You're assembling IKEA furniture. The box is open, the instructions are just pictures of a cartoon man looking mildly concerned, and you're three hours in. You've got the monitor arm bracket in one hand, the hex key in the other, and you're thinking — this is fine. Then you reach for the frame and realize you now need a Phillips driver. And the leveling feet? Three different drives, one desk. And you're standing there holding a little bag of leftover hardware wondering how we ended up here. How did the humble screw — the thing that holds literally everything together — spawn this menagerie of incompatible heads? Was there ever a plan, or did we just...
That moment is so specific and so universal. The hex key for the monitor arm, the Phillips for the frame, the flathead for the feet — it's like a tiny museum of twentieth-century engineering decisions sitting right there in your living room, and you're the one who has to curate it with a cramped hand and a mounting sense of despair.
Daniel sent us this one, and he's asking exactly that. How did we get here? He points out that for most of us, the essential screw types have been the essential screw types our whole lives — Phillips, flathead, Torx, hex — and we just assume it's always been this way. But has it? If he got teleported to the fifteenth century with a Phillips screwdriver in his pocket, would the innkeeper look at it like a useful tool or like something that fell off a passing star?
The short answer is the innkeeper would have absolutely no idea what he was looking at. But the long answer is where all the good stuff is — because the story of screw drives is basically a story of deliberate design failures, patent stubbornness, and physics that nobody bothers to explain in hardware store aisles.
Also, apparently, a story where Canada almost won and then didn't.
Canada's great what-if in fastener history. We're going to get there. But before we do, I want to set the scale here, because it's easy to overlook. Screws are genuinely the invisible infrastructure of the modern world. We produce trillions of them every year. They're holding together iPhones and skyscrapers and the chair you're sitting on and probably the fillings in someone's teeth. They're everywhere, and most people never think about them until one strips.
Which is exactly the moment you realize you've been using the wrong bit for the last twenty minutes.
And here's the puzzle Daniel's getting at. In most areas of technology, we converge toward a single standard. USB-C has eaten almost everything. The three-point-five millimeter headphone jack held on for decades. Shipping containers are the same everywhere on earth. But screw drives? They've proliferated. We've got Phillips and flathead and Torx and Robertson and hex and Pozidriv and about a dozen others that most people have never heard of. And they're all still in active use.
It's like if every country had its own electrical outlet, and also some furniture came with its own proprietary outlet that only worked with a little L-shaped key they included in the box.
That's actually a perfect analogy, and the fact that it's only barely an exaggeration tells you how weird this ecosystem is. So what we're going to trace here is the whole arc — from hand-filed iron screws made one at a time by craftsmen, through the first standardization push that almost nobody listened to, into the battle between two competing visions of what a screw drive should be, and then the post-war explosion that gave us the fragmented landscape we're still living with today.
Somewhere in there, the deliberate design failure that became the most common screw head on the planet.
That's my favorite part. Henry Phillips looked at the problem of screws camming out and said — what if that's not a bug, what if that's the feature? And he was right, and now we all live with the consequences.
Let's go back to the beginning — before screws even had standardized threads, before power tools existed, before anyone had ever stripped a Phillips head and muttered a curse word in a Swedish furniture aisle.
The pre-industrial screw is one of those things that makes you appreciate how much we take for granted. Before the seventeen sixties, every single screw was hand-filed by a craftsman. One at a time. The threads were cut with a file and a steady hand, which meant no two screws were exactly alike. The nut that fit one screw wouldn't fit the one next to it.
If you lost the nut, you didn't go to the hardware store. You went back to the same blacksmith and hoped he remembered what he'd done.
And that blacksmith might be dead. The Wyatt brothers in England and a guy named Jesse Ramsden developed the first screw-cutting lathes in the seventeen sixties, and that was the moment screws stopped being handmade artifacts and started being something you could produce in quantity. But even then, every workshop had its own thread profile. Your screw from Birmingham wouldn't fit a nut from Manchester.
Which is the kind of thing that sounds like a joke but was actually a genuine logistical nightmare for anything involving more than one supplier.
It stayed that way until eighteen forty-one, when a British engineer named Joseph Whitworth stood up in front of the Institution of Civil Engineers and presented a paper proposing uniform thread profiles. A single standard. Same angle, same pitch, same depth. He was basically saying, hey, what if screws were interchangeable?
The kind of idea that makes everyone in the room nod and then ignore for another thirty years.
It took decades to catch on, but Whitworth's standard eventually became the foundation for basically every thread system that followed. The point is, for most of human history, the screw itself wasn't standardized. The idea of a screw drive — the shape of the recess in the head — that's an even later problem. For centuries, there was only one option: the slotted flathead. A single groove cut across the top. And it worked fine when you were turning screws by hand with a manual driver.
Because the problem hadn't been invented yet.
The flathead's fatal flaw only becomes visible when you introduce power tools. As soon as you're driving screws at speed with a motorized driver, the bit wants to slide out of that slot. It's called cam-out, and it's not just annoying — it destroys the screw head, it damages the workpiece, and on an assembly line it injures workers.
The flathead was the only game in town for centuries, and then the twentieth century shows up with factories and power tools and suddenly the thing everyone had been using is actively dangerous.
That's the moment the story gets interesting. Because two different people looked at that problem and came up with two completely different solutions. And the one that won wasn't the better design — it was the better business decision.
The first of those two people was a Canadian named Peter Lymburner Robertson. He was a traveling salesman for a tool company in the early nineteen hundreds, and one day he was demonstrating a spring-loaded screwdriver when the flathead bit slipped and gashed his hand open.
Nothing like a workplace injury to focus the mind on design flaws.
Robertson went home and invented a square-drive socket — a recessed square hole in the screw head that the driver bit fits into perfectly. The geometry is self-centering, which means you can put the screw on the end of the driver and it stays there. One-handed operation. No cam-out. The bit grips the screw like it's holding hands with it.
He solved the problem that flatheads had been causing for four hundred years, and he did it in nineteen oh eight.
Twenty-eight years before Phillips even filed his patent. And the design was objectively brilliant. The square recess has vertical walls, so when you apply torque, the force goes directly into rotation. There's no ramp for the bit to climb out of. Ford put Robertson screws into the Model T — specifically for the wooden body panels — and by some accounts it saved them hours per vehicle.
Ford's using it, it's faster, it's safer, it's one-handed. This should be the origin story of the Robertson empire.
Ford wanted to license the patent so they could manufacture the screws themselves and control their supply chain. Robertson said no. He wanted to retain exclusive manufacturing rights. Ford walked away.
The guy invented a better screw and then refused to let the biggest customer in the world use it at scale.
That's the short version. And it gets worse. Robertson also refused to license internationally. He kept the whole operation in Canada, which is why to this day roughly eighty-five percent of screws sold in Canada are Robertson, and almost nobody else in the world uses them.
That is an astonishing act of business self-sabotage. He had the better product, he had Ford ready to scale it, and he said no thank you, I'd rather keep it small.
The window doesn't stay open. Henry Phillips files his patent in nineteen thirty-six, and his design is completely different in philosophy. Robertson solved cam-out by eliminating it. Phillips solved it by making cam-out predictable.
Right — this is the deliberate design failure you mentioned earlier.
The Phillips head is a cross-shaped recess with tapered flanks. Those flanks are angled, which means as torque increases, the driver is physically pushed upward and out of the screw head. It's designed to cam out at a specific torque threshold. That was intentional.
Instead of fighting the physics, he leaned into it. The screw says, okay, you've tightened me enough, I'm ejecting the driver now.
On an assembly line, that's a feature, not a bug. It prevents over-tightening, which can strip threads or crack the material. It protects workers' hands from sudden torque kickback. And it speeds up the line because the driver pops out automatically and the worker just moves to the next screw without thinking.
Which is exactly what automakers wanted.
General Motors adopted the Phillips drive almost immediately after the patent was issued. Within three years, the entire American auto industry had switched over. The Phillips head didn't win because it was the best at driving screws — it won because it was the best at fitting into an assembly line.
The market didn't pick the better screw. It picked the screw that worked better with the system around it.
That's the core of the whole story. Robertson was technically superior in a vacuum — better grip, no cam-out, one-handed use. But Phillips was superior in context. The assembly line didn't want a screw that held on forever. It wanted a screw that let go at exactly the right moment.
Which is also why the flathead survived as long as it did, by the way. It was the only option, so everyone just lived with the cam-out problem because they didn't know there was an alternative. And then when the alternative arrived, it wasn't designed for the craftsman in his workshop — it was designed for the factory floor.
There's one more layer here that I think is fascinating. The Phillips head's cam-out behavior is actually what makes it so frustrating for DIYers today. We're not on an assembly line. We don't want the driver to pop out. We want to drive the screw flush and move on. But we're using a tool that was optimized for a completely different context — a factory worker in nineteen thirty-seven installing dashboard panels on a Cadillac.
Every time someone strips a Phillips head and swears at it, they're experiencing a feature that was designed before their grandparents were born, for a problem they don't have.
And the Robertson screw is sitting there in Canada, working perfectly, like a ghost of engineering what-ifs.
Phillips wins the patent war, and for about three decades the Phillips head is basically the standard in American manufacturing. But then the patent expires. And that's when things really start splintering.
Because once anyone can make a cross-shaped recess without paying royalties, the incentive to invent something better — and patent it — kicks in.
Nineteen sixty-seven. A company called Camcar Textron introduces Torx. And Torx is the answer to a question Phillips deliberately chose not to solve. What if you need maximum torque transfer with zero cam-out?
The opposite design philosophy entirely.
The Torx head is a six-pointed star, and the critical difference is the geometry of the walls. Phillips has four tapered flanks — they're angled, and that angle is a ramp. As torque increases, the driver rides up that ramp and out of the screw. Torx has vertical walls. The force goes directly into rotation. There's no upward vector.
You can lean into a Torx screw with an impact driver and it just...
The driving angle on Torx is ninety degrees — the bit face is perpendicular to the direction of rotation. On Phillips, it's about thirty degrees. That sixty-degree difference is the entire reason one strips constantly and the other almost never does.
Which also explains why Torx bits don't chew themselves up the way Phillips bits do. The wear is distributed across six vertical contact surfaces instead of four angled ones.
But Torx has its own trade-offs. Those vertical walls mean the bit has to be precisely aligned with the screw head. There's no self-centering ramp to guide it in. And the manufacturing tolerances are tighter, which makes the tooling more expensive. A Torx screw is harder to make than a Phillips screw.
It's better for the person driving the screw, but worse for the person making the screw. And the person making the screw is the one who decides what gets produced.
That's the tension in a nutshell. And it's why Torx found its niche in automotive and aerospace assembly lines where the torque requirements are high and the cost of a stripped screw is a stopped production line. Meanwhile, Phillips stayed dominant in general construction and consumer hardware where the torque demands are lower and the cost sensitivity is higher.
Then there's hex — the Allen key — which seems like it came from a completely different planet.
Hex drives, or Allen screws, emerged for socket head cap screws in machinery. The recess is a simple hexagon, and the driver is that little L-shaped key everyone has a drawer full of. The advantage is that you can apply a lot of torque with a very compact head — the screw sits flush or even recessed into the material. And the hex key gives you leverage in tight spaces where a screwdriver handle wouldn't fit.
Which brings us to IKEA. Because IKEA is basically the reason most people own a hex key.
There's a deliberate engineering decision behind that. IKEA furniture is mostly particle board — low-density material that strips if you over-torque it. A Phillips driver in a power drill will blow right through that stuff. But a hex key gives you a mechanical disadvantage on purpose. You can only tighten it so much with your bare hand and that little L-shaped lever.
The hex key is a torque limiter disguised as a tool.
It's the same principle as the Phillips cam-out, but achieved through user mechanics instead of drive geometry. You physically cannot strip the screw because you can't generate enough force with your wrist and that short lever arm. IKEA includes hex keys not because hex is the best drive for furniture — it's because hex is the best drive for preventing customers from destroying their own furniture.
That's almost paternalistic. The Allen key is saying, I don't trust you with a power tool, here's a little stick instead.
Meanwhile, IKEA uses Phillips screws for the parts that need to be assembled quickly on their own production lines. So even within one company, you get two different drives optimized for two different users — factory workers and confused apartment dwellers.
Which brings us to the regional puzzle. Because if you walk into a hardware store in Toronto, you're seeing Robertson screws everywhere. In Chicago, it's all Phillips. In Berlin, it's Pozidriv.
Pozidriv is the one that looks like Phillips but isn't. It's got four additional radial ribs between the main cross slots, and the flanks are straight instead of tapered. It was developed in Europe as an improvement on Phillips — less cam-out, but still self-centering. And it dominates the European market because the patent landscape was different there.
None of this is about which drive is technically best. It's about who patented what, where, and when.
Robertson refused to license internationally, so his square drive stayed bottled up in Canada. Phillips licensed aggressively in the US and the design became embedded in the entire manufacturing ecosystem. Pozidriv took Europe because the Phillips patents didn't have the same grip there and the European standards bodies went a different direction.
Once those ecosystems are established, they're almost impossible to change. Every screwdriver set in Canada includes Robertson bits. Every hardware store in America stocks Phillips. Every factory in Europe has Pozidriv tooling. You'd have to change all of it simultaneously to switch.
That's why no new drive has broken through since Torx in nineteen sixty-seven. The network effects are brutal. To introduce a new drive, you need screw manufacturers to retool their production lines. You need every hardware store and assembly plant to stock new bits. You need consumers to recognize it and buy the right driver. That's three different constituencies, and if any one of them says no, the whole thing fails.
Someone actually tried this, didn't they? The hybrid thing.
The Phillips Square-Driv. Launched in the early two thousands. It combined a Phillips cross with a Robertson square recess — the idea was you could use either a Phillips bit or a Robertson bit, and the square center would resist cam-out while the Phillips arms provided self-centering.
It was the universal donor of screw drives.
It went nowhere. It's used in some decking screws, but it never broke into general hardware. The problem is that a hybrid drive is worse than a dedicated drive at any specific task. It's fine at everything, excellent at nothing. And in a market where everyone already owns Phillips and Torx and hex bits, "fine at everything" isn't a compelling reason to switch.
Meanwhile, the aerospace industry is off in its own universe with drives most people have never seen.
Aerospace uses twelve-point, Tri-wing, Torq-Set — these are drives designed for completely different constraints. They need to prevent tampering, they need to handle extreme torque in very tight spaces, and they need to be installable with specialized tools that technicians are trained on. A Tri-wing screw looks like a three-bladed propeller recess. You cannot turn it with a flathead. That's the point.
The screw drive ecosystem isn't fragmented because of incompetence or because nobody's gotten around to standardizing. It's fragmented because each drive was an answer to a specific question asked at a specific moment by a specific industry. And then the answer got locked in by patents and supply chains and the sheer inertia of millions of existing screws.
That's the thing I find weirdly beautiful about it. Every time you open that drawer of bits and see Phillips and Torx and hex and flathead all jumbled together, you're looking at a fossil record of engineering decisions. The flathead is pre-industrial. The Phillips is the assembly line revolution. The Torx is the high-torque precision era. The hex is the user-friendly compromise. They're all still here because they all still solve the problem they were designed for.
What does this mean for you, standing in the hardware aisle with a project to finish? Because I think that's where Daniel's question ultimately lands — not just how we got here, but what do I actually do about it?
The practical answer is surprisingly simple. If you're building out a toolbox and you want one set of bits that'll handle the widest range of high-torque situations, get a good Torx set. Torx is the closest thing we have to a universal high-torque drive. It's becoming more common in construction screws, deck screws, automotive work. The geometry doesn't cam out, the bits last longer, and you'll strip far fewer screws.
For furniture assembly?
Use the hex key they gave you. Don't swap to a Phillips driver just because it fits in your drill. The hex key is a torque limiter by design — that little L-shaped lever prevents you from blowing through particle board. The tool is part of the engineering.
The advice is basically: trust the drive that was designed for the job you're actually doing, not the one that happens to be in your hand.
That's the deeper lesson here. We tend to assume standardization is always the goal — that converging on one best thing is progress. But the screw drive ecosystem is fragmented because each design solved a different problem at a different moment. There is no single best drive. Phillips is best when you need deliberate torque limiting and assembly line speed. Robertson is best for one-handed operation. Torx is best for maximum torque transfer. Hex is best for user-friendly furniture assembly where the user is a hazard to themselves.
The best screw drive is the one that matches the constraints. It's not a failure of standardization — it's an optimization for context.
That's a mental model that applies way beyond screws. Every time you see a fragmented standard — electrical outlets, railroad gauges, programming languages — the question isn't "why haven't they fixed this." It's "what specific problem did each version solve that made it worth the fragmentation.
Which brings us to what comes next. Because the screw as we know it may not be eternal. The next disruption might not be a new drive shape at all. It could be magnetic drives that don't need physical recesses. It could be adhesive bonding — we talked about VHB tape in a previous episode, the stuff that holds glass onto skyscrapers. Or it could be three-D-printed fasteners custom-designed for each specific application, where the drive geometry is generated on the fly for that exact joint.
The screw has been around in some form for over two thousand years. But the drive shapes we're arguing about are barely a century old. There's no reason to think we've reached the end of that evolution.
Daniel's fifteenth-century innkeeper would look at a Phillips screwdriver and see something that doesn't make sense. Not because the concept is too advanced — a cross-shaped metal tool isn't incomprehensible — but because the entire ecosystem it belongs to didn't exist yet. No standardized threads, no power tools, no assembly lines. The drive shape is a solution to a problem that wouldn't be invented for another four hundred years.
Which is the thing that gets me. We're living inside our own version of that right now. There are technologies we use every day that future generations will look at and say — wait, they put up with that? The screw drive ecosystem is basically a fossil record of engineering compromises, and we're the ones still excavating it every time we open a bit set.
The fossil record isn't done accumulating. A hundred years from now, someone's going to be standing in whatever replaces a hardware store, looking at whatever replaces a screw, and wondering why we spent an entire century arguing about the shape of a recess in a metal rod.
Now — Hilbert's daily fun fact.
Hilbert: In the 1920s, Italian colonial administrators in Eritrea discovered that local fishermen were using a complex double-loop lashing knot to secure cargo that, when applied to wooden crates, was actually more vibration-resistant than the standard military square lashing. The knot was never formally adopted, but several Italian naval supply officers quietly taught it to their crews.
...I have so many questions about how a knot gets "quietly taught."
The Italian navy, Corn. Not everything gets a white paper.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop, and thanks to Daniel for the question that sent us down this particular rabbit hole. If you enjoyed this episode, leave us a review wherever you listen — it helps other people find the show. I'm Corn.
I'm Herman Poppleberry. We'll be back next week with another weird prompt.
