This quest is called the Search for Extraterrestrial Intelligence (SETI). Let me describe how far we’ve come.

The first SETI program was carried out by Frank Drake at the National Radio Astronomy Observatory in Greenbank, West Virginia, in 1960. He listened to two nearby Sun-like stars for two weeks at one particular frequency. (“Nearby” is a relative term: The nearest was 12 light-years—70 trillion miles—away.)

Almost at the moment Drake pointed the radio telescope and turned the system on, he picked up a very strong signal. Was it a message from alien beings? Then it went away. If the signal disappears, you can’t scrutinize it. You can’t see if, because of the Earth’s rotation, it moves with the sky. If it’s not repeatable, you’ve learned almost nothing from it—it might be terrestrial radio interference, or a failure of your amplifier or detector… or an alien signal. Unrepeatable data, no matter how illustrious the scientist reporting them, are not worth much.

Weeks later, the signal was detected again. It turned out to be a military aircraft broadcasting on an unauthorized frequency. Drake reported negative results. But in science a negative result is not at all the same thing as a failure. His great achievement was to show that modern technology is fully able to listen for signals from hypothetical civilizations on the planets of other stars.

Since then there’ve been a number of attempts, often on time borrowed from other radio telescope observing programs, and almost never for longer than a few months. There’ve been some more false alarms, at Ohio State, in Arecibo, Puerto Rico, in France, Russia, and elsewhere, but nothing that could pass muster with the world scientific community.

Meanwhile, the technology for detection has been getting cheaper; the sensitivity keeps improving; the scientific respectability of SETI has continued to grow; and even NASA and Congress have become a little less afraid to support it. Diverse, complementary search strategies are possible and necessary. It was clear years ago that if the trend continued, the technology for a comprehensive SETI effort would eventually fall within the reach even of private organizations (or wealthy individuals); and sooner or later, the government would be willing to support a major program. After 30 years of work, for some of us it’s been later rather than sooner. But at last the time has come.

The planetary society—a nonprofit membership organization that Bruce Murray, then the Director of JPL, and I founded in 1980—is devoted to planetary exploration and the search for extraterrestrial life. Paul Horowitz, a physicist at Harvard University, had made a number of important innovations for SETI and was eager to try them out. If we could find the money to get him started, we thought we could continue to support the program by donations from our members.

In 1983 Ann Druyan and I suggested to the filmmaker Steven Spielberg that this was an ideal project for him to support. Breaking with Hollywood tradition, he had in two wildly successful movies conveyed the idea that extraterrestrial beings might not be hostile and dangerous. Spielberg agreed. With his initial support through The Planetary Society, Project META began.

META is an acronym for “Megachannel ExtraTerrestrial Assay.” The single frequency of Drake’s first system grew to 8.4 million. But each channel, each “station,” we tune to has an exceptionally narrow frequency range. There are no known processes out among the stars and galaxies that can generate such sharp radio “lines.” If we pick up anything falling into so narrow a channel, it must, we think, be a token of intelligence and technology.

What’s more, the Earth turns—which means that any distant radio source will have a sizable apparent motion, like the rising and setting of the stars. Just as the steady tone of a car’s horn dips as it drives by, so any authentic extraterrestrial radio source will exhibit a steady drift in frequency due to the Earth’s rotation. In contrast, any source of radio interference at the Earth’s surface will be rotating at the same speed as the META receiver. META’s listening frequencies are continuously changed to compensate for the Earth’s rotation, so that any narrow-band Signals from the sky will always appear in a single channel. But any radio interference down here on Earth will give itself away by racing through adjacent channels.

The META radio telescope at Harvard, Massachusetts, is 26 meters (84 feet) in diameter. Each day, as the Earth rotates the telescope beneath the sky, a swath of stars narrower than the full moon is swept out and examined. Next day, it’s an adjacent swath. Over a year, all of the northern sky and part of the southern is observed. An identical system, also sponsored by The Planetary Society, is in operation just outside Buenos Aires, Argentina, to examine the southern sky. So together the two META systems have been exploring the entire sky.

The radio telescope, gravitationally glued to the spinning Earth, looks at any given star for about two minutes. Then it’s on to the next. 8.4 million channels sounds like a lot, but remember, each channel is very narrow. All of them together constitute only a few parts in 100,000 of the available radio spectrum. So we have to park our 8.4 million channels somewhere in the radio spectrum for each year of observation, near some frequency that an alien civilization, knowing nothing about us, might nevertheless conclude we’re listening to.

Hydrogen is by far the most abundant kind of atom in the Universe. It’s distributed in clouds and as diffuse gas throughout interstellar space. When it acquires energy, it releases some of it by giving off radio waves at a precise frequency of 1420.405751768 megahertz. (One hertz means the crest and trough of a wave arriving at your detection instrument each second. So 1420 megahertz means 1.420 billion waves entering your detector every second. Since the wavelength of light is just the speed of light divided by the frequency of the wave, 1420 megahertz corresponds to a wavelength of 21 centimeters.) Radio astronomers anywhere in the Galaxy will be studying the Universe at 1420 megahertz and can anticipate that other radio astronomers, no matter how different they may look, will do the same.

It’s as if someone told you that there’s only one station on your home radio set’s frequency band, but that no one knows its frequency. Oh yes, one other thing: Your set’s frequency dial, kith its thin marker you adjust by turning a knob, happens to reach from the Earth to the Moon. To search systematically through this vast radio spectrum, patiently turning the knob, is going to be very time-consuming. Your problem is to set the dial correctly from the beginning, to choose the right frequency. If you can correctly guess what frequencies that extraterrestrials are broadcasting to us on—the “magic” frequencies—then you can save yourself much time and trouble. These are the sorts of reasons that we first listened, as Drake did, at frequencies near 1420 megahertz, the hydrogen “magic” frequency.

Horowitz and I have published detailed results from five years of full-time searching with Project META and two years of follow-up. We can’t report that we found a signal from alien beings. But we did find something puzzling, something that for me in quiet moments, every now and then, raises goose bumps:

Of course, there’s a background level of radio noise from Earth—radio and television stations, aircraft, portable telephones, nearby and more distant spacecraft. Also, as with all radio receivers, the longer you wait, the more likely it is that there’ll be some random fluctuation in the electronics so strong that it generates a spurious signal. So we ignore anything that isn’t much louder than the background.

Any strong narrow-band signal that remains in a single channel we take very seriously. As it logs in the data, META automatically tells the human operators to pay attention to certain signals. Over five years we made some 60 trillion observations at various frequencies, while examining the entire accessible sky. A few dozen signals survive the culling. These are subjected to further scrutiny, and almost all of them are rejected-for example, because an error has been found by fault-detection microprocessors that examine the signal-detection microprocessors.