on how we measure it, the precision and reach of our instruments dictating how 'far' we can see. As a consequence, no theory that attempts to unify current knowledge can be seriously considered a 'final' theory or a TOE given that we can't ever be sure that we aren't missing a huge piece of evidence."

Essentially, he's arguing that a TOE is impossible because all theories will always be incomplete. That is to say, there will always be some relevant data missing if we try and use it to explain everything.

But this is just an uninteresting feature of induction itself. No inductive inference can ever be proven true in the exact sense of a deductive statement. Inferences can only be stronger or weaker, never "proven".

This is not what prevents us from having a TOE, however. If a theory is shown to be sufficiently strong, then we say it is true. This is what we mean when we say that classical mechanics or relativity is true. It's certainly not true in the deductive and exact sense, but that doesn't mean it isn't true in a different sense.

Is Gleiser committed to the idea that our greatest scientific theories are not true? That is to say, not true in any sense of the word? I would hope not, because that argument showing that any TOE cannot be true necessarily also proves that any theory cannot be true necessarily. Regardless of whether or not this is the correct way to think of scientific theories (it's not), it clearly is not an idea in favor of his argument since it would admit that a "true TOE" is not any more or less sensible than the truth of any other scientific theory.

Outside influences may not integrate into theory

Gleiser's argument still allows the possibility of formulating a theory that could explain any sort of circumstance, had we enough data about that circumstance. And that is exactly what a TOE would supposedly do. It would allow us to explain anything we wanted, so long as we had enough information about it.

It is this idea about a TOE that is actually wrong. For a theory to work in the world there must not be any relevant outside influences acting on the system that is being studied. For physics, these are often things like thick shielding so as to prevent radiation or other similar influences. Making the theory work means making sure no influences like this could affect the system. Without these shielding conditions, our best theories could only come up with 'educated guesses' at best.

So how to make sense of these outside influences? Most who give it some thought might think that these outside influences could be understood within the theory if we knew enough about them.

Here's a mental picture: some scientists are studying fluid dynamics of a substance that is not well understood. If someone breaks the shielding conditions by reaching into an apparatus and waving their hand around, sufficiently disrupting the experiment, we might say that we could in principle factor their influence into the theory if we only knew the exact motions of their hand, etc.

This may be true, but it misidentifies the problem. The problem is not the fact that there is not enough information, but that the information we have is not the right kind. Presumably, we could learn many things about how this person waved their hand in the apparatus, like why they did it or the sensations they felt on their hand when they did it. None of these facts would be helpful, however. We would need to have the right kind of info: we would need to know the exact motion of their hand in the apparatus and the material properties of their hand, etc.

This is why a TOE is such an absurd notion: outside influences on a system might not make sense in the theory.

This is not to say that a TOE is impossible. It may be possible, but it seems highly unlikely we will find one. More importantly, however: searching for such a vain and unlikely scientific theory is very likely to hurt our efforts to understand less "symmetric" or "beautiful" phenomena. Not every scientific discoveries are "beautiful", but this does not make them insignificant by any stretch of the imagination. We should not spend our time seeking out these ideal theories when our attention is best put into doing what science does best: helping us to understand the world around us.


(It should be noted that this discussion is based heavily on the ideas found in Nancy Cartwright's book, The Dappled World.)