What can we learn from flakes?


What is a flake?

Basically, a flake is the simplest unit of reduction when stones are chipped into tools. When the flake is created, a negative impression of the flake remains on the parent rock, but subsequent flakes can remove substantial portions of the scar. Only the flake itself retains the original surface contour of the parent rock. Scars are useful for showing the order in which flakes were removed. Flakes and their scars vary considerably through the reduction process and care should be taken to associate them with the appropriate stage. Cultural affiliation is most easily determined by flake scars that demonstrate habitual organization indicative of a highly organized tradition.

By no means is this guide meant to be comprehensive, it is intended to “prime the pump”. We eagerly await professional contributions.

Why should we pay attention to flakes?

The vast majority of archaeological evidence consists of flakes. At a minimum, they signal human activity at a particular place and time. Unfortunately, there is no universal guide for interpreting the information available from flakes. If we could learn to pay attention though, we could learn much about how early people applied technology. Flakes are the residue from one of humanities earliest technological advances and certainly its most persistent. It is difficult to imagine what our world might be like if people had not figured out how to use stone to fashion yet other tools.

How do flakes inform archaeologists?

Generally, flakes and their scars are associated with knapped tools presumed to be responsible for their creation. Unfortunately, the call usually gets made by someone with limited experience in flintknapping. In fact, identifying for certain how a stone tool was made is complicated by the equifinality problem. The equifinality problem refers to the difficulty in discerning the distinction between separate means of achieving the same result.

How can the equifinality problem be resolved?

There are three basic modes of archaeological flake creation: direct percussion, indirect percussion, and pressure. Each mode can be accomplished by myriad tools and their characteristics may overlap. By focusing on unique conditions imposed by the combination of tool and mode of use we should be able to say with a degree of assurance how flakes from a site were produced. Sometimes the analysis is made easier when a knapping tool is found in direct association with the flakes. Finding the knapping tool still does not insure that we know how it was manipulated.

What are analyses based on?

Since stone tools have few present-day practitioners, we rely on associating attributes of flakes made with known tools and mode of use. Optimally, this would involve an exhaustive survey of all possible combinations but no one has taken up the challenge. Practically, analysts rely on their limited personal experience. Furthermore, professional archaeologists lack the time to become proficient in diverse flintknapping techniques. However, guidelines may be extracted from some important studies. Cotterell and Kamminga (1987), established a theoretical basis for understanding knapping fracture and Baker’s computer modeling (1998-2003) of fracture behavior refined the earlier work. Controlled experiments seek to isolate active variables. Real-life knapping has to deal with an unpredictable mix of variables.


Guidelines for reading flake attributes


Initiation: Fracture starts when an external load stresses the rock beyond its strength.

  • How the load is distributed determines how quickly critical stress is achieved. Small platforms dictate high load per square inch. Hard hammers also restrict contact to small areas, while soft hammers deform to spread their load over a larger area.

  • The separation of flake from parent rock is guided by the distribution of the impacting load. Hard impactors start with a small ring crack that connects to the core edge in a sweeping recurve that looks like a bird in flight, hence alar shaped. Soft impactors distribute the load so the separation looks like a compass arc, hence arcuate shaped.

  • It is possible to force a perpendicular blow with such intensity that the initial cone is collapsed and split to cause a wedge detachment. Wedging presumably drives debris beneath the hammer to split the core in two.

  • The vector direction of load application effects the way a crack starts. Loads perpendicular to the core surface cave a small cone into the surface, the same as a pebble pits a windshield. The remaining flake surface propagates from that initial Hertzian cone. If the load is applied at an angle to the edge, it may actually start a crack by bending a portion away from the core edge to cause a lip.

  • A bulbous swelling just after the first crack formation is known as the bulb of force. Computer modeling predicts the most noticeable bulbs from impacts perpendicular to the core face. Knapping experience suggests that high rates of loading contribute to bulbar swelling, but an inclined blow can be expected to deliver a lower rate of loading. Gentle rates of loading, such as pressure or soft hammers can leave flakes nearly bulb free.

  • Prominent bulbing can leave overhangs at the edge of a flake scar. Platform isolation can emphasize bulbing.

  • Tool hardness effects local stress levels and can cause secondary fracture planes to develop. The effect is particularly noticeable with brittle materials like glass, where small tear lines, perpendicular to the flake face, are found around the edge of a flake. Tear lines point back to the point of impact, even when the platform is missing. Compression waves can also initiate cracks deep inside the rock when they encounter an inclusion. Soft knapping tools lessen the problem. Note that hardness of hammer is not so important as intensity of load.

  • Offset of the load in regard to the face being flaked determines how deep a path the flake wants to follow. Soft baton blows on sharp edges tend to produce shallow, thin flakes. Thick flakes are usually caused by hard hammers impacting far from the core margin.

  • Width of detachment from a core is usually a function of the area contacted by the knapping tool. Platforms can be used to isolate where the blow lands or soft tools can be used to distribute the load broadly.

  • The strength of a platform greatly effects when critical stress is reached. Consequently, knappers often prepare platforms very carefully to control flake formation. Grinding strengthens a platform, isolation concentrates stress, and faceting controls contact placement. Battering serves to introduce microcracks that allow cracks to start at a lower minimum stress.

Travel: Fracture progresses as long as critical stress of the rock is exceeded.

  • Bending stress in the parent rock promotes curved flake paths. Straight flakes indicate deliberate support of the parent rock to avoid bending.

  • Stability of the fracture path rests somewhat on the depth of the fracture, but depends on the load being constant over time. Unsteady support or vibrating knapping tools can force the fracture path to undulate.

  • When radial stress at one edge of a flake initiation penetrates to the opposite face, a perverse fracture may split a biface in two. Clovis knappers depended on the action to separate large platter-like bifaces into usable portions.

  • Strength of the stone is often indicated by the texture of the fractured surface and tear lines that reveal when the rock does not separate easily. The stronger the rock, the more durable the hammer will have to be.

  • A flake can only extend while tool and core are pushing against each other, so short flakes indicate loss of contact with the tool.

  • Surface morphology is responsible for most of flake morphology. Flakes progress most readily when surface contours give the flake stiffness. Lacking ridges, a flake will naturally be circular. However, maintaining the load helps drive the flake forward, as in leverage pressure. Restraining the core against an outward blow also promotes spreading flake geometry.

  • Ripples are caused by slight disturbances in the load, and arc radially away from the point of impact.

  • A special sort of disturbance, called a Wallner wake, shows up in brittle materials like obsidian when the compression wave interacts with the developing fracture surface to leave short lines, sometime known as gull wings, that diverge from a central imperfection. The speed of fracture can be deduced by measuring the angle between the Wallner wake lines.

  • Adding stiffness to the system promotes relatively flat flakes. Stiffness comes from added support, harder hammers, or faster blows. Increasing support of the core causes the core to behave as if it possessed greater mass.

  • Shock introduced by impactor can manifest as concentric rings with low amplitude.

Termination: Fracture ceases when stress levels fall below the critical level.

  • Feathered flakes indicate that contact was maintained between tool and core for the entire duration of the fracture. Hinge flakes indicate that contact was interrupted when the tool and core bounced away from each other. Step flakes may indicate that the blow lost force before the flake could finish, but too much outward force in relation to the forward push can cause a step.

  • The termination of a flake often shows how the core was supported against the knapping load. Severe overshots occur when too much bending is present, but light overshots may be deliberately introduced by pinching the far edge to provide a soft anvil support. Overly severe support can cause flakes to undulate in increasingly severe waves as they reach the end of their travel. Ideally, a stable support and blow will allow the flake to feather as it finishes.

  • Flake thickness impacts the evidence from flake termination. Thick flakes tend to be stable, but thin flakes are highly sensitive to external interruptions of the knapping load. Indirect percussion tends to take very thin, far-reaching flakes that waver as they terminate. The effect is caused by flake stiffness reflecting the core surface morphology.

  • Hard mineral hammers tend to leave upturned terminations where flakes meet at the biface midline.

Common indicators of tool use

Hard hammer

  • Heavily ground platforms

  • Alar platform detachments

  • Tendency for step and hinge termination, especially with copper

  • Continuous platforms

  • Thick, wedge-shaped flakes

  • Distinct ripple marks

  • Tendency for hinge termination

  • Overshots are unusual with thin flakes

  • Impacts offset from the core face plane

  • Generally responsible for the thickest flakes

Soft hammer

  • Arcuate platform detachment

  • Light platform strengthening

  • Isolated platforms

  • Oval flake shape at late stage

  • Subdued ripple marks

  • Overshots are easily controlled

Indirect percussion (by rocker)

  • Acute platform edge

  • Wavered flake termination

  • Broad flake initiation

  • Feathered terminations on long flakes

  • Low frequency of step or hinge termination

  • Flat flake trajectory

  • Thin flakes

  • Overshots are possible, but are seldom prominent

Indirect percussion (by cylindrical or peg punch)

  • Usually to produce blades from cores

  • Lip initiation on blades

  • Biface edges tend to stay in a plane

  • Broad flake initiation

  • Feathered flake termination

  • Thin flakes at late stage

Pressure (by antler)

  • Isolated spur platforms

  • Light platform grinding

  • Small step terminations

  • Selective flake patterning common

  • Subdued bulb of force

  • Tight flake spacing possible

Pressure (by copper)

  • Continuous platforms

  • Ground platforms

  • Serial flake patterning common

  • Prominent bulb of force

  • Tight flake spacing is possible

Technologic purpose served by flakes

Remove arris

One of the first requirement in reducing a quarry core is to eliminate prominent arrises from corners. These flakes are often the basis for robust scraping tools. Continuing to follow successive arrises creates blades. Camp tools usually are thicker than 6-mm, and must be created relatively early in the reduction process.

Thinning

The final series of flakes on formal camp tools are often designed simply to achieve a requisite thickness.

Surface contouring

Some flakes are specially planned to give projectiles aerodynamic symmetry.

Clip corners

When tabular material is obtained at the quarry, corners are often clipped by wedge-shaped flakes to provide striking platforms near a core surface. The clipped corners may be used to estimate thickness of the source material.

Thin base

Hafting requires basal taper and thinning that is sometimes produced by special flakes. Channel flakes taken from the base, called flutes, are an extreme example.

Position impact offset

Similar to corner clipping, short bevel flakes are often used to make sure that the flake starts deep enough to achieve its desired thickness.

Serration

Edges of knives and projectiles are sometimes serrated by specialized flakes related to notching.

Notching

Notching is accomplished by various means, but distinctive flakes looking like flattened cones are useful for recognizing when notching has taken place.

Beveling

Sharpening is typically performed repeatedly from the same face, producing beveled edges. The flakes can be remarkably uniform in size and shape.


GLOSSARY

alar detachment: Wing-shaped separation of flake from core.

arcuate detachment: Arc-shaped detachment of flake from core.

arris: Architectural term describing intersection of two surfaces.

bulb of force: Bulbar swelling of a flake, just beyond fracture initiation.

compression waves: Particle movement waves, also known as Raleigh waves.

concentric ripples: Low amplitude waves causes by compression waves.

hard hammer: Hammer that behaves like a very stiff spring regardless of composition, typically mineral.

Hertzian cone: Fracture caused when a load collapses a cone-shaped volume below the surrounding surface.

lip detachment: Lip-like remnant of the core edge caused by bending the flake away from the core.

Overhang: Cavity at the leading edge of a flake scar, where the bulb of force has been removed.

perverse fracture: Spiral fracture that starts from one edge of a flake and splits a biface.

Raleigh waves: Particle-movement or compression waves, like those associated with earthquakes.

soft hammer: Hammer that exhibits spring-like qualities regardless of composition, typically antler.

Contact and Feedback

©2010 by Bob Patten. All Rights Reserved.