How a Spearthrower Throws a Spear.


John Whittaker

  Feb 2002



Throwing a dart with an atlatl is a complex sequence of motions. It is difficult to visualize and explain exactly what happens, and a number of fallacies and odd theories have made their way into the literature. Unlike pitching or tennis, which have been studied for years, no one has looked at the details with a concern for teaching new atlatlists and improving the skills of experienced throwers, although many individuals within the sport have come up with tips that they find helpful.


With the help of modern photography, we can study the motion and begin to make some practical suggestions. This is an ongoing project of myself and Chuck Hilton in the Department of Anthropology at Grinnell College. We have two main motivations: 1) The throwing motions in using an atlatl are potentially important in human physiology and motion studies (Chuck’s area of physical anthropology) and of course relevant to understanding the use of an important early hunting technology (my archaeological orientation). 2) I also teach students to use atlatls and take them to competitions, boasting that Grinnell has the worlds first collegiate atlatl team. The teaching purpose behind this is to expose people to prehistoric technology both for fun and for the purposes of experimental archaeology.


The following document illustrates some of our initial ideas from several sources. 1) My explanation of the throwing motion comes from examining slow motion video tapes Chuck made of me throwing. The photos illustrating it were taken by Jeff Lindow at Fort Osage, May 2000. If you scroll down through them real fast, you get a sort of motion picture. 2) These are followed by selected photos cribbed from the World Atlatl Association webpage albums of atlatl competitions and other events, and illustrate what I consider good throwing form, and some of the common mistakes.




I think in terms of four phases. From a balanced aiming posture, a throw begins with 1) a step, then 2) the body rotates and the arm and shoulder begin to move the atlatl and dart, 3) the wrist snaps to provide the leverage of the atlatl, and 4) you follow through. All together, a throw takes about one second from balanced beginning to follow through, although we have not yet been able to time it accurately or break down the timing of the different phases of the throw. That’s the next step when we get better recording and analysis equipment.



1) The step. I start with a standing posture, feet close together, and left foot forward, leaning back very slightly (less than I would have thought), with my arm cocked back , body turned not quite 90 degrees from target so that my left arm can be raised and pointed at it. The dart is level or a bit above horizontal, at or above eye level depending on range. I visually align it with the target, even though I can’t actually sight down it like a gun barrel.







The throw begins visibly with a slight bend of the left knee as I rock back fractionally (first photo), then bring the left foot forward in a full step, which brings body, arm, and dart forward, but without moving arm or rotating torso until the full step is complete, with the left foot flat or almost on the ground.







2) Arm and body.

As the step is completed, the torso begins to rotate and the upper arm to flex at the shoulder, bringing the hand and the atlatl forward until it is about even with the back of the head.







The atlatl throughout this remains horizontal. The shoulder flexion seems to me to be small at this point, and the wrist must be rotating to keep the dart pointed at the target.


3) The Wrist Snap. Slightly before the hand reaches the back of the head, the hand and forearm begin to rise.




Then at the point when the hand passes the head as the torso continues to rotate, the wrist must rotate back in the opposite direction from its previous rotation, and then flex violently, swinging the atlatl up to vertical to flick the dart away.

This is what flexes the dart as the point remains aimed at target, while the nock is rapidly raised by the atlatl.








At the same time, the arm is extended straight out. [The wrist motion seems essentially the same as that in cocking and throwing a ball, with the only real difference being that the fingers remain closed to grip the atlatl.] Note how high above the head the dart is as it is just about to leave the atlatl.








With the atlatl in vertical position, the dart has recovered from its initial flexing, and is about to spring away from the atlatl and flex in the opposite direction.



4) Follow-through.

As the dart leaves the atlatl with the atlatl vertical or slightly past, the wrist continues to flex, as does the shoulder, and I bend forward and swing the right arm and atlatl down and across my body, ending outside my left leg.







My chin remains up, and my head at almost the same level throughout the throw, with my eyes fixed on the target (this is conscious “good form”).












1) Atlatl motion. Calvin Howard, in a 1974 article “Atlatl: Function and Performance” in American Antiquity 39(1), says “the atlatl is not a catapult or flipping device. During a proper throw, the spur reaches no greater elevation than that of the handle.” He thinks hook on a level atlatl simply remains in contact and delivers thrust longer than a hand throwing the same spear with same motion. Several other people have the same idea, and it is plainly and unequivocably wrong. Both my photos, and ethnographic photos show that it is the flip of a high, vertical atlatl that provides the lever action and gives the atlatl its force. It is true that the atlatl moves horizontally for most of the throw, with only a very fast snap of wrist bringing it vertical. During this horizontal part of the throw, the step and the rotation of the upper body are starting the forward motion of the atlatl and dart.

Howard also says “Hooking results when thrower fails to keep the atlatl level during the thrust.” This implies to me that he didn’t have an effective hook/nock system – and that may be why he misunderstood how the atlatl works.


2) Dart and atlatl flex. A lot of people feel that the flex of the atlatl and the dart is very important, and provides much of the power of a throw by storing energy like a spring. In the photos above, it can be seen that the dart does flex a lot, but although my atlatl is somewhat flexible, little if any flex is visible in the throw. There are of course lots of ethnographic atlatls that are completely rigid – atlatl flex adds almost nothing to the power of the throw.

            The dart, however, does flex a great deal. This is because the dart stays horizontal until the wrist snap part of the throw, when on our video film it begins to point upward at a slight angle. During the wrist snap is also when the dart flexes, and the snap is a very short, fast, motion, so it must be at this time that most of the energy is imparted to the dart. As the hand rises, and especially as the atlatl begins to move to the vertical, the dart flexes. At the top of the throw you can see that the fletched end is flexed upward and the center bending downward. The next photo shows a straight dart about to spring away from the atlatl, and the photo after that has the dart flexing in the opposite direction (tip and feathers down). I think the flex is largely caused by the sudden vertical movement of the hook of the atlatl with the tail of the dart, while the dart is kept on target, and flexes to compensate. This is why a rigid dart doesn’t work. Our observations of the video seem to confirm this – the dart point and forward 2/3 stay in a straight line and in the same position on the tape, while the tail of the dart is moved upward. Once the dart leaves the atlatl it probably oscillates back and forth rapidly for a bit until it stabilizes, but we have not yet filmed the path to the target, so we don’t know how many flexions it makes and have not timed the beat.

            Although the dart does flex, this is necessary for accuracy, not for power. The spring-like flex of the dart adds almost no forward motion. If you don’t believe this, try flexing a dart by pressing it against the ground and then suddenly releasing it. It will jump into the air very little. Similarly, you can try nocking a dart against a fixed atlatl and flexing only the atlatl to spring the dart away – that too will not get you very far. Conclusions: flex in the atlatl is of no importance, flex in the dart is necessary only for an accurate throw.


Further Examples of Form, Mostly Bad, taken from the WAA webpage albums.

My apologies to anyone who might feel ridiculed by my comments – I mostly don’t know the individuals, so I don’t know their skills, or the outcome of the pictured throw, so I am only evaluating the apparent form in terms of what I find works and what I tell people to do or to avoid.


Start with good normal form:


14th Annual WAA, Flint Ridge, Ohio, 2001. A good solid stance, atlatl high, dart aligned with the target but above the head.








Spring Fling, Ligonier, Pennsylvania 2001 Terry Keefer shows good normal motion, beginning follow through as dart leaves the atlatl with strong flick of the wrist. Note that dart and atlatl are above head level, and he has rotated his upper body for power.








Mas d’Azil 2001 Now compare this one. Atlatl is too low, arm is way down, making it hard to get enough wrist snap leverage. No upper body rotation also reduces possible power. This seems to be common in the photos – I wonder if the atlatlists aren’t trying to sight along the dart and keep it at eye level.








Mariemont, Belgium 2001 Terrible! He’s sidearming it, which makes the flex very hard to control. Now not only is it hard to gauge the arc and drop of the dart on its way to target, you also lose the advantage of being able to visually align the dart with the target and keep it in line. His right foot behind him is on its side – it looks like he pushed off strongly, but now he’s off balance and resting only on front foot, so his weight and momentum have been stopped dead or even pushed back instead of flowing forward.
















Mariemont Belgium 2002 Avoid leaning forward too early, it tends to send the dart low. Get the arm up in a full swing for power – the elbow is too low here.









Montana Mammoth Hunt 2001 Leaning badly into the follow-through – the dart probably went down and left. Right handers tend to throw down and left if they lean, or wide right if they let the dart swing as they cock their arm.







Slope-A-Dope , Wintersville, Ohio 2001 Whee, looks like fun, but probably not accurate. Can’t be consistent while jumping. Maintain balance at all times.













14th Annual WAA meeting, Flint Ridge, Ohio 2001 A large step can add power, but it’s hard to remain consistent and balanced with this much.





14th Annual WAA meeting, 2001 Chuck Bujorak’s unusual hair style is quite irrelevant, but legendary.









Spring Fling, Ligonier PA 2001 Roy Strischek has an unusual wide stance, uses mostly arm motion, with relatively slow lobbing throw. I don’t recommend it, but he is very accurate with it. His atlatls are also distinctive, with a series of pegs in the grips and a V rest for the dart.






Gary Fogelman, also a world champ, has a normal stance and throwing style.

Note that in aiming, his dart is above eye level.
















14th Annual WAA meeting, Flint Ridge, Ohio 2001 Terry Keefer knows what he’s doing. Normal motion with smooth follow through, shoulder rotation, not distracted by muddy ground on bare feet.










Roma, Italy 2001 Compare this to Keefer above. A very static throw, almost no body motion evident – no foot forward, no upper body rotation. He probably had very little power.










Ramioul, Belgium Christian Casseyasa well balanced on a rock. He can’t lean too far forward! Note that atlatl and dart are well above head. It also looks like a relatively long atlatl, and Europeans also favor long darts.








Mas d’Azil, France 2001 Wimpy! Very low arm, hand is forward but hasn’t started wrist snap to elevate the atlatl and get some leverage. This shot couldn’t have gone very far, although at least the shoulders are coming around.








Mas d’Azil, France 2001 Good normal follow through. Note that although he has bent slightly at the waist, his head is level and eyes on the target.











Tautavel, France Feb 2002. Ooh La La! What in the world is she doing ? Trying too hard perhaps? Watched too much American baseball? In the next frame, does she fall on her face? Apparently not, since there was a similar photo of another shot with her distinctive style. Since I don’t know her, I can’t tell you whether she is any good. I would be surprised if she hit the target, and I hope not, because folk who do things all wrong but still are good at them (Louis Armstrong on trumpet comes to mind) make coaching hard. Just because they make bad form work, doesn’t mean you should try it.