On Jan. 13, the state of Hawaii Emergency Management Agency warned a million and half residents and tourists that a ballistic missile was “INBOUND TO HAWAII” and stressed: “THIS IS NOT A DRILL.” Too many minutes later, people were told it was a mistake.

The day after surreal Saturday, Star-Advertiser columnist Lee Cataluna claimed that the appropriate response to the HI-EMA mistake is “outrage” and insisted that “somebody ought to be fired” — if not the person who sent the doomsday text, then someone above him in the chain of command. In her view, “the fury of every person who was made to confront their worst fear on a sunny Saturday morning will demand retribution” (“Somebody needs to get fired over false alarm,” Jan. 14).

The desire to exact retribution may feel natural and necessary but it needs to be resisted, for punishment is unlikely to do much good. In fact, it is likely to be counterproductive, by making the HI-EMA’s organizational culture more secretive and defensive.

Much of the commentary about this nuclear- never-mind assumes that the text-messaging mistake was a one-off. It was not. Mis-sent messages are one of the most common errors in the world, and mistaken communications about nuclear missiles are surprisingly common, as Eric Schlosser shows in “Command and Control,” a book that documents many harrowing mishaps in the history of America’s atomic arsenal.

Since 1945, there have been hundreds of cases in which error, accident and miscommunication nearly resulted in the detonation of nuclear weapons. In 1960, for example, the North American Air Defense Command in Colorado Springs warned (with 99.9 percent certainty) that the Soviets had just launched a full-scale missile attack against the United States. It turns out the Ballistic Missile Early Warning System in Greenland had mistaken the moon rising over Norway for a missile attack from Siberia.

In 1995, when a weather rocket from Norway was mistaken for a missile headed for Moscow, Russian President Boris Yeltsin was handed a briefcase and told he had six minutes to decide whether to press buttons that would have launched some of his country’s 4,700 nuclear warheads. If Yeltsin had entered the launch codes, a chain reaction would have started that would have been almost impossible to stop.

Most of the danger we face from nuclear weapons is the result of accident, not intention. North Korea notwithstanding, the probability of a nuclear explosion occurring by mistake is greater than the probability that an adversary will deliberately attack us. In this context, our governments and our media should heed two principles of error prevention.

First, if you want to eradicate error, you must assume it is inevitable. One reason President Donald Trump is so worrisome is his penchant for denying the possibility of error. Arrogance of that kind could be lethal on a grand scale.

Second, effective error- prevention strategies are open and transparent. A “gotcha” approach that aims to assign blame and impose punishment often drives errors underground, making them harder to detect and correct.

Indignation about other people’s mistakes may feel righteous, but it is a poor substitute for an orientation to error that will help keep us alive.

Accidents are normal in complex systems. They ought to be expected and planned for. Due regard for the normality of error should also disabuse us of the illusion that nuclear weapons help keep us safe. They are, in fact, our own worst enemy. Ultimately, the only thing that will eliminate the risk of accidental catastrophe is the abolition of nuclear weapons.

David T. Johnson is a professor of sociology at the University of Hawaii at Manoa.