The Chance of Finding Aliens:

Reevaluating the Drake Equation

In 1961 astronomer Frank Drake wrote the equation that put the search for alien civilizations on a scientific footing and launched the modern SETI movement.  How do the equation's numbers look today?

By Govert Schilling

(Adapted from Sky & Telescope December 1998. Updated and expanded

by Sky & Telescope, May 1999.)

Searching for extraterrestrial life has become a hot topic among astronomers, biologists, and the general public, but few remember how the subject was jump-started 40 years ago. In September 1959, physicists Giuseppe Cocconi and Philip Morrison published a landmark article in the British weekly Nature with the provocative title, "Searching for Interstellar Communications." Cocconi and Morrison argued that radio telescopes had become sensitive enough to pick up transmissions that might be broadcast into space by civilizations orbiting distant stars. Such messages, they suggested, might be transmitted at a wavelength of 21 centimeters (1,420.4 megahertz). This is the characteristic wavelength of radio emission by neutral hydrogen, the most common element in the universe. Other intelligences might see this as a logical landmark in the radio spectrum where searchers like us would think to look.

Seven months later, in April 1960, radio astronomer Frank Drake became the first person to carry out a systematic search for intelligent signals from the cosmos. Using the 25-meter dish of the National Radio Astronomy Observatory in Green Bank, West Virginia, Drake "listened in" on two nearby Sunlike stars: Epsilon Eridani and Tau Ceti. His Project Ozma (named after the main character in L. Frank Baum's book Ozma of Oz) was cheap, simple, and unsuccessful.

Following the Ozma experience, Drake organized a meeting with a select group of scientists to discuss the prospects and pitfalls of the search for extraterrestrial intelligence -- nowadays abbreviated SETI. In November 1961, ten radio technicians, astronomers, and biologists convened for two days at Green Bank. Young Carl Sagan was there, as was Berkeley chemist Melvin Calvin, who received news during the meeting that he had won the Nobel Prize in chemistry.

It was in preparing for this meeting that Drake came up with his famous equation:

N = R x f_p x n_e x f_l x f_i x f_c x L

Today this string of letters and symbols can be found on T-shirts, coffee mugs, and bumper stickers. It is simpler than it looks. It expresses the number (N) of "observable civilizations" that exist in our Milky Way galaxy as a simple multiplication of several, more approachable unknowns.

R is the rate at which stars have been born in the Milky Way per year, f_p is the fraction of these stars that have solar systems of planets, n_e is the average number of "Earthlike" planets (potentially suitable for life) in the typical solar system, f_l is the fraction of those planets on which life actually forms, f_i is the fraction of life-bearing planets where biological evolution produces an intelligent species, f_c is the fraction of intelligent species that become capable of interstellar radio communication, and L is the average lifetime of a communicating civilization in years.

The Drake equation is as straightforward as it is fascinating. By breaking down a great unknown into a series of smaller, more addressable questions, the formula made SETI a tangible effort and gave the question of life elsewhere a basis for scientific analysis.

Astronomers and biologists alike have tried to "solve" the equation ever since. At first sight, deriving a good estimate for the answer might seem fairly straightforward. But in reality, the number of communicating intelligences can't be computed so easily. Several of the variables in the equation have been firmed up in recent years. But at least three remain very unknown.

The rate of star formation in our galaxy is approximately one per year, R = 1. The next factor, f_p, is probably smaller than one: not every star can have planets. On the other hand, if a star has a planetary system, it seems plausible that two or three of its planets and moons will have liquid water and be potentially suitable for the origin of life, so maybe the product of f_p and n_e is close to 1.

Astronomers estimate that R, the number of stars born in our galaxy each year, is currently about 1. Newborn stars emerge from giant gaseous pillars in the Eagle Nebula, as the stellar winds from their brilliant, older siblings (outside the frame) blow away the new stars' cocoons of gas and dust. Credit: STScI/NASA.

Optimists would argue that life will form wherever it can (f_l = 1), that the Darwinian process of natural selection eventually favors the evolution of intelligence (f_i = 1), and that no intelligent civilization would exist for long without discovering electricity and radio and feeling the urge to communicate (f_c = 1).  In this most optimistic case, the Drake equation boils down to the simple observation that N = L (the average lifetime of technological civilizations in years).  If L is, say, 10,000 years, there would be about 10,000 chatty civilizations in our galaxy.  And that's assuming that only one arises during a planet's entire multi-billion-year lifetime.

That figure of 10,000 would mean there is one radio-emitting civilization per 40 million stars -- reason enough to tune in on the heavens and start hunting for them. If they were scattered at random throughout the Milky Way, the nearest one would probably be about 1,000 light-years from us. A two-way conversation would require a time equal to a large part of recorded human history, but a one-way broadcast might be audible.

However, 40 years of SETI have failed to find anything, even though radio telescope apertures, receiver techniques, and computational abilities have improved enormously since the early 1960s. Granted, the "parameter space" of possible radio signals (the possible frequencies, locations on the sky, signal strengths, on-off duty cycles, etc.) is vastly larger than the tiny bit that has yet been searched. But we have discovered, at least, that our galaxy is not teeming with powerful alien transmitters continuously broadcasting in our direction near the 21-centimeter line. No one could say this in 1961.

Have we overestimated the values of one or more of the Drake parameters? Is the average lifetime of technological civilizations quite short? Or have astronomers overlooked some other, more subtle aspect?

Let's reevaluate the Drake equation by analyzing each term separately. R, the number of stars born in the Milky Way each year, is indeed approximately 1 -- astronomers are quite sure about that. In fact, astronomers have recently determined that the rate of star formation was higher several billion years ago, when the stars that might now bear intelligent life were being born. So a value of R = 3 or 5 might be more realistic.

However, astronomers and biologists are much less certain about the subsequent terms in the equation.

How Many Planets? f_p

The second variable is f_p, the fraction of stars that have planetary systems. Recent discoveries that many young stars are surrounded by planet-forming disks, along with detections of actual planets orbiting nearby Sunlike stars, confirm what astronomers had already suspected: planets are common. 

So-called "protoplanetary disks" have been detected by various infrared observations and are seen directly in Hubble Space Telescope photographs of the Orion Nebula, one of the most prolific star-forming regions in the Milky Way today. Such observations seem to imply that at least 50 percent of all newborn stars will be accompanied by planets. Recent submillimeter-wave observations have shown much more tenuous dust disks around a number of older stars, including Drake's first target, Epsilon Eridani. Many of these disks are doughnut shaped. According to some theorists, the central holes can only be swept clear by planets accreting gas and dust from the inner portion of the disk.

Observations of the Orion Nebula are helping astronomers close in on the value of f_p, the number of stars in our galaxy with planets. At least half of the young stars in this region are surrounded by thick, dusty disks -- excellent planet-forming material. 

As for actual planet detections, the most productive teams of planet hunters have found at least one giant world orbiting about 5 percent of the single (non-binary) Sunlike stars they have searched so far. At face value, this might imply that about 5 percent of stars have planets, so f_p would be 0.05. However, there's a catch: the current search techniques are sensitive only to massive planets, especially those in very tight orbits. Solar systems like ours cannot yet be recognized. Quite probably the real fraction of single Sunlike stars with planets of some kind is much higher than 5 percent. It could be as high as 50 or even 100 percent.

So what do these new observations tell us about f_p? Although we don't yet have a final value, it's now clear that f_p is substantial and is not a bottleneck in the Drake equation.

How Many Good Planets? n_e

There's less encouraging news when we turn to the equation's next term, n_e. This factor represents the average number of worlds in a typical solar system that have environments suitable for the origin of life (the "e" stands for "Earthlike"). In his 1992 book Is Anyone Out There?, Drake recalls that the participants in the Green Bank meeting concluded that the minimum value of n_e lay between one and five. In other words, every planetary system was expected to contain at least one minimally Earthlike place (defined as where liquid water is possible), and that there might easily be three, four or five hospitable worlds per system.

This optimistic view was based on the assumption that our own solar system is typical. Today Mars and Jupiter's moon Europa are being considered as possible sites of early biology, making three possible "Earths" (by the Drake-equation definition) in our solar system. However, the extrasolar planets found in the last few years have taught us a humbling lesson. Our solar system, with lots of worlds and moons in nice, circular, stable orbits, may be the exception rather than the rule.

How Many Origins of Life? f_l

In scientific circles there's much less concern now than in the past about the value of f_l, the fraction of habitable planets on which life evolves. The molecular building blocks of life -- complex organic hydrocarbons and even amino acids -- are abundant in the universe. They have been discovered in meteorites, comets, and interstellar gas and dust. There are vastly more amounts of amino acids in interstellar space than in the Earth's biosphere. Although hydrocarbons and amino acids are not living organisms, there's little doubt that a lot of prebiotic evolution is going on in the dark galactic clouds between the stars.

Amino acids, the organic molecules that are the building blocks of life, are common in gaseous nebulae, comets, and meteorites. Their abundance suggests that most planets will have raw materials for life.  Some scientists now believe that life will arise in any hospitable environment.  A 5-mm piece of the Allende meteorite, found in Mexico, contains formaldehyde.  The spectra of Comet Hale-Bopp, the spectacular comet of 1997, revealed many organic compounds.

Most significant are the surprising recent discoveries that microorganisms appeared on Earth only moments (geologically speaking) after the last devastating, ocean-boiling impacts of the planet's youth some 3.8 billion years ago. Apparently, given the right conditions, the origin of life is a rather straightforward (yet unexplained) process that happens easily -- at least when given a planet-sized laboratory and millions of years for the experiment to run. If the process were rare or difficult, one would not expect it to have happened at the first possible opportunity on our home planet, but somewhat later in Earth's history instead. Biologists now discuss whether life might have arisen several times on Earth separately. There's every reason to think that all living things today have a common ancestry, but other, independent lines could have formed and been wiped out early. If life does form wherever it can, then presumably f_l = 1.

Intelligence f_i

That leaves us with three remaining unknowns. How likely is the evolution of intelligence (f_i)? How confident can we be that intelligent extraterrestrials will broadcast radio (f_c)? And what is the average lifetime of radio-capable civilizations (L)? These biological and sociological factors in the Drake equation are subject to greater scientific debate and uncertainty than the astronomical ones.

According to many life scientists, it is naive to suppose that the evolution on another planet should necessarily result in intelligence as we know it. In his bestseller Wonderful Life, paleontologist Stephen Jay Gould (Harvard University) asserts, "We probably owe our own existence to . . . good fortune.  Homo sapiens is an entity, not a tendency." Evolution is unpredictable and chaotic. Gould has pointed out again and again that if we could rewind the tape of biological evolution on Earth and start over, it is impossible that humans would again appear on the scene.

Others counter, of course, that we are not looking for humans.  No one expects to find men among the stars (little green ones or otherwise). Rather, the issue is whether any life forms evolve the capability to use tools, manipulate information, and organize societies that are large and complex enough to discover the principles of electronics. To optimists this seems like a difference only in degree, not in kind, from the levels of intelligence and purposeful behavior that have emerged independently in widely divergent species of animals on Earth, from apes to octopi.

But Gould notes that there is no overall pattern in evolution, no preferred direction. Our notion that the increase of biological diversity is necessarily accompanied by an increase of mental capabilities may be dead wrong. If some recently evolved animals are bigger and smarter than any earlier ones, that could just be a fluke. Human levels of planning and technology may be even more so. There are no firm indications that increased intelligence is an inevitable product of biological evolution.

To some biologists and SETI proponents, the phrase "survival
of the fittest" implies that greater intelligence inevitably boosts a species' chance to survive and spread through natural selection. But the renowned biologist Ernst Mayr (Harvard University) argues that many astronomers and physicists are much too optimistic concerning the emergence of intelligence.  "Physicists still tend to think more deterministically than biologists," wrote Mayr in the May 1996 issue of The Planetary Report. "They tend to say that if life has originated somewhere, it will also develop intelligence in due time. The biologist, on the other hand, is impressed by the improbability of such a development."

Strangely enough, optimists and pessimists base their claims on the same observation -- namely that technology has appeared on this planet in 4 billion years. Pessimists (or realists, as they would prefer to be called) like Mayr see this as evidence of the unlikeliness of intelligence as an evolutionary given. For optimists, it strengthens their belief in the existence of extraterrestrial civilizations.

This divergence stems in part from different specialists' intellectual backgrounds. To a biologist, something that happened once in 4 billion years is terribly rare. Astronomers take a wider view: something that happened once in less than a single planet's lifetime seems reasonable for planets generally.

Optimists point out that Earth has more than a billion good years ahead before it will get cooked by the expanding Sun.  This is more than twice the time that has gone by since the first simple creatures crawled out of the sea onto land. If the emergence of intelligence were difficult and rare, the optimists argue, it would probably not have happened so early in the time available for it to do so on Earth. Given our early arrival in the long era expected for land life, it seems likely that entirely different intelligent creatures will emerge a few more times in the coming billion years. This argument echoes the point made from the rapid emergence of microorganisms on the young Earth.

Pessimists reply that we don't really know how long the Earth will remain clement. The Earth's seemingly stable climate may actually be the result of a long series of lucky flukes. So in fact we may have arisen late in the span of time available. Given the fact that we are here at all to ponder the question, a late emergence in the time span available would suggest that the birth of intelligence is a very improbable event.

Contrary to popular belief, the fact that it has happened once tells us absolutely nothing about how often it happens -- for the simple reason that we ourselves are the one case! We are a self-selected sample of one. Even if intelligent life is so rare that it appears just a single time in one remote corner of the universe, we will necessarily find ourselves right there, because we are it.

Strangely enough, both camps accept the so-called Copernican principle, which claims that humankind enjoys no preferred position in time or space. Skeptics like Mayr say it is anthropocentric to believe that humanlike intelligence has appeared over and over again in the universe. Believers like Drake are unwilling to accept our uniqueness, because this would put us on a very un-Copernican pedestal.

Evidently, f_i is the most controversial factor in the Drake equation. Some scientists believe it is almost certainly next to zero; others are convinced it's close to one. There seems to be no middle ground. The question of the inevitability of intelligence is currently what most polarizes the discussion about SETI.

Planetary Catastrophes

Even if intelligence is a likely consequence of evolution, f_i will probably be much lower than 1, based on recent insights into the stability of solar systems and planetary climates. Just because a planet starts out good for life doesn't mean it will stay that way forever.

Computer simulations by Fred Rasio and Eric Ford (Massachusetts Institute of Technology) among others show that Earthlike planets are probably unable to survive the gravitational tug-of-war in a system with two (or more) massive, Jupiter-like giants. They would be slung out of the system or sent careening into the central star.

One factor determining f_i, the fraction of life-bearing planets on which intelligence evolves, must be how long evolution can continue without life getting wiped out. In the case of Earth, Jupiter's immense gravitational pull snags most stray comets  and captures them or flings them away before they can collide with Earth.

Conversely, systems with no giant planets at all might also have dire consequences for possible life-bearing planets. Computer simulations by George Wetherill (Carnegie Institution of Washington) indicate that Jupiter acts as the solar system's gravitational vacuum cleaner, efficiently thinning out the population of hazardous comets that venture into Earth-crossing orbits. Without a Jupiter the current impact rate of comets would be 1,000 times higher, says Wetherill, with truly catastrophic collisions (like the one 65 million years ago) happening once every 100,000 years. This would surely frustrate any slow evolutionary progress from simple life forms to higher intelligences

Also, dynamical studies by Jacques Laskar and Philip Robutel (Bureau des Longitudes, Paris) have shown that rocky, Earthlike planets show chaotic variations in orbital tilt that could lead to drastic climate changes. Fortunately, Earth's chaotic tendencies are damped by tidal interaction with the Moon.  Without a relatively large satellite, Earth might have experienced variations in axial tilt similar to those of Mars, possibly as large as 20 ° to 60 °. This would cause extreme variations in the patterns of the seasons.

It's anyone's guess how such an upset would influence the evolution of life and the chance for the emergence of intelligence. Change and stress actually promote the emergence of new, versatile, adaptable species, biologists believe. For instance, Paul F. Hoffman (Harvard University) and three colleagues proposed in 1998 that the series of extreme global ice ages between 760 and 550 million years ago -- which may have frozen every ocean surface even at the equator -- were the crisis that drove the remarkable "Precambrian explosion" of new life forms around or shortly after that time.  The disastrous great extinctions in Earth's later geologic record were always followed by vigorous recoveries, eventually spawning more species than existed before. Our own emergence as a species during an unusual run of ice ages is sometimes cited as a possible example of stress-driven evolution leading to adaptability and intelligence. A planet with a tippy axis might actually speed things along.

But planetary crises that are too extreme or frequent would kill off everything, or keep life beaten down to a low level. In any case, our existence here and now seems to be the accidental result of a number of astronomical coincidences that were unimagined in 1961.

How Would Aliens Communicate?

Suppose that extraterrestrial intelligences are rare but do exist.  Could we expect them to communicate with us through radio signals? What fraction of civilizations are able -- and willing -- to broadcast in a way we can detect? In other words: what is the value of f_c? SETI advocates tend to believe it is large: sooner or later, any technological civilization will discover that radio is an efficient way to communicate over astronomical distances, and will choose to do so.

Might there be a naive form of anthropocentrism at play here?  Is it reasonable to expect that wildly different beings on another planet, even if very smart and versatile, will choose to build radio telescopes?  Maybe we just don't appreciate the true diversity of biological evolution, or the scope of sciences and technologies that remain unexplored by humans. Radio may be hopelessly primitive compared to something we have yet to discover.

With f_i and f_c completely undetermined, we're left with the last term of the Drake equation: L, the average lifetime of communicating civilizations. Here also, optimists and pessimists are far apart.

The optimists claim that a stable, intelligent society could last for tens of millions of years, if not fo