It's an amazing, electrifying, even terrifying feeling: Stare into a starry night sky and ask yourself if someone, or something, might be staring back. Just the thought is enough to make the hair on the back of your neck stand up. But even now, after centuries of wondering, and decades of actively looking, humans still dont know the answer.

But that may be because they havent been asking the right questions.

We Earthlings know only one type of life, in only one place. The very idea of looking for life elsewhere confronts us with a series of questions that are maddeningly undefined. Where should we look? Can we even be sure what were looking for? Or how to find it?

"It seems to me the worst people to look for life are biologists," Nealson, also a senior scientist at the NASA/Caltech Jet Propulsion Laboratory, says. "Because theyre so darned sure of what life is that if it happened to be different, theyd miss it." Nealson admits this point of view isnt very popular with his colleagues. "Oh, they love me," he laughs. "At least I'm a biologist saying it, so they cant just claim I'm some mad physicist whos trying to take their funding away."

As a cautionary tale, Nealson points to the only previous mission to look for life on another planet. In 1976 two Viking landers arrived on the Red Planet, each outfitted with a miniature biological laboratory. Scoops of Martian soil were fed into a series of experiments, where they were exposed to a puff of humidity, or a drop of water, or a nutrient broth. If any microbes were present, the Viking biologists hoped, the instruments would detect their metabolic processes.

But the experiments didnt discover microbes, and Nealson says thats because they were based on the characteristics of Earthly life. Thats something he wants to avoid on future missions: "Id like to be one of the chorus of voices saying, Dont do this again."

But that doesnt mean Nealson wants to avoid Mars far from it. If anything, Mars has become even more alluring since the Mars Global Surveyor probe found signs that liquid water has recently flowed on the planets surface and may still exist underground.

More than anything, Nealson wants to send a new life-detection mission to the Red Planet. "If I get one thing done before I retire," Nealson says, "it would be to have a mission that went to Mars, because its the one place we probably can go in my lifetime that would do a serious attempt at what I would call non-Earth-centric life detection."

If you call that living

Even our understanding of terrestrial life is changing. Recently, scientists discovered a thriving community of microbes deep in a hot spring in Idaho. These microbes, ancient relatives of bacteria called Archea, make energy by combining hydrogen from rocks with carbon dioxide, and exist completely apart from the Sun. And yet, they may be as numerous as living things on Earths surface, an entirely independent biosphere within the crust of our planet.

And some researchers have speculated that the same kind of life may exist in the crusts of other planets, provided there is enough geologic activity to supply the necessary hydrogen. But designing a space mission with probes to look for such life, many hundreds of feet down, is too ambitious in the near term. Even so, the find reinforces a watchword for life-hunters: Expect the unexpected.

For Nealson, it boils down to one question: "Are there ways that you could detect life without the requirement that it look or smell or feel like Earthly life?"

Thats not an easy idea for biologists to embrace, Nealson says. "Theyre so ingrained to using these methods to find life that it wouldnt occur to them that they wouldnt work [absolutely] anywhere."

And its not only professionals who find the idea difficult to swallow.

In seminars to students about life-detection, Nealson likes to show a diagram that has "a bunch of circles with lines through them, like no smoking signs." In this case, the signs say "no" to the most basic building blocks of terrestrial life: DNA, RNA, protein, lipids.

"None of these are fair game," Nealson says. "People in the audience say, Whats left? I say, This is the toughest challenge that anybodys ever given a biologist: Go find life but dont use anything that youve learned in the past."

There's another pitfall for life-hunters: The desire to find what they're after. In 1996 NASA scientists claimed to have found evidence for fossil microbes inside a meteorite from Mars. The claim was extremely controversial, and since then most researchers have come to believe the features in the Martian meteorite can be explained by non-biologic processes. Nealson says the NASA team was seduced by their own data.

"This wasnt a bunch of amateurs; these were good people. Anybody can be fooled by their data; Ive been fooled by mine in the past." But the seduction, Nealson notes, was understandable.

"Almost everything they found was consistent with life," he says. "And when you find enough things that are consistent with life, you start feeling that youve found life. But theres a big difference between five or six things being consistent with life and five or six things needing life to explain their existence. That to me is the big difference: Is there anything there that absolutely couldnt be there if there wasnt life?"

Indeed, Nealson says, it may be that robots are better life-hunters than people. "A really good information package," he says, "run by a very smart but non-opinionated robotic system might actually do much better than a human system."

At USC, Nealson and his colleagues have been working to develop the kinds of technology that an automated life-hunter might employ. After nosing up to a Mars rock, for example, the robot would use x-rays to scan the rock in three dimensions, then analyze the scan to look for complex structures inside the rock. It would learn to recognize minerals and avoid tagging them as potential living things, Nealson says.

If that sounds far-fetched, Nealson's group has already made progress: "Our robot is able to see complexity about as well as our best technicians can see it now."

The power of such a system, as Nealson describes it, is truly remarkable. "We have a CAT-scan device in the lab now that can see a colony of three or four bacteria inside a rock, without ever opening up the rock. Its fantastic."

"If you can get that kind of information in a few minutes, thats one of the things that youd like to consider taking to Mars or wherever you go," Nealson adds. "Because you can now look inside that rock and if it looks like theres something interesting, then you invest the time of cracking that rock open and looking at it with the next technology," such as a multispectral imager.

But before any life-detection instruments are sent to Mars, Nealson says, theyll have to be tested extensively in some of the most barren places on Earth, like Antarctica. "We need to prove that we always get the right answer. We have every opportunity in the next decade or so to make this work."

Next page: Europas ocean and SETI