Seeing isn't believing

"Observation" isn't a direct action.

We see things because light bounces off of them and reflects into our eyes (biological processes aside). We often think of this as "seeing the object", but the key here is that light itself is a thing, and it's really the light you're seeing: you have no direct interaction with the object. You only interact with it through the intermediation of light.

All observation takes place like this: whether it's photons (light), electrons, or anything else, the only way to "observe" something remotely is to bounce things off of it. Most of the time, this doesn't matter much: photons bouncing off of "normal"-sized objects don't do much to the object. When we're talking about atomic or subatomic levels, however, the relative sizes between the thing we're trying to observe and the particle we're using to "see" it are much closer. As such, "bouncing" anything off of them knocks them around, similar to pool balls on a table.

The result, as Werner Heisenberg determined, is that we can only take one measurement - one observation - of a particle at any point, because as soon as you take one (usually position or velocity), you change the particle's other properties. This has been formally labelled "Heisenberg's Uncertainty Principle", and it forever broke down the myth of the "impartial observer" - someone (usually a scientist) who could sit back and "just watch" without interfering with a process.

Heisenberg's was the first blow to the notion of the independent observer, but it wasn't the last. Quantum theory, as it has developed, has focused more and more on particles themselves existing in what we identify today as some kind of probability wave. In short, if you shoot an electron out of a tube pointed in a given direction, you can have a general idea about where the electron may be at any given moment but you can't be sure. That's not surprising in itself for most people; the part that is hard to understand is that for observed results to be correct, saying the electron is actually anywhere doesn't work. The uncertainty in its location isn't just a factor of us not knowing - it seems to be intrinsic to the particle itself.

Here's the classical experiment. We have an electron gun pointed at a wall with two slits in it; logically, the electron can only travel through one or the other slit. However, if we place a piece of electron-sensitive paper on the other side of the slits to see where the electron ended up, we get something odd: the pattern that results can only occur if the electron takes both possible paths, not just one or the other. This is true even if the electrons are fired one at a time with enough time for the first to hit before the second is released.

Let me restate that, because it's possible to miss the implication: in order to account for what we actually see happen, we have to assume that part of the electron - something indivisible under normal circumstances - passes through each slit. The notion that we can't know its position isn't just a factor of our not having the precision to determine it: according to results, the particle has no definite position until it reaches the paper and, it seems, has to take all potential paths to get there. Our best understanding of this is that it exists only as a probability, and that somehow those potential probabilities are what interact and cause the observed pattern.

It gets weirder: if we set up a system to watch the electron as it passes through the slits so that we can observe which one it goes through, the odd pattern requiring the probability explanation disappears; the result is explanable with classical physics models. So, it appears that the electron behaves like a single particle once we observe it, but until we do, it behaves like a set of probabilities. Furthermore, this is true not just of electrons but of every particle the process has been tested with: as near as we can tell, all matter exists probabilistally until observed and classically after observation.

Now, in all likelihood, what's occurring here is probably not and "actual" change in the nature of the particle but some kind of observational distortion similar to Heisenberg's Uncertainty; that being said, scientists have tried to disprove this and other weird aspects about quantum theory for years to no avail. Philosophers and religious types, of course, have taken this "elevation of the observer" to dizzying heights [pardon the pun] in an attempt to "scientifically" validate free will, all manners of Gods, or whatever unprovable phenomenon they prefer. The use of probability as a model for understanding quantum theory also led to Einstein's famous admonition that God "does not play dice with the universe"; unfortunate for the good professor, observational evidence seems to disagree.

Even if we discount the philosophical extensions, the increased awareness of observation and its participation in the process can be very useful. In "real world" terms, there are far more practical impacts to observation than simply photons kareening around like billiard balls: namely, the fact that everything we observe not only has to pass through intermediary objects and multiple analog processing systems, but that it then has to be interpreted by us. Bringing into focus how our biology and psychology can affect our observations (and, thus, our interactions with and even thoughts about everything around us) is the focus of general semantics, but that's for another day.

For now, just try to keep in mind that you aren't really reacting to the world - you're reacting to reactions to the world, and even then through a layer of interpretation.