Sunday, August 30, 2009

Theories (again!)

A while ago, I posted a rhyme about the nature of scientific theories. In thinking back, though, I think some of the ideas I was trying to convey don't come across very clearly in rhyme. So I decided to clean up the notes I made while researching and writing that rhyme, and post them here.

It's very tempting to think that science is the best means for understanding physical reality ... what's really going on in the universe. After all, science has given us so many useful by-products. Who can argue with the light bulb, the car, or the computer? But is it really reality we're observing?

Think of the fable of the blind men and the elephant. Each touched a different part of the elephant, and came away with a completely different idea of what an elephant was. The man who touched the trunk thought the elephant was like a snake, but the man who touched the legs thought it was like a tree, and the tail man thought the elephant was like a rope. Each of them drew conclusions based on observations of a very limited part of the whole elephant. Moreover, their ideas of an elephant are shaped by where they came from, as well as by what parts of the elephant they touched. The man from the swamp thought the elephant very dry and rough. The man from the north thought the elephant very bald and unprotected. Our perception of reality is shaped as much by past experiences as by the current observations.

The other point usually overlooked with the blind men and the elephant parable is that each man understood the elephant by likening it to something already familiar ... snakes, trees, rope, etc. Understanding is the process of fitting new knowledge into an established, intuitive framework. For centuries, physics could be understood by thinking of forces acting on particles.

Theories are cognitive inventions which seem to explain the world. But explanations are simply that which leads to understanding, and understanding is a very subjective, human phenomenon. To understand something means to have a model of how it behaves that enables us both to account for past observations and to predict future behavior. So explanation is the process of creating a metaphor ... a mental model which corresponds to the observed world in important ways.

Like the blind men, we only perceive a tiny fraction of all the information available about anything in the universe. Our whole ideas of objects, planets, stars, etc. is just a mental model we've constructed from the scant bits of evidence we've been able to observe. So any conjectures about what these things are like or how they behave is really a stretch.

Viewed in this way, the heliocentric model of the solar system bears the same relationship to the actual solar system as the abstract idea line does to any roughly straight object or edge in reality. It is an abstraction and simplification, to allow us to grasp cognitively what is, in fact, a complex phenomenon.

There is little doubt the earth moves around the sun, but what this actually means is that the mathematics of describing the motion of the planets relative to each other and the sun is simplest when we think of the sun as being at one focus of the elliptical orbits. Both simplicity and mathematics are themselves products of how our brains work. So all we really know is that our brains find it easier to understand and predict the motion of the planets if we think of it as a heliocentric system. This is a far cry from saying "This is real." If someone discovers a simpler or more descriptive mathematical model, than that will become "the way it is," and our present heliocentric model will seem quaint and naive. In fact, as the moon orbits the earth, the earth orbits the sun and the sun itself orbits the center of the galaxy, which is itself moving. So there are no neat little elliptical paths as Kepler would have had us believe. The paths could just as (more?) correctly be described as helical.

Imagine an alternate universe in which the sun and planets all revolve around the earth, just as the pre-Copernicans believed. Imagine these orbits are composed of cycles and epicycles, just as the pre-Copernicans supposed. The effect of the combination of cycles and epicycles is that the planets and sun are always in the same positions, relative to each other, as they are in our universe. What, then, is the difference between that universe and ours? None! They are identical in every respect. Planets and stars do not obey laws. The laws are just our attempts to understand and describe their behavior. But understanding and description are cognitive tools, not characteristics of nature.

You can never find a line in real life, because a line is an idea ... an abstraction from actual edges and boundaries we find in reality. In the same way, a scientific explanation is an abstraction that fits the observed facts. The planets don't revolve around the sun in elliptical orbits. They are all just moving bodies (or more accurately, just concentrations of particles) in space, whose distances and orientations change in relation to one another. The ellipse is an idea we invented to make things predictable. If another theory fit all the observations just as well, it would be just as true. Two theories which explain the same phenomena with the same verifiability are equally valid.

When you try out a new piece of computer software, you can find out all about it by watching how it behaves, experimenting with different actions to see what reaction the software has, etc. By doing this, you can come up with a very detailed and complete description of every aspect of the software's behavior. Yet this says nothing about what the software is actually doing internally, or what its developers intended. In fact, this is done often, and is called reverse engineering ... the processing of trying to figure out what something does by controlled experiments to observe its behavior.

NOTE: This does not imply there's any validity to the so-called intelligent design! Currently, it's very difficult to make any critical assessment of science without being regarded as a defender of creationism or the so-called intelligent design theory. Nothing could be further from the truth. But science itself has become an authority, and we should always question authority.

One of the cornerstones of science is the experiment. The scientific method, including the experiment, is often thought of as a foolproof means for ascertaining truth. But there are flaws in this. For one, the experiment proves that a given result is possible ... that it might happen. It doesn't prove that the result will always happen. No matter how many times we demonstrate that oil and water don't mix, we can't prove that they never mix.

Moreover, in the scientific method, the experiment is a hypothesis test. We make a conjecture about some phenomenon, and then experiment to see if the conjecture holds up. But we only do experiments on what we already consider to be plausible ideas. In a sense, we're filtering out things that don't seem scientific, and only experimenting with those things that fit the belief framework. In particular, we assume causality to apply in all situations, since causality is the basis of experimentation. So if there could possibly be phenomena which don't behave causally, they will simply be dismissed as observational errors.

Within a belief framework, the theory seems plausible and valid, and consistent with the rest of the framework. Outside the framework, however, the theory may seem irrational. This is not unlike fantasy fiction. We accept that Frodo may have a ring that makes him invisible because it's within the natural order of the universe Tolkien created for his stories. So saying the ring works by powerful magic is a perfectly sound explanation in that universe. Our own universe is, of course, not completely knowable, so we create stories of how things behave to satisfy our own curiosity. In one set of stories, everything may be purely mechanical. The universe is a giant machine that runs according to definite, if unknown, principles. Another set of stories may have the universe a place of mystery and unpredictability, subject to the whims or moods of some controlling being or beings.

If you happen to believe a supreme being created the earth, than the earth becomes evidence of that being's existence. We can see the circularity of the logic when it's stated this way, but in general, belief systems form a context in which we interpret all experiences, and the experiences themselves then reinforce the context. Each of us lives in a universe that's an ''idea'' we formed from our observations. That's why something that violates our expectations of how the world behaves is so surprising.

But that's pretty much the history of science, isn't it? Violated expectations? We're constantly revising theories and replacing them with new ones. That fact alone should convince us that, as persuasive and practical as our scientific knowledge is, we don't have a lock on reality yet.

Well, now that I've reread that, the rhyme was probably clearer.

There were a number of references, but here are two of the most helpful:

No comments: