Monday, June 15, 2009

Is SETI falsifiable?

Overheard in a conversation:
"I don't believe that SETI is science because it is unfalsifiable. If we don't find any aliens, they can simply say that we haven't searched for long enough."

SETI is the Search for Extra-Terrestrial Intelligence. It is a science program which looks at radio waves from space in order to search for signals from alien civilizations. The idea is simple. There are a tremendously large amount of stars out there. Some of those stars have planets, and some of those planets are capable of holding life. Some of those life-friendly planets actually have life, and some of them have intelligent life with civilizations, and some of those civilizations send out radio signals. We know that this must be true, because we already have one example: us. The question is, are there any such civilizations (besides us) close enough that we can see them?

However, some critics (as seen above) say that SETI is not science, because it is unfalsifiable. I disagree. SETI is science, and it is falsifiable.

Falsifiability is an idea created by philosopher of science Karl Popper. According to Popper, it is one of the things which distinguishes science from non-science. If we have a scientific hypothesis, then there must exist some experiment which might produce results to disprove that hypothesis. For example, the statement "All swans are white" could theoretically be disproven if we found a black swan.

The hypothesis "There are intelligent alien civilizations (besides us) which send out radio signals" does not seem to be similarly falsifiable. What observation could we possibly make that would falsify this hypothesis? Theoretically, we could look at every single bit of space throughout the entire universe and find no such civilization. But this is not a remotely realistic experiment. So SETI is unfalsifiable and not science.

But what happens if we restate SETI's hypothesis: "There are no intelligent alien civilizations (besides us) which send out radio signals". Now, this hypothesis can be falsified! We could theoretically make an observation which disproves the hypothesis. All we need to do is find a radio signal from an alien civilization. So now SETI is science.

So how about that? We state SETI's hypothesis one way and it's not science. We state SETI's hypothesis another way and it is science. This is absurd! Either we've misunderstood Popper's idea of "falsifiability", or it was never a very good idea in the first place (I would say both).

In any case, I've misstated SETI's central hypothesis. SETI does not tell us whether there are alien civilizations out there. Rather, it only tells us whether there are any nearby. This, in turn, tells us something about the frequency of such alien civilizations. Are there so many alien civilizations that everyone has at least one in their own backyard? Or are they so sparse that you'd have to go to the next galaxy to find another one? There are a ton of different estimates for the density of extraterrestrial civilizations. Personally, I agree with the more pessimistic estimates.

But of course, I would first defer to the evidence. If we want to distinguish between a universe where extraterrestrial civilizations are common, and a universe where they are sparse, then we have to look at a small sample of the universe and count the number of extraterrestrial civilizations. That's what SETI is doing. And since SETI hasn't found anything yet, the sparse hypothesis is looking more and more attractive.

We never really disprove the idea that alien civilizations are common. Maybe we're just unlucky. Or maybe the aliens are transmitting signals at the wrong frequencies, or none at all. Or their signals are to weak to reach us. Of course, few things are flatly disproven in science. Usually, they become less and less likely (or more and more likely) as we slowly accumulate evidence.


Alan said...

Falsification is a good idea, but in practice, most scientists don't strictly abide by it. A lot of science is purely exploratory in nature, without even a precise hypothesis. I would argue that currently SETI falls into this category -- they've just got their satellites up listening for interesting things. If they find a promising signal, then they can start to make genuinely testable hypotheses about what the signal means and whether it is a statistical anomaly.

Anonymous said...

Good post. I completely agree. I had the same debate with some friends about the simulation scenario (the hypothesis that we are living in a computer simulation) and got the same response from others--not falsifiable. I provided the same counterargument--when this hypothesis is stated one way it is not 'science' but when the negation is stated is magically becomes 'science' so I disagree with Popper's take on falsifiability--I believe that hypotheses need to be testable, which I define as either verifiable OR falsifiable.

Scott said...

The phrase "SETI isn't falsifiable" doesn't make a lot of sense because at least in its more mature forms, the idea of falsification was supposed to apply to theories. SETI isn't a theory -- I would say that it's (at best) an experiment. Even "there are no intelligent extraterrestrial species" isn't really a theory, it's just a statement.

As a side note, Anonymous: I don't see how the negation of the "simulation scenario" is any more falsifiable, but I'm interested in why you think so.

Peter Ozzie Jones said...

. . . the statement "All swans are white" could theoretically be disproven if we found a black swan.

Down here in Western Australia we say "All swans are black" until we find a white one.
Scientific Name: Cygnus atratus.