News notice: The American Physical Society’s Forum on the History and Philosophy of Physics is accepting submissions to its annual History of Physics Essay Contest. Single-author essays from undergraduate and graduate students should be submitted by July 15. The winning author will receive a cash prize of $1,000. For further instructions, see the APS FHPP web page.

---

The Nobel Prize in Physics for 2019 was awarded in part to Michel Mayor and Didier Queloz “for the discovery of an exoplanet orbiting a solar-type star.” The precise phrasing—not the first exoplanet but an exoplanet—quietly acknowledged, without naming, an earlier discovery of planets around an altogether different kind of star. Three years before Mayor and Queloz detected 51 Pegasi b, Alex Wolszczan from Arecibo Observatory and Dale Frail from the National Radio Astronomy Observatory announced the discovery of two Earth-mass planets orbiting a millisecond pulsar, PSR B1257+12.

The marginalization is not limited to the Nobel Prize. Textbooks, popular accounts, and even professional review articles routinely begin the exoplanet story in 1995, treating the pulsar planets as an exotic curiosity, interesting perhaps, but not part of the main narrative. Wolszczan and Frail’s discovery has become, in a sense, the field’s open secret: acknowledged when pressed, but rarely centered.

Why is it that this event is rarely treated as the beginning of the exoplanet story? The reasons for this are tangled up in questions about what counts as a “real” planet, what kind of star it ought to orbit, and which scientific community gets to claim the discovery as its own.

Interviews offer first-hand look at the pulsar planet discovery

I recently conducted oral history interviews with Wolszczan and Frail. Together, they offer a crucial perspective on how one of the most important and underappreciated astronomical discoveries of the twentieth century was made and how the community chooses to remember it.

The discovery itself was unexpected, like so many in the history of astronomy. In early 1990, the Arecibo Observatory in Puerto Rico, then the largest single-dish radio telescope in the world, was shut down for maintenance, its azimuth locked at a fixed position each day. For visiting astronomers, this was a frustrating interruption. For Wolszczan, a Polish-born resident astronomer at the facility, it was the opportunity of a lifetime. He proposed using the immobilized telescope as a transit instrument to sweep for millisecond pulsars at high galactic latitudes—a speculative survey that, under normal circumstances, might never have been approved.

In his interview, Wolszczan describes the atmosphere around millisecond pulsar searches at the time, when only four were known. “There was a little bit of desperation in this,” he recalls, “and the question, ‘Why are we not finding more?’ was always hanging in the air.”

The survey worked. Among the pulsars Wolszczan found was PSR B1257+12, whose timing residuals refused to settle into any standard pattern. After months of puzzling over the data, he requested a high-cadence observing campaign. What emerged were two superimposed periodicities, each with an amplitude translating to planetary-mass companions. “That was the first time that I suddenly got this illumination: ‘It has to be planets,’” Wolszczan recalls.

But illumination is not proof. Wolszczan needed a precise position for the pulsar to rule out the possibility that his signal was an artifact of positional error in the timing model. He turned to Frail, who as a graduate student had worked on phasing up the Very Large Array in New Mexico as a single dish. The VLA provided a position accurate to 0.15 arcseconds—a factor of roughly six hundred improvement over Arecibo’s uncertainty.

In his interview, Frail describes the VLA’s contribution as that of “a supporting actor,” but one without which the lead could not have performed. The VLA position, locked into the timing model, caused the data to snap into focus. The timing of the VLA observations coincided almost exactly with the July 1991 publication by Matthew Bailes, Andrew Lyne, and Setnam Shemar at Jodrell Bank, claiming the detection of a planet around a different pulsar, PSR 1829–10, a result that seemed to be the first exoplanet discovery.

Wolszczan, who had been working toward his own announcement, remembers his reaction with characteristic understatement: “I would certainly lie if I said that I was not a little bit disappointed. But being the second is still much better than being the 21st.” When Frail sent the new position to Wolszczan by fax, he scribbled a note referencing the Jodrell Bank result: “And don’t go finding any planets, ha-ha.” Wolszczan faxed back almost immediately with two orbital periods and two planetary masses.

Memories of not getting scooped and a discovery out of place

In January 1992, both Wolszczan and Frail’s and the Jodrell Bank team’s results were put before the astronomical community at a packed session of the American Astronomical Society meeting in Atlanta. A representative for the Jodrell Bank team spoke first; not to present his planet, but to retract it. A failure to properly account for Earth’s orbital eccentricity had introduced a spurious signal into the timing data. The planet did not exist. Frail, seated in the front row, remembers the moment vividly: “I remember the intake of breath from the audience, and it propagated like a sound wave to the back of the room.”

Wolszczan spoke next. He presented evidence for not one but two planets orbiting a dead star thousands of light-years from Earth. The data were unambiguous, and the positional error that had undone the Jodrell Bank result had been ruled out by Frail’s VLA measurement. The audience, having just watched one pulsar planet disappear, was now being told that two real ones had been found—around a different pulsar, by a different method, with a degree of certainty that left no room for doubt. Wolszczan would later confirm a third, smaller companion as well, a body roughly the mass of our Moon.

For a moment, it might have seemed as though pulsar planet astronomy was on the verge of becoming a vibrant subfield. Here was an entire planetary system—Earth-mass bodies, orbital resonances, a lunar-mass companion—discovered around a single millisecond pulsar. If one pulsar could host such a system, how many others might?

But more than three decades later, the answer has been sobering. A 2025 review by Laycock and Christodoulou in The Astrophysical Journal found only eight confirmed pulsar planets distributed among six pulsars, three orders of magnitude fewer than the thousands of exoplanets found around main-sequence stars. Several large-scale surveys have searched for more and come up empty. The pulsar planets of PSR B1257+12 remain, in many ways, as singular now as they were in 1992.

The new oral history interviews shed light on why. The answer, both men suggest, lies not in the quality of the evidence but in the expectations the community brought to the search. As Frail puts it: “The idea that you could find Earth-mass planets was decades down the way. So, in one paper, we found Earth-mass planets, a hint of a Moon-mass planet, orbital resonances, a whole orbital system.” The discovery was categorically unexpected.

Wolszczan describes the reaction from the nascent exoplanet community: “The common argument was, ‘Well, yeah, that’s interesting, but those planets, if they’re planets at all, they go around a pulsar,’” he said. NASA’s TOPS (Toward Other Planetary Systems) program had laid out a precise roadmap for how extrasolar planets would be discovered, and a radio astronomer finding what would later be known as super-Earths around a neutron star, Wolszczan suggests, “kind of ruined the whole idea.”

Both interviews speak to themes that resonate well beyond the specific case of the pulsar planets: the role of serendipity in scientific discovery, the allocation of scientific credit, and the gap between how discoveries are actually made and how they are later remembered. As Frail observes: “I think you’re doing a disservice for young scientists by not showing them that research is very messy. Most of the time, you don’t know what you’re doing.” He adds: “Perhaps people are uncomfortable with the idea that these two clueless radio astronomers, who were poking about their data, looking at this strange signal, happened upon the idea of finding an entire solar system of planets.”

These oral histories are the start of a study on the history of the pulsar planet discovery that is itself part of a larger AIP-led initiative on the history of exoplanet science that is itself part of a larger survey of the history of astrophysics. While our picture is still very incomplete, these interviews offer a vivid, first-person window into one of the most consequential and underappreciated discoveries in modern astronomy. They are also a reminder that the first exoplanets were found not by following a roadmap, but by two radio astronomers who were, by their own cheerful admission, poking about in the data and paying attention to what they found.

—

Rebecca Charbonneau
American Institute of Physics
rcharbonneau@aip.org

---

You can sign up to receive the Weekly Edition and other AIP newsletters by email here.

---