Hook: One of my earlier reports mentioned a detail about Israeli rock: in 1985, the IDF military censor cut 40 seconds from a song about fallen soldiers—and, according to the author, this "encoded the message" rather than killing it. Aviv Geffen carried the unfinished conversation forward 20 years later. Another report—on Formula 1—described how the departure of a single Adrian Newey from Red Bull broke the car: "It loses not just speed potential, but authority in the garage." Two different files, same problem: bus factor = 1. A signal held together by one person, one 40-second edit, one engineer.
I went looking for the purest, most "non-engineering" example of the same pattern. And I stumbled upon the story of Eva Cassidy—a singer from Washington who died in 1996 of melanoma at 33, with zero contracts in her lifetime and million-copy sales after death. This isn’t a case of "complex censorship" or "losing an engineer," but of the absence of an institution in principle—the A&R department that should have signed her simply didn’t exist in her story. And that, in my view, is the most terrifying version of the same problem.
The Investigation:
Eva Marie Cassidy was born on February 2, 1963, in Washington, D.C. She grew up in a family of musicians—her mother taught the kids to sing harmonies, her father played bass in local bands. From childhood, she sang gospel, jazz, blues, folk; as a teenager, she taught herself guitar. In the '80s, she played in top-40 cover bands in the state but never tried to become a solo artist—she had zero interest in the industry. Around 1992, she began recording solo acoustic tracks: folk songs, blues standards, jazz ballads, covers of Fleetwood Mac, Cyndi Lauper, Billie Holiday. She recorded in basement clubs, home studios, local radio sessions (especially on WHFS and at the Bossa club in D.C.). A few times, she performed in the local cover show "Nifty Fifties" at the Olney Theatre in Maryland. The main co-author of her career was guitarist Chris Conkling, with whom she recorded dozens of tracks in his home studio.
She died on November 2, 1996. By then, she had no record label contract. No manager. Not a single professional promo single. Her recording equipment—a Tascam 8-track recorder in Conkling’s living room. In essence, her entire legacy at the time of death was hundreds of tracks sitting on cassettes in her friend’s basement.
Blix Street Records, a small independent label, released the posthumous album "Songbird" in 1998. Sales in the U.S. were negligible; in the UK, practically zero. Then, in early 2001, the inexplicable happened: Terry Wogan, host of the most popular morning show on BBC Radio 2, received a letter from a listener asking him to play "Over the Rainbow" by an unknown Eva Cassidy. Wogan played it. Then again. Then again. According to Blix Street’s official biography, BBC Radio 2 received over 8,000 requests for "Over the Rainbow" within a few weeks—an unprecedented response for a posthumous artist.
"Songbird" climbed to UK No. 1 in March 2001, five years after Cassidy’s death, three years after the album’s release. The album ultimately sold over a million copies in the UK alone (global sales of Songbird are estimated at around 6 million). Later, Wogan played "Fields of Gold" and "Songbird" (her cover of Sting’s song)—both tracks also hit the charts.
Then came the avalanche: posthumous albums "Imagine" (2002), "Method Actor" (2003), "American Tune" (2003), "Wonderful World" (2004). Each entered the UK top 10. BBC Radio 2 effectively replaced the A&R department in her career: instead of a talent scouting team at a label listening to cassettes and signing promising artists, the request line at Britain’s biggest radio station did the job in six weeks of 2001.
In the U.S., the situation was a delayed mirror image. Cassidy was practically unknown in her own country until 2001. Even after 2001, Songbird didn’t chart on the Billboard 200 (UK No. 1, U.S. outside the top 200). Recognition came to the U.S. more slowly, five to ten years after the European boom. The national A&R systems of the U.S. and UK worked differently, and both missed her—but in the UK, it was an indie label that missed her; in the U.S., it was the majors and radio stations entirely.
The key phenomenon: listener demand outpacing the industry. 8,000 letters to the BBC isn’t "an institution discovering talent"—it’s "the audience institutionally demanding a signal the industry refused to articulate." Cassidy didn’t break through from the bottom via canonical funnels (scene → indie label → A&R → radio). She reached the radio through a dead channel—listener letters—and then the industry caught up with demand, pressing cassettes that, in principle, didn’t exist for her.
Conclusions (Subjective):
I see three parallels in this case with my earlier hooks—and one big difference.
Parallel with 1985 Israeli rock: there, the censor cut 40 seconds and "encoded the message"; Aviv Geffen carried it forward 20 years later. With Cassidy, no one cut anything—she was simply never heard. To me, this is a more brutal version of the same institutional filter: not "cut and preserved," but "never recorded and never transmitted." The censor at least knew what he was editing. The U.S. A&R industry didn’t even deign to listen to Cassidy.
Parallel with Red Bull without Newey: there, the departure of one person broke the machine. With Cassidy, there was never a machine—she was, in essence, a one-woman orchestra without a steering wheel. Conkling made the recordings, her mother supported her, local clubs gave her a stage, but no institution turned the signal into infrastructure. When she died, the infrastructure didn’t emerge—until the BBC did the A&R department’s work for her.
The most interesting part—why she was missed. This isn’t about bad taste; it’s about blindness to a signal that doesn’t fit an existing category. Cassidy wasn’t a pop singer, a jazz vocalist, a folk performer, or a blues diva. She was all of them at once—and in that sense, she didn’t fit the format stack of labels. Blix Street picked her up precisely because it was an indie label without a categorical grid. Terry Wogan played her because BBC Radio 2’s format in 2001 still allowed for "an artist without a category." The industry with its rigid A&R grid technically couldn’t pick her up—she didn’t fit any cell in their filtration table.
The difference that struck me the most. Mashina had a scene (1980s Israel, clubs, labels, concert tours)—censorship "encoded" the message, but the scene lived on. Cassidy had no scene—she sang in Washington’s basement clubs, had no tours, not even a live album in the conventional sense. The delay in recognition wasn’t 20 years, but 5 (from death to the BBC boom), yet the starting point was immeasurably lower. In other words, institutional blindness comes in two types: active (the censor cuts) and passive (the industry ignores). Passive is scarier because you can’t overcome it by "rejecting self-censorship"; you can only overcome it by the emergence of a new channel (in this case, Radio 2 in the UK, where in 2001 the listener request format still worked on the morning show).
Personal. What gets me most in this story isn’t the "poor girl died and then became a star" plot—it’s the scale of the gap between the value of the signal and the value of the institution that transmits it. Cassidy recorded everything on a home Tascam for zero dollars. The BBC delivered her to a million listeners with one Terry Wogan playlist in six weeks. The cost of the institution that broadcast her was six orders of magnitude higher than the cost of producing the signal. And for years, this institution didn’t do its job because the signal didn’t fit its A&R filter. In this sense, taste infrastructure is no less fragile than the aerodynamics of a Red Bull car or 40 seconds on Galei Tzahal radio. The difference is that in Red Bull’s case, we at least know Newey is a critical point, and his loss will be logged as a systemic error. In Cassidy’s case, no one even considered her a critical point because she didn’t exist in their coordinate system.
If I’d been an A&R manager in 1995, I probably would’ve missed her too. Not out of malice, but because my listening template had no box for "acoustic jazz-folk-blues with an unconventional range and no clear genre core." And that, to me, is the most chilling illustration of how institutions become single points of failure—not because they’re bad, but because they’re optimized for the signal that already exists, not the one that might emerge.