No SETI program has ever found a verifiable alien radio signal. What does that null result mean? Any answer must be highly qualified, because the searches have been so incomplete. Nevertheless, researchers can draw some preliminary conclusions about the number and technological sophistication of other civilizations.

The most thoroughly examined frequency channel to date, around 1.42 gigahertz, corresponds to the emission line of the most common element in the universe, hydrogen--on the premise that if extraterrestrials had to pick some frequency to attract our attention, this would be a natural choice. The diagram, the first of its kind, shows exactly how thoroughly the universe has been searched for signals at or near this frequency. No signal has ever been detected, which means that any civilizations either are out of range or do not transmit with enough power to register on our instruments. The null results therefore rule out certain types of civilizations, including primitive ones close to Earth and advanced ones farther away.

The chart quantifies this conclusion. The horizontal axis shows the distance from Earth. The vertical axis gives the effective isotropic radiated power (EIRP) of the transmitters. The EIRP is essentially the transmitter power divided by the fraction of the sky the antenna covers. In the case of an omnidirectional transmitter, the EIRP is equal to the transmitter power itself. The most powerful on this planet is currently the Arecibo radio telescope in Puerto Rico, which could be used as a narrowly beamed radar system with an EIRP of nearly 1014 watts.

The EIRP can serve as a crude proxy for the technological level of an advanced civilization, according to a scheme devised by Russian SETI pioneer Nikolai S. Kardashev in the early 1960s and later extended by Carl Sagan. Type I civilizations could transmit signals with a power equivalent to all the sunlight striking an Earth-like planet, about 1016 watts. Type II civilizations could harness the entire power output of a sunlike star, about 1027 watts. Still mightier type III civilizations command an entire galaxy, about 1038 watts. If the capability of a civilization falls in between these values, its type is interpolated logarithmically. For example, based on the Arecibo output, humanity rates as a type 0.7 civilization.

For any combination of distance and transmitter power, the diagram indicates what fraction of stars has been scanned so far without success. The white and colored areas represent the civilizations whose existence we therefore can rule out with varying degrees of confidence. The black area represents civilizations that could have evaded the searches. The size of the black area increases toward the right--that is, going farther away from Earth. SETI programs completely exclude Arecibo-level radio transmissions out to 50 or so light-years.

Farther away, they can rule out the most powerful transmitters. Far beyond the Milky Way, SETI fails altogether, because the relative motions of galaxies would shift any signals out of the detection band. These are not trivial results. Before scientists began to look, they thought that type II or III civilizations might actually be quite common. That does not appear to be the case. This conclusion agrees with other astronomical data. Unless supercivilizations have miraculously repealed the second law of thermodynamics, they would need to dump their waste heat, which would show up at infrared wavelengths. Yet searches performed by Jun Jugaku of the Research Institute of Civilization in Japan and his colleagues have seen no such offal out to a distance of about 80 light-years. Assuming that civilizations are scattered randomly, these findings also put limits on the average spacing of civilizations and thus on their inferred prevalence in unprobed areas of the galaxy.

On the other hand, millions of undetected civilizations only slightly more advanced than our own could fill the Milky Way. A hundred or more type I civilizations could also share the galaxy with us. To complicate matters further, extraterrestrials might be using another frequency or transmitting sporadically. Indeed, SETI programs have logged numerous "extrastatistical events," signals too strong to be noise but never reobserved. Such transmissions might have been wayward radio waves from nearby cell phones--or they might have been intermittent extraterrestrial broadcasts. No one yet knows. Although the cutting edge of technology has made SETI ever more powerful, we have explored only a mere fraction of the possibilities.