This is a ten-minute excerpt from that talk, where Bell describes how Einstein's hope fails - that "hidden variables" could eliminate "nonlocality," with its strange "action at a distance."

Actually, it is only non-physical information about probabilities that changes instantaneously over large regions of space when a measurement is made and a wave function collapses.

Click here to see it on YouTube, where you can turn on a line-by-line version of the transcript below (CC subtitles), which will follow along as Bell talks.

(For a transcription of the complete talk, click here. Video of the full talk is available here in Windows Media WMV format, but the sound is hopelessly out of sync, making it almost impossible to understand what Bell is saying, unless you read the transcript while watching. We have corrected the audio sync for this excerpt.)

It just is a fact that quantum mechanical predictions and experiments, in so far as they have been done, do not agree with [my] inequality. And that's just a brutal fact of nature. The genetic hypothesis, which seems absolutely compelling for parallel devices, simply doesn't work for nonparallel devices. You can't get away with the genetic hypothesis, and therefore the Einsteinian argument fails. No action at a distance led you to determinism, in the case of parallel polarisers, but determinism, in the case of off parallel polarisers, leads you back to action at a distance:

Now, in my opinion, in answer to the question that you posed at the beginning, I don't know this phrase is too strong and active an assertion, I cannot say that action at a distance is required in physics. But I can say that you cannot get away with no action at a distance. You cannot separate off what happens in one place and what happens in another. Somehow they have to be described and explained jointly. Well, that's just the fact of the situation; the Einstein program fails, that's too bad for Einstein, but should we worry about that? So what?

Now, there are three replies to the question “So what?” One is that the whole idea of action at a distance is very repugnant to physicists. If I were speaking for an hour..., I would bombard you with quotations from Newton, and Einstein, and Bohr, and all the other great men, telling you how unthinkable it is that by doing something here, we can change the situation in a removed place. I think that the founding fathers of quantum mechanics did not so much need Einstein's arguments about the desirability of no action at a distance, as they looked away. The whole idea that, either there might be determinism, or action at a distance, was so repugnant to them that they looked away. Well that's tradition, and we have to learn in life sometimes to learn new traditions. And it might be that we have to learn to accept not so much action at a distance, but inadequacy of no action at a distance.

There are two more professional reasons for being discontented with the situation. Now one is relativity. According to relativity, the notion of simultaneity is relative. And events which are simultaneous for one observer are not simultaneous for another. So it does not make sense for very distant situations, to say that one event has occurred before or after another. So if we allow the result at one of these experimental set-ups to depend on what an experimenter does at the other, we have a puzzle, because we would not like what he does here to have an effect there, before it is done here. But if I say that this is affecting that, I can find some observer for whom this comes after that. So if I set up a traditional causal model, which the cause effects are allowed to be nonlocal, in the sense of propagating instantaneously over large distances, in some frame of reference the cause will come before the effect. So we have to be a bit more subtle than that somehow. I have to find some way out of this situation, which allows something somehow to go from one place to another, very quickly, but without being in conflict with special relativity. And that has not been done. We have the statistical predictions of quantum mechanics, and they seem to be right. The correlations seem to cry out for an explanation, and we don't have one.

The other reason is no signals. It is a fact that I cannot use whatever this nonlocal connection is to send signals. When you look at what quantum mechanics predicts, it predicts so long as you look at just one side or other of this experiment, you will simply have no information about what is happening in the other place. No matter what that other fellow does with his equipment, you will not notice anything funny happening in your side. As an analogy of that, I could say, supposing we were tossing coins, I here and one of you people down here. And supposing I had the power to say that your coin will turn an extra time before it falls on the table. Now you are looking at your coins and you see heads tails heads tails. And you don't know when I have exercised my power to turn it once more, because you didn't know whether it was going to fall heads or tails. So we have the curious situation, that to explain the correlations between my results and yours, we have to invoke some such mysterious power. But it is one which I absolutely cannot use to send you a message. I got here a demonstration of that. This is a computer simulation of such an experiment in which people are calling heads and tails. And when it comes up heads I have written 'H', and for tails I have written blank, so that you can see it from where you're sitting. And they're a whole series of random heads and tails, you can see it there (Fig 2).

Fig. 2. Result of computer simulation of a random series of heads ‘H’ and tails (blank)

Now at some point I exercised my power. My remote power to turn a head to a tail. And here is the result (Fig. 3).

Fig. 3. Inverted display of Fig.2: heads (blank), tails ‘H’. In some places the random code is changed.

So somewhere in there I have done something to the random code. I exchanged heads for tails, but you absolutely cannot see that. This message, as far as you're concerned, is as meaningless as the other. It's only if you have two copies of that, that you can compare it, that you can get something (Fig. 4).

Fig. 4. Superposed images of Figs. 2 and 3 (after B. Julesz)

That curious situation has inspired a musical composition. There is a musical composition called `The Bell's theorem blues'. I'm not going to sing it, I'll say the words:

Doctor Bell says we’re connected,
He called me on the phone,
But if we’re really together baby,
How can I feel so all alone?