I Started Life Going to the Encyclopedia
Where to go to get your facts questioned — or, the comment section is a free design review
I started life going to the encyclopedia. For a couple of decades after that, I went to Wikipedia. Now I just leave a comment on somebody’s post, and every fact I need comes to me.
I want to walk that arc slowly, because it runs backward from the way we usually tell it — and the backward version is the true one.
A few days ago a man posted a short video of himself making a prosthetic socket. No lab, no autoclave, no drama. He wrapped carbon fiber over a mandrel, pulled a plastic sleeve down over the whole thing, and milked the resin through by hand until the weave went wet and dark. Closed molding at about its most basic. Basic and clever as it gets.
Then the comments arrived, and the comments were the real show.
You could read each commenter’s home discipline by what they were anxious about. The audit-and-certification man asked, essentially, is this even real — where’s your outer-mold-line control, where’s your resin transfer system, you’ll end up with dry spots. The aerospace stress engineer worried about epoxy sensitization and bare skin. A metallurgist admitted, with genuine puzzlement, that he’d worn prosthetic legs for twenty-five years and had never once seen the thing get made. A chorus of good people said PPE — gloves, respirator, read the safety data sheet — and said it about eight separate times. One old hand recognized the resin family and warned that if it was the nasty stuff he remembered, it was carcinogenic as hell. He wasn’t wrong.
And then, quietly, two working prosthetists read the whole thing correctly in about four seconds: carbon over a mandrel that was probably printed from a scan of the patient’s leg, milked down under a bag, nice work. The people closest to the actual application had the least to correct.
That’s the tell, every time. The volume of correction a piece of work draws is not a measure of how flawed the work is. It’s a measure of how many places the work gave people to stand. Simple, visible, clean work hands everyone a foothold. Everyone can see it well enough to plant a flag that says I’m in this trade too. Saying “huh, that’s elegant” needs no credential. Saying “you should’ve run a resin transfer system” plants a flag. The medium over-samples the flag-planting. That’s not a verdict on anybody. That’s physics.
But I didn’t sit down to write about them. I sat down to write about the arc — because that comment section, the one that looked like a pile-on, was actually the most useful thing in the room.
Start at the beginning.
The encyclopedia had all the authority in the world and could not test my claim at all. Authority flowed one direction. I went to it, it spoke, I received, and the conversation was over before I arrived. If I had a specific question — not what is a composite but will this specific layup hold under this specific load — the encyclopedia had nothing to say to me. It was finished. It was a monument. You don’t argue with a monument.
Then came Wikipedia, and the thing was alive. Still me going to it — but now the fight was visible. Click the talk page and you could watch the seams: two people arguing over a sentence, a third citing a source, a fourth reverting the lot. For the first time the authority was contested in the open, and contested authority is worth more than the sealed kind, because you can see what it survived. But it still couldn’t cross-examine me. I could read the argument. I couldn’t enter it with my own claim and watch strangers come at it.
And now the third beat, which turns the whole thing inside out. I don’t go and fetch the fact anymore. I emit a claim and the facts come to me. I stopped being a seeker and became a lightning rod. I post the thing and wait four days, and a panel I could never have afforded to convene shows up unbidden and cross-examines my claim from five directions at once — and bills me nothing.
Look at what actually happened across those three steps. Each one got less authoritative and more useful. The encyclopedia had every ounce of authority and zero ability to test my particular fact. The comment section has no authority whatsoever and tests my particular fact harder than anything in the history of the species. The “worst” source is the best cross-examiner. Authority and usefulness came apart, and for the one job of getting my own claim stress-tested, usefulness is what I needed.
Because that’s what a comment section is, if you have the stomach to read it that way. If I genuinely wanted my facts questioned — deliberately, on purpose — where would I even go? I’d have to build a panel. Find a metallurgist, a certification auditor, an aerospace stress man, a couple of working clinicians. Get them in a room. Buy the coffee. Slow, expensive, mostly impossible. Or I post the socket video and wait until Thursday. The panel assembles itself. It arrives disguised as annoyance, which is exactly why most people throw it away instead of reading it.
I spent enough years signing my name to analyses to know that a claim isn’t worth much until somebody has tried to break it and failed. That’s not cynicism. That’s the whole job. A fact that has never been attacked is just an opinion that got lucky. Publishing one is voluntary cross-examination — you are handing your claim to strangers and inviting them to swing.
Here’s the honest limit, though, and it’s the part that keeps this from turning glib.
The crowd is a magnificent cross-examiner and a terrible judge. It will hand you challenges from five directions and adjudicate exactly none of them. It tells you where the weak points might be. It does not tell you which ones are real. That’s still your desk. The authority didn’t vanish across those three beats — it migrated. It moved out of the monument and into you. You became your own editor, and the price of the free panel is that you have to run it yourself.
Which means what you’re actually looking at is a free design review. Every engineer who’s sat in one knows the drill. A real review costs you a conference room, a dozen calendars, a week of prep, and somebody’s travel budget — and the whole expensive ritual exists to buy you one thing: eyes that will find what you can’t. Leave a comment on a technical thread and you get the same eyes for the price of a paragraph. The auditor checks your process, the stress man checks your materials, the clinician checks your application. That’s a review board. It just convened itself in your mentions instead of a booked room.
And every engineer already knows the other half: a review produces findings, not gospel. You don’t obey a review board — you disposition it. Some findings are action items you have to close out. Some are the same concern raised by three people who never heard each other. Some are a reviewer riding his own hobby-horse into a room where it doesn’t apply. You already sort review comments by whether they’re load-bearing. So the sorting isn’t a new skill I’m handing you. It’s the review-board instinct you’ve had for years, pointed at a comment thread.
So when the panel convenes, you sort before you feel. Read the comment as evidence of where the commenter lives before you decide whether it’s a verdict on your work. Roughly, they fall into five piles:
The load-bearing correction — someone caught a real error. You answer, you thank, you fix. The bare-hands-in-wet-resin note is this one. You do not let it roll; letting it roll would be dishonest, and saying so out loud is what keeps you credible.
The redundant-but-right — the same true note struck eight times. One gracious acknowledgment covers all eight. You don’t owe each of them a separate reply.
The genuinely curious — what’s the mandrel material? why not just infuse it? Those aren’t corrections at all. That’s the good stuff. Answer freely; that’s the conversation you were hoping for.
The out-of-domain correction — someone applying his shop’s rules to your context. Right in his world, not applicable in yours. A one-liner, or nothing. And a note to yourself: his misfire tells you which context you didn’t wall off clearly enough.
The reflexive flag-plant — the person demonstrating he’d have done it better. That comment tells you where he lives, not where your work failed. Let it roll, and don’t take the bait, because there’s no fact in it to catch.
The maker who drowns is the one who lets every comment land with equal weight. The one who lasts learns to watch the swarm like weather instead of taking it personally — and then reads it, because underneath the weather is a free panel of experts you could never have booked.
I started life going to the encyclopedia, and it never once questioned a single thing I believed. Now I put a claim into somebody’s comment thread and get more cross-examination before lunch than the encyclopedia offered me in thirty years. That’s not decline. That’s the cheapest design review I’ll ever sit through.
And here’s where I land, after all of it — after the noise, the eight PPE reminders, the flag-plants, the shop rules that don’t apply in my shop. My real answer is thank you. Because buried in the noise, every single time, is somebody who brings an angle I never thought of. There is a lot of noise. There is also a lot of information. You have to be willing to run the review to find it — but it’s in there, and it’s free, and it came looking for me.
So to everyone who ever corrected me in a comment section: thank you. Truly.
And wear your PPE and gloves.
Herbert Roberts, P.E. is a licensed professional engineer with 32 years in aviation research and development across two companies, and has spent eight years analyzing accidents for attorneys under his P.E. license.