Paul Schrader and Brian De Palma made "Obsession" together in 1976. It destroyed them both in exactly the way it had to.
The film succeeded. The partnership did not, and this is not a tragedy with a lesson—it is a contradiction with a mechanism.
The received story treats the triumph and the rupture as separate events, as though artistic lightning could strike without the current running through both of them simultaneously, killing as it illuminates. But this version misses what actually happened.
The film exists because Schrader and De Palma had irreconcilable visions and neither was wrong. De Palma wanted to make a thriller, a formal experiment in suspense and visual grammar—Schrader was writing a psychological study of obsession itself, where the camera's gaze becomes the character's pathology. In production, in the edit, one of them had to surrender. The work we see carries both intentions in permanent tension. It's alive because neither man got his film.
The moment the work was done, that same friction became unmistakable resentment. Each could see the film they'd wanted just beneath the surface of what they'd made. Each knew the other had prevented it. They'd spent months in perfect disagreement. The resulting depth registered as proof of their incompatibility rather than evidence of how their conflict had made the work possible. This pattern is not unique to them—every meaningful collaboration carries this seed. The same intensity that makes the work transmissible makes the partnership unsustainable.
The question isn't why they broke apart. It's why anyone expects shared vision to survive once you realize the vision was never shared at all, only productively at war.