Don't Throw Out Tribal Knowledge
The government just agreed with the thing I've been doing on a shop floor
For years, "tribal knowledge" has been treated inside most engineering organizations as a polite word for a problem — the stuff that lives in one person's head, doesn't show up in a drawing or a spec, and disappears the day that person retires or walks out the door. The instinct, almost everywhere I've worked, has been to try to eliminate the need for it: standardize the procedure, write the perfect work instruction, make the human judgment unnecessary. Get rid of the tribe, keep the knowledge.
That instinct is backwards, and I've spent the last stretch of my career at Quest Defense trying to prove it.
What we're actually doing
The work is simple to describe and hard to do well: sit down with the people who hold the knowledge that never made it into a document, and capture it before it walks out the door. Our pilot session covered a combustion chamber weld repair procedure — the kind of thing where the spec tells you the parameters, but the actual judgment about when a weld pass looks right, when to stop, when something's about to go wrong, lives in the hands and eyes of someone who's done it a thousand times, not in the paperwork. We sat with the person who holds that judgment and worked to get it captured in a form the next generation can actually learn from.
I've also been mentoring some materials engineers — which is its own version of the same principle. Not everything that matters transfers through a training module. Some of it only transfers person to person, in the room, over time.
None of this is glamorous. It's slow, it's specific, and it doesn't scale the way a policy memo scales. But it's the actual mechanism by which knowledge survives an organization's turnover instead of leaving with it.
Then the White House said the same thing
Science: A New Golden Age, the July 2026 OSTP report to the President, has a chapter — "The Marriage of Science and Craft" — that reads like it was written by someone who'd sat in on that weld repair session. The report states plainly that technology is encoded "not just in papers and patents, but in the tacit knowledge passed from mentor to mentee," and that decades of offshored manufacturing broke exactly that transmission line.
Its proposed fixes aren't abstract, either. The report calls for practitioner-in-residence programs that embed machinists and technicians directly alongside PhD researchers. It calls for portable, industry-recognized credentials in advanced manufacturing and lab techniques — so that a skill proven on one shop floor travels with the person who holds it, instead of staying trapped inside one company's walls. It calls for modernized apprenticeships in science and technology fields, not just the traditional trades.
That's not a new bureaucratic layer being proposed. That's the Federal Government looking at what companies like the one I work with are already doing on a shop floor and saying: keep doing that, and let's build the infrastructure to do more of it, everywhere.
Don't throw out the tribe
Here's the diagnosis versus the remedy, side by side. The bureaucratic diagnosis is macro — it looks at a national economy and sees that manufacturing capacity walked overseas, and with it, the mentor-to-mentee chain that used to carry knowledge forward. My remedy is micro — one weld procedure, one mentee, one session at a time, on the assumption that the chain doesn't get rebuilt at the policy level. It gets rebuilt one deliberate capture session at a time, by people willing to sit down and do the unglamorous work of getting judgment out of someone's head and into a form the next person can actually use.
The government's report validates that instinct instead of contradicting it. It's not telling companies to standardize their way out of needing experienced people. It's telling them the opposite: the tacit knowledge is the asset, not the liability, and the fix is building better ways to pass it on, not better ways to make it unnecessary.
Don't throw out tribal knowledge. The government finally sees its value. Now it's a matter of who actually does the work of capturing it before the people who hold it are gone.
Herbert Roberts, P.E. is a licensed professional engineer with 32 years in aviation research and development