This is a very deep question, and one which is still much in dispute among physicists and philosophers. I will give a fairly conventional modern interpretation, but not everyone would agree. Sorry if it's a little long-winded.
The quantum state of something changes in time following a linear
equation, just like many familiar waves. That means that it obeys something called the superposition principle: the sum of any two solutions to the equation is also a solution.
If you have a small system (maybe a single spin) and measure some property of it (maybe whether the spin points north or south), what youíre doing is setting up (or noticing) some physical situation where the behavior of a big thing depends on the property youíre measuring, e.g. you may have a needle that will point up if the spin is north and down if the spin is south. Then in a typical measurement the spin starts off in a superposition of the N and S states, but the overall quantum state ends up a superposition of (N, up) and (S, down), where the letter describes the spin and the word describes the needle.
Now no-one has ever seen a needle that's both up and down. If the whole world obeys the quantum equation, what really comes out is a superposition of:
(N, up, you seeing up) and (S, down, you seeing down).
So the 'observation' splits you into two separate versions! In more realistic pictures, you would constantly be splitting into many versions. Some people think that somehow or other only one version survives, with some random process pruning away all the others.
So to get back to your question:
Observations require definite values for big things (needles, minds, etc.), and in 'measurement' situations that requires that the 'measured' quantity take on a definite value for each version of the observer, even though the initial spin (or whatever) may not have had a definite value for the observed quantity.
We donít know of any special physical role played by ímindí, but the nature of our minds does limit our observations to pieces of the quantum state, not the whole thing.
I wouldnít be surprised if this incomplete answer called for a follow-up question, so please donít hesitate to ask one.
(published on 01/02/2008)