Last year, when The Complete Works of Primo Levi were released in English, New Yorker critic James Wood offered some sobering thoughts about the demise of the chemist and writer who endured Auschwitz and became the most celebrated chronicler of its torments:

He survived for a very long time, and then chose not to survive, the terminal act perhaps not at odds with survival but continuous with it: a decision to leave the prison on his own terms, in his own time. His friend Edith Bruck, herself a survivor of Auschwitz and Dachau, said, ‘There are no howls in Primo’s writing—all emotion is controlled—but Primo gave such a howl of freedom at his death.’ This is moving, certainly, and perhaps true. Thus one consoles oneself, and consolation is necessary: like much suicide, Levi’s death is only a silent howl, because it voids its own echo.

Musing on Levi’s suicide may indeed be moving. But it could also be idle, because it is not clear that he took his own life on April 11, 1987, when he was found dead from a fall in the staircase of his apartment building. In 1999, I reviewed the evidence in these pages and found it inconclusive.

This is not to say that speculation regarding suicide is outlandish. For instance, there is no doubt that Levi was depressed. Robert Weil, the editor responsible for the complete works, includes in the volume harrowing confirmation from Levi’s own words:

Two days before Levi’s death on April 11, 1987, Alvin Rosenfeld received a letter from his friend, in which Levi reported, ‘I am suffering from a severe depression, and I am struggling at no avail to escape it. . . . Please forgive me for being so short; the mere fact of writing a letter is a trial for me, but the will to recover is strong. . . . Let us see what the next months will bring to all of us, but my present situation is the worst I ever experienced, Auschwitz included.’

The fact of depression, however, is hardly proof of suicide. As I document extensively in my earlier writing, proof did not exist when the police declared that Levi had taken his own life.

Nor has this proof materialized in the nearly thirty years since that fateful day. The Turin Tribunal recently granted me access to their file on Levi’s death, which was denied me in 1999. The nine-page document comprises the police report, the pathologist’s report, and testimony from the concierge in Levi’s apartment building and from Levi’s mother’s nurse. There were neither witnesses nor a suicide note. The report rules out foul play but does not touch on the possibility of an accident.

Death by accident is a literary dead-end; it offers little grist for the critical mill. An intentional act, by contrast, is a rich vein for interpreters.

Indeed, the tribunal’s documents further complicate the narrative of suicide. The timing of that morning’s events, which first made me suspicious of the official storyline, is even more compressed than the press had reported, suggesting that, had Levi killed himself, he must have done so following a precipitous decision. The concierge said that around 10:00 a.m. she rang the bell of Levi’s third-floor apartment and “personally [handed] over the mail to him.” This is further established by the nurse: “Around 10:00 a.m., Levi called me and asked me to answer the phone since he had to go to the concierge’s lodgings,” which are located on the ground floor. Less than fifteen minutes later, alerted by a call from the building’s cleaner, police rushed to Corso Re Umberto 75 and found Levi’s body at the bottom of the stairs.

Missed phone calls and concierge visits hardly seem like priorities for one intent on committing suicide. While Levi’s depression may in that short span have produced a raptus and the decision to die, it is similarly possible that once in the stairwell he leant forward to see if the concierge was still descending the steps and fell over. The height of the bannister was, after all, relatively low, and the medication he was taking after a recent prostate operation caused spells of dizziness. Based on the evidence, accident, no less than suicide, is a plausible explanation.

In other words, we do not know whether he fell or jumped. A more thorough inquiry would have reached an open verdict.

• • •

What prompted me to research Levi’s death was not the belief that the manner of his end might mar the value of his writings. I do not think it does. It was rather the painful realization that, in interpreting his death, observers reached unwarranted conclusions that contrasted sharply with Levi’s own virtues: the prudent attention to facts that permeates all his writings; his understated description of the evil he witnessed, for which he sought neither solace nor closure. When evidence demanded uncertainty, he did not shy away—a stance no less important to his work as a writer than as a scientist. He did not fear doubt.

Consider Wood’s account. He knows that the circumstances of Levi’s death are unclear: “On the morning of April 11, 1987,” he writes, “this healthily humane man, age sixty-seven, walked out of his fourth-floor apartment and either fell or threw himself over the bannister of the building’s staircase.” Yet he proceeds as if suicide were assured.

Wood has considerable company. All four of Levi’s biographers—Myriam Anissimov, Ian Thomson, Carol Angiers, and Berel Lang—agree that the writer killed himself. In The New York Review of Books, Tim Parks reviews the Complete Works without even mentioning the possibility of an accident. In a later rejoinder to a more skeptical reader, he dismisses that possibility as “wishful thinking.” In another letter, he claims that the height of the handrail in Levi’s building, which would have reached the man’s navel, convinced investigators that a fall was “impossible.” But the Turin Tribunal makes no mention of the handrail. And while that barrier is intended to prevent accidents, the assertion that it unfailingly thwarts falls is not credible. A person hanging his upper body over the rail at navel height would have more than half his bodyweight positioned over the void.

Why is it so hard for these writers to learn from Levi and bear the burden of such doubts?

Finding assurance where none exists is an all-too human error. Our relation to evidence is assailed from all sides by irrational forces. Many of us have a strong need for closure; we cannot bear uncertainty and prefer to settle on a shaky verdict than on none at all. This proven tendency, however, would not explain why so many, in Levi’s case, affirm suicide rather than accident. Other biases are at work.

One of them, sometimes called agency bias, is the tendency to insist on designed rather than accidental causes, especially when something negative, complex, or momentous occurs. Many people are inclined to believe in “intelligent design” rather than mindless natural selection, in conspiracies rather than cock-ups, in witchcraft or the devil’s work if an earthquake fells their home. Some are convinced that Levi was too remarkable a man to die as the result of a banal accident, and the disturbing conditions of his end leave these observers that much more sure of themselves.

On the other hand, wishful thinking, the cognitive bias Parks alleges, can orient the need for closure in the opposite direction, toward an accident. Many people simply refuse to believe that Levi could have killed himself. He was too precious a witness, and, by surviving the evil of the Nazis with his humanity intact, he gave crucial strength to his words and his readers. Thus some admirers felt orphaned by his death, and the idea that it might have been an accident brings solace to them.

This leaves us with a more circumscribed question: Why are biographers and literary critics impervious to evidence favoring uncertainty? Why stubbornly stick to the explanation of suicide? These writers seem too intelligent and subtle to fall victim to agency bias and the yearning for closure.

Yet, since the day Levi died, presumption has pushed toward the conclusion of suicide: one who has experienced Auschwitz, the thinking goes, cannot but crash against the unbearableness of it, all the more so when bearing witness constantly rekindles memories of its horror. It is a curse. Everything that happens to the survivor is explained as a consequence of the catastrophe. After a divorce, say, or the death of a child, one loses the freedom to go involuntarily under a bus.

But there is more. Accidents are tragic but uninspiring events. They prompt no new stories, only desolate muttering: what a shame, such bad luck. That is to say, death by accident is a literary dead-end; it offers little grist for the critical mill. An intentional act, by contrast, is a rich vein for interpreters eager to display their prowess in imagining and deciphering the minds of others. They never need prove the accuracy of their claims—indeed, they never could. In this sense, a suicide is productive.