: Ah! But when you put a sensor over one of the slits, all
: the electrons land in the same place on the film
: behind it? I don't quite get it.
When you put a sensor over one of the slits, you don't get any interference pattern. You just get two overlapping regions of brightness, one opposite each slit, which is what you'd expect classically. Most electrons go straight through one of the slits and hit right opposite; fewer go through at an angle and hit the screen farther away; etc. Basically, the sensor forces each electron to go through only one slit rather than both at once, so the self-interference pattern doesn't appear.
: (Another odd thing about this experiment that I've heard:
: if the computer sensing the particles is set to erase
: the data after sensing them, they appear as waves. But
: if it is set to keep the data, or a human is looking
: at it as it happens, they appear as particles. So an
: event in the future - the reading or erasure of the
: data - effects an event in the past - whether the
: thing appears as particle or wave).
Yes, quantum theory requires "time travel" of this sort; it just bars the propagation of *information* at speeds faster than light. "Correlation, not causation" is the catchphrase.
If you've never read Bell's Theorem, I recommend that you read it. It's the coolest thing I've learned about in physics to this date--completely destroys the possibility of a locally-deterministic theory (in other words, what Einstein was hoping for where we discover that all the quantum-random processes are actually determined completely by properties we just weren't measuring before), and it *proves* that certain influences propagate faster than light.
--SiliconDream
