William H. Calvin, "Rediscovery and the cognitive aspects of toolmaking:  Lessons from the handaxe.
 Behavioral and Brain Sciences. 25(3):403-404.  See also
 Webbed Reprint Collection
This 'tree' is really a pyramidal neuron of cerebral cortex.  The axon exiting at bottom goes long distances, eventually splitting up into 10,000 small branchlets to make synapses with other brain cells.
William H. Calvin

University of Washington
Seattle WA 98195-1800 USA

Rediscovery and the cognitive aspects of toolmaking:  Lessons from the handaxe.


William H. Calvin


Abstract.  Long before signs of staged toolmaking appeared, Homo erectus made symmetrical tools.  The handaxe is a flattened tear-drop shape, but often with edges sharpened all around.  Before we assign their obsession with symmetry to an esthetic judgment, we must consider whether it is possible that the symmetry is simply very pragmatic for one particular use of the many suggested.


I like Wynn’s analysis (2002) but let me play devil’s advocate for a thousand words and consider whether the biface’s symmetry was initially pragmatic – and only developed into an esthetic after a million years of proving its usefulness.

With its flattened-teardrop symmetry, the Achulean handaxe has long invited cognitive explanations.  It is the earliest hominid tool that seems “designed” in some modern sense.  Yet for most of the “Swiss Army knife” multipurpose suite of proposed uses (defleshing, scraping, pounding roots, and flake source), an easy-to-make shape would suffice - and indeed the simpler tools continued to be made.  None of these uses adequately addresses the “design aspects.”  Why is the handaxe mostly symmetric, why mostly flattened, why the seldom-sharp point, why sharpened all around (when that interferes with gripping the tool for pounding uses)?

Nor does a suite of uses suggest why this form could remain the same from southern Africa to northern Europe to eastern Asia – and resist cultural drift for so long.  The handaxe technique and its rationale were surely lost many times, just as Tasmanians lost fishing and fire-starting practices.  So how did Homo erectus keep rediscovering the enigmatic handaxe shape, over and over for nearly 1.5 million years?  Was there a constraining primary function, in addition to a Swiss-Army-knife collection of secondary uses?

In Calvin (1993, expanded in 2002), I describe the handaxe’s extraordinary suitability for one special-purpose case of projectile predation:  attacking herds at waterholes on those occasions when they are tightly packed together and present a large, stampede-prone target.  Briefly, in the beginner’s version that uses a tree branch rather than a handaxe, the hunters hide near a waterhole.  When the herd is within range, the branch is flung into their midst.  The lob causes the herd to wheel about and begin to stampede.  But some animal trips or becomes entangled by the branch.  Because of jostling and injury by others as they flee, the animal fails to get up before hunters arrive to dispatch it.

Chimpanzees often threaten by waving and flinging branches but, if such are not handy, they will toss rocks or even clumps of dirt in the same general direction.  One can imagine that tree branches were soon in short supply near waterholes.  If our waterhole hominids resorted to second best, lobbing a rock into the herd’s midst, it would not trip animals but it might knock one down.  Because of the delaying action of the stampeding herd, this too might allow an animal to be caught.  Even when you miss, the herd will be more tightly packed together on its next cautious visit to the water’s edge.

What rocks would work best?  Large rocks, but also rocks whose shape had less air resistance.  Most rocks tumble, but flat rocks (say, from a shale outcrop) will sometimes rotate in the style of a discus or frisbee, keeping the thin profile aligned to the direction of travel and thereby minimizing drag.  Because approach distance will increase with heavy predation, range would become important.

Hunters might also have noticed that stones with sharp edges were more effective in knocking an animal off its feet, even when not heavy.  Withdrawal reflexes from painful stimuli, such as a sharp prick from an overhanging thorn tree, cause a four-legged animal to involuntarily squat.  Even if the spinning stone were to hit atop the animal’s back and bounce free, it might cause the animal to sit down.  It is the sudden pain which is relevant, not any actual penetration of the skin.

Handaxes, whether thrown by amateurs or experts, whether lobbed or thrown more directly, usually turn into vertical-plane spinners.  Unlike a frisbee which rolls endlessly after landing, handaxes rotate to bury their point and abruptly halt.  If the point momentarily snagged on a pushed-up roll of skin, it would both augment the pain and transfer all of its momentum to the animal, pushing it sideways.  Ordinarily, righting reflexes would catch the animal before it toppled, but a simultaneous sit-down withdrawal reflex can override this customary protection.

So this is a beginner’s technique for a commonplace high-payoff situation, not a general-purpose hunting technique (it strongly depends on a herd-sized target and the consequent stampede).  This proposed path of discovery would also work well in cases of loss of shaping technique, promoting a return to flattened rocks with an all-around edge and something of a point.

Consider also the “life history” of a handaxe.  Some new sharp ones would be lost in the mud.  Of the ones retrieved, some would have been trampled.  A broken classic handaxe may make an excellent cleaver, now having a grip that no longer bites the hand that holds it.  Many lost handaxes would be tumbled by a flood and then later discovered in the river bed, with some edges smoothed enough to hold comfort­ably.  So (notwithstanding Whittaker & McCall 2001) I see the shape-defining use as special-purpose, but with broken and tumbled handaxes having many secondary uses, including the “Swiss Army Knife suite.”

Channel-cutting floods even set up rediscovery of the best shape by the clueless of a lost generation.  In watercourses where the animals come to drink, some of the easily-grabbed stones throw farther than others and have better knock-down properties.  By the time that this objet trouvé supply is exhausted, toolmakers know what the most effective shape is, from having recycled some lost handaxes.

Clearly an ability to imagine a series of blades in a prepared core was present 50,000 years ago - and equally clearly, little cognitive ability was needed 2.5 million years ago for Glynn Isaac’s shatter-and-search method for producing the sharp split cobbles.  The latter suffice for getting through the skin and amputating limbs at a joint before the competition arrives; they also allow the limb to be swung clublike against tree trunks to produce spiral fractures and extract marrow.  Indeed, shatter-and-search and the handaxe together largely solve the major savanna problems of scavenging and waterhole hunting.

So what cognitive ability was needed by early Homo erectus for handaxe design?  Not much more than for shatter-and-search.  Rather than being seen as an embarrassing exception to 50,000-year modernity, the handaxe can be seen - once the singular controlling use is appreciated - as having a very pragmatic shape, where deviations from the flattened teardrop are more likely to result in dinner running away.  The step up to staged toolmaking (first shape a core, then knock off flakes) at 400,000 years ago is far more impressive as evidence of enhanced cognition.



Calvin, W. H. (1993).  The unitary hypothesis:  A common neural circuitry for novel manipulations, language, plan-ahead, and throwing?  In Tools, Language, and Cognition in Human Evolution, edited by Kathleen R. Gibson and Tim Ingold.  Cambridge University Press, pp. 230-250.

Calvin, W. H. (2002).  A Brain for All Seasons:  Human Evolution and Abrupt Climate Change.  University of Chicago Press.  Available at

Whittaker, J. C., McCall, G. (2001), Handaxe-hurling hominids: an unlikely story, Current Anthropology 42(4):563-572. || Home Page || Calvin publication list || The Bookshelf ||