Sell Your Collection!

We pay top dollar for your collection.
Great Value


By Jeb A. Taylor, G.I.R.S. Associate Editor 

In Volume XLIII, Number 2, 2009, Prehistoric American, Michael J. O’Brien, Ph.D. and R. Lee Lyman, Ph.D. contributed one more in a long line of theories regarding the use of beveled bone rods by Clovis culture big game hunters.

We may never know for certain what these objects were used for. However, replication studies and common sense can help us to determine which are the least and most likely theories regarding their use.

As O’Brien and Lyman pointed out, the oldest theory is that, they are foreshafts. I believe this theory shows the most merit, so it will be addressed last. A second theory was posited by Wilke, Flenniken, and Ozburn in Clovis Technology at the Anzick Site, Montana. They argued that the rods were handle portions of composite pressure flaking tools used to remove flutes from Clovis points.

They demonstrated that a flute could be removed using this method. However their test was performed on opalite, a material that is extremely weak and easy to pressure flake. Even though the theory was tested successfully, it is unlikely that this was the intended use of these objects. Why would a Clovis knapper, for example, need 14 such specialized flaker handles, as were found at Wenatchee? Also, if they were part of a Clovis knapper’s tool kit, why weren’t other tools associated with knapping found with them: hammerstones, abraders, billets, etc.? And lastly, Clovis points are easily replicated using more conventional tools, so why go to all of the trouble of fashioning very labor-intensive bone flaker handles when there is no apparent advantage to them?


The next theory was offered by Dr. Richard Michael Gramly in The Richey Clovis Cache. There he claimed that beveled bone rods were used as “sled shoes.”

Before continuing, I should mention that for a good portion of my life I lived in remote areas of Alaska and Minnesota. There I built my home and supported a family mostly from what the land offered. In the summer we traveled in birch bark canoes and in the winter we used sleds and toboggans, all of which I made from components split out of birch, ash, and cedar trees with relatively primitive tools, axes, draw knives, etc. And more importantly, paddled and pulled them many hundreds of miles through varieties of conditions, mostly, in hindsight, adverse!

Bone and/or ivory sled shoes were used in the Arctic by Eskimos, to protect precious wooden runners, and to create a wide smooth surface. However, their runners were flat in cross section to displace the weight of the sled, and were attached to the runners through counter-sunk holes.

If Clovis bone rods had been intended to be used as sled shoes, they would not have been trapezoidal , they would have been parallelograms. As they are, every other joint would snag on obstacles encountered.

Snow is very abrasive. If the bone rods recovered at Wenatchee (or anywhere else) had been used as sled shoes, their bottoms would have exhibited obvious flattening, polish, and striations. They don’t.

And lastly, round is the worst possible cross section for sled shoes. They would have been extremely difficult to attach, and impossible to maintain.

Dr. Gramly provided a drawing of a sled equipped with these hypothetical sled shoes that looked very convincing. However, archaeologists appear to believe that if something can be drawn, it can be built and used, and that is not the case in real life.

(Dr. Gramly explained to me that, based on the position of the bone rods in situ, their being sled runners was the only logical conclusion that fit all the conditions of the Wenatchee site. – Editor)


This last statement is also true concerning the proposed hypothesis offered by O’Brien and Lyman that Clovis bone rods were used as “levered hafted wedges”.

I must interrupt the flow of this discussion to explain that I have spent literally thousands of hours making and using stone tools, killing deer with stone tipped arrows, butchering them with stone knives, processing their hides with bone fleshers and stone scrapers, and making clothing with the tanned hides with bone awls and sinew. There may be others out there with more experience making, hafting, and using stone tools. If so, I am not aware of them. In any event, in this discussion, I am speaking from experience, not from hypothesis.

In positing their hypothesis, O’Brien and Lyman made a number of assumptions that if not outright incorrect, are certainly open to doubt. Initially, they made the assumption that the Wenatchee Clovis points were butchering tools because they were larger than most Clovis points recovered from Clovis kill sites. Their observations regarding their size was certainly correct, but their assumption regarding their use, although not necessarily incorrect, was at least presumptuous.

They then described the procedures required for hafting such large points/blades, stating that it took them almost two hours to apply 20 strips of sinew, waiting for them to dry, and then applying tree resin as a sealer. Judging frommy experience, sinew should never such as knife blades or scraper bits. Nor would any lashing, especially sinew, be covered with tree resin. Sinew is universally sized with hide glue, not pitch. Pitch is useful to seat points/blades into in their hafts, but not to apply to the surface of the lashing.

Blades the size of those from Wenatchee can easily be hafted in less than one minute with a stretched length of tanned hide or fiber cordage (hide is better). When the haft loosens, which it invariably does, it simply needs to be unlashed and re-lashed tighter. This can also be accomplished in less than a minute because, unlike sinew, tanned hide does not need to be masticated to reapply it and it is not covered with sticky pitch.

As regards the physics of their concept, I think it is impractical. Unfortunately, as was the case with Gramly and the sled shoe theory, their levered hafting wedge theory was also only illustrated with hypothetical drawings, as opposed to actual images or scale drawings of actual models. In any event, the drawings are of a hafted lanceolate point rather than a broad based point/blade like those from Wenatchee, so I assume that no actual models were ever made or tested. If they were, a stronger argument could be made in their defense.


In regard to the 1936-37 excavation at Blackwater Draw #1, Marie Wormington states in Ancient Man in North America:

In addition to the stone artifacts, two extremely important. polished bone pieces were found in situ. One was lying near the foreleg of one of the mammoths; the second near the tusk of another. These are tapering cylindrical bone shafts with a beveled end.

The fact that two beveled bone rods and two projectile points were found with mammoth remains is not proof that beveled bone rods were used as foreshafts. However, I would think that is does fall into the category of evidence, and it is clearly worthy of further consideration.

(TAYLOR, 2006)


Clovis projectile points exhibit a high degree of variability and size for a single projectile point type. This was undoubtedly due to a combination of temporal and spatial considerations: the range in size of their intended prey, accessibility to raw materials, attrition, or personal preferences. On the High Plains, Clovis points were generally fluted or basally thinned for a distance of 15mm – 25mm. Thinning was accomplished by single strikes, multiple strikes, pressure thinning, or any combination of these. Interestingly, even though Clovis points exhibit considerable morphological variability, they exhibit a surprising amount of conformity through their hafting areas.

The hafting area on Clovis points begins with a relatively sharp basal margin, and as a result of fluting and/or basal thinning, a wedge shaped hafting area is created that expands up the stem at a rate of about 12° – 16° (Figure 2). Uniform hafting areas on projectile points are desirable because they are easier to replace, but possibly of even greater importance to mammoth hunters may have been that foreshafts with corresponding “V” shaped notches were considerably stronger than foreshafts with mortised notches.

A “V” notch is hypothetically stronger than a mortise notch because less mass is removed from the critical notch termination area. Experiments performed by the author to test this hypothesis, though, were inconsistent due to variations in shaft diameters and the inherent strengths of different species of wood. However, the results suggested that foreshafts with “V” shaped notches required 18% to 44% greater sideways force applied to them to initiate shaft failures.


Figure 3. The split “V” shaped notch hafting strategy 3A: Illustration showing how a “V” notch is created in a split shaft

3B and 3C: Profile and face views a point hafted in a split “V” notch. The point is a cast of one of the original Clovis points from the Dent Mammoth Kill site, and the foreshaft is chokecherry.

Interestingly, the surface between the point and shaft mated perfectly without any trimming or fitting. Only the outside of the shaft was trimmed so that it tapered nicely into the surface of the point (Figure 3B)

Replication studies have shown that lashing on spears and darts is best accomplished in two stages. The shaft reinforcing lashing (Figure 3B) is separate and does not need to be removed when replacing points, but the haft lashing (Figure 3C) usually has to be cut off and is consequently destroyed.

This type of notch can be created very simply by splitting the shaft and inserting a projectile point into it. With this approach, no mass is removed from the inside of the notch; instead, it is trimmed off outside to create a smooth taper into the face of the projectile point (Figure 3A and 3B).


Another method for creating a “V” notch appears to have been used by Clovis craftsmen on bone. Bone cannot be split and bent the way wood can, so bone shafts were created using a composite method. This required that the end of the shaft be cut at an angle that corresponded to one half the angle of a Clovis point’s hafting area (6°- 8°), and that a corresponding beveled piece be lashed to it to make up the other half of the “V” (Figure 4A).


Figure 4. Composite “V” shaped notch Clovis
hafting strategy

4A: Components of a composite “V” shaped notch haft

4B: Profile view of a beveled bone rod, Blackwater Draw, New Mexico (cast) 

4C: Rotated view of Figure 4B (notice distal portion is missing)

4D and 4E: Two views of a chokecherry reconstruction of the beveled bone rod shown in Figures 4B and 4C with a cast of one of the original Clovis points recovered from the same site.


What is known about beveled bone rods is that they are generally:

– made from mammoth bone or ivory

– 150mm to 250mm long

– wider than they are thick

– 1 Omm — 30mm in width

– 10mm — 22mm in thickness

– beveled on at least one end (some are uni­beveled and some are bi-beveled)

– scored on the beveled surfaces with cross hatching

– found with Clovis points and bifaces in caches and in kill sites

Most Clovis beveled bone rods exhibit some degree of disintegration and are therefore difficult to measure accurately (Figure 4B and 4C). However, after studying a number of originals, casts, and illustrations, it appears that the angles of the bevels on many of them are about 6°- 8° off their vertical axes. Therefore, they would mate quite well with the hafting areas on most Clovis projectile points (Figures 4D and 4E).

Initially, bi-beveled bone rods were thought to be foreshafts, and bone rods with bevels on one end and points on the other end were thought to be projectile points. This is of course a possibility, although personal experience in hunting mammals with pointed arrow shafts has led me to the conclusion that arrows without points that cut are almost worthless. It is more likely that both types of bone rods were foreshafts, and that bi-beveled varieties were attached to their main shafts with the overlapping splice method (Figure 5A) and the pointed varieties were attached to their main shafts by the socket splice method (Figure 5B).


Fig. 5

In light of the fact that beveled bone rods are typically found associated with Clovis points in caches and kill sites, and that, as demonstrated above, they work admirably well as foreshafts, it seems quite plausible that this was their intended use.