The Sunday Sol · one thought, once a week

From The Decisional Pass

On June 24, 2004, Mike Fincke pushed out of the Pirs hatch on the International Space Station in a Russian Orlan suit and got fourteen minutes into a spacewalk before flight controllers in Moscow saw the oxygen pressure in his primary tank doing something it should not have been doing.

Fourteen minutes. Then back through the hatch, hatch closed, over.

I have been thinking about that fourteen-minute spacewalk all week, because on Wednesday Fincke retired from NASA after exactly thirty years. Four flights. Three different vehicles — two Soyuz, one Space Shuttle, one Crew Dragon — which Spaceflight Now notes puts him among only thirteen people who have ever launched from Earth aboard three different spacecraft. 548 days off the planet. Nine spacewalks, 48 hours and 37 minutes outside.

And here is the thing that struck me: that career is bracketed by two decisions to stop.

The second one

On January 7, 2026, Fincke was eating dinner aboard the station. He was scheduled to go outside the next day with Crew-11 commander Zena Cardman. Mid-meal, without warning, he lost the ability to speak. It lasted about twenty minutes. Then it stopped, and he felt fine.

He did not go outside the next day. Eight days later, on January 15, all four Crew-11 members splashed down off the coast of California — the first time in more than twenty-five years of continuous ISS occupancy that a crew came home early for a medical reason.

NASA has never released a diagnosis, citing medical privacy. Fincke has since said publicly that after a battery of post-flight testing, the doctors don't know either. He has declined to say more, and his stated reason for declining is the most Fincke thing in the entire story: he doesn't want other astronauts worrying that whatever it was might be lurking in them.

So: an event with no known cause, no recurrence, and no explanation, twenty-two years after a suit reading that also resolved into nothing catastrophic.

Both times, the answer was come back inside.

The part the coverage skipped

Most of what I read in January was some version of what happened to him? That is a natural question and it is still unanswered. But it is the wrong question to learn anything from, because the decision never depended on the answer.

Nobody on the ground in January knew what had happened. Nobody knows now. The call to stand down the spacewalk and bring the crew home was made anyway, correctly, and fast.

That is fault protection, and it is the single most counterintuitive idea in vehicle operations. When I was testing OPTIMISM — the Earth-bound twin of the Perseverance rover — in JPL's Mars Yard, the rover would sometimes stop dead and refuse to do anything else. It had not figured out what was wrong. It had noticed that a number crossed a limit, and it had a rule: when a number crosses a limit, stop and phone home.

Diagnosis is a ground activity. Ground has coffee, sleep, a hundred people, and all the time in the world. The vehicle's only job is to still be intact and reachable when the diagnosis eventually arrives. A rover that tried to reason its way through an anomaly before safing would be a rover that occasionally reasoned its way into a rock.

A crew is not a rover. But the logic transfers exactly. You do not need to know why a man briefly could not speak in order to know he should not spend six hours in a vacuum the following morning wearing a suit he has to operate by voice and hand.

Why this is genuinely hard

Because the ledger is rigged against the abort.

The cost of stopping is immediate, itemized, and embarrassing. In 2004 it was five lost days and a rescheduled task. In 2026 it was a truncated increment, an early undocking, science left unfinished, and a very public spectacle.

The cost of not stopping is hypothetical. It is a thing that mostly doesn't happen. It never shows up on a schedule chart. It has no line item and no owner, and in almost every case, pressing on would have been fine — which is precisely what makes the habit so dangerous. The failure mode of aborting is a bad week. The failure mode of not aborting is the thing we name centers after.

Any organization can write the rule. What is rare is an organization where the rule survives contact with the schedule, on the specific night when someone senior would rather not invoke it.

There is a phrase from January worth keeping. NASA Administrator Jared Isaacman pointedly declined to call it an emergency, describing it instead as a controlled medical evacuation rather than an emergency deorbit.

That is not public-affairs softening. It is a real distinction, and it is the whole point. Controlled means you took a path somebody had already built, rehearsed, and stocked. Emergency means you are inventing one at three in the morning. Programs spend decades and enormous sums constructing paths they sincerely hope never to walk, and the payoff for all that work is that on the day it finally happens, it is boring. A capability you use once in twenty-five years is not overhead. It is the reason the twenty-five years look uneventful.

Back out the hatch

Five days after that fourteen-minute spacewalk, Fincke and Gennady Padalka went out again and stayed out for five hours and forty minutes. They replaced a failed power controller, and Control Moment Gyroscope 2 spun back up to 6,600 rpm and returned the station its full attitude control.

The abort did not cost them the gyro. The abort is how they got the gyro.

Seven years later, Fincke was outside again on STS-134 — the last spacewalk of the shuttle program — and by the end of it he was the American who had spent more time in space than anyone before him.

Stopping is not the opposite of doing the job. On a good team, it is the maneuver that keeps the job available to you. Thirty years, three spacecraft, 548 days, and both ends of it are a man agreeing to come back inside without knowing why yet.

Godspeed, Mike.

References

• Mike Fincke, a 30-year astronaut with experience across four human-rated spacecraft, retires from NASA — Spaceflight Now, August 12, 2026
• NASA astronaut Mike Fincke leaves agency months after medical emergency in space — CNN, August 12, 2026
• After 30 years of service, NASA astronaut Mike Fincke retires — UPI, August 13, 2026
• Mike Fincke Shares More About His Medical Event — SpacePolicyOnline, March 28, 2026
• NASA reveals the astronaut who required 1st medical evacuation from the International Space Station — Space.com, January 2026
• NASA brings Crew-11 home early in rare medical evacuation — ScienceDaily, January 16, 2026
• Expedition 9 — EVA log — Wikipedia (June 24, 2004 EVA terminated at 14 minutes; June 29–30 RPCM replacement restoring CMG-2)
• STS-134 — Wikipedia (EVA 4, May 27, 2011, final spacewalk of the Space Shuttle program)

Forwarded this? Get The Decisional Pass in your inbox: