Monday, October 12, 2009

A Prehistory of Throwing Things

A long post in honor of Major League Baseball’s upcoming World Series…

Music, language, and calculus have always kept us atop the animal kingdom, but what about baseball?

Before there could be baseball, people had to get good at throwing things. The evolution of throwing ability has been the subject of much more academic speculation than many Yankees fans may realize.

Back when people were still playing “rounders,” Chuck Darwin proposed that throwing may have developed for hunting, self-defense or both, by permitting action at a distance. Walking upright frees the hands to do all sorts of sinister things, leading others to speculate that we throw as a means to punish, as a way to perpetuate peace. And, others have credited the evolution of throwing with brain enlargement, language, and human intelligence. (Yes, intelligence.)

There is real genius in baseball.

Dude throws like a chimp
Besides being the only creatures to calculate batting averages and grow playoff beards, humans are the only animals that can throw accurately and fast with one hand overhead in such a complex motion. Some non-human primates have been known to throw on occasion in the wild. Those who have bravely traipsed through the rain forests of Costa Rica learn this all too well if they’ve been used for target practice by feces-flinging spider monkeys.

Probably victimized occasionally herself, Jane Goodall observed throwing behavior in the chimpanzees at Gombe. She found that chimps threw many objects, including rocks and sticks with both underarm and overarm, and one and two-handed styles. They could throw towards their targets, but they rarely hit them.

No chimp or any other great ape has been observed to throw with the calculated skill or accuracy of a human. If you’ve seen footage of chimps throwing you recognize right away that they look like toddlers when they do it, but toddlers that could bench 400 pounds. This has to do with their different musculoskeletal anatomy and motor control, and both are reasons why throwing is not implemented in any of their hunting forays. Instead, for chimps, throwing things is just a way to threaten other chimpanzees, baboons, bushpigs, or humans, with no real effort to maim or kill their target.

So if chimps are our guide for our earliest ancestors, then baseball didn’t emerge from a hunting history so much as it descended from a display tactic. Under this hypothesis, throwing was originally used to parade strength and skill to enemies or symbolized health and vigor and even intelligence to potential mates (without need for dagger-like canines). This same hypothesis has been put forward for why we walk upright as well and these traits can be linked together considering how it helps to be a balanced biped in order to throw well.

The littlest league
If our closest living relatives, chimpanzees, cannot throw overarm-style efficiently or accurately, then why and how did humans develop the ability to throw? What was the evolutionary pressure for throwing behavior in its earliest stages? I already mentioned the display and threat hypothesis, but accuracy had to develop at some point and greater throwing ability was probably linked to increasing the amount of meat in the diet.

The earliest hominins probably took breaks from their mostly vegetarian diets to hunt prey, as chimps do now. They likely chose species that can be seized with their bare hands or with the aid of handheld tools, like termite fishing sticks and spears for killing tree-dwelling bushbabies. Once regular meat-eating became important, hominins were scavengers before they were great hunters and this is probably when throwing ability started to surpass that of the chimpanzee.

Throwing stones or clods of earth, as a means of defense or aggression against larger carnivores, could have compensated for early hominin vulnerability – not just when preyed upon but also when directly competing with predators and other scavengers.

Imagine yourself, but about 2.5 million years ago, with a brain that has just started getting bigger than a chimpanzee’s, fighting lions or hyenas for access to a gazelle carcass, standing barely four feet tall, with no claws, no formidable canine teeth, no tools more sophisticated than bashed rocks, and possibly no thick coat of fur for protection. On top of all that, this may have been right around the time when the human lineage started to become physically weaker than chimpanzees are now, since there may be a neurological tradeoff between brains and brawn. Throwing things would have been an ingenious way to scare fangorious competitors away from dinner.

It also may have been a novel way to rebuke or punish cheaters without risking as much as in hand-to-hand combat. Like shepherds who throw rocks to steer wandering sheep, throwing could have initially been a way to keep your family, friends and neighbors on the straight-and-narrow. When Fenway revelers hurl hotdogs at rival fans, they are making their point at a much lower risk to their own personal safety than if they threw knuckle sandwiches instead.

This is the basis behind Paul Bingham’s idea of coalitional enforcement, where throwing in groups against cheaters of the system (as in stoning them) provided the mechanism for us to succeed at living rather peacefully within large cooperative groups.

Like in the scavenging scenario, throwing with a group to punish cheaters or criminals would not require much accuracy for our early hominin ancestors. They needed to merely lob with decent accuracy at best, and the need for much accuracy would be diminished if groups were cooperating in the throwing activity. But with hunting, accuracy is far more important. As time progressed, hominin hunting abilities grew increasingly more effective and were probably very well established as early as 1.8 million years ago when the much taller Homo erectus showed up on the evolutionary scene. This is when throwing with speed and accuracy would have been even more crucial to survival because hunting was now a way of life.

Pitcher’s got a big gluteus maximus
There were plenty of anatomical adaptations that occurred in our evolutionary history that made fast, accurate throwing possible. Since we evolved bipedalism our hands have gotten much better at manipulating objects. We have short, straight fingers and we are capable of various precise and strong grips.

Surely our hand anatomy has been crucial for throwing ability but there are few hands in the hominin fossil record and it is thus far impossible to separate out bony evidence for throwing abilities from stone-tool making abilities. Clearly, making and using stone tools was important and, clearly, throwing was a “go” once we could make stone tools. Generally this is agreed upon, by both anatomical and stone tool evidence appearing, by about 2.5 million years ago.

Another equally important trait for throwing, which also changed after we began walking upright, is the bulbous human buttocks. The human-like condition of the gluteus maximus muscle differs from chimpanzees in its size and mechanical functions. In throwing, the activation of the gluteus maximus can increase the velocity on a projectile and stabilize the trunk over the legs for effectively completing the motion. Although muscles don’t fossilize, their locations of attachment, their biomechanical properties, and their sizes can be estimated from fossil bones and it is clear that by 2 million years ago, human bodies, especially human tookuses, were built for walking and running upright, for swinging tools, and for throwing.

Bringing the heat to the Ice Ages
Once you’ve got a hand that can grip the ball, and once you’ve got enough junk-in-the-trunk to make cheese, are you all set to pitch against a roided-up bat-corker? Not yet.

Ultimately it’s the modifications in arm length and shoulder anatomy (and brains, below) that greatly affected throwing and those changes lagged quite a bit behind the hands and glutes.

Anatomically, things get complicated at about 2.5 million years ago. Although the body seems to have all the right stuff for walking upright, there’s a little stint between 2.5 and 1.8 million years ago when small-brained hominins had achieved modern bodies but kept their longish, apish arms, particularly in the forearm.

Sometimes this evolutionary baggage - which is even seen in Lucy over 3 million years ago - is used to argue that hominins were still partly arboreal even after adopting bipedalism. The long arms would have affected throwing ability since long arms can throw a projectile further than short ones. A long forearm is capable of putting more velocity on a projectile, similar to the way atlatls or spear-throwers do so by artificially lengthening the arm.

Even with this potential, early hominins must have lacked accuracy due to their small brains (i.e. less complex neural circuitry and more primitive motor control). However, they may have been able to compensate. A long forearm has a larger range of angular velocities than a shorter forearm; that is, it is moving faster at more limb positions, or angles, than a shorter forearm. Therefore a longer forearm has a “bigger window of opportunity” to release a projectile and hit the target in arcing, lobbed throws.

With little brain power, but with a long arm on a bipedal body, these little hominins would have been better hurlers than chimpanzees. But Major League scouts need not set their flux capacitors for the early Pleistocene. Homo sapiens are much better baseball prospects because we can throw much further and more accurately than our ancestors with the aid of our shorter forearms and our bigger brains. Accuracy is increased with a shorter forearm which, with the right motor control, can tolerate relatively small errors in timing better than longer ones in flinging fastballs dead-on target.

Beginning around 1.8 million years ago, with Homo erectus, hominins spread out beyond Africa and they began to look and behave like actual predators compared to any of their predecessors. When throwing was assimilated into hunting strategies, speed and accuracy became essential because a moving animal can react to the arrival of the projectile. Hunting requires the use of a flattened trajectory throw that is more direct than an arced throw and requires more accuracy.

The brain does enlarge with Homo erectus, but it is still only half to three-quarters the size of ours, so motor control would not have been stellar. Plus, although the arm proportions look modern at this point, the clavicle may still be too short for full human throwing ability. Homo erectus did not have quite as broad shoulders, with the shoulder blades far on the back, as us and these features are necessary to throw like us. When the shoulders are placed far on the back, and when the shoulders are broad enough, they allow a greater range of motion for the full throwing motion. So the first hunters were not anatomically equipped to throw like humans yet.

Modern shoulder anatomy doesn’t emerge until much later, at about 500,000 years ago with what are termed “archaic” humans. Coincidentally, at the point in time when shoulders are finally geared for throwing is the same time when brains reach modern size. Not surprisingly this is when hominins must have been excellent hunters, made obvious by the sophisticated weapons, culinary ware, and dinner leftovers they left behind.

There’s no thinking in baseball
Thinking through a throw is the surefire way to ruin it. It doesn’t take big league pitching experience to understand this point. Anyone who has tossed a wad of paper into the office bin or who has seen Bull Durham understands. This does not mean, however, that the brain is not crucial for throwing.

The cascade of actions that comprise a throw, from start to finish, cannot be changed once they begin. Because the motion is too fast for feedback to reach the brain during the event, the brain plans out the throwing motion before it is executed and then fine-tuned timings as well as momentum carry the action through until the very last moment of finger release. Despite our inability to improve a throw by thinking it through, throwing is a fantastically cerebral activity.

William Calvin proposed that hominins with bigger-than-average brains might have been able to apply more timing neurons to throwing tasks - with the success (hunting or otherwise) of the faster, more accurate throws then selecting for encephalization trends. In other words, he suggests that our big brains should thank our penchant for throwing things for making them that way.

So easy a caveman could do it?
Neanderthals, the quintessential cavemen and women, had anatomy like ours and even had brains just as big as, and sometimes bigger than, ours, but were our evolutionary cousins throwing like us? Debate continues as to whether or not these guys who sometimes lived in very close proximity to humans were actually our occasional bedmates or our enemies. People siding with the former probably root for both the Mets and the Yankees.

There are two main lines of evidence that suggest Neanderthals weren’t big on throwing as a hunting strategy. First of all, many of their fossils have healed bone fractures and the overall pattern of trauma resembles that seen in professional rodeo athletes. Animals weren’t domesticated at this point yet so the best explanation, far better than clumsiness, is that Neanderthals were coming into very close contact with large dangerous prey, like reindeer and mammoth. The thickness and strength of their arm bones also fits better with habitual thrusting of spears rather than throwing.

For now, throwing things, as well as we can today, should be considered a very recent phenomenon, something beginning around 500,000 years ago with archaic humans, after nearly five million years of evolutionary time since the split from the common ancestor we share with chimpanzees.

Most things that make humans strange or fantastic are rooted in either our bipedal, upright-walking anatomy or in the power of our humongous brains. Getting good at throwing took the evolution of both of those traits. But after our bodies mastered bipedalism there was up to one million years of lag time before the brain got to modern size and before the shoulders were tweaked for exquisite throwing. Until this point, hominin throwers may have looked no better than chimpanzees or than Andy Pettitte using his right hand, assuming he’s pretty bad at that.

Throwing ability requires some of the highest levels of brain functioning - There is real genius in baseball.

Players and fans owe a big debt of gratitude to the apeman and the caveman in all of us.

Short Bibliography
Bingham, Paul M. 1999. Human Uniqueness: A General Theory. The Quarterly Review of Biology 74(2):133-169.

Calvin, William H. 1983b. The Throwing Madonna: Essays on the Brain. New York: Bantam.

Darwin, Charles. 1871. Descent of Man. London: John Murray.

Dunsworth, Holly, John Challis, and Alan Walker. 2003. Throwing and bipedalism: A new look at an old idea. In Franzen JL, Dohler M, Moya-Sola S (editors). Upright Walking. Senckenberg Institute, Frankfurt, pp. 105-110.

Goodall, Jane. 1986. The Chimpanzees of Gombe: Patterns of Behavior. Cambridge: Harvard University Press.

Larson, Susan G. et al. 2007.
Homo floresiensis and the evolution of the hominin shoulder. Journal of Human Evolution 53(6): 718-731.

Walker, Alan. 2009. The strength of great apes and the speed of humans. Current Anthropology 50(2): 229-234.

Note: If you liked this, you'll love "Hurling words and turds, an evolutionary link" and you might like "Can you throw with half a brain?" too.

Update: This post has been cited in "How high can a human throw something?"


Anne Buchanan said...

Nice, Holly! And it brings to mind the old joke: I don't know if I can play the violin, I haven't tried it yet.

Holly Dunsworth said...

Thanks Anne!

Author's note: I have no idea if likening a chimp's strength to a toddler benching 400 pounds is anywhere near the right ballpark (heheh). But the point is, they're really strong.

Jason Hodgson said...

Holly, this was a fun and informative entry. It was especially nice to read after the Yankees swept their way into (and the Redsox swept their way out of) the ALCS! GO YANKEES!!!

I find two things remarkable about throwing. First, throwing is so much fun! Second, we are remarkably variable in our ability to throw.

On the first point, I absolutely love throwing things. I would throw a baseball all day long if I could. And I always keep a ball near by to throw off the wall when I get tired of my actual work (I've a blue wiffle ball and a rubber bouncy ball next to my computer right now). I'm certainly not the only one who enjoys throwing either. There are whole industries devoted to throwing things. Walk through Central Park on any given day and there will be throngs of people throwing things. It is not just that we have evolved to be good at throwing things, we seem to have evolved to really love doing it!

On the second point, we all know that the end result of natural selection is a reduction in genetic variability and ultimately phenotypic variability. Holly makes a strong argument that throwing ability was selected for in our lineage. Yet there seems to be a lot of variation in throwing ability left in our species.

Major League pitchers are certainly the extreme right tail of the throwing ability distribution. Yet even within this highly selected group there are pitchers like Barry Zito who's average fastball is about 83 mph and pitchers like Justin Verlander who's average fastball is about 97 mph. That is a big difference! If we were to look across the population as a whole the range in top speed probably ranges from about 35 mph to about 102 mph. That is a massive difference!

Clearly a lot of the variation in ability in the general populace can be accounted for with different childhood experience and learning. Probably none of the variability seen in major league pitchers is learned though. They are all probably maximizing what their bodies can actually do.

So, why so much variability in a selected trait? I am not trying to argue that throwing was not selected for. After all, there is still a lot of variability in foot arches though these must have been selected for. Rather I think it says more about the nature of selection in most traits.

Most traits that we are interested in - and I imagine throwing is one of them - are polygenic. Lots of genes contribute to the phenotype. Selection then probably affects many loci around the genome and the selection is probably weak at any one loci because it is probably only combinations of alleles that segregate independently that are beneficial. Selection for a given allele might then be one step forward and two steps drift depending upon the combination of alleles in any given individual. Thus variation persists.

Holly Dunsworth said...

Thanks Jason and great point about loving it!!! Making it fun certainly must have helped us get good at it... thanks Mother Nature for inventing happiness!

On variation you provide some nice insights too. Since it's not just various musculoskeletal features, but also muscle biology AND brains... I'd say it's very complicated. But it's also something you've got to do since childhood to be any good at. I dare say it's almost like language... if you don't learn at an early age, and then keep up the skill, you don't throw too well later in life. People who study the differences in boys' and girls' abilities and techniques (e.g. "throwing like a girl") have argued that it's just a matter of practicing early during development and more boys than girls actually do that, hence more of them have the tell-tale motion which is probably better for throwing.

Ken Weiss said...

I have three points to make:
1. First, and foremost, and the hell with what Jason thinks: Damn Yankees!

2. Just because we evolved an ability to throw doesn't by itself imply that we evolved _to_throw. That is, doesn't mean that our ancestors who could throw had more kids than those who couldn't. It's easily imaginable that that _could_ have been the case but it's a Just-So story in terms of direct evidence.

For example, it could be that our ability to throw evolved in some other context but that those who could throw used that ability and,say, hunted with spears. That's the story with the violin and calculus: we evolved a capacity for some reason that, once we have it, we use it.

3. There can still be variation today if many genes contribute to our ability to throw, as must be the case. Why? Because if there are many genes, there can be a mutation-selection balance that keeps variation around, but keeps it constrained. Each generation the total throwing-incompetents had fewer children than the accurate spearmen, but mutations continued to introduce new variation.

That, too, is a Just-So scenario, but is at least within general population genetics theory.

4. I repeat: Damn Yankees!

Jason Hodgson said...

Reasonable people might think that Ken had delivered a shit sandwich, but it is actually a good sandwich with really shitty bread (points 1 and 4). Anyone that was turned off by the sacrilege of Ken's point 1 really ought to persevere.

Evolving to throw is a just-so story. This doesn't mean its incorrect, it just means we need good science to test the hypothesis. After all, some percentage of just-so stories must actually be just so. The hard (and fun) part is thinking of ways to test these hypotheses!


Holly Dunsworth said...

Go Pirates!!! (looks at shoes)

Ken Weiss said...

I'll agree with Jason, who is very misguided in other ways, about Just-So stories. They may be true, in whole or in part, but they aren't true just because some Yankee fan says they are.

As to the Pirates....what Pirates?

Ken Weiss said...

Actually, I got so distracted trying to respond in a civil manner to a Yankee fan that I neglected to say that one reason they're called Just-So stories is that they're damnded hard to prove, by any clear criterion of 'proof'.

If, for example, the selective advantage of a slightly more throw-worthy thumb or muscle-fibre structure may only be very small--estimates often are .1% or less for (average) selective values.

When that's the case, it's very problematic even to assert that one thing, such as 'throwing', can account for it, and what would the evidence be? At any given time, perhaps at _every_ given time, samples would not have been adequate to demonstrate the effect.

There are other possibilities, such as occasionally strong, or very local selection, or some functional tests. But usually multiple explanations, or perhaps in reality multiple kinds of cause, are plausible.

Perhaps the wisest thing is to be careful to label hypotheses as such and don't over-state them, even if a newspaper reporter might be listening.

Jason Hodgson said...

What to do about just-so stories?

I became interested in evolutionary anthropology because the just-so stories of human behavioral ecology (or sociobiology, or evolutionary psychology, whatever you want to call it) are so deceptively explanatory. They are like candy. You start to think you really understand human behavior and can predict what people will do given the situation. But, of course you can't. And you realize that for every just-so story there is a just-so explanation for why it is wrong.

I think I gravitated to genetics from human behavior because hypothesis testing in genetics is usually straight forward and because if behavior is adaptive we will eventually find genetic evidence for it. Hypothesis testing becomes easier and easier the more reductionist you are as a scientist. But the hypotheses you're testing explain less and less.

A well-supported just-so story would explain much of what I am really interested in. I hope we can get at them some day! Perhaps we just need to understand all the little bits involved better before we can put it all back together again.

...and what's wrong with the Pirates? They are the great altruists of MLB. They always provide their talent to the rest of the league with no expectation of a return!

Ken Weiss said...

I grew up following baseball's ultimate altruists, the old Washington Senators. Whenever any of their players got any good, they would trade them to the (damn) Yankees or the Bosox. Kansas City (then the A's) were like that, too. And I agree about the Bucs.

As to supporting Just-So stories, I agree with what you say. The problem will be especially tough if, as I think will often be the case, traits of interest--say, tool-useful thumbs--develop using a host of signaling genes in the embryo, whose effects are individually very small. Then you have a polygenic trait and little way to identify the individual contributing genes. Statue is an excellent exemplar of that problem.

Unlike astronomy where, at least according to some, the past can be post-extrapolated theoretically from the present, evolution seems less theory-rigorous for that (too much stochasticity, for example). So it's a challenge.

Finding DNA signals of selection, regardless of what the selection was 'for', is notoriously difficult.

Unknown said...

Excellent article. Other than darts and javelins; most modern sports involve throwing rock-like balls. This allows killing, maiming, and threatening at a safe distance. And is why we duck and blink to protect our vision if a rock is heading towards you. It is so instinctive that even the TV Superman flinched when a gun was going to be thrown at him in the opening credits; the same gun whose bullets he sneered at. And the audience didn't even realize the absurdity and importance of it.

A curve ball could fool prey into thinking it would pass behind them and miss them. This might cause them to slow down and be hit. Hunters who place seeds on one side of a hill or blind could throw large spheres over the hill (barely clearing it) using two hands like a foul shooter. Birds or small mammals would have little warning and even a "minor" injury to wing or leg would prevent the prey from escaping.
When a catcher stands up, he becomes bi-pedal and able to pick off a runner. Our widespread inherited baseball talent has little practical value today. Therefore, it must have been a genetic advantage long ago.

Holly Dunsworth said...

Thanks so much for these thoughts.

I'd argue that there's still lots of practical value. Look how much social learning takes place on the athletic fields and playgrounds!!

Steve G said...

I agree that throwing may have been very important early in the evolution of humans. In fact, I've built a web site about this. Your readers might be interested --

Holly Dunsworth said...

Thanks. And I see Nature Ed was useful--great! You know, you can cite Pontzer for that figure just like any other article you cite, because that's what it is: a peer-reviewed article. Thanks!

Holly Dunsworth said...

You might also want to check out Neil Roach's work (I wrote about it here:

Ronsonic said...

Yeah, I found this by way of XKCD too.

One aspect that didn't get mentioned in this article and being I'm not any sort of anthropologist (aside from having lived among the species for decades) is the relationship between throwing and "normal" tool use.

The ability to precisely strike off a flake of flint is not unrelated to the ability to throw the spear you've lashed it to. I haven't worked much flint, but I drive nails and the same neural and physical skills apply to throwing that hammer as striking something with it. The guy who can stand at one end of a sledge hammer and drive a spike in the ground has the neurology and dexterity to throw.

Watch any carpenter drive nails with a single blow. It's the same thing, just with less shoulder. Same eye-brain-hand action. The work I do is much finer than that, but the same evolutionary features are required.

If I had to find the birth of the human arm as a throwing implement based on the archeological record, I'd look for tools and techniques that involved working at a remove from the tool or work piece. Hafted hammers and axes instead of hand-held and precisely shaped flints rather than rocks beaten into shape.

I would be completely unsurprised if adaptation to tool use was the link between primate flinging and our precision arm.