by Carl Sagan
We have to have some feeling for the number, N, of extant technical civilizations in the Galaxy–that is, civilizations vastly in advance of our own, civilizations that are able, by whatever means, to perform interstellar space flight. (While the means are difficult, they don’t enter into this discussion, just as reindeer propulsion mechanisms don’t affect our discussion of the Santa Claus hypothesis.)
An attempt has been made to specify explicitly the factors that enter a determination of the number of such technical civilizations in the Galaxy. I will not here run through what numbers have been assigned to the various quantities involved–it’s a multiplication of many probabilities, and the likelihood that we can make a good judgment decreases as we proceed down the list. N depends first on the mean rate at which stars are formed in the Galaxy, a number that is known reasonably well. It depends on the number of stars that have planets, which is less well known, but there are some data on that. It depends on the fraction of such planets that are so suitably located with respect to their star that the environment is a feasible one for the origin of life. It depends on the fraction of such otherwise feasible planets on which the origin of life, in fact, occurs. It depends on the fraction of those planets on which the origin of life occurs in which, after life has arisen, an intelligent form comes into being. It depends on the fraction of those planets in which intelligent forms have arisen that evolve a technical civilization substantially in advance of our own. And it depends on the average lifetime of such a technical civilization.
It is clear that we are rapidly running out of examples as we go farther and farther along. We have many stars, but only one instance of the origin of life, and only a very limited number–some would say only one–of instances of the evolution of intelligent beings and technical civilizations on this planet. And we have no cases whatever to make a judgment on the mean lifetime of a technical civilization. Nevertheless, there is an entertainment that some of us have been engaged in, making our best estimates about these numbers and coming out with a value of N. The result that emerges is that N roughly equals one tenth the average lifetime of a technical civilization in years.
If we put in a number like ten million (107) years for the average lifetime of advanced technical civilizations, we come out with a number for such technical civilizations in the Galaxy of about a million (106)–that is, a million other stars with planets on which today there are advanced civilizations. This is quite a difficult calculation to do accurately. The choice of ten million years for the average lifetime of a technical civilization is rather optimistic. But let’s take these optimistic numbers and see where they lead us.
Let’s assume that each of these million technical civilizations launches Q interstellar space vehicles a year, so that 106Q interstellar space vehicles are launched per year. Let’s assume that there’s only one contact made per journey. In the steady-state situation, there are something like 106Q arrivals somewhere or other per year. Now, there surely are something like 1010 interesting places in the Galaxy to go visit (we have several times 1011 stars) and, therefore, an average of 1/104=10–4 arrivals at a given interesting place (let’s say a planet) per year. So if only one UFO is to visit the Earth each year, we can calculate what mean launch rate is required at each of these million worlds. The number turns out to be ten thousand launches per year per civilization, and ten billion launches in the Galaxy per year. This seems excessive. Even if we imagine a civilization much more advanced than ours, to launch ten thousand such vehicles for only one to appear here is probably asking too much. And if we were more pessimistic on the lifetime of advanced civilizations, we would require a proportionately larger launch rate. But as the lifetime decreases, the probability that a civilization would develop interstellar flight very likely decreases as well.
There is a related point made by the American physicist Hong-Yee Chiu; he takes more than one UFO arriving at Earth per year, but his argument follows along the same lines as the one I have just presented. He calculates the total mass of metals involved in all of these space vehicles during this history of the Galaxy. The vehicle has to be of some size–it should be bigger than the Apollo capsule, let’s say–and we can calculate how much metal is required. It turns out that the total mass of half a million stars has to be processed and all their metals extracted. Or if we extend the argument and assume that only the outer few hundred miles or so of stars like the Sun can be mined by advanced technologies (farther in, it’s too hot), we find that two billion such stars must be processed, or about 1 percent of the stars in the Galaxy. This also sounds unlikely.
Now you may say, “Well, that’s a very parochial approach; maybe they have plastic spaceships.” Yes, I suppose that’s possible. But the plastic has to come from somewhere, and plastics vs. metals changes the conclusions very little. This calculation gives some feeling for the magnitude of the task when we are asked to believe that there are routine and frequent interstellar visits to our planet.
What about possible counterarguments? For example, it might be argued that we are the object of special attention–we have just developed all sorts of signs of civilization and high intelligence like nuclear weapons, and maybe, therefore, we are of particular interest to interstellar anthropologists. Perhaps. But we have only signaled the presence of our technical civilization in the past few decades. The news can be only some tens of light-years from us. Also, all the anthropologists in the world do not converge on the Andaman Islands because the fish net has just been invented there. There are a few fish net specialists and a few Andaman specialists; and these guys say, “Well, there’s something terrific going on in the Andaman Islands. I’ve got to spend a year there right away because if I don’t go now, I’ll miss out.” But the pottery experts and the specialists in Australian aborigines don’t pack up their bags and leave for the Indian Ocean.
To imagine that there is something absolutely fascinating about what is happening right here is precisely contrary to the idea that there are lots of civilizations around. Because if the latter is true, the development of our sort of civilization must be pretty common. And if we are not pretty common, then there are not going to be many civilizations advanced enough to send visitors.
Even so, is it not possible that the second UFO hypothesis is true–that in historical or recent prehistoric times an extraterrestrial space vehicle made landfall on Earth? There is surely no way in which we can exclude such a contingency. How could we prove it?
A number of popular books have recently been written that allege to demonstrate such a visitation. The arguments are of two sorts, legend and artifact. I broached this subject in the book Intelligent Life in the Universe, written with the Soviet astrophysicist I. S. Shklovskii and published in 1966. I examined a typical legend suggestive of contact between our ancestors and an apparent representative of a superior society. The legend, taken from the earliest Sumerian mythology, is important because the Sumerians are the direct cultural antecedents of our own civilization. A superior being was supposed to have taught the Sumerians mathematics, astronomy, agriculture, social and political organization, and written language–all the arts necessary for making the transition from a hunter-gatherer society to the first civilization.
But as provocative as this and similar legends were, I concluded that it was impossible to demonstrate extraterrestrial contact from such legends: There are plausible alternative explanations. We can understand why priests might make myths about superior beings who inhabit the skies and give directions to human beings on how to order their affairs. Among other “advantages,” such legends permit the priests to control the people.
There is only one category of legend that would be convincing: When information is contained in the legend that could not possibly have been generated by the civilization that created the legend–if, for example, a number transmitted from thousands of years ago as holy turns out to be the nuclear fine structure constant. This would be a case worthy of some considerable attention.
/> Also convincing would be a certain class of artifact. If an artifact of technology were passed on from an ancient civilization–an artifact that is far beyond the technological capabilities of the originating civilization–we would have an interesting prima facie case for extraterrestrial visitation. An example would be an illuminated manuscript, rescued from an Irish monastery, that contains the electronic circuit diagram for a superheterodyne radio receiver. Great care would have to be taken about the provenance of this artifact, just as art collectors are cautious about a newly discovered Raphael. We would make sure that no contemporary Irish prankster was the source of the circuit diagram.
To the best of my knowledge, there are no such legends and no such artifacts. All the ancient artifacts put forward, for example, by Erik von Danniken in his book Chariots of the Gods? have a variety of plausible, alternative explanations. Representations of beings with large, elongated heads, alleged to resemble space helmets, could equally well be inelegant artistic renditions, depictions of ceremonial head masks or expressions of rampant hydrocephalia. In fact, the expectation that extraterrestrial astronauts would look precisely like American or Soviet astronauts, down to their space suits and eyeballs, is probably less credible than the idea of a visitation itself. Likewise, the idea expressed by von Danniken and others that ancient astronauts erected airfields, employed rockets, and exploded nuclear weapons on Earth is implausible in the extreme, precisely because we ourselves have just developed this technology. A visitor from space will not be so close to us in time. It is as if, framing such an idea in 1870, we concluded that extraterrestrials use hot-air balloons for space exploration. Far from being too daring, such ideas are stodgy in their unimaginativeness. Most popular accounts of alleged contact with extraterrestrials are strikingly chauvinistic.
An American author named Richard Shaver claims that ordinary rocks, sliced fine, contain a set of still photographs left by an ancient civilization, which can be run as a movie film. Just pick up any rock and slice it fine, he says.
In the great high plain of Nasca in Peru, there is a set of enormous geometrical figures. They are quite difficult to discern when standing among them, but quite discernible from the air. It is easy to see how an early human civilization could have made such figures. But why, it is asked, should such constructs be made except for or by an extraterrestrial civilization? If people believe in the existence of gods in the sky, it is not straining credulity to imagine them making messages to communicate with those gods. The markings may be a kind of collective graphical prayer. But they do not necessarily demonstrate the reality of the intended recipient of the prayer.
There are other cases that seem to be quite convincing at first, such as a perfectly machined steel cube, said to reside in the Salzburg Museum and to have been recovered from geological strata millions of years old, or the receipt of the television call signals of a television station off the air for three years. These cases are almost certainly hoaxes.
There are equally provocative archaeological circumstances that the writers of such sensational books have somehow missed. For example, in the frieze of the great Aztec pyramids at San Juan Teotihuacan, outside Mexico City, there is a repeated figure, described as a rain god, but looking for all the world like an amphibious tracked vehicle with four headlights (see page 201). I do not for a moment believe that such amphibious vehicles were indigenous in Aztec times–among other reasons, because they are too close to what we have today rather than too far from it.
These artifacts are, in fact, psychological projective tests. People can see in them what they wish. There is nothing to prevent anyone from seeing signs of past extraterrestrial visitations all about him. But to a person with an even mildly skeptical mind, the evidence is unconvincing. Because the significance of such a discovery would be so enormous, we must employ the most critical reasoning and the most skeptical attitudes in approaching such data. The data do not pass such tests. Pondering wall paintings, for this purpose, like looking for UFOs, remains an unprofitable investment of terrestrial intelligence–if we are truly interested in the quest for extraterrestrial intelligence.
29. A Search Strategy for Detecting Extraterrestrial Intelligence
Suppose we have arranged a meeting at an unspecified place in New York City with a stranger we have never met and about whom we know nothing–a rather foolish arrangement, but one that is useful for our purposes. We are looking for him, and he is looking for us. What is our search strategy? We probably would not stand for a week on the corner of Seventy-eighth Street and Madison Avenue. Instead, we would recall that there are a number of widely known landmarks in New York City–as well known to the stranger as to us. He knows we know them, we know he knows we know them, and so on. We then shuttle among these landmarks: The Statue of Liberty, the Empire State Building, Grand Central Station, Radio City Music Hall, Lincoln Center, the United Nations, Times Square, and just conceivably, City Hall. We might even indulge ourselves in a few less likely possibilities, such as Yankee Stadium or the Manhattan entrance to the Staten Island Ferry. But there are not an infinite number of possibilities. There are not millions of possibilities; there are only a few dozen possibilities, and in time we can cover them all.
The situation is just the same in the frequency-search strategy for interstellar radio communication. In the absence of any prior contact, how do we know precisely where to search? How do we know which frequency or “station” to tune in on? There are at least millions of possible frequencies with reasonable radio bandpasses. But a civilization interested in communicating with us shares with us a common knowledge about radio astronomy and about our Galaxy. They know, for example, that the most abundant atom in the universe, hydrogen, characteristically emits at a frequency of 1,420 Megahertz. They know we know it. They know we know they know it. And so on. There are a few other abundant interstellar molecules, such as water or ammonia, which have their own characteristic frequencies of emission and absorption. Some of these lie in a region of the galactic radio spectrum where there is less background noise than others. This is also shared information. Students of this problem have come up with a short list of possibly a dozen frequencies that seem to be the obvious ones to examine. It is even conceivable that water-based life will communicate at water frequencies, ammonia-based life at ammonia frequencies, etc.
There appears to be a fair chance that advanced extraterrestrial civilizations are sending radio signals our way, and that we have the technology to receive such signals. How should a search for these signals be organized? Existing radio telescopes, even very small ones, would be adequate for a preliminary search. Indeed, the ongoing search at the Gorky Radiophysical Institute, in the Soviet Union, involves telescopes and instrumentation that are quite modest by contemporary standards.
The amiable and capable president of the Soviet Academy of Sciences, M. V. Keldysh, once told me, with a twinkle in his eye, that “when extraterrestrial intelligence is discovered, then it will become an important scientific problem.” A leading American physicist has argued forcefully with me that the best method to search for extraterrestrial intelligence is simply to do ordinary astronomy; if the discovery is to be made, it will be made serendipitously. But it seems to me that we can do something to enhance the likelihood of success in such a search, and that the ordinary pursuit of radio astronomy is not quite the same as an explicit search of certain stars, frequencies, bandpasses, and time constants for extraterrestrial intelligence.
But there are enormous numbers of stars to investigate, and many possible frequencies. A reasonable search program will almost certainly be a very long one. Such a search, using a large telescope full time, should take at least decades, by conservative estimates. The radio observers in such an enterprise, no matter how enthusiastic they may be about the search for extraterrestrial intelligence, would very likely become bored after many years of unsuccessful searching. A radio astronomer, like any other scientist, is interested in working on problems that have a high p
robability of more immediate results.
The ideal strategy would involve a large telescope that could devote something like half time to the search for extraterrestrial intelligent radio signals and about half time to the study of more conventional radioastronomical objectives, such as planets, radio stars, pulsars, interstellar molecules, and quasars. The difficulty in using several existing radio observatories, each for, say, 1 percent of their time, is that these activities would have to be pursued for many centuries to have a reasonable probability of success. Since the time on existing radio telescopes is mainly spoken for, larger allocations of time seem unlikely.
A wide variety of objects obviously should be examined: G-type stars, like our own; M-type stars, which are older; and exotic objects, which may be black holes or possible manifestations of astroengineering activities. The number of stars and other objects in our own Milky Way Galaxy is about two hundred billion, and the number that we must examine to have a fair chance of detecting such signals seems to be at least millions.
There is an alternative strategy to searching painfully each of millions of stars for the signals from a civilization not much more advanced than our own. We might examine an entire galaxy all at once for signals from civilizations much more advanced than ours (see Chapters 34 and 35). A small radio telescope can point at the nearest spiral galaxy to our own, the great galaxy M31 in the constellation Andromeda, and simultaneously observe some two hundred billion stars. Even if many of these stars were broadcasting with a technology only slightly in advance of our own, we would not pick them up. But if only a few are broadcasting with the power of a much more advanced civilization, we would detect them easily. In addition to examining nearby stars only slightly in advance of us, it therefore makes sense to examine, simultaneously, many stars in neighboring galaxies, only a few of which may have civilizations greatly in advance of our technology.