Why does this trick with my arm work?

October 31, 2009 3:15 AM

Why does this trick work? Turn your palm face up. To turn your palm face down you must twist your wrist, right? But if you do six simple moves you can turn your palm face down without twisting your wrist. The trick appears at the beginning of this talk by Lennart Green but can also be done backwards.
posted by twoleftfeet to grab bag (23 comments total)

My guess is that when you "twist your wrist," you don't really twist your wrist... that's misdirection. You actually rotate your whole arm. I can't twist my wrist without moving my elbow, anyway. So there's no surprise that a chain of smaller elbow rotations can get you to the same place.

Before watching the video, I thought, "that doesn't sound hard," and kept rotated my whole arm at the shoulder windmill-style while keeping my arm straight. It worked without a problem.
posted by painquale at 3:39 AM on October 31

It seems to me the trick substitutes upper-arm movement for lower arm movement. When you move your arm across your heart, you rotate your upper arm. Your lower arm has moved into another plane of momement. Try it with a stick or a pen or something and it seems more obvious. You are rotating your wrist, just not flipping it over.
posted by BrokenEnglish at 4:26 AM on October 31

Yeah, this "trick" didn't make any sense at all when I read your description and watched the first link.

To turn your palm face down you must twist your wrist, right?

No. Wrong.

You must twist your ARM. Specifically, I think the radius and ulna bones rotate around each other pivoting near the elbow, but I'm no doctor.

If you watch your arm near your elbow as you turn your palm over, you can see the movement. Now do the 6-move "trick" and watch the same spot. You're still turning it over, just in a more gradual way.

Here's a test: grab your arm halfway between your wrist and elbow with your opposite hand and stop it from moving. Now the ONLY way to turn your palm over is to twist your wrist. Try it and see what happens.

Now show me someone who can turn their palm over entirely by twisting their wrist and I'll be impressed...
posted by mmoncur at 4:32 AM on October 31 [3 favorites]

Why did the cameraman cut away from his demonstration? :-(
posted by Taken Outtacontext at 5:03 AM on October 31 [2 favorites]

It works because he begs the question.
posted by weapons-grade pandemonium at 8:08 AM on October 31

BrokenEnglish has the key to the movement. The deception occurs on the movement across the chest. A one-quarter twist to the wrist occurs each time you do this. The reason this is deceptive is because it is also combined with a rotation of the upper arm humerus so that you don't notice that you are also rotating your wrist.

Here's how to separate out the upper arm rotation from the wrist rotation. Start with your right arm extended with palm up in front of you. Bring your right palm straight up until it is about 18 inches in front of your right shoulder. Now do the upper arm rotation. This will be like doing a karate chop to your left shoulder. Now drop your arm to your waist. This time the the palm remains upright because you have isolated the upper arm rotation so that you are more aware of preventing a wrist rotation.

When you do it in one step you are unconsciously making the arm rotation and a wrist rotation at the same time so that the palm doesn't remain facing up, but instead rotates a quarter turn so that the palm faces your stomach. If you repeat the process twice, you go from palm up to palm down.
posted by JackFlash at 12:02 PM on October 31 [1 favorite]

It's interesting to see how people explain this. Of course, it really has nothing whatsoever to do with your wrist or your arm; the same trick works with a book or a piece of paper. Place a book in front of you on a table, front cover up, spine to your left. Now, if you want to turn the book so the front cover is down, the most obvious way to do this is a 180 degree rotation around a axis parallel to the spine. But that same rotation can be accomplished as follows: lift the top of the book forward to stand it up on its bottom, then tilt the book onto its spine, then slide the spine along the table so the book is perpendicular to you. At this point you've rotated the book a quarter turn. Repeat the same three moves - up, tilt, slide - and the book will be back where it started, but with the front cover down.

So the trick is really just a fact about how a sequence of rotations in three dimensions results in a single rotation. But I was also distracted, thinking about arms and wrists, when I asked the question.
posted by twoleftfeet at 7:05 PM on October 31

Of course, it really has nothing whatsoever to do with your wrist or your arm; the same trick works with a book or a piece of paper.

As presented by both videos, it certainly does have to do with the wrist and arm. As both present it, the mystery is how you can do this without rotating the wrist. To most people, when they first do this, it seems they don't rotate the wrist. The seeming paradox is explained by the combination arm and wrist rotation that unconsciously conceals the actual wrist rotation.
posted by JackFlash at 7:28 PM on October 31

There's no twist of the wrist involved - no supination by the biceps brachii or pronation by the pronator teres. It turns out you really don't need to twist your wrist in order to turn your palm face down.

Try it with a book as suggested above.
posted by twoleftfeet at 8:25 PM on October 31

Sorry, the link should have been pronator teres.
posted by twoleftfeet at 8:31 PM on October 31

There's no twist of the wrist involved - no supination by the biceps brachii or pronation by the pronator teres. It turns out you really don't need to twist your wrist in order to turn your palm face down.

There must be a twist of the wrist. Look at your arm and wrist before the first step and after the 6th step. They are in exactly the same anatomical positions as if you started with your hand face up and ended with it face down by twisting your wrist in a single step. All of the bones are in the same position.

Your book analogy is not quite right because your elbow joint doesn't have the same freedom of movement as your book. When you stick your right arm straight out palm up the ulna is the bone on the inside (left) and the radius is the bone on the outside (right). The radius and ulna at the elbow are in the same position at the start of your six steps as at the end. However, the radius and ulna at the wrist are in flipped. This requires a twist of the radius and ulna.

Do your book analogy with a sheet of paper. Assume that your starting position is with the right hand palm up, so mark the sheet of paper little finger top left corner, thumb top right corner, ulna elbow at lower left and radius elbow at lower right. Now go through your six step maneuver. Peeking underneath at the bottom of your sheet of paper you see that the thumb and little finger have correctly swapped places as expected. But look at the elbow end. You have the radius on the inside and the ulna on the outside. But that is wrong. With your real arm at the elbow, the ulna remains on the inside (left) and the radius remains on the outside (right).

With the piece of paper or book, you flip the elbow end of the paper completely upside down. You can't do that with your elbow because the two bones are connected to the upper arm which prevents them from flipping over. In order to make the four positions of the paper line up properly with your hand at the end, you have to twist the paper longitudinally so that the thumb ends up on the left, the little finger on the right, but the ulna and radius remain fixed in their original positions.

If it helps, write R on the inside of your wrist next to your thumb and write U on the inside of your wrist next to your little finger. Write R on the inside of your elbow on the right and write U on the inside of your elbow on the left. Then compare the movements to your piece of paper. The only way to make the paper line up with your arm is to twist the paper.

I hope this helps.
posted by JackFlash at 12:29 AM on November 1

If you stick a thin strip of paper in the book, so it juts out, bookmark-like, from the bottom of the book and you anchor the strip of paper with a pin to the table, then sure enough after doing the six moves you will get a twist in the paper strip. So in that sense you are right... there has to be a twist somewhere. But when people say "twist of the wrist" they mean a rotation around an axis that runs from the elbow to the hand. The trick shows that you can accomplish the result of this rotation without rotating around that axis, as seen when using the book.
posted by twoleftfeet at 4:44 AM on November 1

But when people say "twist of the wrist" they mean a rotation around an axis that runs from the elbow to the hand.

But that is exactly what happens. The radius and ulna bones rotate around each other. You can simulate the radius and ulna by putting two plastic straws or two thin paper strips so that most of the straws jut out of the book with only the tips in the book. The straw/book interface is the wrist joint and the fixed end of the straws are the elbow joint and the book is your palm. You will see that the two straws rotate around each other lengthwise from elbow to wrist.

If you don't believe that a twist of the wrist is involved, try the exact same moves starting with the palm down and the first move is to bring the back of your hand to your shoulder. You can't do it this way because your wrist won't twist that way.

Maybe this will help.
posted by JackFlash at 8:34 AM on November 1

Well, I still think we're just disagreeing over what it means to "twist the wrist". In the conventional sense, when you turn your palm face up to face down you can see the rotation of the wrist, just below the hand. But (as palinquale pointed out) you can do all of the moves without bending your elbow (start palm up with a straight right arm, bring the straight arm up, then in down to your left, etc.). You still end up palm down, but if you keep your eye on your wrist while you're doing it this way you don't see that same rotation just below the wrist. I'll grant you that there must be a twist somewhere (as BrokenEnglish said, we substitute upper-arm movement for lower arm movement) but I don't think it's a subconscious twist near the hand itself. It certainly doesn't look or feel that way.

It is interesting that this little trick is harder to convincingly understand than it is to do.
posted by twoleftfeet at 1:33 PM on November 1

Well, I still think we're just disagreeing over what it means to "twist the wrist". In the conventional sense, when you turn your palm face up to face down you can see the rotation of the wrist, just below the hand.

Well, I'm game to try again if you are. Maybe there is confusion about what it means to rotate the wrist. In actuality, there is only one way of rotating the wrist. There is no swivel joint just above the wrist. Rotation of the wrist involves twisting the whole forearm.

Go back to the two soda straws. Stick the two ends into the book with the two straws side by side. Hold the free ends fixed while you slowly flip the book over. Notice how the two straws rotate and cross over each other. The rotation occurs along the entire length of both straws. This is exactly what happens to the two long bones in your forearm connecting the elbow and wrist.

You may not be aware of the twisting of the two bones across each other because it is covered by the muscles in your arm but you can feel it. Extend your right arm with the palm up. With your left hand grasp your wrist about 2 inches above the wrist so that the your left thumb is on the right side of your wrist and your fore finger is on the right side of the wrist. The two bones you feel are the ends of your ulna and radius bones that connect to your elbow. Now slowly flip your palm over. You can feel the one bone rotate over the other bone just like the soda straws in the book. If you move your grip a little farther up your forearm you will feel the bones rotating a little less and so forth up to your elbow where the ends are fixed, just like with the soda straws.

When you rotate your wrist, this is the only movement that happens. There is no separate twisting at the wrist -- it is the whole forearm that causes the motion and only the two long bones of the forearm are involved. It is the same if you just flip your wrist over or if you go through the magic motions. The only difference is that when you go through the motions, the rotation of the long bones is more gradual so you don't notice it. But exactly the same rotation takes place in both cases. That is what causes your palm to turn over.
posted by JackFlash at 2:58 PM on November 1

Here's another way to think about this: Rotation Matrices.

First, some notation. With the right hand in the initial position, we'll say that the right forearm lies in the positive x-axis, the right upper arm lies along the positive z-axis, and the y-axis runs perpendicular to the x- and z-axis, with the positive direction to the left. In other words, the right elbow is at the origin (0,0,0) and the right hand starts at, say, (1,0,0). In this notation, a straightforward 180° rotation of the right forearm corresponds to a 180° rotation around the x-axis.

The first three rotations in the trick as shown in the TED video are: a 90° rotation around the y-axis, a 90° rotation around the x-axis, and finally a 90° rotation around the z-axis The product of these three rotation matrices is:

 1  0  0
 0  0 -1
 0  1  0

as you can see here. Note that this matrix corresponds to a -90° rotation around the x-axis (negative degrees means counter-clockwise). The next three rotations are merely the first three repeated, so we end up rotating another -90° around the x-axis, leading to a full 180° rotation around the x-axis, as desired.

The basic idea, as far as I can see, is equating "turning the wrist" with a rotation along the same axis as the forearm. None of the rotations in this trick are done along the same axis as the forearm.

1. The forearm starts along the x-axis.
The first rotation is along the y-axis
2. The forearm is now along the z-axis.
3. The second rotation is along the x-axis
4. The forearm is now along the y-axis.
5. The final rotation is along the z-axis.

posted by mhum at 8:42 PM on November 1 [1 favorite]

Mhum, you are doing exactly the same thing as twoleftfeet did with the book but using matrices. The problem is that the book and the matrices don't account for the fact that the elbow joint is a hinge that restricts the rotation around the axes. The elbow cannot independently make a rotation around the x-axis in step two because the hinge of the elbow joint is aligned with the y-axis. Therefore step two consists of two sub-steps. First in step 2A you rotate the upper arm about the z-axis in the minus direction. This changes the elbow hinge to the x-axis. In the second sub-step 2B you rotate the forearm about the x-axis.

So lets do your same matrix multiplication but instead we insert the minus z rotation between the y and x rotations and see here.

Lo and behold you get the identity matrix:

1 0 0
0 1 0
0 0 1

In other words, unless you twist the forearm in between steps 2a and 2b, you get right back where you started.

You may not realize that you are doing steps 2A and 2B because you are actually doing them simultaneously in one smooth motion. But if you consider the hinge joint of the elbow, you see that they both 2A and 2B must take place.

So if you don't twist the forearm, you get right back to your original position. However in the trick, there is an unconscious twist of the forearm so that actually causes the hand rotation. The matrix rotations don't do anything by themselves.
posted by JackFlash at 11:24 PM on November 1

Of course the forearm twists, as others have demonstrated.

Here's another way to see it: try to continue the respective movements and see what happens.

If you try and rotate the wrist a second 180° while holding it straight out, you can't because the configuration of the bones in your forearm and the way they're attached to your elbow prevents it.

Now try continuing the whole arm movements after you've done the first three that turn the hand over once. You'll find it doesn't work because you can't keep the elbow in position whilst moving the hand across your chest when the hand starts with the palm facing forwards at your shoulder: the movement is trying to rotate the elbow joint further than it can go.
posted by pharm at 1:48 AM on November 2

I see there is a lot of confusion here about whether wrists twist or what. I haven't thoroughly read all the replies here, but I thought I should point out that there are in fact two ways to get your hand to rotate on an axis normal to your chest.

One involves "twisting the wrist", as some are calling it here. Others are calling it "twisting at the elbow". They're both the same thing. Your ulna and your radius physically cross during this process. The easiest way to see this is to stick our your arm in front of you with the elbow facing up, then bend your elbow so your hand is in the air. You can then rotate your whole forearm. Unbend your arm and grab your upper arm with your other hand to keep it steady, then use this same motion to "rotate your wrist".

However, with your arm straight out in front of you, you can also rotate your wrist a little bit using different muscles. You do this by rotating your entire upper arm, which you can feel by grabbing your arm around the elbow as above and rotating your entire upper arm counterclockwise. I can't get my palm to go completely face down using this method, but my palm does rotate, and without crossing the ulna and radius bones.

The analogy to turning over a book using 6 rotations is not applicable to the human arm, because your elbow can only bend along one plane. When you rotate the book, you'll see that in steps 1 and 4, you bend the equivalent "elbow" (which would be the axis running horizontally from your viewpoint) twice, in directions that are perpendicular in the "book-fixed" frame of reference. Your body cannot do this. When you get to step 4 in Lennart Green's trick, you cheat by rotating your forearm a little bit along with your upper arm, in order to get the elbow to bend the right way. The extra rotation of your upper arm is what makes up for that "last little bit" that you can't do with your upper arm alone, as you noticed in the paragraph above.
posted by RobotNinja at 10:52 AM on November 2

Anatomically, a twist of the wrist is a pronation and is accomplished by the pronator teres and pronator quadratus muscles. This is what most people mean when they talk about twisting their wrist. However, as this trick shows, one can accomplish the same end result through a sequence of moves involving other muscles. Specifically:
 MOVE 1 - Flex the elbow using the brachialis and brachioradialis.
 MOVE 2 - Medially rotate using the deltoid and pectoralis major.
 MOVE 3 - Medially rotate using the teres major, latissmus dorsi, subscapularis, and pectoralis major.

These moves however do not directly effect pronation. There is no rotation around the axis of the forearm itself via pronator muscles.
posted by twoleftfeet at 3:49 PM on November 2

These moves however do not directly effect pronation. There is no rotation around the axis of the forearm itself via pronator muscles.

If you do not pronate the wrist, your hand ends up back exactly where it started with the palm up.

1. Flex the elbow.
2. Medially rotate. You do this by rotating the upper arm. This move is like a karate chop to your left collarbone. The palm will be up toward your face.
3. Lower the forearm to the chest. The palm is still facing up.
4. Rotate the forearm horizontally to the starting position with the palm still facing up. No rotation of the palm has taken place.

When you do the trick, you combine steps 2 and 3 into one smooth movement, but you also insert a wrist pronation.

Repeat the steps above very slowly. When you go from step 2, at the collarbone, to step 3 at the chest, instead of just lowering your arm with the palm up, twist your wrist gradually until the palm is facing the chest. That is the twist that is concealed in the trick. You just aren't aware of that twist occurring when you make moves 2 and 3 together.
posted by JackFlash at 5:48 PM on November 2

I admire your persistence, JackFlash, but I don't understand what you mean when you say:
2. Medially rotate. You do this by rotating the upper arm. This move is like a karate chop to your left collarbone. The palm will be up toward your face.
3. Lower the forearm to the chest. The palm is still facing up.

After I flex my elbow my palm is toward my face, but it isn't "up" (not the way it was "up" to start with). After I lower the forearm to the chest my palm is still facing toward me, not "up". The entire movement of the palm seems to occur within a plane parallel to my chest, with no slippage of the palm out of that plane. During this move I can certainly pronate or supinate my wrist back and forth in and out of the plane, but I don't see or feel it doing so while I'm doing the move as in the trick. I'd be happy to be convinced of your point of view because I would learn something, but at the moment I don't understand it.

If I hold a pair of straws parallel using both my hands I can put a twist in the straws by moving either my right or left hand. If the pair of straws runs from my elbow to my wrist, why can't I twist them by moving my elbow instead of my wrist?
posted by twoleftfeet at 10:18 PM on November 2

When you move your arm from the vertical position down to your chest you are pronating. You may not notice this but it is occurring. The pronator teres is activated.

This is how you can make it more noticeable. In the starting position with the palm up you are in the fully supinated position. You can feel the slight strain in your forearm because the supinator is straining to hold the palm facing upwards. (The relaxed position is halfway between fully supinated palm up and fully pronated palm down -- the handshake position). So you can feel that strain holding the palm fully supinated. After you bring the forearm up with the palm facing you, you can still feel the strain of the supinator. But then as you slowly move the hand down to your chest, notice that supinator strain goes away because you have pronated your hand to the midway (handshake) position. That is the supinator has relaxed and the pronator has slightly tensed. In other words, you have rotated the forearm using the pronator muscle.

Now extend the forearm again. The palm is facing to the left in the handshake position. There is no strain on the pronator or supinator muscles in the forearm because that is the relaxed position between fully pronated and fully supinated.

Now do the second set of motions. Raise the forearm with the palm facing left. The forearm muscles are still relaxed. But when you sweep the arm down to your chest with the palm facing down notice the slight strain in your forearm. The pronator has contracted to rotate the forearm to the fully pronated position. And when you extend the arm to the final position you can still feel that slight strain of the pronator hold the palm face down. Notice that if the pronator wasn't contracting your arm would relax to the handshake position.

In the series of motions you've gone from the forearm muscles holding the hand fully supinated, to mid-way, and finally to fully pronated. You should be able to feel the changes from contracted supinator, to relaxed supinator, to contracted pronator. These are exactly the same muscle actions you feel if you just hold your hand in front of you and flip your hand over. The exact same rotation and twisting of the bones happens in both cases. You just don't feel the twisting motions when you do the trick because it is lost in all the waving of arms.



If I hold a pair of straws parallel using both my hands I can put a twist in the straws by moving either my right or left hand. If the pair of straws runs from my elbow to my wrist, why can't I twist them by moving my elbow instead of my wrist?

You can't move the elbow end because the two forearm bones are attached to the humerus. The pronating and supinating muscles are attached to the humerous just above the elbow and to the radius and ulna near the wrist on the other. So those muscles can only twist the wrist end where the hand is free to rotate. The muscles can't twist the elbow end of the radius and ulna because they don't attach to it. The rotational motion of the wrist is relative to fixed muscle ends on the humerus
posted by JackFlash at 1:49 AM on November 3