Like anyone else who periodically tries to convince their spouse to eventually , I was very excited to hear about a , from NASA, no less. It seemed too good to be true, so I steeled myself for disappointment and asked , Stephen Granade.

I actually asked several physicists about it, and the answer I got from all of them was that the engine described really just couldn't work — it would be the equivalent of trying to make your car go faster by pushing on the steering wheel.

Or, maybe think of it like this: we can sort of imagine a regular rocket motor as a tube with one open end, out of which some demon is flinging bowling balls. The force of the balls leaving would have an equal reaction in the other way, sending the tube through space. In this engine, the demon would be in an enclosed tube, and would be bouncing bowling balls off the walls. That shouldn't propel anything anywhere.

To get to the bottom of this, I left a pewter model of Copernicus riding a dragon inside the first "O" of the Hollywood sign, which lets our captive physicst, , know we needed him. It worked, and here's what he had to say:

I've been getting a lot of questions about the Cannae drive that NASA's tested, and it's easy to understand why. It's got all of the elements that make a fascinating science story. Space travel! Using a drive with a new kind of physics! Discovered by someone whose theory was laughed at! It plays into the narrative of a misunderstood genius whose theory was scoffed at, only to have it later be vindicated.

Unfortunately, many news stories about the Cannae drive have been making statements that just aren't supported by NASA's test. Claims of being able to produce thrust the way that the Cannae drive is supposed to are extraordinary, because they violates what we understand about the conservation of momentum. Extraordinary claims require extraordinary proof. What we've got now is potentially more like the that turned out to be caused by a loose cable. NASA's results are extremely preliminary, and the thrust they saw could have come from things other than new physics.

To see why, let's talk about rockets. Rockets are an example of Newton's third law of motion: for every action, there's an equal and opposite reaction. Rockets burn fuel, creating a hot jet of gas that spews out of one end. The force of the gas particles going one way produces an opposite force on the rocket, creating thrust and pushing the rocket in the other direction. It's a consequence of how energy and momentum are conserved.

But hot gas particles aren't the only thing that you could use to make thrust. You could use light! It turns out that electromagnetic waves like light have momentum. That may seem weird at first: light is made up of massless particles called photons. Since the momentum of a particle is its mass times its velocity, it seems like photons shouldn't have any momentum because they don't have any mass. But it's a consequence of E = mc2, Einstein's equation that ties energy and mass together. Photons don't have mass but they do have energy, and that's good enough in this case for photons to have momentum.

Now, electromagnetic waves don't have a of momentum, which is why your head doesn't hurt when you walk outside into the sunlight. But it's there, and it's enough to push things around in space, where you don't have pesky air slowing you down. Japan and NASA have tested light sails, where you unroll a big reflective sail and let the sun's light bounce off of it. The photons' momentum is transfered to the light sail and pushes it like wind pushes a regular sail.

The Cannae drive is using electromagnetic waves, but it closes them up inside a box called a waveguide. The light bounces back and forth inside the waveguide. This is where we get into the apparent violation of the conservation of momentum. The photons bounce off of the front wall, pushing the box backward. Then the photons bounce off of the back wall, pushing the box forward. You get no real change! Rockets and light sails work because they're throwing particles behind them. This setup isn't throwing any particles anywhere. If you get thrust, you'd be violating the conservation of momentum. That's the first red flag.

Ah, but there's new physics supposedly creating the thrust: momentum is being transfered by the "quantum vacuum virtual plasma". That's the second red flag. What a "quantum vacuum virtual plasma"? It's true that quantum mechanics allows you to have virtual particles. You can think of them as ghost-like particles that appear for extremely short periods of time before disappearing again. This sounds crazy, but virtual particles cause effects that we can measure, like the Casimir effect where two uncharged metal plates brought extremely close together will feel a force between them. In fact, because of virtual particles, Stephen Hawking has predicted that black holes will evaporate! At the edge of the black hole's event horizon, virtual particle and anti-particle pairs are created. One of them falls into the black hole, while the other escapes. This makes the black hole lose energy and shrink. But virtual particles don't form a plasma. A plasma's a gas made up of electrically positive and negative ions. That's not what you get from virtual particles.

Note that NASA's not making any claims about the physics of the drive. The NASA abstract flat-out states, "This paper will not address the physics of the quantum vacuum plasma thruster, but instead will describe the test integration, test operations, and the results obtained from the test campaign." And they're clear that their results need double-checking, as the abstract ends with, "Future test plans include independent verification and validation at other test facilities."

The main reason is because they saw a force . They had two drives, one with slots on the end, and the other without slots. The drive with slots on it was supposed to create thrust, but the one without slots wasn't. But they both produced a force! It'd be like testing a new drug and seeing that it has a tiny effect...but so does the placebo! I wouldn't call that drug a miracle breakthrough without a lot more testing and a much more significant result.

It doesn't help that the effect that NASA measured was tiny. They measured a thrust of less than 50 microNewtons. Pick up a single grain of sand. Feel how heavy it is? The sand grain's force on your hand is the same as the thrust they measured. That's so low that, in the full paper, they say,

If ocean waves from 25 miles away can affect the measurement, you can see how delicate their experimental setup is, and how easy it would be to have gotten something not quite right in their measurement. It's why the researchers are calling for more testing. In my opinion, it's far likelier that the NASA researchers are measuring a force due to some experimental error rather than from some radical new physics.

The NASA researchers are being cautious. This is the abstract from a conference paper, which is how researchers talk about work in progress, and the is "ANOMALIES". They're not going to claim to have upended physics after only eight days of constant fiddling with the drive, only two of which involved testing, when they couldn't test it in a vacuum environment*, and when the "null" drive apparently produced the same force.

I'm disappointed that, despite the researchers' caution, many media have given the story headlines like "Nasa validates 'impossible' space drive" and "NASA Validates 'Physics Defying' Space Drive". It reminds me of last month's lionfish story. A 12-year-old's science fair project about lionfish was spun in the media as "12-year-old makes a breakthrough that stuns researchers!" and then counter-spun into "sixth-grader may have stolen credit from a researcher!", when . In both that case and this one about the EmDrive, there's a ready-made narrative that a lot of news outlets have fallen into. In the lionfish case, it's the presumed story of a young prodigy who discovered what scientists have overlooked. In the EmDrive case, it's the presumed story of how a lone researcher stood against the scientific establishment and has proven his detractors wrong.

I'm glad to talk about the science and the experimental approach that's led people to question the Cannae drive. I just wish there weren't such hype around these very inconclusive results.

* As the abstract says, "Testing was performed on a low-thrust torsion pendulum that is capable of detecting force at a single-digit micronewton level, within a stainless steel vacuum chamber with the door closed ." [emphasis added]