Chances are, you don’t know what you think you know.
I got an email today that criticized academics for “not being open to new ways of thinking” and “only being interested in rooting out heresy.” I’m an academic, and this accusation comes from another universe than the one in which I live. The academics I work with are all about discovery. It is true that scientists believe in evidence-based knowledge. We develop theories and then design experiments to test those theories. A theory that’s not falsifiable–not testable–is pretty useless. In particular, it can’t reliably predict how the universe works. If the data don’t support the hypothesis, then as scholalrs we have to discard the hypothesis, no matter how cherished. That applies to theories in sociology every bit as much as theories in physics.
Now, I suppose it is true that I’m not going to look at the entrails of a goat to decide how to design a 3D display. That means I’m picky about the evidence I’ll accept. But the reason I won’t accept the entrails of a goat is that the evidence shows there’s no correlation between them and the truth of a proposition. Historically, humanity has done that experiment and that’s why we’ve collectively abandoned the entrails method.
I don’t really believe in telekinesis either. I mean, by will power alone I can move my arm. I can even move around my laptop–by picking it up with my hand. But that’s not telekinesis. I can’t lift my model of an X-wing fighter out of the swamp by just thinking about it. I can’t even change the pixels on my laptop screen by the power of thought. Photons and electrons don’t work that way, to say nothing of X-wing fighters.
But maybe that’s not right. There’s this graph from the Noetic Research Institute at Princeton:
Anyway, if you run one of these generators repeatedly over a period of time and record the results, you expect the average value to be zero. Most of the time, this is more or less what happens. You’ll almost never get results of exactly zero, but on average the results will be close to zero. And the longer you extend the period and the more generators you include in your calculations, the closer your average will generally get to zero. Not always, but generally. After all, it’s random. The generators are like those photons and electrons which are random quantum entities.
The above graph shows the running average from these thousands of random number generators as the results came in and the outcome of the 2008 presidential election became clear. That’s the jagged red line. Notice that the results exceed the expected value–the bottom solid line–for several hours.
Now, these are random numbers. That means that sometimes these running averages will differ from zero, just due to random chance. But if the deviation is large enough, or continues for a long enough period, random chance becomes a more dubious explanation. That’s where the blue line comes in. Roughly speaking, when the red line is below the blue line, then we’d believe that random chance accounts for whatever difference we see. If it’s above the blue line, however, then we’d accept the alternative explanation that the differences are not merely random. In fact, when we’re above the blue line, we’re “pretty sure” (as in, 95% sure) that the deviations are not due to random chance.
So, what’s going on here? Do results like this make the case for telekinesis? Well, no. These results are not positive evidence of anything except that the generators weren’t spitting out random numbers for a period of time. The fact that they were spitting out “positive” numbers certainly isn’t evidence that people had “positive” thoughts about the election, as some of my fellow Obama supporters gleefully contended.
What the graph does show is that the generators varied from the expected value of zero. They didn’t vary by a lot, numerically, but it’s possible to calculate the likelihood of seeing this magnitude of variance if the random generators were, in fact, random. That calculation suggests–but doesn’t prove–that some non-random effect was present.
Of course, it’s possible this is just coincidence. To an empiricist–me–that seems the most likely explanation. But the Noetic Institute has been gathering this kind of data for a decade and more. For many large-scale events, such as a presidential election, they see similar deviations. However, they don’t see them for every such event, and they sometimes see them when there is no obvious event to “explain” them. This tends to bolster the coincidence explanation. But I admit the jury is out.
In truth, I’d like to believe in telekinesis and other mystical things. My current work-in-progress, Entangled, is based on another sort of mystical and semi-scientific idea. There is a mystical urge in all of us. The universe has enough mystery to accommodate almost anything.
Except goat entrails. You’ve got to draw the line someplace, and chances are that’s a safe one.