Origins of the "Plane on a Treadmill" Meme

By request of Waxy reader Logan Ingalls, I spent most of the day tracking down the origins of the Plane On A Treadmill physics puzzle/meme. I figured this would be simple, but six hours later, I’ve combed through archives of the web, Usenet, magazines, academic journal, physics websites, and hundreds of discussion forums looking for the source.

There have been a couple thought experiments that came close, but still different enough that they can’t be considered the source. Jason Scott found this Usenet post from September 1990, which sounded promising:

There is nothing special about the ground as a reference plane! Consider a large flatbed truck moving at 80 mph from east to west. Does the truck’s velocity have ANY effect on the airplane’s ability to fly? Now try to land on the truck….the velocity of the truck becomes *real* important!

O.K., I’ll carry my reductio-ad-absurdum one step more. I mount the entire state of california on a conveyor belt. Initially my conveyor belt is at rest. I take off, climb to some high altitude, and then key my mike 4 times which sets the ground into motion. Does my plane suddenly stop flying?

Very similar, but key differences in the question (and a five-year-gap before it caught on) suggest that this didn’t inspire the puzzle.

So, where did it come from? The earliest reference to it that I can find is this post to the PhysOrg Forums, dated July 19, 2005. It seems unlikely that some random poster on a physics message board would have invented it, but as far as I can tell, this is the case. Update: Its appearance online goes back to a Russian discussion forum in 2003. Read the updates at the bottom of the post for more.

I’ve tried to contact “dirak” through the PhysOrg site to get a decisive answer, but since he hasn’t posted a new message since July 2006, I think I’m out of luck. Looking for his username turns up some Slashdot submissions, all anonymous, and that’s about it. I’d even tried to contact the PhysOrg forum creators, but they’re extremely protective about their own privacy, too. As a desperate measure to locate the PhysOrg folks, I then tried to track down the freelance writers that wrote for PhysOrg (through Myspace and Facebook), and am waiting for those leads to respond. Update: Dirak responded! Read the updates below.

There’s no reference to the puzzle anywhere before July 2005, and no reference to it again until October 2005, suggesting that the rapidly-expanding PhysOrg thread led to users cross-posting it to other forums to get advice.

One of the earliest references is from November 15, 2005 on this discussion thread from, a popular aviation forum for serious pilots. Often cited in other messages from 2005, the Flightinfo thread was eventually removed from the site entirely in late 2006. On November 27, it spread further when an AVWeb columnist wrote a long article (also now offline) about how the controversy was spreading in the pilot community.

At this point, it was appearing on several message boards, inspiring heated debate everywhere it went. In December 2005, the question was posed to the Straight Dope discussion forum with no clear answers. A few weeks later, Cecil Adams himself addressed the question in his February 3, 2006 column. Cecil’s syndicated column broke the question into the mainstream. A few days later, Jason Kottke started his obsession with the question and the rest is history. Mythbusters’ elaborate tests were intended to answer the question for good, but it seems to have only strengthened the resolve of its detractors.

So until I hear back from “dirak,” or manage to contact PhysOrg, I’m going to consider this case closed. (Needless to say, if you can find an earlier reference or have any other information, please let me know.)

Updated February 7: Unbelievably, “Dirak” (his real name’s Andrew) responded to my private message through the PhysOrg forums! He says, “Yes, I’m sure I was the first who brought this topic to the English-speaking internet. However, I wasn’t the one who invented this question. I’m half-Russian and I read about it on one of Russian forums back in 2005.”

He continues, “I looked for that thread for you, here it is (it’s in Russian). The first guy (shipwreck) wrote about it in 2003 and in his first post he says that they were arguing about it in his college. shipwreck is an unregistered user on that board, so I’m not sure if it’s possible to contact him at all. Hope this helps you.”

Incredible. So, this brings its appearance online back to August 4, 2003. Any Russians out there want to carry on the search? Trying to contact “Shipwreck” would be the next step.

Also: Want me to research another meme? Email or IM me and I’ll give it a shot. Whee!


    “…the Flightinfo thread was eventually removed from the site entirely in late 1996.”

    Do you mean 2006?

    “Want me to research another meme?”

    The white whale seems to be goatse guy. I know a lot of people have tried to track him down but I’ve never heard of anyone identifying or interviewing the guy (or if he’s still alive, according to rumors that he died long ago).

    Go for goatse, Ahab!

    @mathowie, there’s actually a lengthy interview with the [purported] GoatseGuy on the Body Modification E-zine website ( — I’m not linking directly to the interview because the page is, as you’d expect, filled will all sorts of accompanying artwork. But it’s the first result of this google for ‘plp56’, if you’re hungry for goatse.

    Wow. I could have sworn I’d seen that one long before 2005, probably in, or alt.folklore.urban. However, you’ve done the research and I haven’t, so I’m probably misremembering…

    Whenever I see a meme emerge fully formed in a discussion forum, I tend to agree… But I checked a number of print sources, including Lexis-Nexis, and couldn’t find anything. I’m hoping that if it originated somewhere else, some kind soul out there will speak up and fill in the blanks.

    Also, when a meme originates offline, it tends to appear spontaneously in a few places online at once. In this case, it appeared exclusively on the PhysOrg forum and every appearance in the second wave a couple months later appears to have been by someone exposed to the original PhysOrg thread.

    I just heard back from one of the PhysOrg freelance writers on Facebook, and she’s going to get me in touch with the editor of the site to see if they can help. We’ll see if it goes anywhere.

    When I read that Mythbusters was testing this meme I read a few online discussions which popped up. What struck me was that people had difficulties clearly communicating such a physical and graphical problem in threads of comments. Arguments arose over details which should have been cleared up in strictly defining the problem. There was confusion over the difference between a theoretical problem and any physical implementation. It highlighted that there is still a lot of room for improvement in online communications.

    I’ve heard it mentioned in a physics class at my local community college around that time. But the professor did mention that it was a question he likes to ask all his classes, implying that he’s been asking it for at least a few years.

    See Salon’s Ask the Pilot column on this exact question; he doesn’t trace the question back before late 2006, but his answer does go into the various misinterpretations and convoluted explanations that it seems to spark.

    Q: two beer glasses 1 metre apart are the only things that exist in a universe – how long does it take for the beer glasses to clink (i.e. touch)

    This was a very very long discussion in a pub in London circa 2002 – the option were either days/weeks or years/decades/eons. Anyone care to solve the conundrum

    The beer glasses come together eventually, but to find out how soon, just plug the masses into the equations. They have an innate gravitational pull toward each other, but so minimal that it’s hard to say how long it would take.

    And I can’t believe people can’t visualize the conceptual airplane-on-a-conveyor-belt problem. It seems ridiculous that people would equate an airplane’s wheels with a car’s wheels, completely misunderstanding how an aircraft is powered and the requirements for takeoff.

    Can you explain the actual ‘problem/puzzle’? From what I’ve read, it doesn’t make any sense in that what does the ground moving have to do with a plane in the air? Is there more to it that actually makes it a ‘puzzle’? It just didn’t really seem like that much of a brain teaser for anyone dealing with physics.

    I’m guessing that a lot inquisitive technical minded people thought about the scenario long before the internet. I know I thought about it when I was a kid, specifically that aircraft carriers could use treadmills to shorten the takeoff distance of jets.

    The problem is that people think that a treadmill running in the opposite direction of the desired flight path would somehow prevent a plane from accelerating enough to get in the air. For a plane sitting still on a treadmill with its parking brake on, the treadmill would carry it back. As soon as you release the plane’s parking brake, it would still move backwards, but the second you start the engine, the plane starts moving forward through the air, and the wheels act only as ball bearings. In effect, the ground moving has as little to do with the actual velocity of the plane as the ice under a skater. What matters is only the speed of the air through which the plane is moving, and the wheels only come into play when the wheels are locked or when the engine is off.

    Think of the solution this way. Remember the canals that used to run this country? The direction of water flow when traveling upstream (treadmill) was negated by the use of donkeys or mules (the plane’s propellers or jets) traveling against the current. Except with a plane, the wheels provide much less resistance against movement than the bottom of the boats did in those days.

    With reference to the beer glasses, to nitpick, if the glasses are the only things that exist in the universe, doesn’t that preclude the existence of things like gravity? And time I guess, and observers; (if a tree falls in the forest etc.. does anyone give a s**T?)

    Something else to consider is that half the people thought the question was asking if the plane could take off without moving forward.

    But the Russian page is asking if the plane can move forward.

    Nope, it won’t lift. The only reason it ever would be able to lift would be if it could generate enough speed to create a ‘downwash’ sufficient for lift. And that builds on the assumption that the conveyor band won’t adapt its speed. “This conveyer has a control system that tracks the plane speed and tunes the speed of the conveyer to be exactly the same (but in opposite direction).”

    Semantics of course, but to get it to lift under such circumstances? A infinite conveyor band perhaps? With a long dead pilot in the cockpit?

    Here is what you need for a ‘lift’.

    “The lift of a wing is equal to the change in momentum of the air it is diverting down. Momentum is the product of mass and velocity. The lift of a wing is proportional to the amount of air diverted down times the downward velocity of that air. Its that simple. (Here we have used an alternate form of Newton’s second law that relates the acceleration of an object to its mass and to the force on it; F=ma) For more lift the wing can either divert more air (mass) or increase its downward velocity. This downward velocity behind the wing is called “downwash”.”

    Forget about the The Bernoulli explanation. It is a often cited example of why, ah, wings take flight. But it’s not enough to explain it.

    I forgot to comment in my last post about the self speed adjusting treadmill issue. Even if the treadmill adjusts its speed to increase as the plane’s forward speed is increasing this will make no difference because once the friction has been overcome between the plane’s wheels and the treadmill then these two become independent systems. The increase in the speed of the treadmill will not be communicated to the airplane’s body.

    This is actually a simple problem if you look at it the right way. To get a body moving you just need to provide it with a force greater then other forces acting on it in the opposite direction (assume a one dimensional world for simplicity). Once that force is applied the body will begin to accelerate. The speed of the treadmill is completely irrelevant because the friction between the plane’s wheels and the treadmill is the same no matter what the treadmill speed is. As soon as the airplane engine overcomes the friction between the plane’s wheels and the treadmill the plane will begin to accelerate forward in relation to the surrounding air at a rate that will be the same whether the treadmill is moving at one mile per hour backwards or 1000 miles per hour backwards.

    Now, once the plane starts to accelerate forwards it will initially have a speed equal to and in the same direction as the treadmill’s motion. So for example, if the treadmill is moving backwards at 200 mph then when the plane first starts accelerating in the opposite direction of the treadmill it will have a speed of -200 mph going forwards in relation to the surrounding air. For the plane to take off it will need to overcome this speed of -200 mph and get into the positive air speed it needs in order to takeoff. This will happen because the plane is accelerating forwards.

    In the above argument I am assuming that the plane cannot takeoff when moving backwards no matter how fast the treadmill is moving. This is reasonable because the angle of attack of the wings will not allow the plance to take off if moving backwards.

    Now about the self speed adjusting treadmill issue. Even if the treadmill adjusts its speed to increase as the plane’s forward speed is increasing this will make no difference because once the friction has been overcome between the plane’s wheels and the treadmill then these two become independent systems. The increase in the speed of the treadmill will not be communicated to the airplane’s body.

Comments are closed.