Boletín de la Sociedad Botánica de México 66: 81-91, 2000 DOI: 10.17129/botsci.1614

Bol. Soc. Bot. México 66:81-91 (2000)

MAGUEY ( SPP.) LJTJLIZATION IN MESOAMERICAN CIVILIZATION:

A CASE FOR PRECOLUMBIAN "PASTORALISM"

JEFFREY R. PARSONS 1 y J. ANDREW ÜARLJNc2 'Museum of Anthropology, University of Michigan, Ann Arbor, Michigan, U.S.A. e-mail: jpar@umich. edu 2Mexico-North Research Network, 16 de Septiembre 402, Cd. de Chihuahua, Chihuahua, México.

Abstract. Archaeologists have generally neglected the implications of maguey cultivation in their assessments of prehispanic Mesoamerican cultural development in the highla nds of central and north-central Mexico. Recent ethnographic, ethnohistoric, and archaeological studies provide the basis for the formulation of new hypotheses about the importance of the complementarity of maguey and seed-based cultivation in ancient . We see cultivated maguey as functionally analogous in Mesoamerica to the role of pastoral ism in the development of archaic civilizations in the , , , and . Fu tu re archaeological and ethnobotanical research should provide key new information for refining and testing these hypotheses. Key words: Agave, archaeology, ethnobotany, maguey, mesoamerica. Resumen. Los arqueólogos generalmente han descuidado la importancia del cultivo del maguey (Agave spp.) en sus apreciaciones del desarrollo cultural prehispánico en las tierras altas del centro y norte-centro de México. Estudios etnográficos, etnohistóricos, y arqueológicos recientes han sido Ja base para la formación de nue­ vas hipótesis sobre la importancia de la complementariedad de los sistemas agrícolas, basados tanto en la producción anual de semillas como en la producción vegetativa y continua del maguey. Consideramos el cultivo del maguey en las tierras altas de Ja Mesoamérica prehisp<1nica como análogo del pastoreo en el desarrollo de civi li zaciones antiguas en los Andes, Asia, África, y Europa. Futuras investigaciones arqueológicas y etno­ botánicas deben proveer nuevos datos para refinar y probar estas hipótesis. Palabras clave: Agave, arqueología, etnobotán ic a, maguey, mesoamérica.

rehispanic Mesoamerica was the only one of the Most archaeologists have overlooked the full sig­ Pworld's ancient primary civilizations which lacked nificance of cultivated maguey, Agave salmiana Otto a domestic herbivore. With domestic camelids in the ex Salm-Dyck, Agave nwpisaga Trel., Agave atrovirens Central Andes, and sheep and goats in much of the Karw. ex Salm-Dyck, Agave ferox K. Koch, Agave hook­ Olcl World, food proclucers in virtually ali other re­ eri Jacobi, and Agave americana L. (Gentry, 1982) in gions where Archaic States existed were able to sig­ the prehispanic economy of highlancl central and nificantly extend their productive landscapes in to drier north-central Mexico. An extreme expression of this and colder zones and overa full annual cycle i.e., sorne viewpoint is the assertion by Blanton et al. ( 19.81: 174) of them became full- or part-time herclers, ancl herder­ that "in the highland valleys [of Mesoamerica] the cultivator relationships became important in the long­ surest way of producing a large surplus was to plant term clevelopment of socio-political complexity. How maize everywhere." could ancient Mesoamericans -particularly in the In this paper we argue that cultivatecl maguey highlands (tierras altas) of central and north-central (which reproduces vegetatively and matures overa 7-25 Mexico, where severe winter frosts and highly seasonal year cycle) and annual seed crops were fully comple­ rainfall limited seed-based agriculture to one growing mentary in the Mesoamerican tierras altas. Further­ season per year (figure 1 )-, ha ve attained such a more, we suggest that maguey made available in comparably high leve! of organizational complexity? prehispanic Mesoamerica some of the same kirids of

81 ______Parsons JR, Darling A. 2000. Maguey (Agave spp.) utilization in Mesoamerican civilization: a case for precolumbian "Pastoralism". Boletín de la Sociedad Botánica de México 66: 81-91. jEFFREY R. PARSONS Y J. ANDREW DARLING

adaptive strategies complementary to annual seed­ Chihuahua based agriculture as did domestic camelids in the t Central Andes and sheep and goats in Asia, Africa, ,,-,---,km • N o 100 200 300 and Europe. We attempt to develop the fo llowing inter­ relatecl hypotheses:

1. In the Mesoamerican tierras altas the development of complex society during the Middle and Late Formative (table 1) depended u pon the domes­ tication of maguey as a primary complement to

seed crops for the production of food and fiber Pacmc Ocean (Sauer, 1941).

2. The expansion of ancient Mesoamerican civi liza­ tion into the drier highland of central and Figure l. Middle America, showing the approximate north-central Mexico depended upon the full in­ extent of the tierras altas (s haded) . tegration of seed-based and maguey-based agri­ cultura! production. duction of both seed crops and maguey in com­ paratively well-watered e ore are as ( the irriga ble 3. Prehispanic agricultura! production in the drier river valleys) and more specialized agave (and highland regions of central and (especially) north­ probably nopal [ Opuntia sp.]) production in the central Mexico was based upon generalized pro- drier peripheral zones beyond the reach of effec­ tive irrigation.

Table 1. Va l ley of Mexico, Prehispanic Chro­ 4. The archaeological record hints ata major change nology (adapted from Sanders, Parsons, and in the technology of prehispanic maguey produc­ Santley 1979). tion after the Classic period. This technological change is suggestive of basic differences in Clas­ Date Period Phase sic and Postclassic political econom y.

1520 A.D . Materials and Methods Late Postclassic Aztec 111 1350 A.D. It has long been known that maguey was an impor­ Middl e Postclassic Aztec 1-11 tant prehispanic so urce of food and fiber in highland 1150 A.D. Mexico. Nevertheless, the plants themselves are often Early Postc lassic Maza pan poorly preserved in archaeological contexts, and the Coyotl ate leo technology and organization of prehispanic maguey 950 A.D. Epiclassic Metepec exploitation have remained poorly understood. Re­ Xo lalpan cent ethnographic research (Sánchez, 1980; Ruvalcaba, 1983; Salinas and Bernard, 1983; Patrick, 1985; Ran­ 750 A.D. Tlamimilolpa gel, 1987; Parsons and Parsons, 1990) has provided Classic Miccaotli sorne new insights. 150 A.D. Tzacualli Maguey sap and jlesh. The maguey plant provides a rich 50 B.C. Terminal Formative Pat lachique store of both sap and edible flesh. Today, through­ out the highlands of southern, central and north­ 250 B.C. Late Formative Ti coman central Mexico, maguey sap (aguamiel) is acquired 500 B.C. for human use by means of procedures which inter­ Middle Forrnative La Pastora rupt the final stage of a plant's normal 7-25 year El Arbo lill o maturation process in order to extract the sap through 900 B.C. twice-daily scraping and sucking operations over a Early Formative Bomba period of 3-6 months. Individual plants in cultivated lxtapaluca fields typically approach maturity continuously 1200 B.C. throughout the year: the timing of their planting and

82 MAl;U EY (A GAVE Srr.) UT!l..IZATION IN MESOAMERICAN Cl\'ll.IZATION : A CASE FOR PRECOl..UMlllAN "PASTORALISM" replacement is explicitly managed so as to insure con­ quantities of protein, carbohydrates, Vitamin C, cal­ tinuous productivity when usually no more than 5-10% cium, phosphorous, potassium, iron, magnesium, of a field's maguey plants are producing sap at any par­ selenium, chromium, and zinc. Anderson et al. ticular point in time (Parsons and Parsons 1990:336) . (1946:888) found that in the diets of their study group Over its 3-6 month production period, a single plant of rural Otomí villagers supplied 12% of total typically yields severa! hundred liters of sap, and a calories, 6% of total protein, 10 % of total thiamine, hectare of land planted in maguey typically yields 5000- 24% of total riboflavin, 23% of total niacin, 48% of 9000 liters of sap per year (Parsons and Parsons total Vitamin C, 8% of total calcium, and 20% of total 1990:338). The sap may be allowed to ferment to form iron. Ross (1944, cited in Fish et al., 1986) found that pulque (as it usually is today), or it may be consumed 100 grams of cooked agave fl esh contains 347 calo­ in its unfermented liquid form, or it may be boiled ries and 4.5 gms of protein. down to form thick syrup or solid sugar. Aguamiel and lt appears that in many highland agricultura! con­ pulque are quite unstable, and cannot remain unused texts maguey can produce approximately as many for more than about a week. As syrup or sugar, how­ calories and essential nutrients per hectare as the ever, maguey sap is much more durable, and in these standard seed crops; when the plant's fl esh and sap fonns sap surpluses can readily be stored ancl redis­ are both consumed, maguey can potentially produce tributed over a period of months, or even longer. more calories than seed crops on a given uni t of land The modern Tarahumara of ex­ (Parsons and Parsons, 1990:337-338, 345). Most im­ tract agave sap by simply mashing up the plant's leaves portant of al!, maguey can be interplanted with seed ancl squeezing out the liquid in a single operation (Bye crops in virtually al! agricultura! settings, and when et al., 1975). The leaves, heart, and stalk of the maguey this is done the overall nutritional and energetic plant can also be eaten. The Tarahumara, for exam­ output on a given unit of land is potentially doubled. ple, prepare cakes of baked agave flesh which can be lt is important to emphasize that this potential storecl up to six months (Bye et al., 1975). doubling of agricultura! productivity in the tierras altas Maguey sap and flesh are rich in both nutrients and applies to both prime, highly fertile land as well as calories. Ruvalcaba (1983:89) cites analyses showing to more marginal terrain. Modern scholars have usu­ that one liter of pulque contains 574 calories. David­ ally associatecl maguey production with marginal land so n ancl Ortiz de Montellano (1983:155) repon that (e.g., Simpson 1952; Sanders et al., 1979), but this one tablespoon of maguey sap contains significant viewpoint is simply inaccurate: the ethnographic ancl

•••12.5 grams 22.9 grams 29 grams 38 grams

55.1 grams 61.7 grams 74.8 grams

Figure 2. Prehispanic spindle whorls for maguey fiber, sh owing diversity in size.

83 jEFFREY R. PARSONS y J. ANDRE\\' DARLING historie literature is filled with clear evidence that Because maguey is an ideal famine food in high­ maguey has been a key, primary component of agri­ land Mexico, stands of wild or feral maguey may have cultura! productivity in virtually all types of agricul­ been deliberately extended in prehispanic times so tura! land in the tierras altas, including productive as to provide food for sedentary cultivators during zones of deep soil and adequate humidity (Parsons times of serious crop scarcity, as was commonly done and Parsons, 1990). Cultivated maguey will .also flour­ by prehisto1ic colonists with certain types ofintroduced ish on dry, infertile land where no other cultigens will wild or semi-wild plants in ancient (Kirch, grow: consequently, there are large areas of poor 1984:131-132). The extensive stands of wild maguey agricultura! land in central and north-central Mexi­ and nopal which today occur throughout the most co today where maguey is the dominant or exclusive marginal highland zones in central and north-central cultigen. Mexico might be relicts of such prehispanic practic­ Combining maguey and seed crops would have es. The relationship over time between cultivated and maximized subsistence security for prehispanic agri­ feral/wild maguey is a key problem, still little stud­ culturalists in the tierras altas: annual energy produc­ ied by either archaeologists or botanists (Rzedowski, tivity on many kinds of cultivated land could be 1986). cloubled; agricultura! productivity could be extend­ ed over a full annual cycle; agricultura! productivity Maguey fiber. In prehispanic Mesoamerica there were could be extended in to nearby drier, colder, and less only two major fibers suitable for making textiles: fertile areas; maguey is resistant to clrought, frost, ancl cotton and maguey ( ixtle fiber). Cotton coulcl not be hail, ali common causes of seed-crop failure in the grown at elevations above c. 1800 m asl in the tier­ tierras altas; the year-round procluctivity of maguey ras altas, so there would have been a great need for could be combinecl with the long-term storability of large quantities of locally produced maguey-fiber tex­ seed crops; and most maguey-sap exploitation could tiles in highlancl Mexico. easily be deferred to the winter agricultura! off-sea­ An average maguey leaf produces roughly 75 gm son (because after the initial "castration" operation, of clried fiber (Parsons and Parsons, 1990:157). A when the first stages of a plant's nascent central stalk typical maguey plant has 20-30 leaves, and thus pro­ are removed, the collection of the plant's sap can be vides sorne 2000 grams of ixtle. A modern carrying postponed for up to six months, without any appar­ cloth (ayate) made ofwoven maguey thread measures ent loss in productivity) (Macedo, 1950). about one meter square and weighs about 200 grams.

Figure 3. Mocl e rn iro n mague)' scraper mo un tecl in wooden hanclle.

84 MAGUEY (AGAVE Srr.) UTILIZATION IN MESOAMERICAN CIVILIZATION: A CASE FOR PRECOLUMBIAN "rASTORALISM"

' ''

128 g

' Q ' '' 173 g 101 g

cm 1 1 1 1 o 5

117 g ~

547 g

Figure 4. Examples of prehispanic trapezoidal tabular basalt scrapers.

Thus, one maguey plant provides enough fiber for ple living in the in A.D. 1500) would about 10 square meters of cloth. More precise calcu­ annually require the fiber production (c. 600,000 kilos) lations would have to make allowance for variable from sorne 10,000 hectares (roughly 5% of the total thread thickness, thread spacing, fiber quality, and type potentially arable landscape in the Valley of Mexico) of costume, etc. Nevertheless, these rough estimates of cultivated maguey. This same amount of land could suggest that one maguey plant would have provided potentially, at the same time, have produced large enough fiber for outfitting an average precolumbian amounts of aguamiel, cooked maguey flesh, interplant­ person in highland Mexico with most of the maguey­ ed maize or beans, and dried maguey stumps for use fiber textile required for clothing overa period of a as household fue! (Parsons and Parsons, 1990:337-338; few years. Evans, 1990). On an average hectare of cultivated land in high­ The sub-specific variability of maguey is impressive. land Mexico today, about 30 maguey plants can be Today, for example, there are at least a dozen named exploited each year for both sap and fiber (Parsons varieties of cultivated maguey in the eastern Mezquital and Parsons, 1990:336, 338). Thus, one hectare of (Salinas and Bernard, 1983), and in the late nine­ cultivated maguey could potentially outfit approximate­ teenth century there were more than two dozen dif­ ly 30 people with the maguey cloth they need for a ferent named varieties on the large pulque haciendas few (say three) years. Or, assuming each average per­ in nearby Tlaxcala and Puebla (Blásquez and Blásquez, son requires one-third of his/her wardrobe to be 1897). Different varieties have different properties and replaced each year, then one hectare of cultivated characteristics: sorne are recognized as better sap maguey would provide the annual maguey cloth needs producers, others are better for fiber, while others for sorne 90 people (once again, let's simplify our perform better on certain types of land, etc. The calculations by calling it an even 100). On this basis, origins of this botanical diversity remain poorly un­ 1,000,000 people (approximately the number of peo- derstood, but we suspect that sorne of it ultimately

85 jEFFREY R. P ARSONS y J. ANDRE\\" DARLI NG

Figure 5. Experimental use of a trapezoidal tabular basalt scraper. derives from delibe rate prehistoric human selection position to infer the extent to which different work­ for specific qualities in an increasingly specialized and shops were involved in specialized vs. ge neralized diversified maguey productive system. These key is­ spinning, in the production of cotton vs. maguey sues can only be illuminated with more ethnobotan­ thread, and in tributary vs. market vs. domes tic modes ical and paleo-ethnobotanical investigation. of production. We also suspect that the elaborate stamped, molded, and incised designs so character­ Results istic of Postclassic spindle whorls may relate to spe­ cific social units (e.g., caljJ'lllli, altepetl, etc:) associated Ethnographic and archaeological studies indicate that with particular kinds of whorl, thread, and textile severa! categories of stone and ceramic tools can be production (Carrasco, 1976; Parsons, 1975). associated with sorne aspects of prehispanic maguey Ceramic spindle whorls were extremely scarce in processing. highland central Mexico before the Postclassic. Sev­ era! possible factors might explain this (e.g., spinning Spinning. In central Mexico molcl-made Postclassic without whorls; the use of perishable wooden whorls; ceramic spindle whorls are common, and archaeolo­ the use of simple perforated circular fragments of bro­ gists car1 distinguish between small whorls (almost ken pottery (she rd disks) that are not always recog­ always weighing less than c. 7 grams) used for spin­ nizable as spindle whorls). However, the difference ning cotton fib er and large whorls (almost always between Classic and Postclassic in this regard is so weighing more than c. 11 grams) used for maguey fiber dramatic as to suggest a major reorganization of spin­ ( e.g., Parsons, 1972; Smith and Hirth, 1988) (figure ning after Classic times . We suspect that spinning (ancl 2). Furthennore, we now know that those whorls which possibly weaving as well) became more specialized and weigb be tw een 20-30 gm can be used to produce a more efficient during the Postclassic than it had been wide range of fine to coarse maguey thread, whereas earlier. lighter (c. 11-15 gm) and heavier (c. 35-1 40 gm) whorls coulcl only have been usecl to produce a much nar­ Scraping maguey fiber. Today maguey fibe.rs are usual­ rowe r range of fine or coarse maguey thread, respec­ ly detached from the encasing fl esh.with an iron scrap­ ti\1ely (Parsons ancl Parsons, 1990:329, 331). er mounted in a wooden handle (figure 3). These Once we know more about spindle whorl weights scrapers are clull, even-eclged tools designed to scrape at specific wÓrkshops we should be in a much better away the flesh without cutting or shreclding the fiber.

86 MAc; uEY (Ac;AVE Srr.) UT!L.I ZATION IN MESOAMERICAN C:l\'IL.I ZATIUN : A C:AS I·: FUR PRE COL. LJ ,\ líl!AN "rASTURALISM"

We think the prehispanic analog is a trapezoidal pect that the sc raper plane would have predominated ground-stone too! made of tabular basalt (figure 4). in the context of earlier and/ or more generalized fiber This latter too! is most common in the Postclassic production, where high efficiency was less important. (Tesch and Abascal, 1974; Brumfiel, 1976; Sanders et If so, then increasing specialization and efficiency of al., 1979), but it also occurs in at least one Late-Te r­ maguey fiber processing (manifested archaeologically minal Formative context (Serra, 1988). These imple­ by a progressive shift from scraper planes to trape­ ments have been variously call ed azadas, f elsite knives, zoidal ground-stone scrapers) appears to have paral­ lwes, and des}Zbradores. Recent experimental work (Par­ le led increasing spinning effi ciency (manifested so ns and Parsons, 1990: 175) shows that they are ad­ archaeologicall y by a dramatic increase in quantities mirably suited for scraping maguey fiber (if the and variability of ceramic spindle whorls) during encasing flesh has been softened beforehand by heat­ Postclassic times in the Mesoamerican tierras altas. ing and rotting), and we are fully convinced that this was their primar)', perhaps exclusive function (figure 5). The extrnction of maguey sap. There is a ve ry distinctive These trapezoidal scrapers are widespread through­ and highly specialized modern iron too! usecl d aily out the highlands of central and north-central Mexi­ for scraping the surface of the sap-collecting cavity co and in the North American Southwest (Spence, in the maguey plant's interior. The prehispanic ana­ 1971; Tesch and Abascal, 1974; Brumfiel, 1976; Sanders log of this elliptical or circular iron scraper appears et al., 1979; Sejourne, l 983:Fig. 137; S. Fish et al., 1986; to be a distinctive "strangulatecl" plano-co nvex stone Trombolcl, 1985, 1989; Cabrero, 1989; Mastache et al., sc raper (figure 7), which has a broad clistribution in 1990) , especiall y in Postclassic times. Over time in the highlands of central and north-central Mexico (Ca­ central Mexico they tend to displace another distinctive mio, 1979 [1922] :214; Dibble ancl Anderson, l 963:Fig. too !: the sc raper plane o r "turtleback sc raper" (Tol­ 778; Spence, 1971; Meigham, 1976; Sanclers et al., 1979; stoy, 1971) (figure 6). Experimental wo rk with archae­ Michelet, 1984; Rodríguez, 1985:199; Trombold, 1985, ological scraper-planes in the southern highlancls of 1989; Cabrero, 1989:234, 238-241; Mastache et al., Mexico (Hester and Heizer, 1972) has sh own that 199 0:189; Parsons and Parsons, 1990) . T his implement repeated downwarcl blows with the rounded sicle of seems to appear as early as Late Formative times and this too! (which typically we ighs about 400 grams) are to increase markedly in frequency in the Postclassic. effective to m as h up raw agave leaves, while the fl at These scrapers apparently do not occur archaeolog­ bottom side of the same implement can serve to scrape ically outside the tie1rn fría lending additional support th e mas hed fl esh away from the fibers. to our belief that this artifact was used exclusively in Trapezoidal scrapers were probably used in more the production of maguey sap. spec iali zecl maguey fiber production in which greater We infer that over time maguey sap processing in efficiency in fibe r extraction was ac hieved by means of central and north-central Mexico shifted from ( 1) coo king and rotting leaves to soften the fl esh. We sus- something akin to the ethnographic Tarahumara

==------­ ------...

Planar edge Planar surface

Figure 6. Archaeologica l scraper plane (adapted from Hes ter a nd H e izer, 1972).

87 jEFFREY R. P ARSONS Y J. ANDREW D AR LI NG procedure (in which agave leaves are simply mas hed Mesoamerica without domestic maguey would have up and the sap squeezed out in a single operation), remainecl atan organizational leve! comparable to to (2) something comparable to the historically known that of, for example, highland Colombia, where process in central Mexico in which both the sap and both maguey ancl domes tic camelids were absent fiber of individual plants are extractecl overa period from the economic base of p1'ehispanic chiefcloms of severa! months. This shift may have been linked throughout a developmental sequence of over 2000 to a need for greater overall efficiency of plant us e years in which state-level organization was appar­ in the context of increasingly specialized economies ently never attained (Reichel-Dolmatoff, 1965). stimulated by both high population clensities ancl increasecl tributary cl emands. 2. Maguey procluction, and agricultura! production in general, remainecl generalized throughout most Discussion of the Formative in the ti erras altas, with no sig­ nificant shifts towarcl greater specialization or ef­ O ur present knowleclge about the clistribution in time ficiency until the clevelopment of increasingly ancl space of key archaeological materials related to complex and urbanized society in southern and prehispanic maguey use in central and north-central central Mexico late in the first millennium B.C. Mexico is still too limited to provicle adequate devel­ With their increased overhead cos ts and greater opment or tests of our hypotheses. We predict that spatial se paration between foocl producers and new archeological research will show the following: food consumers, urbanizecl states of the first millennium A.D. woulcl have faced the need to 1. Since at least Middle Formative times maguey cul­ intensify and expand all types of agricultura! pro­ ti vation has been an equal partner with annual cluction. seecl crops in agricultura! procluction in the Me­ soamerican tierra fría. We cl oubt that agriculture 3. The northwarcl expansion of Mesoamerican civi­ without clomestic maguey coulcl have sustainecl li zation into north-central Mexico in Classic times prehispanic state-level society in this pan of Me­ was underwritten, in large part, by the integration soamerica. Perhaps a post-Formative highlancl of specialized maguey-nopal producers (living in

Figure 7. Modern iro n scraper (left) and prehispanic (two at right) obsidian scrapers used in sap collecting. The lowe r handles of both obsidian scrapers have been broke n off.

88 MAGUEY (AGAVE Srr.) UT!LIZATION IN MESOAMERJCAN CJVILIZATION: A CASE FOR PRECOLUMBIAN "rASTORALISM"

more marginal, peripheral landscapes) and more ed spindle whorls in northern Middle A.merica de­ generalized seed crop-maguey cultivators (living in rive from late first millennium and early second restricted, more productive irrigable landscapes) . millennium contexts in north-central and north­ The effective in tegration of these agriculturally­ western Mexico ( e.g., Ekholm, 1942; Kelly, 1945, generalized cores and agriculturally-specialized 1947, 1949; DiPeso et al., 1974; Meigham, 1976; peripheries would have been dependent upon the Foster, 1985), and in the adjacent Southwestern existence of redistributional networks large U.S. (DiPeso, 1951, 1956) although it is not yet enough to move staples over significant distanc­ clear whether these whorls were used for spinning es in a regular and predictable m anner. Because cotton or maguey thread. Formative redistributive networks were too restrict­ ed, Mesoamerican civilization was unable to ex­ The largest prehispanic Mesoamerican empires pand northward into north-central Mexico until were based in highland central Mexico, and most of after c. AD. 200 (Braniff, 1989; Kelley, 1990; Trom­ these imperial systems date to the Postclassic. Do the bold, 1990; Darling, 1997). changes we note in the technology of maguey pro­ duction, saltmaking, and food preparation, and in the 4. The Classic-to-Postclassic transition in central and surviving impressive sub-specific variation in cultivated north-central Mexico saw the development of in­ maguey, reflect primarily the demands of imperial creasingly specialized and efficient economies. Part administrators for greater productive effi ciency and of this shift might relate to the changing charac­ specialization in the core areas of their domains? Will ter of urbanism and the dynamics of urbanization we eventually discove r n o tably less technological e.g., the development of large centers inhabited change or botanical diversity in areas where such very predominantly by non-food producers. An­ imperial demands were weak or absent? other aspect of this change may relate to the Looking further back in time, it should also be development of new status roles and the need to useful to think about the relationships between the distinguish them by implementing new sumptu­ competitive arena of chiefly poli tics ( e.g., Helms, 1979) ary rules, such as regulating the production and and the initial domesticatio n and accompanying bo­ use of pulque and certain types of clothing made tanical diversification of maguey in the tierra fiía dming of differen t types of fibers and finer vs. coarser the Early and Middle Formative at a time when trib­ thread. Most important of ali might have been the al Big Men and aspiring Chiefs in ranked societies changing nature of tribute, exchange, and gov­ throughout southern and central Mexico were seek­ ernance, whereby, for example, different kinds of ing higher leve ls of local productivity to sustain and cloth and beverages assumed new functions as ma­ enhance their prestige. terial symbols of new socioeconomic and socio­ poli tical relationships (Anawalt, 1980; Hicks, Acknowledgments 1987). This paper is based on fi eldwork supported by the 5. Sorne techniques and procedures developed for National Geographic Society, the National Science maguey exploitation in north-central Mexico Foundation, the University of Michigan Phoenix during the Classic period might have been sub­ Memorial Project, and the University of Michigan sequently "imported" from there back into cen­ Museum of Anthropology. We are also grateful for tral Mexico. If, as we have argued, it was in useful discussions about maguey that we have had over north-central Mexico that maguey was especially the years with Russel Bernard, Elizabeth Brumfiel, critica! in the domestic ancl political economy, Thomas Charlton, Robert Cobean, Susan Evans, Paul then it might be logical to expect that it was in Fish, Susan Fish, Leah Mine, Guadalupe Mastache, Ben the context of state expansion into this arid, Nelson, Deborah Nichols, Mary H. Parsons,' Will-iam northernmost part of ancient Mesoamerica that Sanders, Mari Carmen Serra, Michael Smith, and the most effective and efficient maguey exploita­ Charles Trombold. tion first developed. Weintraub ( 1992), for exam­ ple, reports the presence of maguey fiber and leaf Literature Cited fragments in flotation samples from Late Classic­ Epiclassic contexts at La Quemada, Zacatecas Anawalt P. 1980. Costume ancl Control: Aztec Sumptuary -perhaps the earliest known post-Archaic exam­ Laws. Archaeology 33:33-43. ples of such material in northern Mesoamerica. Anderson R.K., Calvo C., Serrano G. y Payne G. 1946. A Furthermore, some of the earliest well-document- Study of the Nutritional Status and Food Habits of Ot-

89 jEFFREY R. PARSONS Y J. ANDREll' DARLI NG

orni Indians in the Mezquital Valley of Mexico. Ame1i­ P. Eds. The A-rchaeology aj West and Northwest Mesoaineri­ ca.n journal of Pu.blic H ea lth and the Nation 's Health ca, 327-352. Westview P ress. Boulcler. 36(8):883-903. Gentry H.S. 1982. of Conti¡iental North A11unica. Uni­ Blanton R., Kowalewski S., Fe inman G. y Appel J. 1981. versity of Arizona Press, Tucson. Ancient Mesoa:merica. Cambridge University Press, Cam­ Helrns M. 1979. Ancient Panmna: Chiefs in Search aj Powe1: bridge. University of Texas Press, Austin. Blásquez P. y Bh1squez l. 1897. Tratado del Maguey. 2a edi­ Hes ter T. y Heizer R. 1972. Problems in the Functional ción. Narciso Bassols, Puebla. Interpretation of Artifacts: Scraper Planes frorn Mi tia and Braniff B. 1989. Oscilación de la frontera norte Mesoame­ Yagu l, Oaxaca. Contributions of the University of California ri cana: un ensayo nuevo. Arqueología 1:99-114. México, A1cltaeological Research Facility 14: 107-123. D.F. Hicks F. 1987. First Steps Toward a Market-Integrated Econ­ Brumfie l E. 1976. Sf;ecialization and Exchange at the Late omy in Aztec Mexico. En: Classen H. y Van ele Velde P. Postclassic (Aztec) Co1!lmunity of Huexotla, Mexi co. PhD, Uni­ Eds. Early Sta.te Dynamics, Stu.dies in Hwnan Society 2:91- versity of Michigan, Ann Arbor. 107. E. Brill, Leiden. Bye R., Burgess D. y Trias A. 1975. Ethnobotany of the Ke ll ey J. C. 1990. The Retarded Formative of th e North­ Wes tern Tarahumara of Chihuahua, Mexico. 1: Notes on west Frontier of Mesoamerica. En: M. Carmena Macias, the Genus Agave. Botanical i\tluseum Leaflets, Harvard Coord. El Preclásico o Formativo - Avances y Persj;ectivas, University 24(5) :85-11 2. 405-42 3. INAH, México, D.F. Cabrero M. T. 1989. Civilización en El Norte de México. UNAM, Kelly l. 1945. Tite Anhaeology aj tite Au.tlan-Tuxcacu.esco !\rea México, D.F. of jalisco, Part J. Ibero-Americana Núm. 26. U niversity Carrasco P. 1976. La Sociedad Mexica antes de la Conquista. of Cali forn ia Press, Berkeley. Historia General de México 1: 165-288. El Colegio ele México, Ke ll y l. 1947. Excavations al Apatzingan, Michoacan . Viking México, D.F. Fund Publ ications in Anthropo logy, num. 7. New York. Dar! in g J.A. 1997. Obsidian Distribution and Exchange in the Kelly l. 1949. The Archaeology of the Autlan-Tuxcacuesco Ana North.-Central frontier of Mesoainerica, A.D. 0-1500. PhD, aj Jalisco, Part JI: The Tuxcacuesco-Zapotitlan Zone. Ibero­ University of Michigan, Ann Arbor. Arnericana num. 27. University of Cali fornia Press, Berkeley. Daviclson J. y Ortiz el e Montellano B. 1983. The Antib ac­ Kirch P. 1984. Tite Evolution of Polynesian Chiefdoms. Cam­ terial Properties of an Aztec Wouncl Remecly. journal of bridge Un iversity Press, Cambridge. Ethnophar11Lacology 8: 149-161. Maceclo M. 19 50. Manual del Magueyero. Bartolomé True­ Dibble C. y Anclerson A. (Trad.). 1963. Florentine Codex: co, México, D.F. General HislO'I)' of th.e Things of New Spain, b)' fray Berna1c Mastache G., Cobean R., Rees C. y J ackson D. 1990. Las

90 MAGUEY (AGAVE Srr. ) UTI LIZATION I N M ESOAMERICAN CI\'ILIZATION: A CASE FOR PRECOL.UMBIAN "rASTORALISM"

Range l S. 1987. Etnobotánica de los Agaves del Valle del Mez­ Serra M. 1988. Los Recursos Lacustres de la Cuenca de México quital. Tesis de L icenciatura, UNAM, México, D.F. durante el Fonnativo. UNAM, México, D.F. Re ichel-Dolmatoff G. 1965. Colombia. F. Praeger, New York. Smith M. y H irth K. 1988. The Development of Prehispanic Rodríguez F. 1985. Les Cltichimeques. Cen tre d'Etudes Me x­ Cotton-Spinning Techno logy in Western More los, Mexico. icai nes et Centra m ericaines, México, D.F. joumal aj Field Arclweology 15:349-358. Ross vV. 1944. The Present-Day Dietmy Habits oj tite Papago Spence M. 1971. Sorne Lithic Asseu1.blages aj Western Zacate­ Jndian.s. Tesis de L icenc iatura, Un iversity of Arizona, cas and Durango. Mesoamerican Stud ies num. 8, University Tucson. M useum, South ern Illino is U niversity, Carbondale. Ruva lcaba J. 1983. El Maguey Manso: Historia y Presente de Tesch M. y Abascal R. 1974. Azadas. Comunicaciones 11:37- Ef;azoyucan, . Universidad Autónom a C ha pi ngo, 40, Puebla. Depto. de Industrias Agrícolas, Serie Ciencias Socia les Tolstoy P. 1971. Util itarian Artifacts of Cen tral Mexico. En: núm. 4, Texcoco. Ekh o lm G. )' Berna! I. Eds. Arcltaeology aj Northern Me­ Rzedowski, J. 1986. Vegetación de México. Editoria l L imusa, soamerica, Pt. 1. Handbook oj Middle American Indians, Méx ico, D.F. 10:270-296. U niv. of Texas P ress, Austin . Sa hagün Fray B. 1969. Historia General de las Cosas de Nueva Trombold C. 1985. A S u mrnary of the Archaeology in the Esj)(l1ta, Garibay A. Ed. 3 vo ls. Editoria l Porrüa, Méxi­ La Quemada . En : Foster M. y Weigan d P. Eds. co, D.F. Tite Arcltaeology aj West and Nortltwest Mesoainerica 237-267. Salin as J. y Bernard R. 1983. Etnografía Otomí. !NI, México, D.F. Westview P ress, Bould er. Sa nchez J. 1980. Etnografía de la Sierra Madre Occidental: Trombo ld C. 1989. Compreh e ns ive Summary of the 1986 Tepe/manes y Mexicaneros. INAH, México, D .F. Excavatio ns of MV-1 38, A Vi ll age Outli er of La Quema­ Sanders W.T., Parsons J. y Santley R. 1979. Tite Basin aj d a in Zacatecas, Mexico. Informe a l INAH, Méx ico, D .F. lvlexico: Ecological Processes in tite Evolution aj a Civilization. Trombold C. 1990. A Reconsideratio n of Chro n o logy for Academic Press, New York. the La Quemada Portio n of th e North ern Mesoameri­ Sauer C. 1941. The Personali ty of Mexico. GeogmjJhical Review can Frontie r. Ame1ican Antiqu.ity 55:308-324. 31:353-364. We in traub P. 1992. The Use oj Wild and Cultivated Plan! Foods Sejo urne L. 1983. Arqueo logía e Historia del Valle de J\lléxico, al tite Site aj La Quemada, Zacatecas, Mexico. Tesis de L i­ de Xo chimilco a Amecameca. Sig lo Ve in t iu no Editores, cenciatura, State U niversity of New York at Bu ffa lo, México, D.F. Buffa lo, New York.

9 1