University of Minnesota
Department of Anthropology


Lithic Raw Materials...

Contributed by Kent Bakken

There are three basic stone tool technologies—flaking, carving, and grinding. Carved stone artifacts include pipes and, in some regions, bowls that are literally carved from soft raw materials, such as pipestone or soapstone. Ground stone artifacts include stone axes, adzes, and mauls. Such objects were shaped by pecking the surface until the desired shape was reached (no doubt a slow and tedious process). The surface was then smoothed and polished with sand or a similar abrasive. Carved and ground stone technologies are relatively uncommon in this region. Most stone artifacts, including projectile points, are instead the products of flaking. This process, also known as flintknapping, is almost like whittling, although individual pieces are not removed by cutting but by breaking. A skilled artisan—a flintknapper —is able to control how rock breaks when it is struck (see Appendix A). By removing one small piece after another, the flintknapper is able to shape what remains into something as elegant and functional as a projectile point.

In Minnesota prehistory, a variety of stone types (lithic raw materials) were used by traditional flintknappers for making flaked stone tools. While some of these materials came from bedrock exposures, most came from glacial sediments. There are three important differences between these kinds of raw material sources—extent of distribution, and amount and variety of material present. Bedrock sources are relatively localized, usually contain large amounts of material, and provide a single kind of raw material. They may consist of a narrow band of suitable outcrops that stretch for a short distance or for several miles, or be confined to a single patch that can be traversed in a few minutes. For example, a good quality material called Grand Meadow chert apparently comes from a single quarry complex in Mower County in southeastern Minnesota. The whole quarry complex covers a few acres.

Glacial sources are quite different. While the amount of raw material available at any one place is usually limited, several different materials are generally available. In addition, the source is not usually confined to a narrow band or single small location. It may cover hundreds or thousands of square miles. For example, Swan River chert and Red River chert may be found from the Canadian border area of northwestern Minnesota through western and central Minnesota to the central Iowa border.

In order to understand the distribution of lithic raw materials in Minnesota, one has to know something about the distribution of bedrock exposures and glacial sediments in the state. This complex picture may be simply summarized this way. In parts of northeastern and southeastern Minnesota, much of the landscape is not covered with glacial sediments and, consequently, bedrock outcrops are relatively common. In other parts of southern and eastern Minnesota, erosion cut through glacial sediments to expose bedrock. In both cases, some outcrops contain suitable “toolstone.” Except for these limited areas, however, the surface of Minnesota is covered with glacial sediments. These sediments were deposited by complex processes that involved multiple glacial movements over hundreds of thousands of years. Since these masses of ice moved southward from different directions, their loads of stone came from multiple sources. Some ice came from the northwest and carried in the kinds of stone found there. Other ice came from the northeast and carried in the kinds of stone found in that region.

In recent years, archaeologists have begun to understand the distribution of the lithic raw materials in the state. This was accomplished by studying the state’s geological history and samples of toolstones from many regions. Three very general “raw material” regions have been identified—a western, an eastern, and a southern (Figure 11). Each contains a different set of raw materials because the materials are from different combinations of sources.

Raw Material Resource Regions. The Western Raw Material Resource Region is apparently the largest. It covers much of western Minnesota, from the Manitoba border to approximately the Minnesota River valley. It also includes a large part of central Minnesota. Most of the region is mantled with glacial sediments whose origin was to the northwest. Characteristic raw materials include Swan River chert, Red River chert, rhyolite, and Tongue River Silica. Swan River and Red River cherts are available throughout the area; rhyolite is more common in the northern ranges; and Tongue River Silica is more common in the southern parts of the region. Less common materials include small pieces of Knife River Flint, chalcedony, jasper, and silicified wood.

The Eastern Raw Material Resource Region includes parts of north central, northeastern, and east central Minnesota. The region is partly mantled by till (glacial sediment), but it also includes large bedrock exposures. The glacial sediments come mostly from northeastern sources. Characteristic raw materials include Jasper Taconite, Gunflint Silica, siltstone, and Lake Superior agate. Less common materials include Kakabeka chert, Biwabik Silica, and Hudson Bay Lowland chert.

The Southern Raw Material Resource Region includes that part of Minnesota beginning at approximately the Minnesota River and continuing southward to the Iowa border. This region also contains both till cover and bedrock exposure. The glacial sediments are a mixture of materials derived from both northwestern and northeastern sources, and from the region’s own underlying bedrock. Characteristic raw materials include Prairie du Chien chert, Galena chert, Grand Meadow chert, Swan River chert, Tongue River Silica, Sioux quartzite, and Cedar Valley chert. Less common materials include Shell Rock chert, silicified wood, Lake Superior agate, and siltstone, as well as small pieces of Knife River Flint, jasper, and chalcedony.

Certain other raw materials must be mentioned. Two raw materials, quartz and quartzite, are so widespread that it makes little sense to discuss them within the confines of raw material resource regions. Quartz pebbles were commonly used for certain kinds of tools, especially in particular cultural-historical periods and in areas where better quality resources were absent. Local quartzites were generally inferior for the production of most tools and were seldom used (in contrast to some “imported” quartzites that are discussed below). In parts of the state, other coarse, laminated, or otherwise inferior materials were used to produce specific kinds of tools. Materials like basalt, schist, and slate, which lack most of the qualities of a good toolstone, are unsuitable for manufacturing sophisticated tools like projectile points. Nonetheless, they were used at times to make large, crude tools called “choppers.” These small cobbles were sharpened along all or part of an edge by the removal of a few flakes. Such tools was probably used on the spot then discarded.

Many artisans had access, too, to supplies of exceptional stone from distant sources. In Minnesota, four such materials (Knife River Flint, Hixton Quartzite, Burlington chert, and obsidian) are found on a regular basis in archaeological sites. Each was apparently quarried at a distant source and traded over wide areas.

The primary source area for Knife River Flint (KRF) is an enormous series of quarries in west central North Dakota. KRF is found at sites throughout the state (and as far away as Ohio), although it is most common at sites in west central and northwestern Minnesota. In the latter areas, some sites contain 20 to 40 percent KRF by count. Sites in eastern Minnesota, by contrast, typically contain only one to two percent KRF. Although small pebbles of KRF are present in glacial sediments in Minnesota, nearly all archaeological KRF in the state came from the North Dakota quarries. This conclusion is based on considerations of rarity, quality, and artifact size.

Hixton Quartzite, on the other hand, comes from a single source, commonly called Silver Mound, in west central Wisconsin. In contrast to locally available (glacial till) quartzite, Hixton is a very easy to flake, high quality material. In Minnesota, it is found mainly at sites in the east central region of the state. Even in these sites, it seldom accounts for more than five percent of all lithic artifacts.

Burlington Chert is available in parts of Iowa, Illinois, and Missouri. Its primary quarry complexes were apparently located near St. Louis. Burlington is common at sites in southern Minnesota near the St. Croix River, where it also seldom accounts for more than five percent of all lithic artifacts.

Obsidian comes from a number of western, Rocky Mountain, sources. Trace element analysis suggests that the Yellowstone source provided most of Minnesota’s obsidian. The archaeological distribution of obsidian in the state is not well understood. However, it does not seem to be abundant anywhere, and it may be absent or rare in some regions. It is almost never found, for example, in the Red River valley of northwestern Minnesota. In contrast, most Late Woodland sites in central and east central Minnesota contain at least one or two obsidian waste flakes.

Raw Material Identification. Experience has demonstrated that it is usually not possible to accurately identify lithic raw materials without access to a comparative collection. As a natural material, stone is variable. Stone like Tongue River Silica and Knife River Flint, which shows minimal variation from piece to piece, is the exception. The rule is quite different—most materials show a substantial range of variation in essential characteristics, such as color, texture, and inclusions. Since the range of variation defies memory, access to identified, provenienced samples is essential.

Experience has also demonstrated that it is risky to depend on single characteristics to identify raw materials. This is particularly true of color, which is a notoriously fickle characteristic of raw materials. Paradoxically, most beginners seem to intuitively rely on color to form their initial categorizations, perhaps because color is an easily observable characteristic. However, its use is rife with problems. Most raw materials have a wide color range. Nonetheless, the overall range of these colors is restricted, which means that many materials share the same color range.

In identifying raw materials, all observable characteristics must be considered. Among these are range of color, texture, homogeneity or lack of homogeneity (color and texture), degree of translucency or opacity, presence or absence of inclusions, nature of inclusions (including identification of fossils), color of transmitted light, graininess (coarse to fine), form of cortex, nature and color of patination, fracture pattern or texture of fracture surface, presence or absence of vugs (hollows in the rock), and presence or absence of druse (crystal lining of vugs). It is frequently necessary to examine these characteristics under low power magnification, using either a hand lens (a “loupe”) or a low power microscope (such as a binocular dissection microscope). Some researchers also use ultraviolet (UV) light in identifying raw materials. Which characteristic or combination of characteristics proves to be most diagnostic varies from material to material. This is where an identified, comparative collection becomes indispensable: it allows regular calibration of one’s observations (in addition to the more obvious function of direct comparison).

A related matter should be mentioned. Traditional artisans commonly heated the stone they used to fashion tools. When done in a controlled and deliberate fashion, heating often makes poorer quality material easier to flake. However, it also alters some visible characteristics, often dramatically. For example, the type of Tongue River Silica found in Minnesota has a natural yellow-brown (“ochre”) color and a coarse texture. When the material is heated, the texture becomes finer and the color alters to a distinctive red or orange-red. Many other materials develop a pink or red tint, or a waxy texture that was absent before heat treatment. Overheating produces even more drastic changes in the stone's physical appearance. Galena Chert normally has a relatively chalky texture. When overheated, the texture can become glassy and the material excessively brittle. Knife River Flint, which is normally brown and translucent, can become nearly opaque and bluish gray. (For these reasons, a good comparative collection should include samples of raw materials that have been heat-treated; an exceptional collection would include “burned” samples.)

Like all identification methods, macroscopic visual identification has its limits, too. Given the amount of variation in natural materials, as well as the widespread movement of materials by both natural and human activities, it is not realistic to reach 100% identification in any except unusual circumstances. Failure to identify every raw material in a collection is currently viewed as a sign of integrity in a raw material analysis. A corollary of this lesson is that some identifications nearly always remain suspect because of the limits of visual inspection. Where possible, visual identification should be complemented by more objective methods of identification, such as trace element analysis.

Lithic Raw Material Economies: Selection and Use. Raw materials suitable for flintknapping share a few basic characteristics. They are relatively fine grained, homogeneous, isotropic (equally likely to break in any direction), and hard (without being too brittle). Most materials that meet these basic qualifications are composed mostly or entirely of silica—the major ingredient of glass. A great deal of variation remains, however, within these broad parameters. A material like Grand Meadow chert (GMC) is very fine grained, very homogeneous, and highly isotropic. It occurs as well as cobbles of a suitable size for tool manufacture. GMC is a nearly ideal material for making nearly any kind of flaked stone tool. Siltstone, on the other hand, is often coarser grained, tends to have cracks, and is relatively soft compared to other toolstones. Its chief virtues are that it is common in its source area and available in relatively large pieces. Quartz has a crystalline rather than grainy structure, which means that it is not highly isotropic and that it tends to break along crystal planes. In addition, it usually occurs as pebbles or small cobbles, which means that it cannot be used to make large tools.

Traditional flintknappers were faced with a host of challenges and decisions. Besides determining which raw materials were available in a given region and assessing their characteristics, they had to know what tools were needed and how they would be used. They then had to match available raw material to final product in a way that made the best use of the piece of material. This series of challenges, decisions, and activities, which involves managing raw material use and tool production, is called “raw material economies.” The growing understanding of the distribution of lithic raw materials in the state has made this kind of study possible. While only preliminary results are available, they are interesting and informative.

Several studies have demonstrated that Paleoindians preferred certain raw materials. To the northeast, this was either siltstone or Jasper Taconite, depending on location. In western and central Minnesota, it was Swan River chert. To the south, it was most likely Prairie du Chien chert. Although these materials are generally dissimilar, they are all available in relatively large cobbles or blocks. The Paleoindians produced large, lanceolate projectile points. Since large bifaces cannot be made from small rocks, it seems that they were selecting the largest available pieces of raw material (which was different in each region) to produce their large tools. Characteristics like relative hardness and homogeneity were secondary considerations.

Other studies have concentrated on end scrapers, a small scraping tool common in the region throughout prehistory. End scrapers, which are believed to have been used for preparing hides, are usually made of the finest grained, most easily flaked, highest quality raw material. The reason may be that scrapers must have a regular, durable edge. Scrapers made of inferior material would have a rough or irregular edge. For instance, those made of soft or brittle material might break during use, which generally produces a jagged edge. Such inferior tools would cut, gouge, or scratch the hide, which was an unacceptable result. The finest grained, most easily flaked, highest quality raw material was probably set aside, then, for making end scrapers. If there was an abundant supply of the material, some might be used to make projectile points or other tools. The priority, however, went to end scrapers. (Note also that these are small tools with a relatively simple form. They could be made from small cobbles or even pebbles; size of the raw material stock was not a particular concern.)

As mentioned above, quartz is difficult to flake and ordinarily occurs in the form of pebbles. Because of these characteristics, it is not an especially useful stone for making large, lanceolate projectile points or end scrapers. It does have its virtues, however. Two of the most important are that it is very common and breaks with a sharp edge. Traditional flintknappers devised a specialized technique to take advantage of these characteristics. Known as “bipolar reduction,” the technique involves setting the quartz pebble on a solid rock, then striking it from above with another rock. The pebble fractures from both above and below (ergo “bipolar”). Although bipolar reduction is not a suitable procedure for producing sophisticated tools like projectile points, it can easily produce large numbers of sharp-edged fragments that are suitable for ordinary cutting tasks. Such “expedient tools” were used until the edge became dull, then discarded. This technique allowed useful (if simple) tools to be made from an otherwise useless raw material. Furthermore, it reduced the need for making cutting tools from limited stocks of better quality material. The better quality material could be saved for making scrapers, projectile points, or other technically more demanding tool forms.

These examples briefly illustrate why strategies of raw material use and tool production varied from one period to another and from one place to another. Besides the kinds and amounts of stone available (either locally or by trade), the kind of tool required was a major consideration. By analyzing stone tools and other artifacts in this manner, we can begin to appreciate why quartz projectile points are relatively rare and why the makers of fluted points imported high quality materials from distant sources.

Suggested Reading:

Bakken, K. 1999. Lithic Raw Material Resources in Minnesota. Minnesota Archaeologist 55-58:51-83.