This chert tula adze slug is from near Camooweal in Northwest Queensland.  The adze dates to the Late Holocene, after ca. 5000 BP.  See the annotations for technological details about this stone tool.

---

The tula adze in this model was made by the gull wing method.  The adze was resharpened right up to the point of force application on the ventral surface.  The U-shaped profile is clearly visible by viewing the model edge-on.  Retouching in this case was by hard-hammer percussion.

---

Stone adzes are a common stone tool type in Australia.  They emerged as part of a suite of technological innovations in the Holocene, and proliferated after ca. 5000 BP.  In many regions they were replaced by metal when it became available, but stone adzes were made and used into the 1960s in remote areas.  Stone adzes were hafted with plant resin (usually spinifex resin or grass tree resin) or native beeswax onto a wood handle.  In Central Australia, stone adzes were sometimes hafted onto the handle of of a spearthrower.  The flake blank for ‘tula’ adzes was retouched across the distal end to form the working edge, and hafted with the handle parallel to the percussion axis of the blank.  The flake blank for ‘burren’ adzes was retouched on the lateral margin and hafted with the percussion axis at right-angles to the handle.  Tula adzes were about  30-50 mm long before resharpening, although ‘microadzes’ made on smaller flakes are common in some regions, and unusually large adzes—between 50 mm and 100 mm long—were sometimes made in the Barkly Tableland region of Northern Australia.  Stone adzes were items of exchange between Aboriginal groups, and caches of adzes have been discovered by archaeologists in Queensland.

Adze flakes were typically used to fashion wooden implements such as boomerangs and other objects from Australia’s exceptionally hard acacia and eucalyptus tree species, although residue studies suggest that they adzes were sometimes used for butchering and other tasks as well.  In woodworking,  the worker sat on the ground with the wood blank held down by the feet.  The hafted adze was held in both hands and used in a chipping motion (as opposed to scraping), moving in lines towards the body.  The motion was repetitive and carefully controlled—too hard a strike or a blow at the wrong angle could break the stone from the resin.  Incremental amounts of wood were removed with each blow.  Adzing was most effective on green and fresh wood.  When the stone became dull, damage, or embedded with plant fibres, it was resharpened by unifacial percussion.

Resharpening was normally done by freehand percussion using a hammerstone.  However, the use of boomerangs to resharpen adzes was documented by many European observers.  This is clear evidence that the soft-hammer percussion technique was widely practiced by Aboriginal flintknappers across much of Australia.  Historical photographs show that the flat face of the boomerang was used in retouching, rather than the boomerang’s edge.  The boomerang was brought down in a chopping motion so that the flat face brushed the platform edge; the wood ‘grabbed’ the edge, removing small retouching flakes.  The use of a soft hammer resulted in bending initiations and created a more even and acute edge profile than the use of a hard hammer.  Adzes resharpened by hard hammer and soft hammer are found together in archaeological assemblages, but hard-hammer retouching predominates.  It is possible that the different techniques were used to produce task-specific edge profiles on these tools, although this has yet to be verified by use-wear or residue studies.

Resharpening was done while the adze was still in the handle, and with each resharpening episode, the adze grew shorter, and attrition progressively moved the working edge towards the resin.  It was sometimes adjusted in the resin by re-heating it and moving the stone, and in one historical account the stone was repositioned on the adze-handle for resharpening with a boomerang, and then replaced in its former postion.  Burren adzes could be rotated and the opposite margin used and resharpened.  Eventually the adze became too small for further resharpening and it was taken from the handle, discarded, and a new stone was inserted into the resin.  The worn-out adze is called a ‘slug’ by archaeologists.  Tula adze slugs are among the most common retouched tools found on archaeological sites in parts of Northern and Central Australia, and burren adze slugs are relatively common on Cape York.

Flintknappers at Camooweal in Northwest Queensland—the Indjilandji-Dhidanu people—used their intimate knowledge of fracture mechanics to create tula adzes with an ideal edge shape for woodworking; this is referred to by archaeologists as the ‘Camooweal Gull Wing Method’.  One of the rules for making flakes is to strike behind a ridge or lump on the core face as a way of manipulating the crack path and producing a flake of the desired size and shape.  The ‘Camooweal Multiple Facet Method’ and ‘Camooweal Standard Method’ were practiced by these flintknappers to produce macroblades by creating ridges on the core face, and removing them in series.  However, in making tula adze blanks by hard-hammer percussion, the Indjilanji-Dhidanu flintknappers deliberately struck behind hollows on the core face, rather than ridges.  The hollow was created by the removal of the prior flake.  The intent was to produce as pronounced a bulb of percussion as possible by striking behind the hollow on the core face created by the prior tula flake removal. As the crack propagated under the concave surface, it took on a similar but slightly more exaggerated shape, and each subsequent flake in the series possessed a more pronounced bulb than the one that preceded it. Tulas made using this method have the profile of a flying bird when viewed end-on, hence the name ‘gull wing’.  As the adze was retouched and resharpened by unifacial flaking, the edge retreated from attrition until it encroached onto the curved bulb of percussion, thus taking on a U-shaped edge profile. This profile is ideal for woodworking, and resembles the edge on a modern steel gouge.

The gull wing method is technically challenging because much of the mass of the resulting flake is distributed on the edges rather than down the centre.  As the crack propagates through the stone, the peripheral mass causes intense bending forces to act on the flake’s centre, and it often splits in what archaeologists call a ‘siret fracture’.  The Indjilandji-Dhidanu flintknappers adjusted to this by 1) slightly staggering the locations they struck subsequent flakes, by 2) periodically striking off the high-mass ‘wings’ that developed on the core, and by 3) using as little force as possible to initiate the flakes  (resulting in ‘double-tap’ ring cracks on the flake platforms).  Nevertheless, the failure rate from siret fracture was very high compared to other reduction methods.

Aside from Australia, a ‘gull wing’ method was independently invented three other times in prehistory.  In parts of Egypt and the Middle East, a gull wing method was used to make blanks for ‘side blow knives’, ca. 5000-7500 BP.  In Texas, Native Americans used a gull wing method to make ’sequent flake unifaces’—used as hafted knives for plant processing—during the Baker Interval at Hinds Cave, ca. 5000-8500 BP.  In Japan, the gull-wing Setouchi method was used to make blanks for retouched ‘Kou’ knives at the Suichoen site, ca. 20,000-30,000 BP.  The pronounced bulb is thought to have been a deliberate design feature for side blow knives, and perhaps Kou knives, to make them easier to hold in the hand.