Monday, December 08, 2008

Who knows what ET is thinking?

[My early New Year resolution is to stop giving my Nature colleagues a hard time by forcing them to edit stories that are twice as long as they should be. It won’t stop me writing them that way (so that I can stick them up here), but at least I should do the surgery myself. Here is the initial version of my latest Muse column, before it was given a much-needed shave.]

Attempts to identify the signs of astro-engineering by advanced civilizations aren’t exactly scientific. But it would be sad to rule them out on that score.

“Where is everybody?” Fermi’s famous question about intelligent extraterrestrials still taunts us. Even if the appearance of intelligent life is rare, the vast numbers of Sun-like stars in the Milky Way alone should compensate overwhelmingly, and make it a near certainty that we are not alone. So why does it look that way?

Everyone likes a good Fermi story, but it seems that the origins of the ‘Fermi Paradox’ are true [1]. In the summer of 1950, Fermi was walking to lunch at Los Alamos with Edward Teller, Emil Konopinski and Herbert York. They were discussing a recent spate of UFO reports, and Konopinski recalled a cartoon he had seen in the New Yorker blaming the disappearance of garbage bins from the streets of New York City on extraterrestrials. And so the group fell to debating the feasibility of faster-than-light travel (which Fermi considered quite likely to be found soon). Then they sat down to lunch and spoke of other things.

Suddenly, Fermi piped up, out of the blue, with his question. Everyone knew what he meant, and they laughed. Fermi apparently then did a back-of-the-envelope calculation (his forte) to show that we should have been visited by aliens long ago. Since we haven’t been (nobody mention Erich von Daniken, please), this must mean either that interstellar travel is impossible, or deemed not worthwhile, or that technological civilizations don’t last long.

Fermi’s thinking was formalized and fleshed out in the 1960s by astronomer Frank Drake of Cornell University, whose celebrated equation estimates the probability of extraterrestrial technological civilizations in our galaxy by breaking it down into the product of the various factors involved: the fraction of habitable planets, the number of them on which life appears, and so on.

Meanwhile, the question of extraterrestrial visits was broadened into the problem of whether we can see signs of technological civilizations from afar, for example via radio broadcasts of the sort that are currently sought by the SETI Project, based in Mountain View, California. This raises the issue of whether we would know signs of intelligence if we saw them. The usual assumption is that a civilization aiming to communicate would broadcast some distinctive universal pattern such as an encoding of the mathematical constant pi.

A new angle on that issue is now provided in a preprint [2] by physicist Richard Carrigan of (appropriately enough) the Fermi National Accelerator Laboratory in Batavia, Illinois. He has combed through the data from 250,000 astronomical sources found by the IRAS infrared satellite – which scanned 96 percent of the sky – to look for the signature of solar systems that have been technologically manipulated after a fashion proposed in the 1960s by physicist Freeman Dyson.

Dyson suggested that a sufficiently advanced civilization would baulk at the prospect of its star’s energy being mostly radiated uselessly into space. They could capture it, he said, by breaking up other planets in the solar system into rubble that formed a spherical shell around the star, creating a surface on which the solar energy could be harvested [3].

Can we see a Dyson Sphere from outside? It would be warm, re-radiating some of the star’s energy at a much lower temperature – for a shell with a radius of the Earth’s orbit around a Sun-like star, the temperature should be around 300 K. This would show up as a far-infrared object unlike any other currently known. If Dyson spheres exist in our galaxy, said Dyson, we should be able to see them – and he proposed that we look.

That’s what Carrigan has done. He reported a preliminary search in 2004 [4], but the new data set is sufficient to spot any Dyson Spheres around sun-like bodies out to 300 parsecs – a volume that encompasses a million such stars. It will probably surprise no one that Carrigan finds no compelling candidates. One complication is that some types of star that might resemble a Dyson Sphere, such as those in the late stage of their evolution when they become surrounded by thick dust clouds. But there are ways to weed these out, for example by looking at the spectral signatures such objects are expected to exhibit. Winnowing out such false positives left just 17 candidate objects, of which most, indeed perhaps all, could be given more conventional interpretations. It’s not quite the same as saying that the results are wholly negative – Carrigan argues that the handful of remaining candidates warrant closer inspection – but there’s currently no reason to suppose that there are indeed Dyson Spheres out there.

Dyson says that he didn’t imagine in 1960 that a search like this would be complicated by so many natural mimics of Dyson Spheres. “I had no idea that the sky would be crawling with millions of natural infrared sources”, he says. “So a search for artificial sources seemed reasonable. But after IRAS scanned the sky and found a huge number of natural sources, a search for artificial sources based on infrared data alone was obviously hopeless.”

All the same, he feels that Carrigan may be rather too stringent in whittling down the list of candidates. Carrigan basically excludes any source that doesn’t radiate energy pretty much like a ‘black body’. “I see no reason to expect that an artificial source should have a Planck [black-body] spectrum”, says Dyson. “The spectrum will depend on many unpredictable factors, such as the paint on the outside of the radiating surface.”

So although he agrees that there is no evidence that any of the IRAS sources is artificial, he says that “I do not agree that there is evidence that all of them are natural. There are many IRAS sources for which there is no evidence either way.”

Yet the obvious question hanging over all of this is: who says advanced extraterrestrials will want to make Dyson Spheres anyway? Dyson’s proposal carries a raft of assumptions about the energy requirements and sources of such a civilization. It seems an enormously hubristic assumption that we can second-guess what beings considerably more technologically advanced than us will choose to do (which, in fairness, was never Dyson’s aim). After all, history shows that we find it hard enough to predict where technology will take us in just a hundred years’ time.

Carrigan concedes that it’s a long shot: “It is hard to predict anything about some other civilization”. But he says that the attraction of looking for the Dyson Sphere signature is that “it is a fairly clean case of an astroengineering project that could be observable.”

Yet the fact is that we know absolutely nothing about civilizations more technologically advanced than ours. In that sense, while it might be fun to speculate about what is physically possible, one might charge that this strays beyond science. The Drake equation has itself been criticized as being unfalsifiable, even a ‘religion’ according to Michael Crichton, the late science-fiction writer.

All that is an old debate. But it might be more accurate to say that what we really have here is an attempt to extract knowledge from ignorance: to apply the trappings of science, such as equations and data sets, to an arena where there is nothing to build on.

There are, however, some conceptual – one might say philosophical – underpinnings to the argument. By assuming that human reasoning and agendas can be extrapolated to extraterrestrials, Dyson was in a sense leaning on the Copernican principle, which assumes that the human situation is representative rather than extraordinary. It has recently been proposed [5,6] that this principle may be put to the experimental test in a different context, to examine whether our cosmic neighbourhood is or is not unusual – whether we are, say, at the centre of a large void, which might provide a prosaic, ‘local’ explanation for the apparent cosmic acceleration that motivates the idea of dark energy.

But the Copernican principle can be considered to have a broader application than merely the geographical. Astrophysicist George Ellis has pointed out how arguments over the apparent fine-tuning of the universe – the fact, for example, that ratio of the observed to the theoretical ‘vacuum energy’ is the absurdly small 10**-120 rather than the more understandable zero – entails an assumption that our universe should not be ‘extraordinary’. With a sample of one, says Ellis, there is no logical justification for that belief: ‘there simply is no proof the universe is probable’ [7]. He argues that cosmological theories that use the fine-tuning as justification are therefore drawing on philosophical rather than scientific arguments.

It would be wrong to imagine that a question lies beyond the grasp of science just because it seems very remote and difficult – we now have well-motivated accounts of the origins of the moon, the solar system, and the universe itself from just a fraction of a second onward. But when contingency is involved – in the origin of life, say, or some aspects of evolution, or predictions of the future – the dangers of trying to do science in the absence of discriminating evidence are real. It becomes a little like trying to figure out the language of Neanderthals, or the thoughts of Moses.

It is hard to see that a survey like Carrigan’s could ever claim definitive, or even persuasive, proof of a Dyson Sphere; in that sense, the hypothesis that the paper probes might indeed be called ‘unscientific’ in a Popperian sense. And in the end, the Fermi Paradox that motivates it is not a scientific proposition either, because we know precisely nothing about the motives of other civilizations. Astronomer Glen David Brin suggested in 1983, for example, that they might opt to stay hidden from less advanced worlds, like adults speaking softly in a nursery ‘lest they disturb the infant’s extravagant and colourful time of dreaming’ [8]. We simply don’t know if there is a paradox at all.

But how sad it would be to declare out of scientific bounds speculations like Dyson’s, or experimental searches like Carrigan’s. So long as we see them for what they are, efforts to gain a foothold on metaphysical questions are surely a valid part of the playful creativity of the sciences.


1. E. M. Jones, Los Alamos National Laboratory LA-10311-MS (1985).
2. Carrigan, R.
3. Dyson, F. J. Science 131, 1667-1668 (1960).
4. Carrigan, R. IAC-04-IAA-, 55th International Astronautical Congress, Vancouver (2004).
5. Caldwell, R. R. & Stebbins, A. Phys. Rev. Lett. 100, 191302 (2008).
6. Clifton, T., Ferreira, P. G. & Land, K. Phys. Rev. Lett. 101, 131302 (2008).
7. Ellis, G. F. R. (2008).
8. Brin, G. D. Q. J. R. Astr. Soc. 24, 283-309 (1983).