The Roasted and the Boiled: Food Composition and Heat Treatment with Special Emphasis on Pit-Hearth Cooking

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Anthropology Faculty Publications Anthropology, Department of

1-1997

LuAnn Wandsnider University of Nebraska - Lincoln, [email protected]

Follow this and additional works at: https://digitalcommons.unl.edu/anthropologyfacpub

Part of the Anthropology Commons

Wandsnider, LuAnn, "The Roasted and the Boiled: Food Composition and Heat Treatment with Special Emphasis on Pit-Hearth Cooking" (1997). Anthropology Faculty Publications. 28. https://digitalcommons.unl.edu/anthropologyfacpub/28

This Article is brought to you for free and open access by the Anthropology, Department of at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Anthropology Faculty Publications by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Published in Journal of Anthropological Archaeology 16 (1997) , pp. 1–48. Copyright © 1997 Academic Press. Used by permission. http://www.elsevier.com/locate/jaa Submitted March 1, 1996; revised October 17, 1996; accepted October 20, 1996.

The Roasted and the Boiled: Food Composition and Heat Treatment with Special Emphasis on Pit-Hearth Cooking

LuAnn Wandsnider Department of Anthropology, University of Nebraska–Lincoln, Lincoln, Nebraska 68588-0368

Abstract Heat treatment is one of the major ways humans change the composition and chemistry of food tissues, making them more digestible, less toxic, and more durable. This paper reviews salient features of food chemistry and food composition and how heat treatment, especially pit-hearth cooking, af- fects that composition. Ethnographic accounts of cooking indicate that traditional populations relied on pit-hearth cooking especially to alter the composition of foods high in either lipids or complex carbohydrates. Historically, pit hearths were also used to process large quantities of food. Various kinds of pit hearths figure prominently in the archaeological deposits of the American Great Plains and elsewhere. The implications of this recurring phenomenon are discussed in terms of the coevolution of diet, cooking systems, and the appearance of Neel’s ‘‘thrifty’’ genotype.

Introduction tion (Brace et al. 1987; Jackson 1991; Larsen 1995; Stahl 1984), subsistence change and the origins It seems to be a general rule that sciences begin of food production (contributors to Harris and their development with the study of the unusual. Hillman 1989; Hayden 1990), the appearance of They have to develop considerable sophistication ceramic container technology (Braun 1983, 1987; before they interest themselves in the common Brown 1989; Schiffer and Skibo 1987), and status place. differences in access to foods with different qual- Linton 1994:369 ities (e.g., Welch and Scarry 1995). Recently, Ann Stahl (1989) has considered the In 1944, Ralph Linton decried the lack of inter- fundamental relationship between food chemis- est shown by anthropologists and archaeologists try and food processing technology, especially in common archaeological remains, ceramics, highlighting the changes in food processing that and the associated, equally common pursuits of might occur in lieu of subsistence change. This cooking and food processing. By Linton’s reck- paper attempts to continue Stahl’s work, focus- oning, 50 years later, either our sciences have ing expressly on the relationship between food become much more sophisticated or the com- chemistry and cooking strategies that involve monplace is now perceived as the unusual and heat treatment. Where Stahl addressed her work therefore worthy of our scrutiny. Certainly to- to the naive assumptions made about the costs day, research on a wide range of anthropologi- and benefits of food processing in optimal forag- cal questions relies on or considers variation in ing calculations, I am interested in assembling a food exploitation and processing strategies, in- body of reference knowledge (1) that will further cluding studies of hominid and human evolu- our appreciation of the selective use of specific

L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) cooking technologies and (2) that will make more from any one mouthful of food. They may re- accessible to anthropology the richness of the ar- duce the chance of illness by killing food-borne chaeological record, dominated by the durable bacteria and parasites and by eliminating tox- remains of cooking systems (i.e., ceramic, stone, ins that occur or develop in some tissues. And, and basket containers; varieties of hearths, ov- by cooking, spoilage bacteria are eliminated and ens, and heat reservoir elements), for approach- water, needed by bacteria to grow, is reduced, so ing important anthropological issues. Thus, this that the storage life of food may be extended. paper focuses on the food processing strategy of Cooking encompasses many activities heat treatment and considers the circumstances (Medved 1986; Penfield and Campbell 1990; in which, for example, root baking is selected Stahl 1989). It may entail mechanically grinding over root boiling, pit-roasting over spit-roasting. or pulverizing tissues to increase the surface area Heat treatment is just one of a number of cook- over which various physical or chemical pro- ing strategies available to people and so I begin cesses operate. Food preparation may involve with an overview of cooking and cooking sys- flushing and rinsing plant tissues. Foods may tems. One of several factors that influences the be fermented or partially digested by cultivated design of cooking systems is that of food com- bacteria to break larger components into smaller, position and an abstract of this body of knowl- digestible components. People may bathe food edge, as it relates to thermal food processing, is tissues in salty or acidic solutions to induce developed in the succeeding section. Given this chemical changes. Parching, broiling, spit-roast- understanding of food chemistry and in light of ing, baking, and frying are examples of dry heat the digestive capabilities of anatomically mod- cooking. Moist thermal treatments include brais- ern humans, an evaluation of several expecta- ing, stewing, or boiling. Given this range of food tions for the heat treatment of food by traditional preparation techniques and tools, why is one populations follows. In order to understand the suite or sequence of cooking activities executed special conditions under which they are used, I rather than another? next focus on the pit hearth and related cooking Undoubtedly, many different factors influ- technologies. The paper concludes by highlight- ence the design of a specific cooking system and ing the special role of one heat treatment system, the implementation of specific cooking strategies pit-hearth cooking, for understanding the ap- (Figure 1). For example, the BTUs provided by pearance and maintenance of Neel’s (1962, 1982) an available fuel source, be it seal blubber, grass ‘‘thrifty’’ genotype in association with complexes blades, or wood, will constrain the cooking en- of noncereal foods and related cooking systems, vironment, as will the availability and nature of especially on the Great Plains of North America. containers. Whether food is to be processed for storage or for immediate consumption stipu- lates the sequence and nature of cooking activi- Cooking and Cooking Systems ties (Stahl 1989; Turner et al. 1990:30). If a large number of hands are available, the cooking sys- Ultimately, cooking entails manipulating the tem may be designed to accommodate simulta- temperature, moisture, and pH regime of food as neous root harvesting, fuel collection, and oven well as its surface area so that specific, desirable construction. If other duties must be met and lit- physical and chemical changes can take place. tle time for food preparation is available, peo- Nutritionists and others (Penfield and Campbell ple may draw from previously processed ready- 1990; Stahl 1989) tell us that by processing tissues to-eat foods. And, as Stahl (1989) has recently in this way, people may do one (or all) of several emphasized in her review of food chemistry of things. They advance the digestion process, so cooking strategies, in designing their cooking that more energy and nutrients can be obtained systems, people may be attempting to maximize Food Composition and Heat Treatment: Pit-Hearth Cooking

Figure 1. Cooking system design parameters. nutrient value, minimize toxins, or compromise importantly, on the physical and chemical prop- these two ends. erties of the raw tissues. Food chemistry is but one of the factors that people consider in designing their cooking sys- Heat Treatment Effects tems. Heat treatment is but one of the strategies available for altering tissues. The following sec- Extrasomatically advance digestion. To un- tions consider the relationship between the two. derstand how cooking advances digestion, we need first to understand raw tissue composition and the form in which those tissue components Food Composition and Thermal Processing must be in order to be absorbed into the human bloodstream. Food tissue is composed of pro- As already mentioned, the thermal processing teins, carbohydrates, lipids, minerals, enzymes, has three effects on food (Penfield and Campbell and water. The tissue polymers of protein, car- 1990): increasing its nutrient density; detoxifying bohydrates, and lipids within food vary in their it and removing pathogens and other sources of complexity and in the nature and strength of illness; and, extending its shelf life. (Of course, the links between constituent units. For exam- cooking also enhances the flavor of foods, an ob- ple, 98% of the carbohydrates found in dried figs servation that begs the question of how selection are in the form of simple sugar, while only 6% operated to make some processed foods palat- of the carbohydrates in white potatoes are in this able to humans and others not.) The selection of a form (Briggs and Calloway 1984:27, Table 2-1). particular heat treatment depends on which sev- Significantly, only small and simple polymers eral of these three effects is desired and, equally are absorbed within the small intestine. There- L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) fore, complex or large carbohydrates, proteins, tion. Both the microorganism and its associated and lipids must be broken down by chemical or toxin, however, may be controlled by heat treat- physical processes into smaller polymers. ment or by changing the acidity or salt content of the food product. For Staphylococcus aureus Hydrolysis is one process by which complex and Salmonella sp., in meat products, and Clos- molecules are cleaved into smaller molecules tridium botulinum, in plant and animal products, through the uptake of a water molecule. In the temperatures above 50°C are effective at reduc- gastrointestinal tract, the long, complex mole- ing the risk of illness (Penfield and Campbell cules of the protein, carbohydrates, and lipids 1990:257–263). are hydrolyzed to more fundamental constitu- In addition, plants make use of a wide vari- ents by digestive enzymes. It is these fundamen- ety of secondary compounds to defend them- tal units that are absorbed and used for a vari- selves from predation by insects and other her- ety of tasks, e.g., oxidized to provide energy, or bivores. These compounds, including alkaloids, remodeled for use in enzymes or for structural glycosides, and phenolics, and others have le- purposes. Because, however, of how digestion thal or sublethal effects if ingested (Leopold is regulated by the body and the nature of the and Ardrey 1972; Ames 1983; Jackson 1991). We chemical bonds that join the fundamental units, humans register foods with many of these com- not all proteins, carbohydrates, and fats that pounds as bitter or unpalatable and have de- make it through the small intestine are broken vised an elaborate series of procedures for de- down into the simpler polymers needed by the toxifying them, as reviewed by Johns and Kubo body. In general, carbohydrate-rich foods move (1988). For example, protein-based toxins in through the alimentary tract faster than protein- many seeds may be degraded by exposing them rich foods, which in turn move faster than fatty to heat; the proteins become denatured (see be- foods; foods with a mixture of proteins and fats low) and are unable to perform (Leopold and leave the stomach even more slowly (Briggs and Ardrey 1972:512). Volatile fatty acids, like the Calloway 1984:97). cyanogens occurring in manioc and the cabbage By beginning the hydrolysis process extraso- family, are oxidized when exposed to heat (Leo- matically, more of the complex polymers in plant pold and Ardrey 1972:512). In the case of many and animal tissues can be broken down to the roots (Johns 1989:505), however, the secondary simple form required by the body, leaving the compounds are neither of these and heat treat- digestive system to complete the process. Hy- ment is therefore not useful. Indeed, in some drolysis of some tissues, which the body accom- potatoes, heat does not destroy the compound plishes at low temperatures and with the help of and may in fact enhance the level of glycoal- digestive enzymes, can also be accomplished by kaloids, creating a variety of neurological and exposing tissues to heat, as well as by subject- gastrointestinal problems (Johns 1989; Jackson ing them to acid or salt baths, or by some com- 1991:517). For many root foods, thus, other de- bination of the preceding. In other words, cook- toxification procedures are followed. In the case ing increases the energy value or energy density of agave use throughout northern Mexico, espe- of the tissues. As discussed below, however, dif- cially toxic subspecies were known and avoided ferent food polymers have different temperature (Gentry 1982:6). Tissue portions especially high thresholds at which hydrolysis occurs. in toxins were rubbed or scraped away (Johns and Kubo 1988). And, as in the case of cheeky Reduce pathogens, harmful microorganisms, yam used by Aborigines in northern Australia, and toxins. Food microorganisms or the tox- the toxins may be dissolved and washed from ins they produce may cause illness upon inges- the tissue (O’Dea 1991:237). Food Composition and Heat Treatment: Pit-Hearth Cooking

Food preservation. Food preservation prac- tation highlights (1) the temperatures required tices are designed to reduce or delay enzy- to promote the desired effects, (2) the necessary matic and microbial changes in food and can be the moisture regime, and (3) when available, the achieved in many ways. For example, blanching rate at which these transformational processes (brief immersion in boiling water or steam) re- occur. These parameters are especially important duces enzymatic activity by inactivating the con- for understanding the design of cooking sys- stituent proteins. Pasteurization destroys some tems and have implications for the archaeologi- pathogenic and spoilage organisms while steril- cal manifestation of the cooking regime. It is use- ization kills almost all microorganisms (Penfield ful to consider foods in terms of the constituents and Campbell 1990:266 –271). And, by increas- of carbohydrates, lipids, and proteins. ing or decreasing the acidity of foods, the opti- Carbohydrates. Carbohydrates occur primar- mal pH required by certain microorganisms is ily in plant tissues and comprise sugar, sugar al- changed and so they die. cohols, reserve or storage polysaccharide (used Many seeds and nuts are rich in lipids and also by plants to store energy), and structural poly- lipases, which are naturally occurring enzymes saccharide (which make up cell walls). The lat- that catalyze lipid hydrolysis. While lipid hydro- ter, the structural saccharides, include cellulose lysis results in more digestible shorter-chain fatty and pectin and are commonly glossed as dietary acids, it also results in, for stored nuts or seeds, fiber. Simple sugars include glucose and fruc- the accumulation of free fatty acids and a vari- tose, among others, and are found in fruits and ety of carcinogens (Ames 1983:1258), not to men- vegetables. tion a rancid flavor. Brief exposure to heat- dis The reserve polysaccharides include starch, ables lipase and therefore enables the long-term a glucose polymer, and, in about 15% of the storage of seeds and nuts (Penfield and Camp- world’s flora (Hendry and Wallace 1993:136), bell 1990:341). fructan, a polymer of fructose and a single glu- Water, especially unbound or freely available cose residue. Starches are found in cereals such water, is critical for most of the biological, chem- as rice and corn and in roots such as potatoes ical, and physical changes that tissues undergo. and taro. Jerusalem artichoke and chicory are By reducing tissue water content, or by binding two commercially exploited sources of fruc- water to the chemical structure of the tissue, e.g., tans; Table 1 lists other fructan-bearing plants by adding sucrose, salt, or glycerol, the growth used by humans today and in the past. Fruc- of bacteria, molds, and yeasts is arrested and en- tan is a much more readily tapped energy re- zymatic activity drops (Penfield and Campbell serve in plants than is starch and seems to be 1990:99–101). a preferred storage form in semiarid and tem- Thus, heat treatment directly destroys poten- perate areas where sudden changes in grow- tial spoilage microorganisms, deactivates pro- ing conditions occur (Pollack and Chatterton tein-based enzymes that might cause the degra- 1988:135–136). dation of the tissue, and removes a component The heat treatment of carbohydrates is pre- vital to the actions of spoilage agents, water. dominantly concerned with enhancing digestion and reducing water content so that they can be Food Composition and Thermal Processing stored. In addition, carbohydrate-rich seeds may contain protein-based secondary compounds, Given this overview of the effects of heat treat- which can be detoxified as discussed for protein ment on tissues, let us look more closely at the in general, below. composition of food and how manipulation of The digestibility of a carbohydrate is related its composition through heat treatment results to its polymer size and structure. Simple sugars in these effects. The following technical presen- are small enough polymers that they are read- L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Table 1a. Fructan-Containing Plants in Monocotyledonous Families Identified as Food (or Other) Plants by Incoll and Bonnett (1993: 316, Table 3) and Others (see References) Common Edible (other) Chemistry Plant name part (structure; DP)a Reference Liliaceae Allium ampeoprasum Leek, Leaf, bulb great- headed garlic Allium cepa Onion, Bulb DP3–12 (DP3) Darbyshire and Henry shallot (1981) Allium porrum Leek Bulb DP3–12 (DP12) Darbyshire and Henry (1981) Allium sativum Garlic Bulb Inulin; DP3–50 Darbyshire and Henry (1981) Asparagus racemosus Tuber Arthropodium aff. Tuberous roots Neokestose; DP4 Incoll et al. (1989) milleflorum Asphodelus aestivus Asphodel Tuberous root Bulbine bulbosa Corm Inulin; DP3–5 (DP5) Incoll et al. (1989) Caesia calliantha Tuberous root Inulin; DP3–>5 Incoll et al. (1989) (DP > 5: 41%) Camassia Camas Bulb Inulin Yanovsky and Kingsbury (1938) Dichopogon strictus Neokestose; DP3–>5 Incoll et al.(1989) (DP > 5: 89%) Erythronium Glacier lily Corm Inulin Turner et al. (1990:122) gradiflorum

Agavacae Agave vera cruz Agave Meristem Branched; DP32 Meier and Reid (1982: 447) Cordyline terminalis Palm lily Tuber Branched; DP15 Meier and Reid (1982: 449) Dracaena australis Tuber Yucca filamentosa (Stem) a Underlined values indicate a dominant degree of polymerization in the raw state. ily absorbed into blood vessels in the small in- latinization and pasting. As heat is applied, the testine. In contrast, the structural carbohydrates starch granule swells with water (gelatinization) are undigestible by alimentary enzymes, but and, if held at high temperatures (below 95°C some are broken down by bacteria in the large for most starches) for several minutes, eventu- intestine. The reserve polysaccharides simi- ally implodes (pasting). Upon rupturing, gran- larly resist digestion. With heat treatment, how- ule integrity is lost and amylose, one of two ever, starches and fructans are affected both starch fractions, is released. Ruptured starch physically and chemically and become more molecules are much more susceptible to hydro- digestible. lysis by digestive enzymes than are intact starch Upon exposure to moist heat, starch granules granules (Greenwood and Munro 1979b:385– experience the physical transformations of ge- 387). In addition, dextrinization, the hydrolysis Food Composition and Heat Treatment: Pit-Hearth Cooking

Table 1b. Fructan-Containing Plants in Dicotyledonous Families Identified as Food (or Other) Plants by Incoll and Bonnett (1993: 318, Table 4) and Others (See References) Edible (other) Chemistry Plant Common name part (structure; DP) Reference Asteraceae Arctium lappa Edible burdock Tap root Cichorium intybus Chicory Taproot Inulin; in cultivars, Vukov et al. (1993:341) DP > 10 (70% mass) Cirsium oleraceum Meadow cabbage Taproot Helianthus tuberosus Jerusalem Tuber Inulin; in cultivars, Vukov et al. (1993:341) artichoke DP > 10 (50% mass) Inula helenium Elecampane Root Inulin Microseris lanceolata Murnong Tuberous root DP > 5 (91% mass) Incoll et al. (1989) Polymnia sonchifolia Yacon Tuber DP3–10 Ohyama et al. (1990) Scorzonera hispanica Black salsify Taproot Taraxacum officccinale Dandelion Taproot Inulin Yanovsky and Kingsbury (1938) Campanulaceae Campanula rapanculus Rampion Taproot Boraginaceae Symphytum officinale Comfrey Leaf of starch molecules into dextrin (shorter starch require extensive heat treatment to be maxi- molecules) and maltose, occurs upon exposure mally digestible. to an intense, dry heat. Fructan is the other, less familiar, reserve car- A similar kind of hydrolysis occurs during bohydrate. In humans, two of the four forms it digestion through the actions of enzymes in sa- is commonly found in, inulin and levan, cannot liva and pancreatic fluids. Thermal process- be digested by acid hydrolysis or by enzymes in ing, however, appears to significantly advance the human stomach or small intestine (Hendry gelatinization and dextrinization of the root and Wallace 1993:128). Instead, these fructans starches, thereby increasing their energy value pass into the colon where they are almost com- (Greenwood and Munro 1979b). For example, pletely metabolized by bifidobacteria. Neverthe- Hellendoorn and colleagues (1970) report that less, only about 50% of the potential energy in only 10% of raw potato starch is digestible but uncooked fructan is available to humans. There that after 15 min of cooking, 75% is digestible; are several benefits, however, to consuming raw after 40 min of cooking, 90% is digestible. In fructan-rich foods, which may have been impor- the case of foods with higher proportions of the tant to traditional populations. First, its use by amylose (as opposed to the other starch faction, bifidobacteria in the colon suppresses the pro- amylopectin), gelatinization occurs at relatively duction of intestinal putrefactive substances, re- high temperatures (compare data presented in lieving constipation (but also producing flatu- Greenwood and Munro 1979b) and these same lence). Second, for modern populations, the fact foods appear to have longer digestion times that it is sweet to the taste but contributes rela- (Thorburn et al. 1987:105). As discussed below, tively few calories (Incoll and Bonnett 1993:309; many bushfoods from arid and semiarid areas Roberfroid 1993:105, 132–133; Vogel 1993:66–67) appear to be of this variety and may therefore makes it an attractive food component. L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Figure 2. Percentage composition of Jerusalem artichoke syrup heated for varying lengths of time at 100°C and pH 2.5; data from Yamazaki and Matsumoto (1993:357).

Additionally, we can also speak of the nutri- In addition to advancing tissue digestion, the tional benefits of consuming hydrolyzed inulin. heat treatment of carbohydrate-rich foods also Through the application of heat (and manipula- results in a more storable food product because tion of pH), inulin can be depolymerized (i.e., its of water loss. Through dehydration, water is ei- polymer length reduced) and hydrolyzed. Fig- ther lost or so bound to other molecules that it is ure 2, for example, illustrates the hydrolysis of a effectively unavailable to bacteria and other mi- Jerusalem artichoke syrup as heat is applied for croorganisms. For this reason, dehydrated fruits varying lengths of time. The degree of polymer- and other carbohydrate-rich foods can be stored ization (DP; polymer length) decreases over time without spoiling for great lengths of time. while the number of fructose (and glucose) res- Lipids. Simple fats are composed of glycerol idues simultaneously increases. Pit-baking pro- and up to three fatty acids and are found in both duces the same result; Malouf (1979:37) reported plant and animal tissues. Polymer length and the that pit-baked camas resulted in a fructose- degree to which that fatty acids have bonded or rich product. Fructose is one of the simple sug- become saturated with hydrogen are especially ars (along with glucose) that is readily digested important in determining their digestibility. by humans. It is about twice as sweet as glucose Short-chain fatty acids are better absorbed than and 1.6–1.8 times sweeter than sucrose (common long-chain fatty acids, as are unsaturated fats table sugar; Greenwood and Munro 1979a:40); compared with saturated fats (Briggs and Callo- many accounts of the traditional processing of in- way 1984:104). ulin-rich foods emphasize the very sweet nature Heat treatment of fatty foods seems to have of the final product. In the past, people made use four goals, which are consistent with the over- of these same principles of food chemistry to re- all effects of food preparation. First, subject- alize up to a 100% increase in the energy avail- ing moist foods to high temperatures, as in pan able from fructan-bearing plants. frying, results in lipid hydrolysis (Penfield and Food Composition and Heat Treatment: Pit-Hearth Cooking

Campbell 1990:341–342). Lipid hydrolysis in- Finally, upon heating, the fat in tissues dis- volves breaking the ester linkage between glyc- perses. This dispersion appears to contribute to erol and a fatty acid and requires one molecule the rapid transfer of heat throughout foods (Hol- of water for every ester linkage that is broken. mes and Woodburn 1981:272– 273). Also, disper- The resulting glycerols and short-chain fatty ac- sion results in meats that are perceived as both ids are more digestible (Briggs and Calloway flavorful and tender, although both the reason 1984:99). for this tenderness and its relationship with lipid Second, as noted above, heat may be used to hydrolysis are still being researched (Penfield inactivate or denature enzymes, such as lipases, and Campbell 1990:196). that catalyze undesired lipid hydrolysis and ran- Proteins. Proteins are large complex chains of cidity. The low heat, dry roasting of piñon nuts amino acids found in animal tissues, milk, eggs, thus conserves the food value and flavor of the and legumes. Again, small chains of amino ac- nuts during storage. ids can be further broken down in the mod- Other effects of heating fats concern fat re- ern human alimentary tract; large chains may covery, reducing cooking time, and prepar- pass through the alimentary tract only partially ing flavorful and more chewable foods. Short- digested. chain fatty acids have lower melting points Heat, unsurprisingly, helps in several ways to than long-chain fatty acids and are therefore make protein-bearing foods more nutritious and more likely to be liquid at room temperatures. also less likely to induce sickness than uncooked And, fatty acids that are maximally saturated protein. Exposed to heat, protein becomes dena- with hydrogen have a higher melting point tured, that is, the protein peptide chain unfolds than do unsaturated fatty acids (McWilliams as weaker bonds responsible for the structure 1993:263 –266). In general, lipids of a plant or- of the protein molecule give way (Penfield and igin are rich in unsaturated fatty acids, while Campbell 1990:111, 190–192). Unfolded or dena- those from domesticated animals contain tured protein is more readily chewed (Brace et substantial amounts of saturated fatty acids al. 1987), hydrolyzed, and has increased solubil- (Medved 1986:108); animals that are free to se- ity. It is thus more easily digested by proteolytic lect their diet, however, appear to deposit rel- enzymes (FAO 1990:75) and resolved to its com- atively more unsaturated fats, compared with ponent amino acids. domesticated animals (Crawford 1968; Craw- Muscle protein becomes denatured with ex- ford et al. 1970). posure to temperatures between 40 and 60°C. The melting point of lipids is of concern here At higher temperatures (for beef, those above because it has implications for the recovery of 60°C), the myofibullar proteins coagulate and those lipids, especially by traditional popula- lose their water-holding capacity; we perceive tions in generally fat-poor environments. Oils muscle fibers cooked at high temperatures as are extracted from plant tissues with presses or tough. Also, at higher temperatures, critical hot baths. To extract solid fats, which are com- amino acids such as cystonine are destroyed. posed of crystals in oil, the crystal-to-liquid ratio High-quality muscle fiber, then, requires little must be reduced using heat so that the lipid can exposure to heat. be expressed from the tissue. Traditional popu- Muscles also contain, however, several connec- lations exploited this feature to render fat from tive tissues that support muscle fiber and attach muscle tissues and bones. Similar techniques are the muscle to bone. Collagen is one of these con- commercially used today to produce pork lard nective tissues and it, too, is composed of pro- using steam and beef tallow using dry heat (Pen- tein. When heated, collagen fibers denature and field and Campbell 1990:348). contract; we perceive lightly cooked collagen- 10 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) rich meat as being tough. But, when heated to the protein-based enzymes responsible for pro- temperatures higher than 65°C in the presence of ducing toxins in seeds especially is a final impor- water, collagen is hydrolyzed to gelatin, thereby tant application (Leopold and Ardrey 1972:512; making the meat more tender overall (Penfield Johns 1989:505). and Campbell 1990:194 –195). In sum, food tissues with different composi- The location of the muscle on the body and tions require cooking systems with specific pa- the degree to which it is exercised influences the rameter values to promote tissue digestion; to amount of connective tissue present and, there- denature and destroy bacteria, microorganisms, fore, the preparation of the tissue. For this rea- and toxins; and to reduce the water content of son, the darker meats in chicken, for example, are food to be stored. Table 2 summarizes the infor- tougher than the white meat. While the amount of mation presented above in terms of cooking re- connective tissue is the same in younger and older gime and the effects on tissues treated in this animals, the number and strength of bonds be- fashion. tween peptide chains are greater in older animals; thus, meat from older animals is tougher than that from younger animals and requires longer Stewed, Roasted, Boiled, Dried: cooking (Penfield and Campbell 1990:194–195). Patterns in Food Processing Other factors such as the amount of bone present, the fiber density (high in young animals, lower in In light of the body of reference knowledge as- adults), and fat content influence the time it takes sembled above, we are in a position to examine for the internal temperature of a piece of meat to and appreciate some of the variation observed increase and, therefore, for protein denaturation in ethnographically documented traditional to occur (Holmes and Woodburn 1981). cooking systems. Relatively comprehensive, Thus, as most cooks know (Rombauer and but underdetailed, surveys of food preparation Becker 1975:450–452), preparation of muscle fiber techniques are available for several North Amer- to optimize its food value depends on the size of ican hunter–gatherer and agricultural groups, the animal, its age, and the amount of connec- as listed in Table 3. An examination of the tradi- tive tissue present. Muscle with little collagen, tional methods used by these groups to prepare such as that attached to the dorsal portion of the animal and plant foods suggests that specific spine on a large quadruped, needs only a brief ex- heat treatment methods were effective in maxi- posure to moderate temperatures; broiling, pan-fry- mizing food energy density and producing other ing, and roasting suffice in this case. When more desirable effects. connective tissue is present, as will be true for Focusing first on animal foods, Table 4 pres- most carcasses of wild species, then a longer ex- ents proximate analyses of muscles by species, fo- posure to moist heat is needed to convert the col- cusing on water, protein, and lipid content. Also lagen to gelatin; i.e., braising and stewing is rec- listed is a lipid-to-protein ratio, calculated by di- ommended (Medved 1986:232). viding the lipid by the protein mass. Two series Besides making muscle protein more chew- are represented. The first, from Anderson (1989), able through denaturation, other applications of is based on raw tissues and involves mean val- protein denaturation are important to traditional ues computed over more than a few specimens. populations. Denaturing lipases, which are com- The second comes from Kuhnlein and colleagues posed of amino acids and which might induce (1994) and is based on samples taken from an- lipid hydrolysis, is one example. In addition, de- imals that had been killed and heat treated us- naturing bacteria and parasites so that they are ing traditional techniques by Sahtú Dene and unable to perform is another. Finally, denaturing Métis individuals. For none of these animals do Food Composition and Heat Treatment: Pit-Hearth Cooking 11

Table 2. Heat Treatment Model Moisture content Heat source temperature Dry Moist Very low (.40°C) Sun drying: detoxification, dehydration Low (40–55°C) Sun/fire drying: detoxification, dehydration Medium (55–95°C) Roasting: protein denaturation Boiling: starch gelatinization, pasting; (large portion; long time) Stewing/Pot Roasting: collagen hydrolysis, lipid hydrolysis High (.95°C) Broiling: protein denaturation Blanching (short exposure): bacteria denaturation; (small portion; short time); Pasteurization (medium exposure): Baking: dextrinization, detoxification, enzyme denaturation; inulin hydrolysis Sterilization (long exposure): detoxification, microorganism denaturation we have information on season of death, age, or Table 3. Ethnographic Surveys of Food and Food sex. Yet, as compiled by Speth and Spielmann Preparation (1983:9–12), we know that carcass fat levels may Group Reference vary as much as 200–500% during an annual cy- cle, depending on individual species, pregnancy, Acoma/Laguna Castetter (1935) and so forth. The degree to which muscle lipid Coeur d’Alêne Teit (1930a, 1930b) levels and total carcass fat are coupled and thus Plains Cree Mandelbaum (1979) the degree to which large seasonal variations oc- Dakota Gilmore (1919) cur in muscle lipids are unclear. Table 4 seems to Flathead/Kutenai Hart (1976) indicate that such variation may occur (compare Isleta Castetter (1935) rabbit values in each series); however, it may be owed to seasonal variation in fat content or com- Mescalero Basehart (1974) positional variation among subspecies.1,* Navajo Castetter (1935) For the purposes of this discussion, I will sim- Nez Perce Hart (1976); Spinden (1964) ply note that several species—bear, muskrat, Omaha/Ponca Gilmore (1919) opossum, horse, and beaver—appear to display Owens Valley Paiute Steward (1993) rather higher lipid-to-protein values; the other Sun Valley Paiute Kelly (1932) species, i.e., elk, moose, and so forth, display Pawnee Gilmore (1919) lower values. In Tables 5 and 6, the designations Pima Russell (1908) of ‘‘high lipid content’’ and ‘‘low lipid content’’ Sanpoi, Nespelem Ray (1932) follow from this pattern. In addition, the ethnographic accounts at times clearly refer to the fatty Northern Shoshone Lowie (1909) nature of specific animals such as ground-hog, Tarahumar Pennington (1963) otter, and badger; this information was used to Tepehuan Pennington (1969) indicate ‘‘suspected lean’’ or ‘‘suspected fatty’’ Tewa, Hopi Castetter (1935) meats when analytic information on lipid con- Thompson Indians Turner et al. (1990) tent could not be found. Winnebago Gilmore (1991)

* See Notes section at end of pages for all footnotes. 12 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Table 4. Proximate Compositional Analysis of Protein-Rich Tissues, Ordered by Lipid/Protein Ratio Tissue component (/100 g muscle)a Raw (or laboratory heat treatment) Traditional selection and heat treatment (Anderson 1989) (Kuhnlein et al. 1994: 151) b b H2O Protein Lipid L/P H2O Protein Lipid L/P Common name N (g) (g) (g) ratio N (g) (g) (g) ratio Raccoon 1 (cooked) 54.3 29.2 14.5 0.49 Bear 1 (cooked) 53.55 32.42 13.39 0.41 1 73 22 4 0.18 Muskrat 4 69.35 20.76 8.1 0.39 2 68 25 4 0.16 Opossum 1 (cooked) 58.3 30.2 10.2 0.34 Horse 119 72.63 21.39 4.6 0.22 Beaver 3 70.97 24.05 4.8 0.20 2 75 19 5 0.26 Boar 2 72.54 21.51 3.33 0.15 Caribou 30 71.45 22.63 3.36 0.15 6c 73 23 1.9 0.08 2d 75 23 1.5 0.07 Squirrel 16 73.83 21.23 3.21 0.15 Deer 33 73.57 22.96 2.42 0.11 Goat 92 75.84 20.6 2.31 0.11 Rabbit 30 74.51 21.79 2.32 0.11 2 72 26 1.2 0.05 Antelope 54 74.08 22.38 2.03 0.09 Bison 33 74.57 21.62 1.84 0.09 Bisone 30 74.5 21.7 1.9 0.09 Water buffalo 80 76.3 20.39 1.37 0.07 Elk 22 74.38 22.95 1.45 0.06 Moose 35 75.55 22.24 0.74 0.03 a Significant digits as presented in the original material. b L/P ratio: lipid/protein ratio. c Barrenland caribou. d Woodland caribou. e Marchello et al. (1989: 178).

How are meats with varying amounts of pro- ethnographic accounts. [Similarly, insects with tein and fats prepared? Table 5 abstracts ethno- very high organism lipid levels, such as bee pa- graphic information on food preparation and pae (lipid/protein ratio, 0.27) and termites (lipid/ this information is summarized in Table 6. Three protein, 1.22), along with crickets and grasshop- general patterns are evident. First, meat that is pers, are often described as being either pit-pro- boiled is primarily from species with low lipid/ cessed or parched (Ebeling 1986:101)]. Addition- protein ratios. In addition, the ethnographic ac- ally, some organs and segments of carcasses are counts made frequent reference to boiling dried especially high in fat. For example, Kuhnlein and meats. Second, both lean and less lean meat may colleagues (1994) report fat proportions of 27 and be roasted in ash, coals, or hot sand and for short 43% for beaver tails and feet, respectively; these amounts of time, i.e., from 10 min to 1 h. Third, were also roasted. One possible exception to this pit-roasting is carried out for meat that has a rel- last trend is the pit-roasting of analytically lean atively high lipid-to-protein ratio as indicated in rabbits and squirrels by the Surprise Valley Paiute Table 4 or as interpreted from comments in the (Kelly 1932:93), about which more is said below. Food Composition and Heat Treatment: Pit-Hearth Cooking 13

Table 5. Meat Tissue by Thermal Processing Technique, Organized by Fat Content (See Table 4) Thermal processing technique Common name (fat content code) Boil Broil Spit roast Coals roast Pit roast Group (reference) Low fat content Antelope X X H S. V. Paiute (Kelly 1932:92) Bison X Nez Perce (Spinden 1964:207) X X Sit Plains Cree (Mandelbaum 1979:59) X Crow (Linderman 1962:253) Deer X Nez Perce (Spinden 1964:207) H S. V. Paiute (Kelly 1932:92) X Tepehuan (Pennington 1969:130) Elk X Nez Perce (Spinden 1964:207) Rabbit X Paiute (Lowie 1924:196–197) X S. V. Paiute (Kelly 1932:93) X X Tarahumar (Pennington 1963:91) Squirrel X O. V. Paiute (Steward 1933:245) X S. V. Paiute (Kelly 1932:93) Suspected low fat content Fawn X Nez Perce (Spinden 1964:207) Mice X N. Shoshone (Lowie 1909:189) X O. V. Paiute (Steward 1933:255) S Tarahumar (Pennington 1963:92) Rodent X S. V. Paiute (Kelly 1932:93) X X Tarahumar (Pennington 1963:92) X X Tepehuan (Pennington 1969:125) X Pima (Russell 1908:82) Sheep X Nez Perce (Spinden 1964:207) Shellfish X Sanpoil, Nespelem (Ray 1932:107) St Nez Perce (Spinden 1964:206) High fat content Bear X Nez Perce (Spinden 1964:207) X Sanpoil, Nespelem (Ray 1932:106) X X S. V. Paiute (Kelly 1932:92) Raccoon X S. V. Paiute (Kelly 1932:92) Suspected high fat content Badger X S. V. Paiute (Kelly 1932:92) X Tarahumar (Pennington 1963:93) Crickets X N. Shoshone (Lowie 1909:188) X S. V. Paiute (Kelly 1932:93) Eel X Sanpoil, Nespele (Ray 1932:107) Grasshoppers X N. Shoshone (Lowie 1909:188) Groundhog X O. V. Paiute (Steward 1933:255) X S. V. Paiute (Kelly 1932:92) Otter X S. V. Paiute (Kelly 1932:93) X X Tarahumar (Pennington 1963:93) Pig X Tepehuan (Pennington 1969:128) Porcupine X S. V. Paiute (Kelly 1932:93) Salmon X Sanpoil, Nespelem (Ray 1932:106) X X X Nez Perce (Spinden 1964:206) Woodchuck X Sanpoil, Nespelem (Ray 1932:107) Unknown fat content Fish X Tepehuan (Pennington 1969:133–134) Fowl X X Tepehuan (Pennington 1969:114–116) Lizard, snake X N. Shoshone (Lowie 1909:189) Turtle X Sanpoil, Nespelem (Ray 1932:106) H, heads only; S, singe; Sit, situationally; St, steam. 14 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Table 6. Summary of Thermal Processing Technique by Tissue Lipid Content

Thermal processing technique

Lipid content Boil Broil Spit roast Ash/coals roast Pit roast

High 1 3 Suspected high 4 1 7 6a Low 6 2 4 2 (rabbit, squirrel) Suspected low 4 1 5 2 (rodent, shellfish) Unknown 2 1 2 3 Total 16 2 5 20 13 a Includes heads, which likely contain fat-rich brains.

These ethnographic results are consistent with changed dramatically in reliability and precision our theoretical understanding of food composi- over the last 50 years (Southgate 1991). For ex- tion and heat treatment. Dried and lean meats ample, of the 66 plant foods they examined, only are boiled to restore moisture, which will assist 6 were reported as containing inulin. Erythro- the action of digestive enzymes. Fatty meat tis- nium grandiflorum was not among these 6, but is sues may be boiled to further lipid hydrolysis mentioned by Turner and colleagues (1990:122) and to melt and ex-press tissue lipids that may as inulin-bearing. It is unclear if this discrep- then be recovered and used for other purposes, ancy is owed to technological advances in com- e.g., to make pemmican (Reeves 1990). Fatty positional analysis or to differences in the work- meats may also be pit-roasted in a moist environ- ing definitions of starch, fructan, inulin, dietary ment for several different reasons, the primary fiber, and other dietary components, which have of which is likely the ease of preparing of a large also seen a great deal of elaboration over the past piece of meat. That is, the fat enables heat to per- 60 years (Southgate 1991). In any event, more re- meate the tissues very rapidly and the protein cent (and therefore, reliable) detailed analyses of is thereby quickly denatured. It is not surpris- these plant foods are not available and the fol- ing, then, that large aggregations of people and lowing analysis makes due with values from Ya- roasts of fatty meats are often associated. Lipid novsky and Kingsbury. and collagen hydrolysis would also be promoted Table 8 again presents ethnographically re- by such treatments, which would, again, serve to ported food preparation techniques—none, dry, increase the energy density of these tissues. boil, dry bake, moist bake, other—by plant food How do food composition and preparation as organized by major component; this infor- correlate in plant tissues? Table 7 presents the mation is distilled in Table 9. (Where question compositional analysis of various plant foods marks occur in Table 8, meaning that mode of utilized by North American native populations preparation could not be distinguished among and summarized by Yanovsky and Kingsbury two or three techniques, these were accordingly (1938). These data are used here with a great deal apportioned; hence the fractional values in Ta- of caution since the techniques and protocols for ble 9.) The few plant foods for which we have analyzing and reporting carbohydrates have both information on composition and process- Food Composition and Heat Treatment: Pit-Hearth Cooking 15 7.6 7.8 0.0 5.3 5.2 5.4 4.2 3.0 5.5 2.4 0.4 8.6 5.5 7.9 7.9 9.1 1.3 10.4 17.5 11.7 12.5 14.4 13.3 24.5 1.0 4.7 5.5 1.6 1.1 0.0 7.6 1.0 0.9 0.3 0.6 53.5 5.0 9.8 7.7 9.2 3.8 10.4 13.9 19.2 18.4 20.2 26.3 17.7 21.1 13.1 15.8 23.0 43.6 15.4 25.8 19.8 14.9 25.9 19.8 17.8 25.9 21.6

0.0 7.5 0.0 0.0 0.0 0.0 4.4 0.0 5.5 0.0 31.9 26.9 20.0 44.8 15.3 22.4 20.2 39.9 33.3 30.0 29.3 36.8 17.5 22.1 6.4 0.0 9.2 9.4 0.0 6.8 1.3 5.3 1.9 4.1 7.1 4.0 0.7 6.0 55.2 25.4 11.6 20.0 13.8 32.0 22.3 36.6 41.1 68.8 11.2 15.8 11.4 3.7 2.7 2.7 4.5 1.6 1.7 2.6 2.0 8.2 1.6 6.9 0.0 0.2 0.8 1.0 3.1 6.2 2.4 4.3 0.9 1.3 0.7 1.5 29.0 12.2 11.6 10.5 O Glucose Sucrose and fructose Starch Hemicellulose Lipids Protein 2 8.2 8.5 7.9 9.5 6.3 6.5 5.1 9.0 7.9 7.3 64.4 43.8 34.3 75.5 20.8 25.2 69.5 63.0 70.2 20.0 10.5 74.5 37.2 18.1 77.2 14.1 72.2

a —Bulbs (1) Nutt. (saskatoon)—Berries (2) —Tubers (1) Torr. (mesquite)—Pods, beans (4) (Hook and Arn.) A. Gray (yampa, squawroot)— (yellow mariposa)—Bulbs (1) Pursh (Indian breadroot)—Roots (2) Salisb.—Bulbs (1) —Bulbs (1) (bitterroot)—Roots (4) (Kous)—Roots (2) (L.) Kuntze (hogpeanut)—Tubers (1) (Wild.) Pers. (American lotus)—Tubers (1) (wild onion)—Bulbs (2) L. (potatobean)—Underground fruit (1) (nut pine)—Nuts (2) Mean Composition of Plant Foods (from Yanovsky and Kingsbury 1938), Organized by Gross Pinus edulis Amelanchier alnifolia Falcata comosa Lomatium cous Lewisia rediviva Nelumbo lutea Psoralea esculenta Calochortus luteus Claytonia lanceolata Fritillaria kamschatensis Glycine apios Fritillaria pudica Hookera coronaria Perieridia gairdneri Roots (3) Prosopis glandulosa Prosopis pubescens (screwbean)—Pods, beans (4) Allium nuttali Balsamarhiza sagittata (arrowleaf balsamroot)—Roots (5) Camassia leichtlinii (Leichtlin camas)—Bulbs (1) Camassia quamash (common camas)—Bulbs (6) Erythronium grandiflorun Pursh. (glacier lily)—Bulbs (1) Helianthus tuberosus (Jerusalem artichoke)—Tubers (1) Alectoria jubata (L) Arch.—Lichen mass (1) Glycyrrhiza lepidota Pursh (licorice)—Roots (3) Pinus contorta Dougl. (shore pine)—Cambium (2) Scirpus validus (great bulrush)—Roots (1) Yucca baccata (banana yucca, datil)—Fruit (2) High glucose or starch Starch and fructan High fructan (not inulin) Fructans (inulin form) High hemicellulose Table 7. % dry weight Plant (common name)—Edible portion (samples analyzed) % H Lipid Carbohydrate fraction a Turner and colleagues (1990:122) report inulin present, however, Yanovsky Kingsbury (1938) do not. 16 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Table 8. Plant Tissue by Thermal Processing Technique, Organized by Major Tissue Component Preparation technique Dry Moist Plant Raw Dry Boil bake bake Other Group (reference) Lipid Pinus—Nuts albicaulisa X Coeur d’Alêne (Teit 1930a:43) edulis P Mescalero (Basehart 1974:54–56) X Navajo (Castetter 1935:42) Carbohydrate fraction High glucose or starch Amelanchier alnifolia—Berries X X X Thompson (Turner et al. 1990:253–254) Falcata comosa—Tubers U Nebraska Tribes (Gilmore 1919:43–44) Lomatium—Roots canbyia X Paiute (Kelly 1932:101) cous X X X Nez Perce (Spinden 1964:203) X X X Flathead (Hart 1976:26) dissectuma X Nez Perce (Hart 1976:26) leptocarpuma X Paiute (Kelly 1932:101) macrocarpuma X X ? ? Paiute (Kelly 1932:101–102) X X Flathead (Hart 1976:26) ? ? Thompson (Turner et al. 1990:155) Lewisia rediviva—Roots X Nez Perce (Spinden 1964:203) X X Paiute (Kelly 1932:102) X X X Thompson (Turner et al. 1990:244) X Salishan (Ray 1932:100) X Flathead/Kutenai (Hart 1976:49) Nelumbo lutea—Tubers X X Nebraska Tribes (Gilmore 1919:27) Psoralea esculenta—Tubers X X ? ? ? Nebraska Tribes (Gilmore 1919:40–41) X X X X E. Montana Groups (Hart 1976:61) Starch and fructan Calochortus—Bulbs aureusa X Navajo, Hopi (Castetter 1935:19) luteus X Navajo (Castetter 1935:19) macrocarpusa X Paiute (Kelly 1932:102) X X Thompson (1990:19) Claytonia lanceolata— Roots X Coeur d’Alêne (Teit 1930a:92) X X X Thompson (Turner et al. 1990:240) Erythronium—Bulbs grandiflorum X Thompson (Turner et al. 1990:123–124) mesochoreuma X Winnebago (Gilmore 1919:19) Fritillaria—Bulbs camschatcensis Unspecified X Coeur d’Alêne (Teit 1930a:92) Glycine apios—Tubers X ? ? Nebraska Tribes (Gilmore 1919:42) High fructan (not inulin) Perieridia—Roots gairdneri ? ? ? Nez Perce (Spinden 1964:204) oreganuma X X X Paiute (Kelly 1932:101) Prosopis—Pods, beans glandulosa X Acoma/Laguna (Castetter 1935:44) X G Mescalero (Basehart 1974:56–57) X Isleta (Castetter 1935:44) pubescens X G Mescalero (Basehart 1974:63) ? ? Pima (Russell 1908:75) ing means that these results must be considered first being dried for storage. They also point to provocative rather than definitive. They suggest, the occurrence of pit-baking in the case of pine however, that plant foods rich in starches are nuts with high lipid content and for foods high primarily boiled prior to consumption, perhaps in inulin. Food Composition and Heat Treatment: Pit-Hearth Cooking 17

Table 8—Continued Preparation technique Dry Moist Plant Raw Dry Boil bake bake Other Group (reference) Fructan [inulin present or likely (given other members of genus)] Allium—Bulbs acuminatuma X Thompson (Turner et al. 1990:117) cernuuma X Thompson (Turner et al. 1990:118) geyeria X U Nez Perce (Spinden 1964) mutabilea X Nebraska Tribes (Gilmore 1919:19) pleiathuma X Paiute (Kelly 1932:102) recurvatuma X X Isleta (Castetter 1935:15) X Tewa, Hopi (Castetter 1935:15) Unspecified X Salishan (Ray 1932:100) Unspecified ? ? Coeur d’Alêne (Teit 1930a:92) Balsamorhiza—Roots sagittata X Flathead (Hart 1976:20) X Thompson (Turner et al. 1990:176–177) terebinthaceaa X X Paiute (Kelly 1932:103) Camassia—Bulbs Quamash X Nez Perce (Spinden 1964:201) Quamash X Paiute (Kelly 1932:102) Unspecified ? ? Coeur d’Alene (Teit 1930a:92) Unspecified ? ? Salishan (Ray 1932:99) Unspecified S Salishan (Ray 1932:99) Unspecified X X Salishan (Ray 1932:99) Helianthus tuberosa—Tubers X Pawnee (Gilmore 1919:79) X X ? ? Dakota, Omaha/Ponca, Winnebago (Gilmore 1919:79) High hemicellulose Alectoria—Lichen jubata Unspecified X Nez Perce (Spinden 1964:205) ? ? Coeur d’Alêne (Teit 1930a:92) X Salishan (Ray 1932:104) Glycyrhiza lepidota M Dakota (Gilmore 1919:40) Pinus—Cambium contorta X Thompson (Turner et al. 1990:102) ponderosa X Nez Perce (Spinden 1964:205) X Salishan (Ray 1932:103) Unspecified X Mescalero (Basehart 1974:47) Scirpus validus—Roots X Dakota (Gilmore 1919:17) Yucca baccata—Fruit C X Pima (Russell 1908:73) X X Navajo (Castetter 1935:54–55) X Mescalero (Basehart 1974:52–54)

Notes: ? = cooking technique cannot be differentiated; C = catharitic; G = grind; M = medicinal poultice; P = parch; S = steam; U = preparation unspecified. a Compositional data unavailable.

The different traditional techniques used in e.g., freeze-drying, adding salt, are available as the preparation of carbohydrate-rich plant food well). Starchy foods are reported as being dried are not surprising in light of the information as- for storage or boiled. In this latter case, starch sembled above. Foods containing simple sugars gelatinization and pasting are likely being pro- require little treatment prior to consumption; to moted. Foods containing fructan are processed extend their storage life, however, dehydration in a variety of ways. It is likely that the mode of through either sun-drying or heat treatment pro- processing depends on the structure and degree vides two appropriate means (and other means, of polymerization of the fructan. For example, 18 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) onion (Allium sp.; DP 3 –?) might be eaten raw each cooking event. As well, some groups, such (e.g., Nebraska tribes; Gilmore 1919:19), yacon as the Havasupai, the Thompson Indians, and the tubers (DP 3–10) sun-dried first (Zardini 1991), Paiute, seem to pit-roast almost everything. For and mesquite bean pods (unknown DP) some- these groups, it is unclear the degree to which a times pit-baked (e.g., Pima; Russell 1908:75) and mobile lifestyle and a lack of large, portable con- other times pounded into flour (e.g., Shoshone tainers is a confounding factor in this analysis. As and Southern Paiute; Fowler 1995:102). well, fuel availability and conservation may be Finally, inulin-rich foods were reported as an issue with some groups, but not others. In the eaten raw or pit-baked. Raw consumption of in- next section, I attempt to deal with some of these ulin-rich foods may have been undertaken sea- shortcomings by looking at pit-hearth cooking in sonally, when the degree of polymerization a more detailed fashion. had decreased because of frost (Fontana et al. 1993:252). For example, the Thompson Indi- ans of British Columbia preferred to consume Pit-Hearth Food Processing the corms of E. grandiflorum (glacier lily) after they had been in cold storage for a while, when All of this information on tissue composi- they would become sweet like candy (Turner et tion and thermal alteration suggests that cook- al. 1990:123). Raw consumption might also pro- ing strategies should be applied selectively and mote good health by enabling the activities of bi- that a specialized heat treatment, like that of pit- fidobacteria residing in the colon. Pit-baking of roasting or baking, should be employed in very inulin-rich foods would have resulted in up to a specific circumstances. To further define these 100% increase in the energy obtained from these circumstances, a more detailed analysis of eth- foods. It also would have denatured some of nographic accounts of pit-hearth cooking is con- the secondary compounds that might otherwise sidered here. A sample of 110 ethnographic ac- have caused ill effects. The final processing of in- counts, derived from the Human Resources Area ulin-rich foods often involved shaping the prod- File and other sources, is abstracted in Appen- uct into loaves or mats that were further dried, dixes 1 –3. This compilation is hardly exhaustive either in the pit hearth or by the sun, for storage and may overrepresent some applications of pit- (e.g., Fowler 1995; Whiting 1985; Malouf 1979; hearth cooking to an unknown degree. For exam- Thoms 1989). Thus, pit hearths also served as de- ple, camas and other root preparation in north- hydraters, in the final phase of processing inulin- western North America is emphasized, as is the derived fructose cakes. preparation of agave, sotol, and cholla buds in From the understanding of food chemistry and the American Southwest. From the western Pa- human digestive physiology abstracted above, we cific come various descriptions of pit-hearth pro- expect that certain kinds of food will be boiled, cessing of a wide variety of foods, including taro, dried, baked, or roasted, if people, knowingly or sweet potato, pig flanks, packets of fish, and not, are attempting to maximize energy density, other vegetable foods. Because of the likelihood minimize disease possibilities, and extend food of bias in these data, graphs and figures, rather storage life. Ethnographically documented food than confirmatory statistics, are used to explore processing appears generally consistent with these these patterns. goals. There is by no means, however, total con- Figure 3 summarizes the incidence of pit- formity between expected and observed. Several hearth cooking accounts according to known reasons might be given for this state, including the and suspected food composition. These pie fact that the ethnographers were ‘‘normalizing’’ charts reaffirm some of the trends identified ear- what was in fact a fairly situational proceeding; lier, namely, that fatty meats and inulin-bearing unfortunately, we rarely have the full details for plant foods were often pit-roasted. The charac- Food Composition and Heat Treatment: Pit-Hearth Cooking 19

Table 9. Summary of Thermal Processing Techniques by Plant Tissue Composition

Food content Raw Dry Boil Dry bake Moist bake Other

High lipid 1 1 Carbohydrate fractions High glucose or starch 5 9 9.3 1.3 1.3 1 Starch/fructan 3 1 4 .5 2.5 High fructan (not inulin) 1 2.3 1.8 .8 2 High fructan (inulin) 3 1 2 8 1 Hemicellulose 5 4 1.5 1.5 1

Only foods for which species level information on composition is known are summarized here. That is, plant foods marked with footnote a in Table 7 are not represented here. teristics of these and other pit-roasted foods are would cut many green boughs of the chokecherry examined in the following. trees and cover the hot stones a foot deep with them. Upon these we would place thick chunks of buffalo meat, fat and fresh from the plains, sprin- Protein Denaturation and Lipid Hydrolysis kling them with water. On top of the meat went [Plenty-coups reflected on his youth:] ‘My mouth another layer of boughs, then more meat, more waters when I remember the meat-holes,’’ he water, and so on, until the hole was full. Finally said. ‘‘We used to dig a hole in the ground about we spread the animal’s paunch over the hole, as deep as my waist. You have seen many of them covered it all with its hide, put gravel on this, and along the creeks and rivers. We would heat little kindled a log fire. Men kept the fire going all day boulders until they were nearly white and cover and all night yet never burned the robe. The next the bottom of the hole with these stones. Then we morning when we opened the hole to feast, even

Figure 3. Frequency of ethnographic accounts of pit hearth cooking according to composition of prepared product. (Tissues with unknown compositions are not included.) 20 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Figure 4. Quantities of pit-processed foods. Small refers to individual portions, medium to portions consumable by several families or for several family meals, and large to quantities larger than that immediately consumable by several families.

the birds of the plains were made hungry by the In the other 18 accounts of pit-hearth cooking smell of the cooked meat. . . . I have made myself of meat listed in Appendix 1, this pattern of fatty hungry telling you this. I will talk of something meat preparation is supported and two other else to forget meat-holes.’ (Linderman 1962:253) patterns are readily apparent. First, medium to large packages of meat—a bear, a bison, a py- In this account, Plenty-coups is describing thon, the heads of large game animals, and mul- the thermal alteration of bison muscle. Plenty- tiple lizards, doves, rabbits, or other small mam- coups emphasized the high fat content of the mals—are prepared this way (Figure 4). Thus, meat and the delicious flavor of the cooked the pit-roasting of meat appears to be a means meat. While bison meat is not noted for its high for processing in bulk medium to large quanti- fat content (Table 4), groups appear to have sea- ties of meat. sonally and situationally targeted specific bison Second, the heat–time profile of the roasting when they were carrying relatively high fat re- pit is one likely characterized initially by moder- serves (Driver 1990; Speth 1983). Adding water ately hot temperatures that rapidly decline and a to the meat as well as sealing the cooking en- comparatively short cooking time. As indicated vironment so as to prevent the escape of tissue in Figure 5, a rock element was used to store and water indicates a moist heat cooking regime. release heat in only 37% of the 14 cases for which The lengthy cooking time suggests that heat such information is available. Bear and Plenty- treatment was carried out to promote collagen coups’ (above) bison were prepared making hydrolysis and lipid hydrolysis undoubtedly use of a rock heating element and the Havasu- occurred as well. And, some degree of fat ren- pai (Whiting 1985:62–63) inserted a heated stone dering likely occurred since in many accounts in the abdominal cavity of larger animals during of pig pit-roasting, the plant insulating material roasting. In most other cases, however, the sedi- was removed from the pit and the fat drippings ment in which the pit was constructed served as sucked from it (Dwyer 1990:132). the heat reservoir. Food Composition and Heat Treatment: Pit-Hearth Cooking 21

Figure 5. Rock element presence/absence.

Finally, as Figure 6a shows, cooking times of 10 rich foods. Second, the expressed fat drippings, min (for lizards) to perhaps 20 h (Plenty-coups’ which were sucked from insulating material by account of bison roasting; see above) are reported some (perhaps underprovisioned or disenfran- for pit-processing of meat, with most cooking chised) people, might allow for the wider distri- times being around 2–4 h. Certainly, the amount bution of valuable fat in a generally fat-poor envi- of food being processed determined to some ex- ronment. It is likely that some fat was lost in this tent the overall cooking time. As indicated in the way and so it is not impossible that the pit-roast- preceding sections, however, protein requires ing of fatty meat tissues may represent a some- only limited exposure to moderate temperatures what ostentatious display of fat wastage, a propo- to enhance its energy and nutrient value. sition that I cannot presently evaluate. In sum, fatty meats were often, but not exclu- sively pit-processed. These data suggest that tra- Inulin Hydrolysis ditional cooks were making use of the heat trans- [From his notes on life in Polynesia, Rev. Wil- fer property of fat in perennially or seasonally liam Wyatt Gill wrote:] … On a given day great fatty meats to rapidly, evenly, and flavorfully logs of firewood and firestones were collected. prepare large quantities of meat. Such process- Next morning an enormous hole or oven was ing might be useful in either of two situations: dug. All being ready, on the third day a great (1) when presented with a large quantity of fatty crowd of men, women, and children came bearing great uncooked ti roots. The firewood was meat or (2) when hosting a feast in which large at a given signal carefully piled up so as com- numbers of people are to be fed. pletely to fill the hollow; and on the top enor- Two other effects follow from preparing meat in mous stones of black basalt were laid, so as to this fashion. For one, the fatty acid chain length is cover the whole. The wood was now lighted in reduced through lipid hydrolysis, producing fatty different places, and speedily the whole was in polymers that are more readily digestible. This ex- a fierce blaze. When the wood (more than half trasomatic hydrolysis might be especially critical of which was green) was burnt out, the fires- tone, now red-hot, sank to the centre of the deep to people who only sporadically consumed fat- hollow. Now was the moment of danger, when 22 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Figure 6. Ethnographically reported minimum cooking times for (a) meat and (b) plant tissues. Bars represent the in- terquartile range and the median values.

the bravest and most adept selected long green men carefully to pack the roots without burning branches of the chestnut with a hook at the end their feet. As the oven was filled with ti, these in order to pull these heated stones into proper sticks were gradually withdrawn. order, so as to form, in fact, a red-hot pavement, on which the ti roots might be thoroughly In packing these roots the least valuable part and equally baked. This dangerous feat accom- is placed downwards, the slender juicy tops up- plished, laminae stripped off banana stalks with wards. Each person ties something on his own abundance of the juiciest leaves were thickly half-dozen or more big roots, so that there may be strewed over these glowing stones. Green limbs no confusion when the oven is opened. The roots of trees were then laid over the oven, to enable require to be packed closely, to prevent heat from Food Composition and Heat Treatment: Pit-Hearth Cooking 23

escaping upwards. When the oven is at last com- relationship between pit-processing and all fruc- pletely filled up with ti, the biggest roots being in tan-bearing plant parts with high DP and/or an the centre, the whole is covered over with a vast inulin or branched structure. That is, plants with quantity of leaves. Finally the oven is covered in these characteristics were, when processed in with earth to the depth of three feet. A large oven requires two whole days and nights for the con- any moderate to large quantity, always prepared tents to get thoroughly done. Upon opening the in a pit hearth. (As discussed below, however, oven, each person is careful to take only his bun- pit hearths were also used to prepare millet pan- dle of ti roots . . . (Gill 1880:28–29; emphasis in cakes and nuts.) original) In general, large quantities of plant foods (Fig- ure 4) were prepared this way, at least around Gill’s very graphic account tells of the mass and just after the time of contact. For example, preparation of Cordyline terminalis (palm lily) Spinden (1964:201–202) reports that 30–40 bush- tubers.2 Palm lily tubers contain fructan with a els of camas were prepared at a time, Apache degree of polymerization of 15 and a branched groups prepared ‘‘a ton of [mescal] tubers’’ (Rea- structure (Table 1). The long cooking time and gan 1930:293), and Pima families processed ‘‘sev- the hot stone pavement (suggesting high tem- eral large baskets, dishpans, and tubs’’ of cholla peratures) are consistent with the effective hy- buds (Thackery and Leding 1929:414). Several drolysis of inulin to fructose and glucose. As Gill factors are likely responsible for this mass food (1880:28) noted, ‘‘[Processed t]i is remarkably preparation. While many of these plants could sweet and agreeable to the European palate.’’ be utilized at any time, the time during which In other ethnographic accounts, it is clear that the maximum amount of energy could be har- people are promoting this hydrolysis by chang- vested from the plant is just before that energy ing the pH value and moisture content of the is directed to reproduction. Camas was exploited cooking environment. For example, the Pima and just after flowering (Hart 1976:16); agave, just Tohono O’odham made use of chuchk onk (seep- prior to its flowering. Moreover, the plant part weed, inkweed; Suaeda sp.), which translates as must be processed soon after harvesting, else ‘‘black salt,’’ to line the pit in which cholla buds spoilage occurs (Turner and Kuhnlein 1983:214). were roasted (Curtin 1949:58 in Bohrer et al. The window of maximum opportunity for har- 1969:8; Greenhouse et al. 1981:229; Thackery and vesting camas or agave or other similar plants at Leding 1929:414). And, in about 50% of the 24 ac- any one locale, thus, was likely on the order of counts of camas roasting, water was added dur- several weeks. Groups that especially depended ing heat treatment. on camas and agave, for example, kept a close Of the 85 accounts of plant foods listed in watch on those plants and the harvesting and Appendix 2, the majority of them involve the processing of the plant parts was under the strict plants listed in Table 1, all of which contain some direction of a specialist [usually female (Reeve amount of fructan or inulin. In a 1983 article on 1986), in the case of camas, and either male or fe- native food use by North American Northwest male (Fowler 1995:106), in the case of agave]. Isa- Coast Native Americans, Turner and Kuhnlein bel Kelly observed southern Great Basin families (1983:214) highlighted the significance of pit- harvesting agave hearts at increasing elevations processing to the use of camas: ‘‘. . . it would be throughout the spring as agaves in turn sent fair to state that without cooking—and probably up their flower stalks (Fowler 1995:105). Mass without pit-cooking—camas could not have at- processing, then, may be a response to the nar- tained the importance it did in native diets.’’ The row small window of availability and the short data presented here suggest that this statement amount of time allowable between harvesting can be generalized to refer to an almost essential and spoilage. 24 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

In addition to a narrow temporal window, the prehistoric sequence (see discussion below of the laborious and resource-demanding nature historical trends). Many accounts note that the of the processing itself presumably played a size of the pit and heating element was readily role in decisions about the quantities to pro- expanded in plan (rather than in depth) to access. Heat treating even a small portion of commodate the amount of product. Older (more many of the inulin-bearing foods apparently hydrolysis-resistant) roots were placed in the requires an extended exposure to high tem- center where oven temperatures were the high- peratures. Similar to that described by Gill for est, as Gill mentions above. ti, very long cooking times are ethnographi- An additional comment on processing time cally documented for the treatment of camas, is important here. The pit-processing times sotol, agave, and other plants (Figure 6b). The given in Figure 6b include heating to promote length of cooking time likely has to do with the dehydration as well as hydrolysis. Detailed ac- length of the fructan polymer and possibly also counts of agave, camas, and sotol processing its structure. Low DP plants were apparently speak of baking the product for several days, processed for short periods, while plants with removing it from the pit, and kneading the hy- higher DP values were processed for longer drolyzed product into cakes or loaves. These times. For some, like yacon (DP 3), the tubers were either sun-dried or further baked, to re- were only exposed to the sun (Zardini 1991), duce water content (and its availability to mi- while Jerusalem artichokes (DP 20) were some- croorganisms) and, therefore, the possibility times boiled and other times pit-processed. On of spoilage. The final product, whether camas the other hand, structure may be more impor- or agave, had the consistency of plug tobacco, tant. In ti and agave, with DP values of 15 and was easy to transport, and preserved well if 32, respectively, their branched structure may kept dry. David Thompson wrote about his require more energy for depolymerization and companion, Lord Metcalfe, who found two hydrolysis to occur. camas loaves that were 36 years old to be well To achieve a moderate to high temperature preserved, although not as tasty as fresh camas over an extended period, people commonly re- (Hart 1976:17). Similarly, David Prescott Bar- lied upon a rock or stone heat reservoir (Figure rows found an agave stalk edible 60 years af- 5) and a top fire was sometimes built. Indeed, in ter it had been harvested and dried (Bean and 61 of 72 reports of pit-processing of known in- Saubel 1972:34). ulin-bearing foods, the presence of a rock ele- In sum, pit-hearth processing of plant tis- ment was mentioned; a majority of the accounts sues is predominantly associated with mass of camas processing report a fire being built over processing of inulin-rich plant parts. By har- the sealed oven, especially at night. Several ac- vesting and pit-processing such foods, people counts (e.g., Tarahumara; Bye et al. 1975:86–89 were able to take advantage of an intensifiable in Parsons and Parsons 1990:279) indicate that and storable energy source that thrived in areas the food product would be tested to see if it had with few other intensifiable resources. Indeed, cooked enough. If more time was required, the Thoms (1989) emphasizes the degree to which oven might be disassembled and then reassem- camas fields were managed and exploited in bled with a renewed heat source. the American Northwest and, in the Ameri- Given this minimum required level of effort, can Southwest, Fish and colleagues (1990) have people undoubtedly realized an economy of documented extensive mulch fields that appar- scale by processing as much food product at any ently supported agave, perhaps for both food one time as possible, at least in the latter part of and fiber. Food Composition and Heat Treatment: Pit-Hearth Cooking 25

Starch Hydrolysis and Meat Protein Denaturation other accounts, a rainstorm is underway as the … Women and one older man from Namosado oven is prepared. Water would be important in prepared family dishes. They took previously pre- starch degradation as well as in collagen and pared banana leaves, folded at the midrib and of- lipid hydrolysis. ten warmed on the oven, and used these to wrap The role of pit-processing of fatty meats has up a mix of greens, meat and fat of pig, and crum- already been addressed above. Pit-processing bled sago. Watercress, hibiscus, acanth spinach, of starchy foods, however, appears inconsistent or pumpkin tips were chosen by different women as their preferred green. with the results of the earlier section reviewing plant food processing in general. The recent anal- As women completed their packages of food, yses of bushfoods alluded to above, however, in- the men and youths turned to the fire. The wood dicate that not all starchy foods are equally di- had burned away, and the stones were hot. Some gestible. A variety of sweet potato grown in the had exploded, but none had hurtled dangerously Pacific islands is two to four times less- digest from the fire. Using long poles and giant wooden ible than the common potato, both of them hav- tongs the men rearranged the stones, working rapidly to retain the heat; and when the stones ing been baked for 1 h (Thorburn et al. 1987). were ready— or even sooner—they were buried Similarly, in their analysis of traditional Pima beneath fern fronds; and the ten sides of pork, in- foods, Brand and colleagues (1990) observed a terspersed with pumpkins, chokos, and a dozen wide range of in vitro starch digestion; corn was or more game mammals, were laid on top and the most digestible and mesquite and acorns the smothered under more fern. No tubers were in- least. Thorburn and colleagues (1987:105) suggest cluded. Stones from the edge of the oven were that the relative proportions of the two starch repositioned among the slabs of pork. People called to the others who they thought were being fractions, amylose and amylopectin, are respon- slow, jostled through the crowd to the oven, or sible for this variation in digestibility, since am- shouted angrily that someone else had taken the ylose resists digestion more than amylopectin. very place they had chosen for a package of food. Thus, pit-roasting may play an especially signifi- The heap grew. Water was poured over it; and as cant role in the hydrolysis of some starchy foods. stem billowed up, everyone, shouting, laughing, In addition, most of these accounts refer to the and calling for others to help, threw huge leaves over the food. The steam was contained and the preparation and immediate consumption of a heap buried beneath dirt. It was done. The ex- communal meal. That is, large quantities of food citement dissipated, and for two hours the food are being heat treated and then eaten. The excep- cooked (Dwyer 1990:131–132). tion to this pattern is the pit oven prepared for Steensberg (1980:197–201) upon his departure. This description of pit-hearth cooking comes Within individual communities, joint food prepa- from Dwyer’s account of a public feast hosted by ration may be carried out using an earth oven. On one Etolo (Highland Papua New Guinea) com- Bellona Island in Polynesia, for example, Chris- munity for another. It illustrates the steam cook- tiansen (1975:74–75) observed the women of sev- ing of foods with a variety of compositions. In eral families collaborating to prepare the evening other accounts (see Appendix 3), fish, sweet po- meal of fish, poultry, taro, sweet potato and other tatoes, and packets of taro or sego (the latter items in an earth oven. He attributed this joint ef- three being high in starch; FAO 1990) are men- fort to the large investment in labor and time in- tioned. Minimum cooking times of from 1 to 3 volved in getting fuel and attending to heating the h are given, which are consistent with the heat rock element (Christiansen 1975: 75–76). treatment of both starch-rich foods and meats Pit-processing of mixed foods, thus, appears with varying amounts of fat and connective tis- to be associated with the mass preparation of sue. Water is sometimes deliberately added; in foods with similar heat and moisture demands. 26 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Starch-rich foods require limited exposure to was mentioned specifically by traditional peo- moist heat for gelatinization and pasting to oc- ples as being pit-processed to remove bitter- cur and a high temperature heat to support dex- ness (Johns and Kubo 1988). In the case of other trinization. As discussed above for fatty meats, roots (e.g., Balsamorrhiza sp.), toxins were re- limited exposure to moderate heat promotes the moved by mechanically peeling the toxic outer heat transfer throughout the meat, resulting in skin. The lichen that was pit-roasted by north- protein denaturation and lipid hydrolysis. For western North American groups was first ex- this reason, a rock heating element is always tensively washed and pounded to express the present (Figure 5) when this kind of cooking is vulpinic acid it contains (Turner et al. 1990:72). carried out. Interestingly, this heating element For other plant genera with toxin-bearing may have two parts, a lower one, upon which members, specific nontoxic species were tar- the meat is placed, and, amidst the packets of geted. For example, the leaves of many agave starchy foods above, other heated rocks will be species contain steroid saponins, and terpenes, distributed. as well as vitamins (Table 10; Gentry 1982:21). Throughout northern Mexico and the American Southwest, notes Gentry (1982:6), agave species Toxin Oxidation and Enzyme Denaturation that are unpalatable because of their high sa- The accounts highlighted above do not ex- ponin content were avoided as food and bev- haust the full spectrum of ends achieved in pit- erage sources. The Apachean groups living in hearth cooking. Two others, toxin oxidation and this area focused on a number of agave species enzyme denaturation, deserve mention. in the group Parryanae; these species (A. parryi, As previously mentioned, plants are rich in A. neomexicana, A. gracilipes, and A. havardiana) a variety of allelochemicals or secondary com- show no or little evidence of steroids (Gentry pounds which, if not removed or treated in some 1982:521 –524). way, have direct or indirect deleterious effects on Pine nuts (with high lipid content) are some- their consumers. Heat treatment is one way to re- times mentioned as pit-roasted. Heat treat- duce these negative effects and pit-roasting may ment has the effect of reducing the bitterness of serve in this capacity. For example, in several the nuts of some species (e.g., Pinus albicaulis; ethnographic accounts, acorns and millet were Thompson Indians; Turner et al. 1990:101–102) described as being processed into a dough and and, as previously mentioned, denaturing the li- then baked. Presumably, the intent of this pro- pases that promote rancidity. The pine nuts processing was to deactivate the tannic acid found in duced by different species of pine trees are not these foods (Table 10; Jackson 1991:524–525). In all equally rich in fats (Farris 1995), but, espe- each of these accounts, a relatively small amount cially for those that are, heat treatment should of food—for an individual or for a family—was result in an extended storage life. As well, cones prepared for immediate consumption. The mini- were heated to express the nuts. mum cooking times of 4–5 h would likely be suf- One final kind of plant food mentioned as pit- ficient to denature the responsible protein-based processed is that of green or sweet maize. Ste- compound. venson (1915:76) reported that Zuni people exca- Generally, the secondary compounds in vated a deep (3–3.5 m) pit in the corn field, filled many roots and bulbs resist degradation by it with cedar, and allowed it to burn to coals. Un- heat treatment and people must resort to other husked corn was placed inside and roasted for means of detoxification (Johns 1989: 505). While more than 12 h. In modern ripened corn, starch many of the species in Table 10 were identified comprises up to 72% (Greenwood and Munro elsewhere as being pit-processed, only onion 1979b:378) of the kernel by weight. In green corn, Food Composition and Heat Treatment: Pit-Hearth Cooking 27 McKey (1979); Jackson (1991:522) McKey (1979:64); Frohne and Pfänder Jackson (1991:254) References (1983:26) Gentry (1982) Frohne and Pfänder McKey (1979) Jackson (1991:522) Unknown Yes Yes Unknown No Traditional heat treatment (Johns and Kubo 1988) (1983:26) Yes Sometimes Yes Unknown Modify steroid hormones Alter permeability of cell membrane Damages skin, central nervous Damage cell membranes; decrease Skin irritant General effects (varies by species and subspecies) system; may work as depressent; can irritate kidneys and bladder metabolic disturbances digestion and absorption of protein; interfere with absorption of iron; depress food intake digestion and absorption of protein; interfere with absorption of iron; depress food intake With chronic ingestion, results in Damage cell membranes; decrease Alter permeability of cell membrane Alter permeability of cell membrane Steroidal saponins Saponins Terpenes Tannic acid Terpenoids Secondary compound Nonprotein amino acids Tannic acid Saponins; protocatechuic acid Saponins

ssp. (balsamroot) sp. Secondary Compound and Heat-Related Detoxification Technique for Selected Plant Species[from Johns and Kubo (1988; Table 1) Unless 1) Table (1988; Kubo and Johns Species[from Plant Selected for Technique Detoxification Heat-Related and Compound Secondary spp. sp. (bitterroot) sp. (onion) sp. (agave; leaves) ssp. Table 10. Table Otherwise Indicated] Plant (common name; part) Seeds for which heat treatment by pit processing is reported Pinus Prosopis pubescens Quercus Sorghum Roots/bulbs for which heat treatment by pit-processing is frequently reported Agave Allium Balsamorrhiza Glycyrrhiza glabra Roots/bulbs for which heat treatment by pit processing is rarely reported Lewisia ‘‘Traditional Heat Treatment’’ refers to processing that involved heat specifically remove the bitterness likely owed one or more allelochemicals. 28 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) however, much of the kernel, which may be 10– possible to infer or monitor the temperature 20% smaller than when ripe, is composed of achieved in a pit hearth with experimental data a sugar. By roasting corn cobs while green, the on hearth rock heat capacities or modified pa- enzyme responsible for converting the sugar to leomagnetic techniques (e.g., Takac and Collins starch was likely denatured. In this dehydrated, 1995). Although crude, the presence/absence high-sugar state, sweet corn may be preserved of hearth rock also indicates to some extent the for up to 6 months. From the several accounts length of time over which an elevated oven tem- presented in Appendix 2, it seems that a large perature was maintained. And, it may be possi- quantity of corn, albeit representing a small pro- ble to experimentally determine heat–time pro- portion of the total yield, was prepared this way. files for ovens of varying shapes and containing As O’Shea (1989:61) notes for the Pawnee, har- varying quantities of rocks. vesting and processing corn in this preripe state The second trend involves the typically large ensured that some of the crop was available for quantity of food that was pit-processed. In the case consumption, even if hail or insects devastated of meat or mixed food preparation, several fam- the remainder. ilies and sometimes an entire community was fed from the pit hearth. In the case of vegetable tissues, mass quantities of food were almost always prepared for storage although family- Summary sized portions of inulin-rich foods were some- From this discussion of the relationship be- times prepared for immediate consumption tween food composition and food preparation, (see Del Barco 1973 account in Parsons and Par- three trends in pit-hearth cooking are clear. sons 1990:277 – 278). Pit-processing of doughs First, pit-hearth food processing is associated with and pastes usually involved small quantities. those situations in which tissues are cooked for ex- Archaeologically, estimates of pit-hearth vol- tended periods of time and subjected to moderate ume are approachable and have been used by to high heat; the moisture regime may vary in researchers (e.g., Reeve 1986; Thoms 1989) to these ovens. In the case of inulin-rich foods like monitor the reliance of past peoples on pit-pro- camas, agave, and ti, an extended cooking time cessed foods. It is not necessarily the case, how- at moderate to high temperatures appears to ever, that pit size translates exactly to volume of be required for inulin hydrolysis to occur (Fig- food processed. In preparing yucca hearts, the ure 7). In the case of high-lipid meats, lipid hy- Kawaiisu would excavate a pit to contain the drolysis and possibly also collagen hydrolysis heating element; the actual cooking structure are promoted through pit-hearth cooking, with extended to a height of more than 1 m above cooking times varying by product amount, but the ground (Zigmond 1981:70). generally being much shorter and at relatively Finally, also telling is what is NOT being pit-pro- lower temperatures than that observed for inu- cessed, namely, lean meats and plants high in sug- lin-rich foods. In mixed food preparation, foods ars and fast-release starch. In the ethnographic with similar heat treatment requirements, i.e., accounts of pit-processing reviewed here, no ac- fatty meats and hydrolysis-resistant starches, counts of lean meats were explicitly noted, al- are prepared together at moderate to high tem- though some animals with generally lean tissues peratures, but for relatively abbreviated cook- are mentioned. In the case of rabbits and squir- ing times. Pit-hearth cooking with the end of en- rels pit-processed by the Surprise Valley Paiute zyme denaturation or toxin oxidation appears (Kelly 1932), it may be that seasonally fatty an- to occur at moderate temperatures for short pe- imals were processed this way or that the sub- riods of time. Archaeologically, it should be strate precluded hot sand cooking but permit- Food Composition and Heat Treatment: Pit-Hearth Cooking 29

Figure 7. Pit hearth cooking summary by length of cooking time and temperature. ted pit-hearth cooking. Furthermore, except for Type II diabetes mellitus, i.e., maturity onset sweet corn, plant foods with high sugar content non-insulin-dependent diabetes mellitus (NI- are not mentioned as pit-roasted or baked. And, DDM), occurs at an alarming frequency and with while some starchy roots and piths like sego were lethal consequences in many indigenous popula- pit-baked, starchy seeds were only mentioned if tions today (Stinson 1992; Szathmáry 1994). The they had first been mechanically altered through above discussion of food composition and pit- grinding or grating. No whole seeds were men- roasting may help to shed light on the origin and tioned as pit-roasted, but may be roasted in hot distribution of NIDDM in these populations in sand and then sifted. at least two ways. It suggests that besides starch, other complex carbohydrates, i.e., those rich in fructan and inulin, be examined for their dietary effects. Second, it suggests that we can use Evolutionary Implications and archaeologically visible roasting features, that Archaeological Observations is, features used in extrasomatically hydrolyz- ing resistant starches and inulin, to monitor the The relationship between diet and the evolu- appearance and relative dependence on these tion of human physiology is a perennially im- foods. In this way, we may be able to track Neel’s portant topic of research involving population (1962, 1982) thrifty genotype through time. Each geneticists, nutritionists, and points is discussed of these points is discussed below. below. other physical anthropologists (Eaton and Konner 1985; Garn and Leonard 1989; Larsen Evolutionary Implications 1995; Stahl 1984; Stinson 1992). The focus here is on the relationship between the precontact di- Diabetes mellitus is manifested as abnormally ets of many indigenous populations and diabe- high levels of blood glucose. In individuals tes mellitus. with type I diabetes, the pancreas fails to pro- 30 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) duce insulin, which in turn works to promote evolutionary process that, in populations with the uptake of glucose, fatty acids, and amino ac- a long history of cereal use, happened perhaps ids by cells. Death results unless insulin is sup- over many generations. Now that recently colo- plied exogenously, hence the term insulin-de- nized populations have reliable access to abun- pendent diabetes mellitus. In NIDDM or type II dant fatty foods and processed carbohydrates, diabetes, hyperglycemia is accompanied by ex- the ‘‘thrifty genotype’’ has become detrimental cessively high levels of insulin, which leads to a to the health and well-being of its bearers. Much variety of complications. Over time, an individ- research has been conducted on diabetes since ual with this disease may lose the ability to pro- 1962, leading Neel (1982) to revise his under- duce insulin and may experience heart disease, standing of the mechanisms responsible for NI- kidney failure, blindness, and poor circulation, DDM; the fundamental basis for a thrifty geno- often resulting in gangrene in injured extremi- type remains unaltered. ties (and, therefore, amputations) (Szathmáry Other evolutionary scenarios have been of- 1994:457–458). fered. Wendorf and Goldfine (1991) suggest that Over the last several decades, an increasing selection for a thrifty genotype would have oc- incidence of NIDDM has been noted for some curred as ancestral Amerindians attempted to Native Americans (Gohdes 1986, 1995), Polyne- subsist on risky faunal resources south of the sians (Collins et al. 1994), Australian Aborigines glacial ice. A thrifty genotype associated with (O’Dea 1991), New Guinea groups (Hodge et al. the ability to reduce hyperglycemia during fast- 1996), and others (see Szathmáry 1994:457). As of ing but allow for energy storage in fat and liver 1987, half of all adult Pima have been diagnosed during feasting would be beneficial to members with this disease (Gohdes 1995), which usually of the hypothesized first and second colonizing occurs at a level of around 1 –3% in North Amer- populations, but not in the latest or third (when ican populations of European extraction. The in- the climate ameliorated). Szathmáry (1994) ar- crease in prevalence rates appears to coincide gues that the high-protein, low-carbohydrate with lifestyle changes that include a reduced diet of Paleoindian immigrants to North Amer- level of activity and also diets composed of foods ica may have selected for thrifty lipid metabo- high in fats and processed carbohydrates, and lization and storage. Their descendants, espe- low in fiber. cially those in the northern latitudes who until In 1962, James Neel proposed that NIDDM recently subsisted primarily on meat, now suf- was one manifestation of a ‘‘thrifty genotype,’’ fer the consequences of glucose intolerance in which, under the ‘‘feast and famine’’ conditions a carbohydrate-rich environment. While either he postulated for early hunter–gatherer groups, of these explanations might explain some of the conferred a selective advantage to its bearers. observed variation in diabetes rates seen in Na- Such individuals, he suggested, were able to rap- tive American populations, neither speaks to the idly process and store glucose as fat rather than high rates of NIDDM observed in other Western- suffer urinary caloric loss. As groups came to izing populations. rely on more reliable cultivated cereals, which Already mentioned above, other recent re- are high in starch and are processed to glucose, search points to important digestive qualities a concomitant change in human physiology oc- of traditional foods. Thorburn and colleagues curred. Historic hunter–gatherer or horticul- (1987) show, for example, that the starchy bush- tural populations that have only recently been foods traditionally used by Pacific Island groups incorporated into the Western economy are suf- and Australian Aboriginal groups can be classed fering, he suggests, the telescoped effects of an along with legumes as slow-digesting, slow-en- Food Composition and Heat Treatment: Pit-Hearth Cooking 31 ergy-release foods, which contrasts with fast- upon along with corn. For Aborigines in south- release Western carbohydrate foods. They re- ern Australia, Gott (1983) highlights the impor- late this slow-release quality to the nature of the tance of murnong (Microseris lanceolata), which dominant starch fraction, amylose, which re- served as a staple until recently; when murnong sists digestion to a much greater degree. Sim- was destroyed by the grazing sheep and rabbits ilar results were reported by Brand and col- of the colonizing Europeans, the Victorian Ab- leagues (1990) for six staple foods of the Pima. originals were forced to seek food from the col- Again, beans, mesquite, acorns, and even native onists (Gott 1983:12–13). Zea mays, each of them traditionally prepared, In other groups, inulin-rich foods functioned showed slow-release profiles compared with the as critical emergency rations. ‘‘Hard foods’’ control of wheat bread. In both studies, the au- was the term used by the Mescalero Apache thors suggest that the thrifty genotype, which is to describe agave, sotol, and yucca. These are manifested today in complications attributable ‘‘foods that [had] to be baked’’ and were re- to NIDDM, may have evolved in circumstances lied on during droughts (Basehart 1974:60). which hydrolysis-resistant starches comprised The Pima and Tohono O’odham similarly used major portion of the diet. agave in difficult circumstances. Even as re- To date, evolutionary scenarios describing the cently as 1988, the Mountain Pima living in appearance of a thrifty genotype have not con- northern Mexico turned to agave and other sidered the role of inulin-bearing foods. Such wild foods when their crops failed (Laferriere foods, however, comprised large portion of the 1992). Gill (1880:22) notes that in Polynesia, ti diet for many groups. In the northwestern por- was used during the winter months when taro tion of North American, for example, camas was unavailable. was a mainstay of many groups (Thoms 1989). In addition to this documented reliance on in- In nearby British Columbia, the Thompson In- ulin-rich foods, we also know that fructose, to dian informants named E. grandiflorum (gla- which inulin and other varieties of fructans are cier lily), an inulin-bearing member of the lily hydrolyzed (along with a relatively low propor- family, as the ‘‘boss’’ root because of its abun- tion of glucose moieties), may perform as an al- dance and importance in their diet (Turner et ternative energy source for diabetic patients al. 1990:122). Similarly, agave served as a sta- today. Various studies have reported both pos- ple food for groups in the southern Great Basin itive and negative effects accruing to diabetics (Fowler 1995). The vast mounds of fire-altered consuming fructose rather than sucrose (Crapo rock found on the southern Plains of North 1994:423). The fear is that fructose consumption America and usually associated with mescal will lead to increased levels of serum lipids, re- processing suggest a major reliance on this food sulting in other diabetic complications. This con- (Greer 1965) and at least one account of an indi- sequence appears to occur in some susceptible vidual who lived for some time with Comanche NIDDM individuals, but not in those with mild and Apache groups (Eastman 1879:115 –116) tes- symptoms. tifies to its extensive usage. Agave was widely Thus, it is reasonable to hypothesize, as sug- utilized all throughout northern Mexico (Gen- gested by O’Dea (1986), that populations with a try 1982). For the Pima and Tohono O’odham of genotype selected to metabolize hydrolyzed inu- southern Arizona (North America), cholla buds, lin and slow-releasing starches may be ill suited a plant food I classed as likely containing inulin to diets rich in fast-release starches and sugars. (because of its postprocessed sweet taste), and If this is the case, then we can outline expecta- mesquite pods (fructan-bearing) were relied tions for the geographic distribution of NIDDM. 32 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Brand and colleagues (1990) have already sug- ent sub populations within an otherwise appar- gested that the wild larder of groups resid- ently homogenous population. ing in arid areas would likely contain legumes The important point here is that inulin-and and other slow-release complex carbohydrates; fructan-rich foods not only provided a crit- groups in these areas ingesting modern Western ical and intensifiable energy source (Thoms diets high in simplified sugars may suffer from 1989) but also that the fructose and inulin na- NIDDM. We can go a step further and hypoth- ture of that energy source itself may be signifi- esize that in those temperate and semiarid areas cant. Populations that evolved while dependent where plants store energy as fructan rather than on fructan, inulin, and other digestion-resis- starch, groups harvesting and consuming these tant foods were subject, it appears, to a selec- bulbs and stalks may have been subjected to a tive force that may be in part responsible for the similar selection. The descendants of groups that high incidence of NIDDM in indigenous popu- evolved in these areas, thus, may also be more lations around the world today. susceptible to NIDDM. Several lines of evidence suggest that this proposition is worthy of further research. First, Archaeological Observations: in general, subarctic groups for which high NI- Examples from the American Great Plains DDM incidences (Gohdes 1986, 1995) are com- and Elsewhere monly reported are those that, at European contact, resided in the western United States, where Archaeologically, we are in a good position to slow-release starches and inulin-rich foods ap- observe the appearance of technologies designed pear to occur. Of course, other factors, includ- to advance the digestion of hydrolysis-resistant ing the differential reporting of NIDDM in starches and fructans or prepare fatty meats. If Native American populations, degree of depen- the speculation entertained in the preceding sec- dence on a Western diet, amount of European tion holds any merit, hot-rock and pit-hearth fea- admixture, and others play a role as well. Sim- tures may inform on the appearance of a thrifty ilarly, as noted above, high and increasing NI- genotype. DDM incidence rates are reported for Austra- On the northwestern Plains of North Amer- lian Aboriginal, Polynesian, Micronesian, and ica, Davis and colleagues (1993) have reported New Guinea indigenous populations with a the presence of Paleoindian (9400 BP) roasting similar historical reliance on complex carbohy- pits and hearths at the Barton Gulch site. The drates. Second, when affected or predisposed roasting pits, which ranged in depth from 26 to NIDDM individuals are placed on a traditional 68 cm, contained charcoal and various macro- diet, their NIDDM symptoms are ameliorated botanical remains but no rock element. They are or abate (O’Dea 1991:239). Third, the consump- found in redundantly occurring arrangements tion of inulin by rat and human subjects appar- of two to four per central hearth. Their lack of ently has beneficial effects on glucose tolerance a rock heating element suggests that they were and serum lipid levels (Gupta et al. 1993; Rober- not used to hydrolyze inulin-bearing plants or froid 1993), although the responsible mecha- fatty meats, although they may have been used nisms are not well understood. Finally, several to process high-lipid seeds, like those of Pinus studies have reported variation in glucose re- flexilis (a few seeds of which were recovered) or sponses within populations (Neel 1982; Omar et to dethorn the meats of Opuntia polycantha (the al. 1994). These responses may be evolutionary seeds of which are relatively abundant here) consequences owed to the differential composi- (Armstrong 1993:13). Following Greenhouse tion of the diets historically enjoyed by differ- and colleagues (1981), the other seeds recov- Food Composition and Heat Treatment: Pit-Hearth Cooking 33 ered from these features may represent the nat- 1965; Sheldon 1905; Wandsnider et al. 1995). ural seed rain that occurred in the area. In addi- These hearths often contain a rock element, but tion, numerous bones of deer and rabbits were to date few other remains have been recovered recovered here and it is not impossible that the to suggest their function. It is unlikely that these pits represent protein denaturing pits. An anal- hearths were used to process inulin-rich foods ysis of burn marks on the bone fragments them- because the rock element, composed of the local selves may be able to inform on this possibility. sandstone, has a relatively low heat capacity, In any case, slow-release starchy foods, fructan- unlike the basalt cobbles frequently mentioned bearing foods, or fatty meats appear not to have in ethnographic texts of agave or ti processing. been processed at Barton Gulch. Bone preservation is generally good in this area This manifestation is in contrast to that on the and the lack of bone here suggests that meat southern North American Plains, where Der- processing was not undertaken. Presently, I hy- ing (1996) reports charred camas and onion frag- pothesize that an as yet unidentified plant comments from roasting pits that date to as early as posed of a low DP fructan or a resistant starch 8000 BP. The presence of likely inulin hydrolysis was pit-processed in quantities as high as 75 pits as early as the late Paleoindian period sug- kg/ pit. The presence and high density of these gests that the coevolution of some Native Ameri- pit hearths suggest that a population bearing can populations with these foods was underway the thrifty genotype was in the area as recently by this time. as 1000 years ago. Elsewhere on the northwestern Plains of In other parts of the world, we find dramatic North American, Frison (1983) notes the ap- archaeological evidence for the exploitation of pearance of cylindrical pit hearths with rock el- hydrolysis-resistant resources. Throughout the ements during the early Archaic (8000 BP). He southern Great Plains, the American Southwest, (1983:89) writes that ‘‘the extent of the oxida- and northern Mexico, thick and extensive aggre- tion rings present in many fire pits argues for gations of fire-cracked rock occur, sometimes in a considerable amount of heat and possibly association with definable pits. These features more than can be explained away as needed are usually attributed to agave processing and in simple food preparation.’’ The examination date from the Archaic into Protohistoric times of food composition and processing presented (Gentry 1982; Greer 1965; Abbott and Freder- above suggests that foods requiring extended ick 1990). The very massive nature of these fea- heat treatment, i.e., those rich in inulin or re- tures testifies to the importance of agave to the sistant starches, or in fat, were likely prepared people in this area, although whether as a sta- here. The location of some of these features in ple or as starvation food remains unclear. As al- high-altitude meadows, where camas and other ready noted, in this area today, NIDDM is recog- bulbs flourished historically (Francis 1995), es- nized as a major health problem among Native pecially supports the interpretation that these American and Mexican populations (Knowler et features performed as inulin hydrolysis pits. al. 1978; Stern et al. 1991). These pits also, then, signal the appearance of a Throughout northwest North America, ar- population that is likely evolving on a diet that chaeological evidence for camas exploitation contains some amount of fructose. abounds. In an exceptional study, Thoms (1989) In the central High Plains south and south- documents the appearance of camas use in inte- east of the Black Hills, deep cylindrical pit rior areas by about 8000 BP. Increasing frequen- hearths dating to the Late Prehistoric (700–1100 cies of roasting ovens suggests an intensifica- AD) are common occurrences (Falk et al. 1978; tion of this resource by around 4000 BP (Thoms Johnson 1996; Meston 1976; Schultz and Smith 1989:463–464) and camas continued in use as a 34 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) staple until just recently. Again, NIDDM is re- Conclusion ported at above average levels in indigenous populations in Oregon and Washington (Free- Pit hearths are common archaeological fea- man et al. 1989). tures with great potential to inform not only on In northwestern Europe, vast mounds of functional aspects of food processing but also burned stone have received professional scru- on evolutionary matters. To further our realiza- tiny for many years (Buckley 1990; Hodder and tion of this potential, this paper has attempted to Barfield 1991; O’Kelly 1953; Ó Drisceoil 1988). build a body of reference knowledge on the rela- As Thoms (1989:476) notes, the geographic dis- tionship between food composition, food chem- tribution of this archaeological phenomenon istry, and heat treatment. is consistent with the range of bulbs and roots Ethnographic accounts indicate that inu- similar to camas and glacier lily. On the basis lin-and lipid-rich foods of moderate and large of the information assembled here, the dense quantities were pit-processed. In addition, hy- and extensive burned rock mounds also sug- drolysis-resistant starches appear to have been gest that a processing strategy applied to a hy- processed this way, especially in the western drolysis-resistant food, likely containing inulin, Pacific. Pit-hearth cooking is an obvious and ef- was in use. If this is the case, than we may ex- ficient way to hydrolyze large amounts of resis- pect to find residual occurrences of NIDDM in tant starches and inulin, which require expos- north European, north Asian, or other popula- ing tissues to high temperatures for, in the case tions that evolved while subsisting on these di- of inulin, long periods of time. The thermal pit- gestion-resistant foods and have only recently processing of fatty meats appears to take advan- begun to rely on processed starches. Archaeol- tage of the heat transfer properties of tissue lip- ogy and medicine may profitably work together ids. In addition, pit hearths were used to reduce to sort out this possibility. toxins in small batches of food and to enhance Of import here is that, at least in the New the durability of foods like sweet corn and pine World, the initial use of inulin-rich foods appears nuts. Archaeologically, we may be able to use to date to the early to middle Holocene. By the pit-hearth morphology, the nature of the rock mid–late Holocene, more intensified use seems element, lipid identification (Evershed 1993), pit to occur, and Thoms (1989) has argued that other contents (Dering 1996; Hather 1991), and other more cost-effective foods would likely have been indicators to more specifically diagnose the pit intensified first, if available. If the dependence hearth’s function. on diets with a large carbohydrate fraction com- Other work on the subject of cooking sys- ing from inulin and fructose is related to the ap- tems and pit-hearth cooking remains to be done. pearance of the thrifty genotype, then selection This paper has primarily focused on the cook- for this genotype was perhaps most intensive ing function of pit hearths. Yet, we know that during the later part of the Holocene. At this pit hearths functioned as kilns, in which ceramic time, both historically and ethnographically in vessels were fired (Blackfoot; Ewers 1968:10); North America, we see a relatively high depen- as medicinal heat com presses (Kawaiisu; Zig- dence on inulin-rich foods, especially in those mond 1981:64); as steam elements in sweat baths areas where cereal agriculture could not be un- (Coeur d’Alêne; Teit 1930a:62); and for extraction dertaken or could not be depended upon. When of yucca fibers (Tewa; Robbins et al. 1916:51). other viable crops, such as the potato, are intro- Hopefully, such facilities retain characteristics duced, or the market economy makes flour and that would allow us to distinguish among these other commodities available, the use of both inu- various possible functions. lin-rich foods and pit hearths plummets. Second, while it appears that pit-hearth cook- Food Composition and Heat Treatment: Pit-Hearth Cooking 35 ing is associated with foods having specific com- Archaeologically, we see through time im- positions, the data presented here to not demon- portant changes in human physiology, such as strate that foods necessarily must be prepared the reduction in tooth size during the Lower Pa- this way. A better understanding of the pit- leolithic, which complements the appearance of hearth cooking regime, i.e., the temperatures artifacts and features that reflect extrasomatic and moisture levels achieved therein (and a sim- tissue processing (Brace et al. 1987). At the end ilar understanding for stone-boiling), would en- of the Middle Paleolithic, there is incontrovert- able us to better comprehend decisions made by ible evidence for the deliberate construction past peoples about food preparation. Such pa- of hearths (James 1989), although whether for rameters are difficult to mathematically model denaturing animal protein, detoxifying plant and therefore replicative experimentation of the foods (Leopold and Ardrey 1972), or other pur- type undertaken by Armstrong (1993) will be poses remains to be addressed. By the early to necessary.3 middle Holocene, there is ample archaeologi- A third step concerns other reference infor- cal evidence in the form of pit hearths, grinding mation. Baseline physiological information for technology, and ceramics, as well as the floral human populations that have not coevolved and faunal remains themselves, indicating that with starchy cereals is necessary. Hopefully, in- people were drawing from a very extensive lar- formation of this sort will be forthcoming from der. And, throughout the rest of the Holocene, the Human Genome Project and from research the archaeological record is dominated by evi- conducted on NIDDM in non-Western popu- dence that people are resorting to increasingly lations. Also necessary is baseline information costly food processing practices to detoxify and on the lipid and carbohydrate fraction of tra- maximize the energy value of foods. Our con- ditional foods. Most contemporary analyses of temporary situation, in which the amount of food composition report the major components energy spent to obtain 100 kcal is many times of food as well as mineral and critical amino that amount (and is heavily subsidized by fossil acid composition. Repeating and expanding the fuel use), and in which we are told daily of the Yanovsky and Kingsbury (1938) study of tradi- health risks of the Western high-fat, highly pro- tional food composition would take advantage cessed diet, is an ironic continuation of trends of recent advances in the precision and reliabil- that began in the distant past (Eaton and Kon- ity of carbohydrate fraction measurement. In- ner 1985). deed, even more detailed studies of traditional An understanding of the coevolution of food foods that provide information on starch frac- resources, processing technologies, and geno- tion or fructan form would be helpful. And, typic variation, thus, is increasingly approach- information on seasonal levels of animal tis- able using the durable parts of the archaeological sue lipids by species and sex may also be use- record, even fire-cracked rock and pit hearths. ful. With these several bodies of knowledge, we With the anchor of knowledge about food chem- would be in a better position to appreciate the istry and heat treatment established here and significance of the various forms of food pro- other such anchors, we can continue to explore cessing that Stahl (1989) has outlined. the evolution of these coupled phenomena.

• • 36 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) Whiting (1985:63) Whiting (1985:181) Whiting (1985:62–63) Kelly (1932:92) Kelly (1932:92) Lewis and Clark Grinnell(1923:255, 256) Whiting (1985:180) Whiting (1985:180) Kelly (1932:93) Kelly (1932:93) Kelly (1932:92) Marshall (1976:89–91) Tanaka (1980:39) Steensberg (1980:189–197) Marshall (1976:144) Holmberg (1957:216–218) Kelly (1932:93) Linderman (1962:253) Reference Cook in wet sand to prevent Bury in wet sand;fire on Hot stone in stomach cavity, cover with wet sand hot sand skunk, etc. Cut meat thin to hasten ‘‘Nothing else was cooked this way’’ In Spinden (1964:207) Saucer-shaped pit fast cooking top Small hot rocks in stomach cavity Hole to 6-foot D if large catch dry to store Eat immediately Cooked quickly Photo, p. 91 Bury meat in embers and 8–10 × 1.25 m wide Scorch on embers, descale, bake Squirrels, rabbits, rat, otter, Comment Multiple 10–15 Various package medium small available 1 large package 1 large package 1 hide At least several Catch varies in size Catch varies Single Large package Small– 28 pork sides Small package Medium– large (picnic) Several packages When 1 bison Quantity 0.6 0.3 0.2 0.0 1.0 (m) Depth 1 1.3 0.5 (m) Width 1.0 9.0 0.8 (m) Length 2.00

5.00 3.00 3.00–4.00 20.00 2.00 4.00 0.08–2.00 0.05 1.50 3.00 2.00 0.10 1.00–3.00 Time (h)

None None Pinyon twigs Grass, dirt Fir boughs Pine boughs Leaves Chokecherry boughs None None Grass Grass None None Banana leaves, ferns, herbs None None Insulation No No No Yes Yes Yes Yes No No Yes No Maybe No No No No fire Top O 2 H No No Yes No Yes Substrate No Substrate Yes No No No No Substrate Yes No No No No No No No Add

Region CA, NV, OR CA, OR, NV Upper Wyoming Montana NW Arizona NW Arizona NW Arizona NW Arizona CA, NV, OR CA, NV, OR NW Arizona CA, NV, OR Kalahari Kalahari Highland Kalahari Peru CA, NV, OR Missouri New Group SV Paiute SV Paiute Upper Cheyenne Crow Havasupai Havasupai Havasupai Havasupai SV Paiute SV Paiute Havasupai SV Paiute !Kung San San NG !Kung San Virunero SV Paiute Missouri Highland Guinea Ethnographic Accounts of Pit-Hearth Food Preparation: Meat Tissues Food Bear meat Bear meat Bear flitches Bison bull hide Buffalo meat Doves, quail Doves Quail Turkey Crickets Caterpillars Porcupine, rock squirrel, woodrat, bobcat, badger Groundhog, badger Animal heads Large piece of meat Fat-rich pork Python Lizards Rabbits, squirrels Appendix 1. Food category Bear Bear Bear Bison Bison Fowl Fowl Fowl Fowl Insect Insect in size Medium mammal Medium mammal Meat Meat Pork Reptile Reptile Small mammal Food Composition and Heat Treatment: Pit-Hearth Cooking 37 Castetter (1935:10–14) (1972:31–36) Reagan (1930:293) (1972:77) (1990:124) Basehart (1974:49–52) (1990:276–279) Spier (1928:105–106) Whiting (1985:49–50) Parsons and Russell (1908:70) Fowler (1995:106) Pennington (1963:130) Pennington (1963:129–131) Bye et al. (1975:86–89) Hough (1959:S46) Ray (1932:103) Teit (1975:237) Underhill (1935:15) Saunders (1914:133) Bean and Saubel Turner et al. Castetter and Bean and Saubel Reference Palmer (1871) Pits 10 to 15-foot (1972:50) pit bake

details in Armstrong (1993) In Greer (1965:52) Detailed description Make mound, not pit Cook longer if necessary; famine food; storage From Kelly notes; some Replace cooled rocks with hot Range of baking times; depends on age of hearts In Parsons and (1990); for storage D:6–20’; slope to center, 1–3’ deep; in Greer (1965:51) Remove spines, later Water chimney for steam; toofast’’ In Bean and Saubel

‘‘Not creosote bush, burns Detailed description Comment Remove spines first bushels For trade, 1 ton food make expedition 30–40 plants Large Seemingly large Large Seemingly large Seemingly large Consume/ store A quantity collection 3 days of Mass Several Quantity 0.9 1.0 1.5 1.0 0.9 0.0 0.9 (m) Depth 3.8 1.5 1.5 1.2 1.5 1.6 (m) Width 3.8 1.5 2.0 1.2 1.5 1.6 (m) Length 24.00

72.00

48.00 12.00 24.0–40.0 24.0–36.0 24.0–48.0 72.00 24.0–120.0 44.0–92.0 48.0–72.0 0.20 12.0–24.0 12.0–24.0 12.00 20.00 72.00 Time (h) 12.00 Grass Wet twigs, brush leaves and Grass grass Green grass Grass None None None Grass, pine needles Grass, pine needles Dirt, palm leaves Grass None Fir branches, pine grass Leaves, Hot ashes Needles Inkweed; Grass Insulation Inkweed No No No No Yes No No No No n/a Yes Maybe No No No No fire Top No O 2 H Insulation Yes No No No No No No No No No No Yes No No Yes No No Insulation No Add No Washington Region New Mexico EC Arizona NW Arizona NW Arizona NW Mexico; Arizona Majova Chihuahua Chihuahua Chihuahua American SW NE British S. CA S. CA

British Columbia S. Arizona S. CA New Mexico CA Desert Columbia Arizona Group Apache Apache Havasupai Havsaupai NW Mexico Pima S. Paiute, Tarahumar Tarahumar Tarahumara Various Sanpoil Thompson Cahuilla Cahuilla Thompson Indians Papago Cahuilla Mescalero Chemehuevi Indians Apache

species Ethnographic Accounts of Pit-Hearth Food Preparation: Plant Tissues grandiflorum undulata Opuntia Food hearts Agave Agave Agave Agave heads Agave Agave Agave A. species A. species Agave hearts Agave Pursh. Gray Prickly pear pads Cactus roots Agave deserti Agave parryi Erythronium Calochortus Hesperocallis Appendix 2. Food category Agave Agave Agave Agave Agave Agave Agave Agave Agave Agave Agave Agave Agave Bulb Bulb Bulb Cactus Cactus Cactus buds Cholla Cactus buds 38 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) Lowie (1954:24) Bean and Barrows (1990:67) (1972:67–68) Wissler (1910:25) Fr. Point (1933:252) Schaeffer (1940:47) (1981:228–232) Grinnell (1962:204) Teit (1930:92) Stubbs (1966:54–55) (Donnelley, 1967:166) Turney-High Reagan (1933:60) DeSmet (1905:488) Spinden (1964:201– 202) McClintock (1910:442) (1929:413–414) Thackery and Leding Reference Bean and Saubel Russell (1908:71) Curtin (1949:58) Greenhouse Saubel (1972:95) Hot rocks in pit; bitter taste Use wet willow; in Malouf (1979) In Malouf (1979); in firewood; woman, cook In Malouf (1979) lined; in Bohrer (1969) lined Pit is stone-lined; cook time proportional to quantity With black moss and onions In Malouf (1979) Johnston (1987:24) In Malouf (1979); men, In Malouf (1979:57) In Malouf (1979) Fire around base of conical heap repeated,’’ mound ‘‘Cooking sometimes Comment Parboil to remove Several layers; bottom-sides Flat-bottom pit, sides rock- quantities large Large Several bushels 20–30 bushels In Malouf (1979) Large Quantity Large Seemingly Two bushels 0.9 0.6 0.2 0.5 0.9 0.9 0.3 (m) Depth 0.5 3 1.2 1.5 1.1 1.5 3.1 (m) Width 1 3.0 2.0 1.5 3.0 1.5 3.1 1.0 1 (m) Length 1.0 36.00 12.0+ 36.00 48.00 24.0–68.0 48.00 68.00 30.0–40.0 12.0–18.0 24.00 50.0–70.0 72.00 12.0–66.0

20.00 Time (h) 3.00 12.0–20.0 20.00 18.00 O) 2 Willow leaves Willow leaves, branches Grass, branches Grass Grass, bark Leaves, grass, tree moss (with H Earth Grass, moist earth Wet leaves, brush, Grass, Willow branch, sunflower stalks, grass Grass, dirt Inkweed Insulation Damp sand, Saltbush Inkweed Seepweed Yes Yes Yes Yes Fire Yes Yes No Yes Yes Yes No fire No Top No No No O 2 H No No Insulation Yes No No Maybe No No Yes No Yes Yes No No

Add No No No

Region S. Alberta W. of Rockies S. Alberta W. of Rockies Idaho NW Rockies NW Rockies NW Rockies Coastal N. Plains N. Plains Idaho S. Arizona S. CA S. CA

Arizona Arizona S. Arizona Washington

Group Blackfoot Blackfoot Blackfoot Blackfoot Coeur d’Alene Flathead Flathead Flathead Hoh/Quileute Kutenai Kutenai Nez Perce Papago Cahuilla Cahuilla Pima Pima Pima

acanthodes O. veriscolor, Opuntia basiaris Echinocactus Food O. echinocarpa Engelm. Lem. buds Camas Camas Camas Camas Camas Camas Camas Camas Camas Camas Camas Camas Appendix 2. Continued Food category Cactus buds Cactus buds Opuntia fruit Cactus buds Cholla Cactus buds Cholla Cactus buds Cactus buds Camas Camas leaves Camas Camas Camas Camas Camas Camas Camas grass Camas lichen Camas Camas Food Composition and Heat Treatment: Pit-Hearth Cooking 39 Downing and Furniss Ray (1932:97–106) (1968:48–59) (1975:74–75) Downing and Furniss Kelly (1932:102) (1968:48–59) Curtis (1907–30:VIII, (1983) 293) Weidfeldt (1914) Griaule (1938) 41) Spinden (1964:201– 202) Malouf (1979:54–55) Collier et al. (1942:38) Olson (1936:52–53) Phillips (1940:102) Suttles (1951) Hilty et al. (1972:7–8) Reagan (1930:292– Whiting (1985:68) Stevenson (1915:76) Christiansen Whiting (1985:65) Brown (1868:383) Reference Greene Photographs Many (5) var. camas and firewood and seasonings Photographs Camassia quamash In Malouf (1979) communication other roots; cook with onions and lichen storage Usually boil, roast Fire around base of conical heap Williams, personal In Malouf (1979) Twice-baked In Malouf (1979) In Turner and Kuhnlein Flavor according to Large implied For storage or immediate consumption For consumption and Steam up to 6 h; add Cover with damp sand Bread baked with red-hot stones Comment sample 7 sacks Short season household seawater 2 sacks; implied preparation 20–30 bushels For sample 6 sacks Seasonal Large quantities 2–3 bushels Seasonal Likely Large Joint Medium Quantity 0.9 0.8 0.6 0.6 0.9 0.5 0.3 0.6 0.6 3.5 0.5 (m) Depth 1.4 1 0.9 3 1.8 1.5 1.2 1 1.3 1 0.5 (m) Width 1.4 1.0 0.9 3.0 1.8 1.5 1.2 1.0 1.3 1.3 0.5 (m) Length 48.00 24.0–72.0 48.0–51.0

12.0–66.0 48.0–72.0 48.00 20.00 18.00 15.00 24.0–36.0 24.00 24.00 12.00 0.25 5.0–6.0 Time (h)

grass, Rye grass, Bracken pine packets green grass Alfalfa, rye willow, mud no sacks Grass Grass, dirt Grass, alder leaves Green tule Ferns Rye grass with water Grass Kelp, vines, fern, fir sk cabbage, needles Wet grass, corn husks, twigs None None Leave Large firm leaves Corn husks, Baobab leaves Leaves, dirt Insulation No Yes Yes Yes Yes Yes Yes Maybe Yes No Yes No No No Yes No No fire Top O 2 H Insulation Yes No Yes Yes No No Insulation St. No Yes No Yes No No Yes No No No Insulation Yes Add Island, Region Idaho Idaho Idaho Washington NW Plateau Coastal CA, NV, OR NE Rocky Washington Oregon EC Arizona NW Arizona W. New Bellona Formosa NW Arizona Subst. Mali, W. California Idaho Washington Washington Mountains Mexico Melanesia Africa Group Nez Perce Nez Perce Nez Perce Nez Perce Okanagon Quinault SV Paiute Sanpoil Shoshoni Straits Salish Warm Springs Indians Apache Havasupai Zuni Bellonese Formosa Havasupai Madougou S. California Nez Perce Lakes Aborignals Food quamash Camas Camas Camas Camas Camas Camas Camas Camas Camas Camas Camassia quamash Green corn Green corn Corn Taro,fish, or poultry Millet pancake Seed/green corn cake Millet pancake Acorn dough Camassia mas Appendix 2. Continued Food category Ca Camas Camas Camas Camas Camas Camas Camas Camas Camas Camas Camas Corn Corn Corn Dough Dough Dough Dough Dough 40 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997) Fowler (1995:102) Kelly (1932:99) Sillitoe (1983:110–112; personal Olson (1936:53) Kelly (1932:103) Kelly (1932:102) Russell (1908:75) Gorman (1896:78–79) (1977:27) Boas (1921:518–523) (1990:72–73) Griaule (1938) 17) Tanaka (1980:39) Whiting (1985:211) Gill (1880:28–29) communication (1996) Jennifer Galindo, Reference Coville (1897:105) Teit (1930a) Robinson in Presland Turner et al. Basehart (1974:52–54) Hilty et al. (1972:16– Wats. Nutt. O chimney 2 From Kelly notes; details not given Little use made of pine nuts Photo, Pl. 13 Pl. 13) Nadjowh for madjingmi

Allium pleianthum Valeriana edulis In Turner et al. (1992)

In Turner et al. (1992) eating Est. dim. from sketch cooled and then cut, frozen,or dried Bury in embers and ashes Starvation food Refers to mythical H roots Description from Timothy Comment Disagreeable odor Steam roots until cooked; In Gott (1983) Dried and then boiled for Datil fruits Large,flat bar of koonts Seemingly large store Small 2–5 sacks at a time Large; consume/ cannibalistic oven Quantity Seasonal 0.3 (m) 0.6 Depth 1.5

(m) 1.1 Width 1.5 (m) 1.1 Length 8.0–12.0 1.50 72.0–96.0 14.0–15.0

15.0–20.0

18.00 48.00 12.00 0.6 Time (h) 20.00 18.00 24.00 1.0

12.00 0.6 leaves, weeds’’ Cocklebur Moss, kelp Seaweed, branches None arrow bush No Grass, earth None Banana leaves ‘‘Water- Insulation Grass Mud, grass Grass Douglasfir None Pine needles Maybe Yes No No No No fire No Yes No No No No Top No n No O 2 H No No Yes No No Yes No No Yes No Yes No Insulatio Add Yes New Australia Region Coast Texas Kalahari Mojave Desert NV, CA, OR Highland NW Arizona CA, OR, NV Mali, W. Rarotonga, Arizona Arnhemland, Ft. Klamah, CA, NV, OR Idaho Victoria, British New Mexico– Australia SE Alaska Pacific NW Coast hemlockbranches Washington Columbia Guinea Africa South Pacific OR Oregon Group Quinault San Timbisha Shoshone Surprise Wola Havasupai SV Paiute Sanga Pima Gindgali? Klamath SV Paiute Coeur d’Alene Victoria Thompson Mescalero Aboriginal Tsimshian Coast Kwakiutl Indians Indians Valley Paiute Warm Springs

)

species

Prospis glandulosa expansa scapigera Food expansa Fern roots fruit Melons pubescens Green honey mesquite ( Pine cones Screw-pine nut cluster Allium palmeri Allium species Onions Ti galbra Dryopteris Dryopteris Yucca baccata Alectoria Bryoria fremontii Prosopis Daucus pusillus Valeriana edulis Valeriana edulis Microseris Amorphophallus Appendix 2. Continued Food category Fern Fern Fern Fruit Lichen Lichen Melon Mesquite Mesquite Nuts Nuts Onion Onion Onion Palm lily Root Root Root Root Root Food Composition and Heat Treatment: Pit-Hearth Cooking 41

Whiting (1985:46) Bean and Saubel (1916:51) (1990:130–133) Dwyer (1990:133) Dwyer (1990:135–136) (1975:183–185) (1980:197–201) LeCoeur (1950:15) Buckelew (1911:72– 73) 70) Pennington (1969:141) (1972:150) Dwyer DuToit Steensberg Reference Reference Robbins et al. Zigmond (1981:69– Earth oven lined with flat Used Spanish Bayonet; pig meat/fat sego Mound 1 m high; many different foods 100 carcasses, corn, ferns, pandanus corn, greens, packets, pork sides (1980:201–202) preparation In Greer (1965:51) rocks cooked, too stalk full of tasy sap just prior to blossoming Packets of greens, Includes tubers, Steensberg Daily food Comment Comment Fiber extraction 5-foot-high mound Available (10 sides Feast Feast house- holds onions Considerable For storage; agave roots number; famine food for several Feast of pork) Several Taro 3 fowl and months Quantity Quantity Large 1.1 0.9 0.2 0.2

(m) (m) 0.9 Depth Depth 1.5 0.9 0.9 1.8 0.5 (m) (m) 1.5 Width Width 1.5 0.9 0.9 3.5 0.5 (m) (m) 1.5 Length Length 12.0–24.0 3.00 2.50 48.0–72.0 60.00 2.00 1.50 1.00 2.00 Time (h) Time (h) 20.00 42.00

Sand Leaves? sacking fig, pitpit and ferns None Brush, leaves Dirt, madrono wood pine needles Ferns Leaves, grass, Cordy-line, Banana leaves Insulation Insulation None Sand/dirt, No No Yes No No No No fire fire Yes No Top Top O O 2 2 H No No No H Yes No No No No No Add Add Guinea Guinea Guinea Guinea Guinea Guinea Region N. Arizona SE Sahara W. Texas Chihuahua S. CA Region Highland New Highland New Highland New Highland New Highland New Highland New New Mexico SW California

Group Havasupai Teda Lipan Apache Tepehuan Cahuilla Group Etolo Etolo Etolo Gadsup NG Highland NG Highland Cochiti Pueblo Kawaiisu

Ethnographic Accounts of Pit-Hearth Food Preparation: Mixed Tissues Food Mutton grass Alkao seeds Sotol Sotol and o’yi Yucca stalk leaves Food Packets, sides, pumpkins Sego packets, pumpkin, pig skulls, etc. Taro, sw pot, pumpkin, etc. Taro/sw pot mush loaf; meat Taro, sweet potato Fowls, onions; Taro with fungus Yucca baccata Yucca whipplei

Appendix 2. Continued Food category Seed Seed Sotol Sotol Yucca Yucca Yucca Appendix 3. Food category Mixed Mixed Mixed Mixed Mixed Mixed 42 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Acknowledgments Brace, C. Loring, Karen R. Rosenberg, and Kevin D. Hunt 1987 Gradual change in human tooth size in the Late Pleis- Portions of this paper were presented at the University of Ar- tocene and post-Pleistocene, Evolution 41(4):705–720. izona, the University of Missouri–Columbia, the 1995 Annual Brand, Janette C., B. Janelle Snow, Gary P. Nabhan, and A. Meeting of the Plains Anthropological Society in Laramie, and Stewart Truswell the 1995 Island-in-the-Plains Conference. The comments I re- 1990 Plasma glucose and insulin responses to traditional ceived at each venue were most helpful. This paper has fur- Pima Indian meals, American Journal of Clinical Nutri- ther benefited from the input of Steve Black, Peter Bleed, Les tion 51:416–420. Davis, Jim Gibson, Jim Gunnerson, Harriet Kuhnlein, Martha Braun, David P. McCullough, Ray Hames (who provided the title, with apolo- 1983 Pots as tools. In Archaeological hammers and theories, gies to Claude Lévi-Strauss), Steve Holen, Alan Osborn, John edited by J. Moore and A. Keene, pp. 107–134. Aca- O’Shea, Karl Reinhard, Michael Schiffer, Marilynn Schnepf, demic Press, New York. Katherine Spielmann, Alston Thoms, Ann Yaktine, and, most especially, from an anonymous reviewer. I appreciate greatly 1987 Coevolution of sedentism, pottery technology, and their help. I also acknowledge the encouragement of Mike Fo- horticulture in the Central Midwest, 200 B.C.–A.D. sha, Julie Francis, Christopher Lintz, Ralph Rowlett, Lisa Sat- 600. In Emergent horticultural economies of the Eastern tenspiel, and Patrice Teltser. None of these individuals are Woodlands, edited by W. F. Keegan, pp. 153–181. Oc- responsible for any inadvertent errors. The University of Ne- casional Paper No. 7, Center for Archaeological Inves- braska—Lincoln Interlibrary Loan facility was called on to tigations. Southern Illinois University, Carbondale. perform exceptionally, which it did; I thank those individuals Briggs, George M., and Doris Howes Calloway responsible. 1984 Nutrition and physical fitness,11th ed. Holt, Rinehart & Winston, New York. References Brown, James A. 1989 The beginnings of pottery as an economic process. Abbott, James, T., and Charles D. Frederick In What’s new? A closer look at the process of innovation, 1990 Proton magnetometer investigations of burned rock edited by S. E. van der Leeuw and R. Torrence, pp. middens in west–central Texas: Clues to formation 203–224. Unwin Hyman, London. processes, Journal of Archaeological Science 17:535–545. Brown, Robert Ames, Bruce, N. 1868 On the vegetable products, used by the northwest 1983 Dietary carcinogens and anticarcinogens, Science American Indians as food and medicine, in the arts, 221:1256–1264. and in superstitious rites, Botanical Society of Edin- Anderson, Barbara A. burgh Transactions 9:378–396. Reprinted in An ethno- 1989 Composition of foods: Lamb, veal, and game products. Ag- biology source book, edited by R. I. Ford, Garland, New riculture Handbook No. 8–17, U.S. Department of York, 1986. Agriculture. Buckelew, F. M. Armstrong, Steven 1911 Buckelew, the Indian captive. Mason Herald, Mason. 1993 Alder Complex kitchens: Experimental replication of Buckley, V. (Ed.) Paleoindian cooking facilities, Archaeology in Montana 1990 Burnt offerings: International contributions to burnt 34(2):1–66. mound archaeology. Worldwell, Dublin. Barrows, David Prescott 1900 The ethno-botany of the Coahuilla Indians of southern Cal- Bye, R. D. Burgess, and A. M. Trias ifornia. University of Chicago Press, Chicago. 1975 Ethnobotany of the Western Tarahumara of Chihua- hua, Mexico. In Notes on the genus Agave. Botanical Basehart, Harry W. Museum Leaflets, Harvard University 24(5):85–112. 1974 Mescalero Apache subsistence patterns and sociopolitical organization. Apachean Indians XII, Garland, NY. Castetter, E. F. 1935 Uncultivated native plants as sources of food. Univer- Bean, Lowell John, and Katherin Siva Saubel sity of New Mexico Bulletin 266, Ethnobotanical Series 1. 1972 Temalpakh: Cahuilla Indian knowledge and usage of plants. Malki Museum Press, Banning, CA. Castetter, E. F., and R. M. Underhill 1935 The Ethnobiology of the Papago Indians, University of Bohrer, Vasila H., Hugh C. Cutler, and Jonathan Sauer New Mexico Bulletin Series 4(3):3–84. 1969 Carbonized plant remains from two Hohokam sites, Arizona BB:13:44 and Arizona BB:13:50, The Kiva Chamberlin, Ralph V. 35(1):1–10. 1964 The Ethno-botany of the Gosiute Indians of Utah, Food Composition and Heat Treatment: Pit-Hearth Cooking 43

Memoirs of the American Anthropological Association tification and assessment of a once archaeologically 11(5):329–405. Kraus Reprint, New York. ‘‘invisible’’ plant resource. In Abstracts of the 61st An- nual Meeting of the Society for American Archaeology. Christiansen, Sofus 1975 Subsistence on Bellona Island (Mungiki). Folia Geo- Downing, Glenn R., and Lloyd S. Furniss graphica Danica, 8. København. 1968 Some observations on camas digging and baking among present-day Nez Perce, Tebiwa 11(1):48–59. Collier, Donald, Alfred E. Hudson, and Arlo Ford 1942 Archaeology of the upper Columbia region, University of Driver, Jonathan C. Washington Publications in Anthropology 9(1):1–178. 1990 Meat in due season: The timing of communal hunts. In Hunters of the recent past, edited by L. B. Davis and Collins, Veronica R., Gary K. Dowse, Palanitina M. Toelupe, B. O. K. Reeves, pp. 11–33. Unwin Hyman, London. Tofaeono T. Imo, Faalh L. Aloaina, Raymond A. Spark, and Paul Z. Zimmet DuToit, Brian M. 1994 Increasing prevalence of NIDDM in the Pacific Island 1975 Akuna: A New Guinea village community. Balkema, population of Western Samoa over a 13-year period, Rotterdam. Diabetes Care 17(4):288–296. Dwyer, Peter D. Coville, Frederick V. 1990 The pigs that ate the garden. University of Michigan 1897 Notes on plants used by the Klamath Indians of Or- Press, Ann Arbor. egon, Contributions from the U.S. National Herbar- Eastman, Edwin ium 5(2):87–108. Reprinted in An ethnobiology source Seven and nine years among the Camanches and Apaches. book, edited by R. I. Ford, Garland, New York, 1986. 1879 Clark Johnson, Jersey City, NJ. Crapo, Phyllis A. Eaton, S. Boyd, and M. Konner 1994 Dietary management. In Joslin’s Diabetes Mellitus, 13th The New England Journal of Medi- ed., edited by C. R. Kahn and G. C. Weir, pp. 415–430. 1985 Paleolithic nutrition, cine Lea and Febiger, Philadelphia. 312(5):283–289. Crawford, M. A. Ebeling, W. Handbook of Indian foods and fibers of arid America. 1968 Fatty-acid ratios in free-living and domestic animals, 1986 Uni- The Lancet 1:1329–1333. versity of California Press, Berkeley. Crawford, M. A., M. M. Gale, M. H. Woodford, and N. M. Evershed, R. P. Casped 1993 Biomolecular archaeology and lipids, World Archaeol- ogy 1970 Comparative studies on fatty acid composition of 25(1):74–93. wild and domestic meats, International Journal of Bio- Ewers, John C. chemistry 1:295–305. 1968 Indian life on the upper Missouri. University of Okla- Curtin, L. S. M. homa Press, Norman. 1949 By the prophet of the earth. San Vincent Incorporation, Falk, Carl R., Steve Holen, and Robert Pepperl Santa Fe. 1978 A preliminary assessment of archaeological resources in Darbyshire, B., and R. J. Henry the vicinity of the proposed White River development, Bad- 1981 Differences in fructan content and synthesis in some lands National Monument, South Dakota. Midwest Ar- Allium species, New Phytology 87:249–256. chaeological Center Occasional Papers in Anthropol- ogy No. 5. Davis, Leslie B., Stephen A. Aaberg, and Linda Scott Cummings FAO 1993 Northern Rocky Mountain paleoecology in archaeo- 1990 Roots, tubers, plantains and bananas in human nutrition. logical, paleobotanical, and paleoethnobotanical per- Food and Agriculture Organization of the United Na- spective. In Plants and their environments: Proceedings tions, Rome. of the First Biennial Scientific Conference on the Greater Farris, Glenn J. Yellowstone Ecosystem, pp. 33–50. National Park Ser- vice, Mammoth, WY. 1995 Pine nuts as an aboriginal food source in California and Nevada: Some contrasts, Journal of Ethnobiology Del Barco, M. 2(2):114–122. 1973 História natural y crónica de la antigua California. Insti- Fish, Suzanne K., Paul R. Fish, and John H. Madsen tuto de Investigaciones Históricas, Univ. Nac. Aut. de México, México, D.F. 1990 Analyzing regional agriculture: A Hohokam example. In The archaeology of regions, edited by S. K. Fish Dering, J. Philip and S. A. Kowalewski, pp. 189–218. Smithsonian In- 1996 Utilization of geophytes on the Southern Plains: Iden- stitution Press, Washington, DC. 44 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Fontana, A., B. Hermann, and J. P. Guiraud Gott, Beth 1993 Production of high-fructose-containing syrups from 1983 Murnong—Microseris scapigera: A study of a staple Jerusalem artichoke extracts with fructose enrich- food of Victorian Aborigines, Australian Aboriginal ment through fermentation. In Inulin and inulin-bear- Studies 1983/2:2–18. ing crops, edited by A. Fuchs, pp. 251–258. Elsevier, Greenhouse, Ruth, Robert E. Gasser, and Jennifer Gish Amsterdam. 1981 Cholla bud roasting pits: An ethnoarchaeological ex- Fowler, Catherine S. ample, The Kiva 46(4):227–242. 1995 Some notes on ethnographic subsistence systems in Greenwood, C. T., and D. N. Munro Mojavean environments in the Great Basin, Journal of Effects of heating on foodstuffs, Ethnobiology 151(1):99–117. 1979a Carbohydrates. In edited by R. J. Priestley, pp. 35–76. Appl. Sci., London. Francis, Julie 1979b Cereals, roots and other starch-based products. In Ef- 1995 Root procurement in the upper Green River basin: fects of heating on foodstuffs, edited by R. J. Priestley, Archaeological investigations at 48SU1002. Paper pp. 373–402. Appl. Sci., London. presented at the 53rd Annual Plains Anthropological Conference, Laramie. Greer, John W. South- Freeman, William L., Gwendolyn M. Hosey, Paula Diehr, and 1965 A typology of midden circles and mescal pits, western Lore Dorothy Gohdes 31(3):41–55. 1989 Diabetes in American Indians of Washington, Ore- Griaule, Marcel gon, and Idaho, Diabetes Care 12(4):282–288. 1938 Jeux dogons. Travaux et memoires de l’Institut Frison, George d’ethnologie, 32. Institut d’ethnologie, Paris. 1983 Stone circles, stone-filled fire pits, grinding stones and High Plains archaeology. In From microcosm to macro- Grinnell, George Bird cosm: Advances in tipi ring investigation and interpreta- 1962 Blackfoot lodge tales. University of Nebraska Press, Lin- tion, edited by L. Davis, pp. 81–91. Plains Anthropol- coln (originally published in 1892). ogist Memoir 19, Lincoln. 1972 The Cheyenne Indians. University of Nebraska Press, Frohne, Dietrich, and Hans Jürgen Pfänder Lincoln (originally published in 1923). 1983 A colour atlas of poisonous plants. Wolfe Publishing, Gupta, Anil K., Narinder Kaur, Maninder Kaur, and Rangil London. Singh Garn, Stanley M., and William R. Leonard 1993 Potential medicinal and nutritional uses of chicory 1989 What did our ancestors eat?, Nutrition Reviews roots and inulin. In Inulin and inulin-containing crops, 47(11):337–345. edited by A. Fuchs, pp. 359–365. Elsevier, Amsterdam. Gentry, Howard Scott Hammat, Hallett H. 1982 Agaves of continental North America. University of Ari- 1976 The Gore Pit site: An Archaic occupation in south- zona Press, Tucson. western Oklahoma and a review of the Archaic stage in the southern Plains, Plains Anthropologist Gill, William Wyett 21(74):245–277. 1880 Savage life in Polynesia. George Didsbury, Wellington. Harris, D. R., and G. C. Hillman (Eds.) Gilmore, Melvin R. 1989 Foraging and farming: The evolution of plant exploitation. 1919 Uses of plants by the Indians of the Missouri River region. Unwin Hyman, London. Republished by the University of Nebraska Press, Lincoln (1975). Hart, Jeff 1976 Montana—Native plants and early peoples. Montana Glander, Kenneth E. Historical Society, Helena. 1982 The impact of plant secondary compounds on pri- mate feeding behavior, Yearbook of Physical Anthropol- Hather, J. G. ogy 25:1–18. 1991 The identification of charred archaeological remains of vegetative parenchymous tissue, Journal of Archaeo- Gohdes, D. M. logical Science 18:661–675. 1986 Diabetes in American Indians: A growing problem, Diabetes Care 9(6):609–613. Hayden, Brian 1990 Nimrods, piscators, pluckers, and planters: The emer- 1995 Diabetes in North American Indians and Alaska na- gence of food production, Journal of Anthropological tives. In Diabetes in America, 2nd ed. National Insti- Archaeology 9:31–69. tutes of Health Publication No. 95-1468. Food Composition and Heat Treatment: Pit-Hearth Cooking 45

Hellendoorn, E. W., M. G. Noordhoff, and J. Slagman ris and G. C. Hillman, pp. 504–519. Unwin Hyman, 1970 Digestibility in vitro of dry mashed potato products, London. Journal of Science and Food Agriculture 26:1461–1468. Johns, Timothy, and Isao Kubo Hendry, George A. F., and Rebecca K. Wallace 1988 A survey of traditional methods employed for the 1993 The origin, distribution, and evolutionary signif- detoxification of plant foods, Journal of Ethnobiology icance of fructans. In Science and Technology of Fruc- 8(1):81–129. tans, edited by M. Suzuki and N. J. Chatterton, pp. Johnson, Ann 119–139. CRC Press, Boca Raton, FL. 1996 The Fog Creek Archaeological Sites, Badlands National Hilty, I., E. M. Peters, M. Edwards, and L. T. Miller Park. Intermountain Field Area, National Park Ser- 1972 Nutritive values of native foods of Warm Springs Indians. vice. Manuscript on file with the author. Extension Circular No. 809. Extension Service, Ore- Johnston, Alex gon State University, Eugene. 1987 Plants and the Blackfoot. Occasional Paper 15, Leth- Hodder, M. A., and L. H. Barfield bridge Historical Society, Lethbridge, Alberta. 1991 Burnt mounds and hot stone technology. Sandwell Met- Kelly, Isabel T. ropolitan Council Borough, West Midlands, Ireland. 1932 Ethnography of the Surprise Valley Paiute, University of Hodge, Allison M., Janice Montgomery, Gary K. Dowse, Barn- California Publications in American Archaeology and abas Mavo, Tahlia Watt, and Paul Z. Zimmet Ethnology 31(3):67–210. 1996 A case-control study of diet in newly diagnosed NI- Knowler, W. C., P. H. Bennet, R. F. Hamman, and M. Miller DDM in the Wanigela people of Papua New Guinea, 1978 Diabetes incidence and prevalence in Pima Indians: Diabetes Care 19(5):457–462. A 19-fold greater incidence than in Rochester, Minn., Holmberg, Allan R. American Journal of Epidemiology 108:497–504. 1957 Lizard hunts on the north coast of Peru, Fieldiana An- Kuhnlein, Harriet V., Donna Appavoo, Natalia Morrison, Rula thropology 36(9):203–220. Soueida, and Patricia Pierrot Holmes, Z. A., and M. Woodburn 1994 Use and nutrient composition of traditional Sahtú 1981 Heat transfer and temperature of foods during pro- (Hareskin) Dene/Métis foods, Journal of Food Compo- cessing, CRC Critical Reviews in Food Science and Nu- sition and Analysis 7:144–157. trition 14:231–294. Laferriere, J. E. Hough, Walter 1992 Cultural and environmental response to drought 1959 Mescal. In Handbook of American Indians north of Mex- among the Mountain Pima, Ecology of Food and Nutri- ico, Part I, edited by F. W. Hodge. Bureau of Ameri- tion 28:1–9. can Ethnology, Bulletin 30. Larsen, Clark Spencer Incoll, L. D., and G. D. Bonnett 1995 Biological changes in human populations with agri- 1993 The occurrence of fructan in food plants. In Inulin and culture, Annual Review of Anthropology 24:185–213. inulin-containing crops, edited by A. Fuchs, pp. 309– LeCoeur, Charles 322. Elsevier, Amsterdam. 1950 Dictionaire ethnographique téda precede d’un lexique fran- Incoll, L. D., G. D. Bonnett, and Beth Gott çais-téda. Institute Français d’Afrique-Noire, Mem- 1989 Fructans in the underground storage organs of some oires 9. Librairie Larose, Paris. Australian plants used for food by Aborigines, Jour- Leopold, A. Carl, and Robert Ardrey nal of Plant Physiology 134:196–202. 1972 Toxic substances in plants and the food habits of early Jackson, Fatimah Linda Collier man, Science 176:512–514. 1991 Secondary compounds in plants (allelochemicals) as Linderman, Frank B. promoters of human biological variability, Annual Re- 1962 Plenty-coups, chief of the Crows. University of Nebraska views in Anthropology 20:505–546. Press, Lincoln. James, Steven R. Linton, Ralph 1989 Hominid use of fire in the Lower and Middle Pleisto- 1944 North American cooking pots, American Antiquity cene, Current Anthropology 30(1):1–16. 9:369–380. Johns, Timothy Lowie, Robert H. 1989 A chemical–ecological model of root and tuber do- 1909 The northern Shoshone, Anthropological Papers of the mestication in the Andes. In Foraging and farming: American Museum of Natural History 2(2):165–307. The evolution of plant exploitation, edited by D. R. Har- 46 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

1924 Notes on Shoshonean ethnography, Anthropologi- O’Dea, Kerin cal Papers of the American Museum of Natural History 1986 Aboriginal health and changes in lifestyles, Australian 20:185–314. Family Physician 15:875–891. 1954 Indians of the Plains. McGraw–Hill, New York. (Re- 1991 Traditional diet and food preferences of Australian printed in 1982 by the University of Nebraska Press, Aboriginal hunter–gatherers, Philosophical Transac- Lincoln.) tions of the Royal Society of London B 334:233–241. Malouf, Richard T. 334:233–241. Ó Drisceoil, Diarmuid A. 1979 Camas and the Flathead Indians of Montana. In Con- 1988 Burnt mounds: Cooking or bathing?, Antiquity tributions to Anthropology No. 7. Department of An- 62:671–680. thropology, University of Montana. 62:671–680. Ohyama, T., O. Ito, S. Yasuyoshi, T. Ikarashi, K. Minamisawa, Mandelbaum, David G. M. Kubota, T. Tsukihashi, and T. Asami 1979 The Plains Cree. Canadian Plains Studies 9. Canadian 1990 Composition of storage carbohydrate in tubers of ya- Plains Research Center, University of Regina, Regina. con (Polymnia sonchifolia), Soil Science and Plant Nu- trition 36(1):167–171. Marchello, M. J., W. D. Slanger, D. B. Milne, A. G. Fisher, and P. T. Berg O’Kelly, Michael J. 1989 Nutrient composition of raw and cooked Bison bison, 1953 Excavations and experiments in ancient Irish cook- Journal of Food Composition and Analysis 2:177–185. ing-places, Royal Society of Antiquaries of Ireland 84:105–155. Marshall, Lorna 1976 The !Kung of Nyae Nyae. Harvard University Press, Olson, Ronald L. Cambridge, MA. 1936 The Quinault Indians. In University of Washington Publications in Anthropology 6. McKey, Doyle 1979 The distribution of secondary compounds within Omar, Mahomed A. K., Mahomed A. Seedat, Robin B. Dyer, plants. In Herbivores: Their interaction with secondary Ayesha A. Motala, Lucinda T. Knight, and Peter J. Becker plant metabolites, edited by G. A. Rosenthal and D. H. 1994 South African Indians show a high prevalence of NI- Janzen, pp. 55–133. Academic Press, New York. DDM and bimodality in plasma glucose distribution patterns, Diabetes Care 17(1):70–73. McWilliams, Margaret 1993 Foods: Experimental perspectives, 2nd ed. Macmillan, O’Shea, John New York. 1989 The role of wild resources in small-scale agricultural systems: Tales from the Lakes and the Plains. In Bad Medved, Eva year economics: Cultural responses to risk and uncer- 1986 Food preparation and theory. Prentice–Hall, Englewood tainty, edited by P. Halstead and J. O’Shea, pp. 57–67. Cliffs, NJ. Cambridge University Press, Cambridge, UK. Meier, H., and J. S. G. Reid Palmer, Edward 1982 Reserve polysaccharides other than starch in higher 1871 Food products of the North American Indians. In plants. In Encyclopedia of plant physiology, New Se- U.S. Department of Agriculture, Report of the Commis- ries Vol. 13A, Plant carbohydrates I: Intercellular carbo- sioner for 1870, pp. 404–428. (Reprinted in An ethnobi- hydrates, edited by F. A. Loewus and W. Tanner, pp. ology source book, edited by R. I. Ford, Garland, New 419–471. Springer-Verlag, Berlin. York, 1986.) Meston, Larry Parsons, Jeffrey R., and Mary H. Parsons 1976 Archaeological investigations in the White River region, 1990 Maguey utilization in Highland Central Mexico: An ar- northwest Nebraska. Department of Anthropology, chaeological ethnography. Anthropological Papers University of Nebraska, Technical Report 76-12. No. 82, Museum of Anthropology, University of Neel, James V. Michigan. 1962 Diabetes mellitus: A ‘‘thrifty’’ genotype rendered Penfield, Marjorie P., and Ada Marie Campbell detrimental by ‘‘progress’’?, American 1990 Journal of 1990 Experimental food science, 3rd ed. Academic Press, San Human Genetics 14:353–362. Diego. 1982 The thrifty genotype revisited. In The genetics of dia- Pennington, Campbell W. betes mellitus, Serono Symposium No. 47, edited by J. 1963 The Tarahumar of Mexico: Their environment and mate- Köbberling and R. Tattersall, pp. 283–293. Academic rial culture. University of Utah Press, Salt Lake City. Press, London. Food Composition and Heat Treatment: Pit-Hearth Cooking 47

1969 The Tepehuan of Chihuahua: Their material culture. Uni- Sillitoe, Paul versity of Utah Press, Salt Lake City. 1983 Roots of the earth. Manchester University Press, Man- chester, UK. Pollock, C. J., and N. J. Chatterton 1988 Fructans. In The biochemistry of plants, Vol. 14, Carbo- Southgate, D. A. T. hydrates, edited by Jack Preiss, pp. 109–140. Academic 1991 Determination of food carbohydrates, 2nd ed. Elsevier Press, San Diego. Applied Science, London. Ray, Verne F. Speth, John D. 1932 The Sanpoil and the Nespelem: Salishan peoples of 1983 Bison kills and bone counts. University of Chicago northeastern Washington, University of Washington Press, Chicago. Publications in Anthropology 5:1–237. Speth, John D., and Katherine A. Spielmann Reagan, Albert B. 1983 Energy source, protein metabolism, and hunter–gath- 1930 Notes on the Indians of the Fort Apache region, An- erer subsistence strategies, Journal of Anthropological thropological Papers of the American Museum of Natural Archaeology 2:1–31. History 31:part 5. Spier, Leslie Reeve, Stuart A. 1928 Havasupai ethnography. Anthropological Papers of the 1986 Food crops and prehistoric social process in the Snake American Museum of Natural History, No. 29. River headquarters, northwestern Wyoming. Unpub- lished Ph.D. dissertation, Department of Anthropol- Spinden, Herbert Joseph ogy, State University of New York, Albany. 1964 The Nez Perce’ Indians, Memoirs of the American An- thropological Association 11(3):165–274. Kraus Reprint, Reeves, Brian O. K. New York. 1990 Communal bison hunters of the Northern Plains. In Hunters of the recent past, edited by L. B. Davis and B. Stahl, Ann B. O. K. Reeves, pp. 168–194. Unwin Hyman, London. 1984 Hominid dietary selection before fire, Current Anthro- pology 25(2):151–168. Robbins, Wilfred W., John P. Harrington, and Barbara Freire-Marreco 1989 Plant-food processing: Implications for dietary qual- Foraging and farming: The evolution of plant ex- 1916 Ethnobotany of the Tewa Indians. Bureau of American ity. In ploitation, Ethnology Bulletin 55, Washington, DC. edited by D. R. Harris and G. C. Hillman, pp. 171–196. Unwin Hyman, London. Roberfroid, M. Steensberg, Axel 1993 Dietary fiber, inulin, and oligofructose: A review New Guinea gardens. comparing their physiological effects, Critical Reviews 1980 Academic Press, London. in Food Science and Nutrition 33(2):103–148. Stern, Michael P., J. Ava Knapp, Helen P. Hazuda, Steven M. Rombauer, Irma S., and Marion Rombauer Becker Haffner, Judith K. Patterson, and Braxton D. Mitchell 1975 Joy of cooking. Bobbs–Merrill, Indianapolis. 1991 Genetic and environmental determinants of type II diabetes in Mexican Americans, Diabetes Care Russell, Frank 14(7):649–654. 1908 The Pima Indians. Bureau of American Ethnology 26 Stevenson, Matilda Coxe (1904–1905), Washington, DC (reprinted in 1975, Uni- Ethnobotany of the Zuñi Indians. versity of Arizona Press). 1915 Bureau of American Ethnology 30 (1908–1909), Washington, DC. Saunders, Charles Francis Steward, Julian H. 1914 With the flowers and trees in California. Robert M. Mc- University Bride and Co., New York. 1933 Ethnography of the Owens Valley Paiute, of California Publications on American Archaeology and Schiffer, Michael B., and James M. Skibo Ethnology 33(3):233–350. 1987 Theory and experiment in the study of technological Stinson, Sara change, Current Anthropology 28(5):595–622. 1992 Nutritional adaptation, Annual Reviews in Anthropol- Schultz, C. Bertrand, and H. T. U. Smith (Eds.) ogy 21:143–170. 1965 INQUA VIIth Congress. Guidebook for Field Conference Szathmáry, Emöke J. E. D Central Great Plains. Nebraska Academy of Science, Lincoln. 1994 Non-insulin dependent diabetes mellitus among aboriginal North Americans, Annual Review of Anthro- Sheldon, A. E. pology 23:457–482. 1905 Ancient Indian fireplaces in South Dakota badlands, American Anthropologist 7:44–48. 48 L. Wandsnider in Journal of Anthropological Archaeology 16 (1997)

Takac, Paul R., and Michael B. Collins ulin-bearing crops, edited by A. Fuchs, pp. 341–346. 1995 New directions in archaeomagnetism: Analyzing the Elsevier, Amsterdam. context of burned rock features. Geological Society of Wandsnider, LuAnn, Heidi Athen, Jennifer Galindo, and America, Abstracts with Programs 27(6):A-301. George MacDonell Tanaka, Jiro 1995 University of Nebraska—Lincoln summer field school 1980 The San hunter–gatherers of the Kalahari. University of in archaeology 1994 report to the Nebraska National Tokyo Press, Tokyo. Forest. Manuscript in the possession of the authors. Teit, James A. Welch, Paul D., and C. Margaret Scarry 1930a Salishan tribes of the Western Plateaus. In 44th An- 1995 Status-related variation in foodways in the Mound- nual Report of the Bureau of American Ethnology, 1927– ville chiefdom, American Antiquity 60(3):397–420. 1928, pp. 23–439. Washington, DC. Wendorf, Michael, and Ira D. Goldfine 1930b Ethnobotany of the Thompson Indians of British Co- 1991 Archaeology of NIDDM: Excavation of the ‘‘thrifty’’ lumbia (edited by Elsie V. Steedman). In 44th Annual genotype, Diabetes 40:161–165. Report of the Bureau of American Ethnology, 1927–1928, Whiting, Alfred F. pp. 441–522. Washington, DC. 1985 Havasupai habitat: A. F. Whiting’s ethnography of a tradi- Thackery, Frank, and A. R. Leding tional Indian culture, edited by S. A. Weber and P. D. 1929 The giant cactus of Arizona: The uses of its fruit and Seaman. University of Arizona Press, Tucson. other cactus fruits of the Indians, Journal of Heredity Wiedfeldt, O. 20:400–414. 1914 Wirtschaftliche, rechtliche und soziale Grundtat- Thoms, Alston V. sachen und Grundformen der Atayalen auf Formosa. 1989 The northern roots of hunter–gatherer intensification: In Mitteilungen der Deutschen Gesellschaft für Natur- und Camas and the Pacific Northwest. Unpublished Ph.D. Völkerkunde Ostasiens, Vol. XV, Part C., pp. 7–55. Tokyo. dissertation, Department of Anthropology, Washing- Yamazaki, H., and K. Matsumoto ton State University, Pullman. 1993 Production of fructo-oligosaccharide-rich fructose Thorburn, Anne W., Janette C. Brand, and A. Stewart Truswell syrup. In Inulin and inulin-bearing crops, edited by A. 1987 Slowly digested and absorbed carbohydrate in tradi- Fuchs, pp. 355–358. Elsevier, Amsterdam. tional bushfoods: A protective factor against diabe- Yanovsky, E., and R. M. Kingsbury tes?, American Journal of Clinical Nutrition 45:98–106. 1938 Analysis of some Indian food plants, Association of Of- Turner, Nancy J., and Harriet V. Kuhnlein ficial Agricultural Chemists 21(4):648–665. 1983 Camas (Camassia spp.) and riceroot (Fritillaria spp.): Zardini, Elsa Two liliaceous ‘‘root’’ foods of the Northwest Coast Polymnia sonchifolia Indians, Ecology of Food and Nutrition 13:199–219. 1991 Ethnobotanical notes on ‘‘yacon,’’ (Asteraceae), Economic Botany 45(1):72–85. Turner, Nancy J., Leslie M. Johnson Gottesfeld, Harriet V. Kuhnlein, and Adolf Ceska Zigmond, Maurice L. Kawaiisu ethnobotany. 1992 Edible wood fern rootstocks of Western North Amer- 1981 University of Utah Press, Salt ica: Solving an ethnobotanical puzzle, Journal of Eth- Lake City. nobiology 12(1):1–34.

Turner, Nancy J., Laurence C. Thompson, M. Terry Thompson, Notes and Annie Z. York 1 It is unlikely that the variation is owed to heat treatment. A 1990 Thompson Ethnobotany. Royal British Columbia Mu- comparison of lipid–protein ratios between raw and cooked seum Memoir No. 3. tissues in the Anderson (1989) series (cooked values not pre- Vogel, Manfred sented here) indicates that while the absolute weights of tis- 1993 A process for the production of inulin and its hydro- sue change with cooking, the lipid–protein ratio remains sta- lysis products from plant material. In Inulin and inu- ble (r = 1.00, p = 0.000, n = 14). lin-bearing crops, edited by A. Fuchs, pp. 65–76. Else- 2 Gill goes on to recount a tale of sabotage, in which a commu- vier, Amsterdam. nity was invited to a ti feast, but then was tumbled into the Vukov, K., M. Erdélyi, and E. Pichler-Magyar oven to be themselves roasted. 1993 Preparation of pure inulin and various inulin-con- 3 The Headwaters Experimental Group out of the University taining products from Jerusalem artichoke for human of Texas at Austin has recently initiated a concerted effort in consumption and for diagnostic use. In Inulin and in- this area.