Posted by In April 1960 a young astronomer at Green Bank pointed an 85-foot dish at Tau Ceti, tuned a receiver near the 1420 MHz hydrogen line, and listened. Frank Drake called it Project Ozma, gave it about 150 hours across two stars, and found nothing. Every serious radio SETI program since has been an argument with that experiment — bigger dishes, more channels, better rejection of the noise humanity makes about itself — and the sequence of radio missions that followed is the closest thing this field has to a history.
Nothing has been confirmed. That statement needs to sit here early and stay uncomfortable, because the interesting part is what six decades of not finding anything has actually taught people about how to look.
Ozma set the rules
Drake’s setup covered a bandwidth that today looks laughably narrow: a few hundred kilohertz, one channel at a time, scanned slowly. He chose the hydrogen line because it is the most abundant element in the universe broadcasting at a frequency any radio astronomer anywhere would know, which remains the standard argument for the so-called waterhole between 1420 and 1720 MHz.
The real contribution was procedural. Drake treated “is anyone transmitting” as an observational question with a null result worth publishing, and that reframing survived long after the specific hardware became a museum piece. A year later he wrote down the equation that still bears his name, less a prediction than a way of organizing everything nobody knew.
Big Ear, and the signal that never came back
Ohio State’s Big Ear ran a survey for years on a shoestring, staffed largely by volunteers, printing its results as columns of characters on fanfold paper. On 15 August 1977 the telescope recorded a narrowband burst about thirty times the background, lasting the 72 seconds it took the beam to sweep past. Jerry Ehman, reviewing the printout days later, circled the string 6EQUJ5 and wrote “Wow!” beside it.
The intensity profile matched what a genuine sky source should produce as the fixed telescope let Earth’s rotation carry the beam across it, which is exactly why the record has never gone away. A passing comet has been proposed as the culprit and largely rejected. Nobody has explained it, and nobody has seen it again. Searches with more sensitive instruments have come back empty, and a signal that cannot be re-observed cannot be verified no matter how good it looked once. That is the lesson the episode actually delivered, and it hardened into policy across every project that came after.
Channels, channels, channels
The 1980s belonged to computing rather than to telescopes. Paul Horowitz at Harvard built META in 1985 — the Megachannel Extraterrestrial Assay, 8.4 million channels at once, running on the Oak Ridge Observatory dish with funding that famously included a donation from Steven Spielberg. BETA replaced it a decade later with a quarter of a billion channels, and then a windstorm knocked the telescope over in 1999.
NASA meanwhile built the High Resolution Microwave Survey, switched it on in October 1992 on the five-hundredth anniversary of Columbus’s landfall, and watched Congress kill the funding almost exactly one year later. The hardware and much of the team were rescued by the SETI Institute as Project Phoenix.
Phoenix ran from 1995 to 2004 and was the most disciplined targeted search anyone had attempted. Roughly 800 Sun-like stars within about 200 light-years, observed from Parkes in Australia, the 140-foot at Green Bank and finally Arecibo, each candidate checked in real time against a second telescope thousands of kilometers away so that local interference would rule itself out. It found no transmitters. It also demonstrated that the two-site verification trick works, which is why the radio missions running today are built around it.
Five million desktops
SETI@home went live in May 1999 out of Berkeley and did something none of the professional programs managed: it made the search participatory. Data recorded piggyback at Arecibo was sliced into work units, mailed out to volunteers’ home computers, chewed on during screensaver time and returned. More than five million people signed up over the project’s life. It stopped distributing new work in March 2020.
Whether it was good science is still debated inside the field — the pipeline was optimized for public engagement as much as for detection — but it built the distributed computing platform that later became BOINC, and it put the phrase “search for extraterrestrial intelligence” into millions of households.
Arrays, and the modern era
The Allen Telescope Array at Hat Creek in northern California took a different structural bet: instead of one enormous dish, use many small ones and combine them electronically. Forty-two 6.1-metre antennas went into service in 2007, funded largely by Paul Allen, well short of the 350 originally planned. An array can form multiple beams, track several targets, and separate a real sky source from interference by checking whether it appears in the same place for every antenna.
Breakthrough Listen changed the scale again. Announced in 2015 with $100 million from Yuri Milner, it bought serious observing time on the Green Bank 100-metre, on Parkes, and later on MeerKAT in South Africa, targeting a million nearby stars, the galactic plane and a hundred nearby galaxies. The bottleneck stopped being telescope access and became data: petabytes of recorded spectra that have to be sifted by software looking for narrowband features that drift in frequency the way a transmitter on a rotating, orbiting planet would.
Its most instructive result so far is a negative one. BLC1, a tone near 982 MHz found in 2019 Parkes data pointed at Proxima Centauri, drifted correctly and looked genuinely promising for months. Careful work traced it to human equipment, and the team published that conclusion rather than quietly dropping it.
What the silence is worth
Radio SETI’s search space has more dimensions than most people picture. Not just which star, but which frequency, what bandwidth, what transmitter power, what polarization, whether the beam is aimed at us, and whether it is on when we happen to be looking. Jill Tarter’s comparison stuck because it is fair: everything searched so far amounts to roughly a glass of water dipped from the ocean.
So the surveys do produce results, just not the kind that make headlines. Each one sets upper limits — no omnidirectional transmitter above a certain power around these particular stars, at these frequencies, at this sensitivity, during these hours. That is genuine astronomical constraint, and it accumulates.
The next decade should accelerate it. China’s 500-metre FAST dish is the most sensitive single aperture ever built. The Square Kilometre Array, under construction in Australia and South Africa, will generate data at rates that make current archives look modest, and much technosignature work can now run commensally — parasitizing observations taken for entirely different astronomy rather than demanding dedicated time. Arecibo’s collapse in December 2020 removed one of the great instruments, and the loss still stings, though the search did not depend on it alone.
Radio missions of this kind carry an unusual burden: they are designed by people who mostly expect to find nothing, and who keep going because the one alternative outcome would be the largest discovery in the history of the subject. SETIworld follows the surveys, the candidate signals and the arguments about what would count as confirmation, if you want to watch the next one develop in something close to real time.
