Jerry Ehman was not at the telescope when it happened. Nobody was — that was the point. The Big Ear radio observatory at Ohio State University ran its SETI survey mostly on autopilot, scanning the sky in 50-channel strips and printing the signal intensities as a column of numbers and letters on continuous-feed paper. Ehman was reviewing the printout a few days later, on August 17, 1977, when he noticed something extraordinary on data collected two nights before. He circled it in red pen and wrote a single word in the margin: Wow!
That annotation gave the signal its permanent name, and the permanence feels fitting — the Wow! signal has remained exactly that, a circle in the margin, a moment of startlement that never resolved into something more legible. Forty-seven years of follow-up observation, hypothesis formation, and technological advancement have not produced a second detection. What they have produced is a clearer picture of what the signal was, what it almost certainly was not, and what its stubborn singularity tells us about the difficulty of the search for extraterrestrial intelligence.
What the data actually showed
Understanding why Ehman circled what he circled requires a brief detour into the logic of SETI radio searches in the 1970s. The theoretical framework for searching for artificial radio signals had been laid out in 1959 by Giuseppe Cocconi and Philip Morrison in a landmark Nature paper, which argued that a technologically advanced civilization wishing to communicate across interstellar distances would likely transmit near the 1420 MHz emission frequency of neutral hydrogen — the most abundant element in the universe, and one any spacefaring civilization would certainly know. This is the so-called "watering hole," a quiet region of the spectrum sheltered between the hydrogen line at 1420 MHz and the hydroxyl radical line at 1666 MHz.
Big Ear was scanning near 1420 MHz when it detected the Wow! signal. The signal lasted for the full 72-second window during which a point source in the sky would pass through the telescope's beam — exactly what a narrowband transmission from a fixed celestial position would produce. Its intensity reached 30 standard deviations above background noise, an almost absurdly clean spike by any measurement standard. It was narrowband, meaning its bandwidth was smaller than the observatory could resolve (less than 10 kHz). It showed the characteristic rise and fall in power expected from a source moving through a parabolic telescope's beam. In other words, it looked exactly like the theoretical profile of a signal of extraterrestrial origin. Ehman later described it as matching "all the predicted hallmarks" of such a signal. The problem was that Big Ear only had a single feed horn pointed at that region of sky. When the telescope returned 72 seconds later to resample the same patch — Big Ear used two feed horns to scan each strip twice — nothing was there.
The years of searching and the failure to find it again
Big Ear itself returned to that patch of sky over 100 times in the following years. Nothing. The region, located in the constellation Sagittarius near the Chi Sagittarii star group, was observed by a succession of instruments afterward: the Very Large Array in New Mexico, Arecibo in Puerto Rico, and numerous other facilities. The Arecibo searches of the 1990s were particularly exhaustive, using an instrument far more sensitive than Big Ear had been. Silence.
The absence of a second detection is the signal's defining characteristic, and it has driven the debate over what it was ever since. One possibility is that it was a genuine narrowband artificial signal that simply isn't continuous — a transmission that happened once, or on a duty cycle longer than any single observing campaign. This is actually the case most consistent with the raw data, uncomfortable as it is. Another possibility is that it was a natural astrophysical phenomenon that nobody had yet characterized — a type of interstellar scintillation, a transient maser emission, or some other brief, bright radio event. A third, now frequently discussed possibility is that it was human-made radio frequency interference (RFI), possibly reflected off space debris or otherwise scattered into the telescope's beam from an unexpected angle.
The RFI hypothesis gained significant traction in 2017 when Antonio Paris, an astronomer at St. Petersburg College in Florida, published a paper suggesting that two comets — 266P/Christensen and P/2008 Y2 (Gibbs) — had been in the vicinity of the sky position where the Wow! signal was detected. Paris argued that hydrogen clouds surrounding cometary nuclei could emit at 1420 MHz, potentially producing a signal consistent with what Big Ear observed. The proposal was almost immediately challenged. Other researchers pointed out that the cometary hydrogen clouds would not produce a signal narrow enough to match the Wow! data, and that the bandwidth mismatch was a fundamental objection, not a technicality. Independent verification of the comet positions was also contested. The hypothesis remains a minority view in the literature.
What modern SETI makes of it
The Wow! signal occupies a peculiar place in modern SETI research: universally acknowledged, frequently invoked, and stubbornly resistant to integration into any broader theoretical framework. Unlike most anomalies in science, it has not been absorbed into the background radiation of "explained events" because the explanation never arrived. It sits in the literature as a single data point with no context — no repeating source, no associated multi-wavelength emission, no catalog counterpart.
The Breakthrough Listen initiative, which since 2015 has applied unprecedented computational resources and observing time to SETI using facilities including the Green Bank Telescope and Parkes Observatory, has not found anything matching the Wow! profile despite scanning billions of frequency channels. This null result does not rule out a genuine artificial origin for the 1977 event; it merely confirms that whatever produced it is not producing a steady detectable signal in the directions and frequency ranges Breakthrough Listen has covered so far. The parameter space for possible signals is enormous, and the Wow! signal's bandwidth, polarization, and temporal structure were only partially characterized to begin with — Big Ear's technology in 1977 did not capture full polarimetric data, for instance.
There's a broader epistemological problem embedded in the Wow! story. SETI as a discipline has historically required repeatable detections before a candidate signal can be taken seriously — the history of false positives, from the Soviet detections of the 1970s to the 2015 Peryton events at Parkes (which turned out to be microwave oven interference), has made the field appropriately cautious. But a genuinely non-repeating signal from an extraterrestrial source would, by definition, fail that repeatability test. The methodology built to protect against false positives would also suppress real positives of a certain type. Ehman himself acknowledged this tension in his later writings, noting that the one thing the Wow! signal most needed — a second detection — was the one thing it never provided.
What remains, 47 years on, is a 72-second trace in an old printout and a margin annotation that has become the most famous two syllables in SETI history. Ehman circled it because it deserved to be circled. The universe has not yet seen fit to offer a clarification.
