Fancy Math Can’t Make Aliens Real

An astrophysicist says extraterrestrial civilizations “almost certainly” existed at one time or another. Here’s what’s wrong with his argument.


Last week, The New York Times published an op-ed titled, “Yes, There Have Been Aliens.” As its headline suggests, the piece makes an extraordinary claim. “While we do not know if any advanced extraterrestrial civilizations currently exist in our galaxy,” its author writes, “we now have enough information to conclude that they almost certainly existed at some point in cosmic history.”

That we could know such a thing is not inconceivable. For decades now, a small group of “interstellar archaeologists” has pored over star surveys, looking for evidence of long-dead civilizations, in the form of enormous technological structures. Reading that headline in the Times, I wondered: had one of these astronomers seen something extraordinary?

Alas, I was disappointed.

Adam Frank, a professor of astrophysics at the University of Rochester, wrote the essay that appeared in the Times. Frank is a gifted scientist, and a thoughtful science writer. He begins the op-ed with an enthusiastic update on the ongoing exoplanet revolution. I must confess I share his enthusiasm. I suspect that future historians of science will wonder what it was like to live in this moment. A little more than two decades ago, we weren’t sure whether there were any planets outside our solar system. Now we have reason to believe that nearly all stars host planets, and that many of them are rocky and wet like our own. No generation of humans has ever gazed up at night skies so pregnant with possibility.

It is precisely this profusion of planets that gives Frank confidence that ours is not the first intelligent civilization. “Given what we now know about the number and orbital positions of the galaxy’s planets,” he tells us, “the degree of pessimism required to doubt the existence, at some point in time, of an advanced extraterrestrial civilization borders on the irrational.” Most of us have heard a version of this argument, late at night, around a campfire: Look at all the stars in the night sky. Is it really possible that all of their planets are sterile, and all of their predecessors, too?

These arguments have their appeal, but it is an appeal to intuition. The simple fact is that no matter how much we wish to live in a universe that teems with life—and many of us wish quite fervently—we haven’t the slightest clue how often it evolves. Indeed, we aren’t even sure how life arose on this planet. We have our just-so stories about lightning strikes and volcanic vents, but no one has come close to duplicating abiogenesis in a lab. Nor do we know whether basic organisms reliably evolve into beings like us.

We can’t extrapolate from our experience on this planet, because it’s only one data point. We could be the only intelligent beings in the universe, or we could be one among trillions, and either way Earth’s natural history would look the exact same. Even if we could draw some crude inferences, the takeaways might not be so reassuring. It took two billion years for simple, single-celled life to spawn our primordial lineage, the eukaryotes. And so far as we can tell, it only happened once. It took another billion years for eukaryotes to bootstrap into complex animal life, and hundreds of millions of years more for the development of language and sophisticated tool-making. And unlike the eye, or bodies with legs—adaptations that have arisen independently on many branches of life’s tree—intelligence of the spaceship-making sort has only emerged once, in all of Earth’s history. It just doesn’t seem like one of evolution’s go-to solutions.

Frank compresses each of these important, billions-of-years-in-the-making leaps in evolution into a single “biotechnical” probability, which is meant to capture the likelihood of the whole sequence. For all we know, each step could be a highly contingent cosmic lottery win. Perhaps eukaryotes “usually” take tens of billions of years to evolve, and we lucked into an early outlier on the distribution curve. Perhaps we have been fortunate at every step of the way. Frank’s argument skips over these probabilities. Or rather, it bundles them up into a single, tidy unknown, that he can hammer with a big italicized number:

“What our calculation revealed is that even if this probability [that technological civilization evolves] is assumed to be extremely low, the odds that we are not the first technological civilization are actually high. Specifically, unless the probability for evolving a civilization on a habitable-zone planet is less than one in 10 billion trillion, then we are not the first.”

Absent a clear account of how often we can expect planets to spawn technological civilizations, we don’t have any way to evaluate that “10 billion trillion” number. We certainly don’t have grounds to say that the “odds are high” that some civilization preceded ours, or enough evidence to suggest that skepticism about the possibility “borders on the irrational.”