Read Chapter 25: MARGIN

Students always asked Luc the wrong question.

By the 1990s he lectured occasionally at engineering schools in France and the United States. After telling the DCP story, someone invariably asked, “How much safety margin is enough?”

They expected a number.

He gave them another question.

“Enough for what uncertainty?”

Margin, he explained, was not cowardice engineered into metal. It was acknowledgment. The world was larger than the model. Materials varied. Humans erred. Instruments lied. Equations simplified. Launch vehicles vibrated in modes nobody predicted perfectly.

A good design contained room for being wrong.

A good organization did too.

That second kind of margin was harder to measure.

Could a technician report a failure without being blamed for delay?

Could an engineer challenge a manager without losing a career?

Could a contractor show NASA bad data without fearing financial ruin?

Could a nation pursuing a historic goal admit that a schedule might slip?

If the answer to those questions was no, Luc told students, then the program had already consumed its most important margin.

Not voltage.

Not mass.

Truth.

Luc’s lectures became less technical as he grew older.

Young engineers could learn contact mechanics from better specialists. What he wanted to teach was how technical truth became organizational truth—or failed to.

He would project two charts from the Galveston campaign.

The first was the official smooth trace.

The second was Charlie Voss’s raw strip chart with the voltage dropout.

“Same hardware. Same day. Same organization,” Luc would say. “Two histories.”

Then he asked students which chart was more dangerous.

Most answered the one with the dropout.

“No,” Luc said. “The smooth one.”

A failure you can see can be redesigned.

A failure removed from the record becomes part of every later assumption.

He never dramatized the burglary, Vale’s conviction or his own role. Students often learned those details afterward and accused him of omitting the exciting parts.

“They are not the important parts,” he said.

“What is?”

“The chart.”

In engineering, drama was usually what happened after information had already failed.