When Donald Rumsfeld delivered his now-famous chestnut about "known unknowns" and "unknown unknowns," he wasn't talking about scientific discovery. But he could have been.

Science is full of known unknowns -- the puzzles scientists set out to solve, the quantities they know they can measure -- but its best prizes are the unknown unknowns, discoveries that were never even imagined before they materialized.

In the first category are those scientific-method experiments we learned about in science class: Ask a question, form a hypothesis, test that hypothesis and report your results. Is there water on Mars? Does that exoplanet have an oxygen atmosphere? How big is the black hole at the center of our galaxy? These are known unknowns.

Then there are the unknown unknowns. Take gamma-ray bursts. In the 1960s, the United States deployed a group of military satellites called Vela, designed to keep Cold War vigil over the planet, watching and waiting for the telltale blasts of X-rays, neutrons and gamma rays that would be produced by (banned) nuclear bomb tests down on Earth. In 1967, two of the Vela satellites detected an unusual burst of gamma rays. When the Vela engineers examined their data closely, they found that their satellites had spotted more than a dozen such gamma-ray surges. But the gamma rays weren't coming from Earth -- they were coming from the sky.


Today, we know these mysterious signals as gamma-ray bursts, and astronomers believe that most of them are set off by the gravitational collapse of massive stars morphing into black holes. But until Vela's serendipitous discovery, we had no idea that gamma-ray bursts even existed: They were unknown unknowns.

There are plenty of other unknown unknowns in astronomy. No one was looking for pulsars, the Cosmic Microwave Background Radiation or the rings of Uranus before they were serendipitously discovered. But that does not mean that astronomers must sit at their desks waiting for the Discovery Fairy to tap them with her magic wand: By considering what these unknown unknowns have in common, astronomers can learn to make their own serendipity.

So how do you make serendipity? You might think of it like searching an orchard for the perfect piece of fruit. Instead of hanging out near the path, where the trees are already picked over, you head deeper inside, where the branches are still untouched. When it comes to astronomy, that means searching in a new part of the electromagnetic spectrum, as Vela did when it started to read gamma rays; picking up signals so weak that previous telescopes were blind to them; or taking data in slow-motion so that even very short-lived signals show up loud and clear.

That's the idea behind a new telescope called the Low Frequency Array for Radio Astronomy (LOFAR). LOFAR will observe the universe all the way down to the far edge of the radio band, picking up electromagnetic waves as long as 20 meters (longer than a city bus) with exceptional sensitivity and resolution. This part of the electromagnetic spectrum is almost totally unexplored, which means that LOFAR is perfectly positioned to yield up a raft of apparently serendipitous discoveries.

LOFAR doesn't look much like a traditional telescope. Made up of clusters of antennas spread over dozens of sites in northern Europe, LOFAR is more like a club for antennas than like the monolithic radio dishes that preceded it. Made up of low, flat tiles, the high-frequency antennas look like some kind of outdoor dance floor, while the low-frequency antennas rise out of the ground like so many tent poles.

What will LOFAR discover? Signals from the very distant, early universe, which have been so stretched by cosmic expansion that they are invisible to other telescopes. Ultra-high-energy cosmic rays. New pulsars. But best of all will be discoveries we haven't even imagined yet, the unknown unknowns of the cosmos.

Kate Becker is a science writer living in Boston. Contact her at spacecrafty.com.