Archaeologists often use ceramic stylistic changes as a way of identifying sociopolitical fluctuations such as state and empire formations, and these stylistic shifts are interpreted to reflect materialization of political identity or authority (Costin Reference Costin2022; DeMarrais et al. Reference DeMarrais, Castillo and Earle1996; Pauketat and Emerson Reference Pauketat and Emerson1991). This can be an initial approach for identifying stylistic changes over time, but it has been demonstrated ethnographically that shifts in vessel form and surface treatment can be superficial—by copying designs from neighboring groups (Gosselain Reference Gosselain2000, Reference Gosselain, Stark, Bowser and Horne2008). Adopting new designs may reflect changes in desired relationships between groups, and such changes may indicate anything from a shared aesthetic appeal to emulation. When trying to understand connections in ceramic production and sociopolitical fluctuations, it is important to consider other aspects of ceramics that are less susceptible to copying, such as clay recipes and formation techniques (Roux and Courty Reference Roux and Courty2013; Womack et al. Reference Womack, Wang, Zhou and Flad2019).
One way of exploring how knowledge of ceramic production is passed over time is by considering the role of communities of practice. Often defined as social networks in which potters share knowledge about production and motor skills (e.g., Eckert Reference Eckert, Cordell and Habicht-Mauche2012; Eckert et al. Reference Eckert, Schleher and James2015; Sassaman and Rudolphi Reference Sassaman and Rudolphi2001; Stark Reference Stark, Habicht-Mauche, Eckert and Huntley2006), a community of practice usually involves hands-on learning. This type of learning, which has been observed ethnographically (Bowser and Patton Reference Bowser, Patton, Stark, Bowser, Horne and Longacre2008; Minar and Crown Reference Minar and Crown2001), may include surface finishing but also temper choice and preparation skills. The identification of ceramic communities of practice suggests that scholars can detect the social networks in which potters learned their craft, and that the study of ceramic technology and any associated social changes should ideally involve analysis of all production steps, including clay recipes and vessel-forming and -finishing techniques.
Understanding continuity and change in ceramic production is particularly salient in contexts where stylistic changes in ceramics have been linked to sociopolitical fluctuations such as state and empire formation. In western Mexico, the precontact Purépecha empire, circa a.d. 1350–1530, has long been associated with imperial-style ceramics that are characterized by spouts, globular supports, polychrome decorations, intricate drawings; and wax-resist firing (Cabrera Castro Reference Cabrera Castro, Crespo, Viramontes and Paz1996; Pollard Reference Pollard1993; Rubín de la Borbolla Reference Rubín de la Borbolla1939, Reference Rubín de la Borbolla1941). The presence of these ceramics is often associated with other material changes, such as increasing numbers of metal artifacts and the appearance of semicircular pyramids called yácatas. These material changes are explained as a reflection of state formation sometime during the Middle Postclassic period (a.d. 1100–1350), culminating in the Purépecha empire during the Late Postclassic (a.d. 1350–1530) (e.g., Pollard Reference Pollard2008). Despite these material changes, previous work on pottery production in the imperial core Lake Pátzcuaro region suggests that paste recipes were relatively consistent for over 1,200 years (~a.d. 250–1530) (Cohen et al. Reference Cohen, Hirshman, Pierce and Ferguson2023; Hirshman Reference Hirshman2008, Reference Hirshman and López Varela2017, Reference Hirshman, Huntington, Arnold and Minich2018). Other ceramic paste and mineralogical data from one site in the core region similarly indicates conservative continuation of technology over the same time period (Cohen et al. Reference Cohen, Galaty and Fisher2018; Cohen et al. Reference Cohen, Pierce and Fisher2019). Moreover, an explanation for continued technological practices of paste recipes and clay procurement within the context of changing ceramic symbols and decorations remains unexplored (however, see Hirshman Reference Hirshman, Huntington, Arnold and Minich2018).
This analysis builds on studies of ceramic communities of practice, which explore the connections between ceramic production, style, and sociopolitical changes. By combining petrographic and geochemical ceramic data, I show that Purépecha potters at the site of Angamuco were part of a community of practice that persisted despite broader political changes. Although these political changes are visible materially in new ceramic forms and surface decoration, I show that ceramics were mainly created using local clays and materials over multiple centuries. I contend that these procurement and production activities were consistent because they reflect a community of practice that resulted in the conservative continuation of technology. This study highlights that although changes in ceramic form and decoration may signal sociopolitical changes such as state and empire formation, information about production techniques and paste recipes can shed light on the technological impacts of these changes.
Social changes in the imperial heartland
Angamuco was an urban landscape mainly located on lava flows in the Lake Pátzcuaro Basin, Michoacán, Mexico (Figure 1). In the centuries before European arrival in Mexico, this lake basin was the center of the Purépecha empire, which at times controlled 750,000 km2 of western Mesoamerica. Although the exact dates of state and empire formation are unclear, beginning at least during the Early to Middle Postclassic periods (ca. a.d. 900–1350), the lake basin and surrounding regions were home to numerous nucleated urban settlements often on raised lava flows (also called a malpaís) (Forest Reference Forest2023; Michelet Reference Michelet, Sanders, de Escobar and Cobean2008; Pollard Reference Pollard2008).
Figure 1. Map showing Angamuco excavation areas (A through G) and excavation overview photo of imperial Area C (map data from Digital Globe 2015).
Angamuco was occupied from at least the Preclassic through Late Postclassic periods (a.d. 250–1530), including during Purépecha state and empire formation sometime during the Middle to Late Postclassic periods (a.d. 1150–1530) (Cohen Reference Cohen2021). Stone and earthen architecture includes not only monumental features such as rectangular and semicircular yácata pyramids, a ballcourt, and plazas, but also domestic features such as rooms, terraces, and roads (Fisher et al. Reference Fisher, Cohen, Solinis-Casparius, Pezzutti, Bush, Forest and Torvinen2019). Survey and excavation data recovered between 2009 and 2014 demonstrates that Angamuco was an urban palimpsest, occupied intensively during certain periods, reorganized, and repurposed during others (Cohen Reference Cohen2021). Some of the earliest occupations date to the Preclassic period, circa a.d. 250, when residents lived on the lower elevations of the site to be close to water resources and cultivatable land. During the Epiclassic and Early Postclassic periods (ca. a.d. 700–1100), the population expanded and settled on the upper elevations of Angamuco, living in plaza groups that were connected by an extensive road system. In the late Middle Postclassic (a.d. 1100–1350), some or all of the upper areas were abandoned, and occupants started repurposing and building large plazas and yácata pyramids in the lower zones. Radiocarbon determinations and imperial Purépecha-style artifacts indicate that there was an imperial presence primarily in the lower elevations of Angamuco during the Late Postclassic period (Cohen Reference Cohen2021). Exploring changes in the pre-imperial and imperial-style pottery is therefore an important step in the larger process of understanding the impact of new political formations and technologies on people living in Angamuco and the broader Lake Pátzcuaro region.
Previous ceramic research
Imperial-style pottery in the Lake Pátzcuaro area is hypothesized to have formed out of earlier versions of forms and styles in the region (García García Reference García García2009; Hirshman et al. Reference Hirshman, Lovis and Pollard2010), and geochemical characterization shows that paste recipes have remained relatively consistent over 1,200+ years (Cohen et al. Reference Cohen, Pierce and Fisher2019; Cohen et al. Reference Cohen, Hirshman, Pierce and Ferguson2023; Hirshman and Ferguson Reference Hirshman and Ferguson2012). Data from the Lake Pátzcuaro area and the nearby Zacapu Basin suggest that pre-state- and empire-period pottery consisted of both coil and mold-built fine and coarse wares (e.g., Jadot Reference Jadot2016; Pollard Reference Pollard1999). Pre-imperial forms included various bowls, incised bowls for grinding (molcajetes), jars, and spoons; decoration was often a reddish or cream-colored slip and occasional painted red stripes and geometric forms. Subsequent Purépecha state- and empire-period ceramics consisted of forms such as bowls, tripod bowls with globular supports, jars, and spouted vessels; decoration included reddish or cream-colored slip, sometimes with red, black, and white paint and wax-resist firing (Figure 2) (Cohen Reference Cohen2016:165–174; Hirshman et al. Reference Hirshman, Lovis and Pollard2010; Pollard Reference Pollard1993, Reference Pollard1999).
Figure 2. Examples of imperial style pottery from excavation Area C (overview photo in Figure 1) at Angamuco. Imperial Purépecha pottery is characterized by globular supports, spouts, and polychrome and wax-resist firing (negative) decorations. Drawings by Daniel Salazar Lama.
Previous ceramic archaeometry in the Lake Pátzcuaro area suggests that most pottery manufacturing was a localized activity (Cohen et al. Reference Cohen, Pierce and Fisher2019; Hirshman and Ferguson Reference Hirshman and Ferguson2012; Pollard et al. Reference Pollard, Hirshman, Neff, Glascock, Williams and Weigand2001). A regional geochemical (Neutron Activation Analysis, or NAA) study summarized previous work by combining 470 ceramic samples from four pre- and post-state and empire sites (Angamuco, Erongarícuaro, Tzintzuntzan, and Urichu; Figure 1) (Cohen et al. Reference Cohen, Hirshman, Pierce and Ferguson2023). The study identified eight compositional groups among the samples and found that they were locally produced and long-lived, and that they represent a complex ceramic economy with differential community participation. At Angamuco, petrographic analysis of ceramics and clay indicate that potters exploited local mineral-rich clay sources, adding various tempers and generating numerous, complex paste recipes (Cohen et al. Reference Cohen, Galaty and Fisher2018). Similar to the regional geochemical study, the mineral data show that multiple small-scale producers structured pottery production at the site, and that techniques of manufacture likely remained relatively stable over long periods.
These existing pottery data align with some of the broader interpretations of occupation and complexity in the lake basin. For example, excavation and survey data indicate that beginning at least during the state period, Purépecha imperial elites began to exploit local, centrally-produced materials rather than non-local, high-status goods (Pollard and Cahue Reference Pollard and Cahue1999). An emphasis on local production would have consolidated political economic power in the imperial core region and required co-option of local labor. Yet, the long-term ceramic data demonstrate that regardless of distinctive changes in styles and forms, ceramic manufacturing was consistently a local activity. Technological conservatism has been considered as an explanation for local manufacturing in the Lake Pátzcuaro Basin (Hirshman Reference Hirshman, Huntington, Arnold and Minich2018), but further discussion is necessary for understanding why techniques persisted in a region with substantial political changes over 1,200+ years.
Methods and materials
The methods and initial results of the Angamuco petrographic and geochemical analyses are reported elsewhere (Cohen et al. Reference Cohen, Galaty and Fisher2018; Cohen et al. Reference Cohen, Pierce and Fisher2019; Cohen et al. Reference Cohen, Hirshman, Pierce and Ferguson2023); consequently, here the focus is on the combined analyses. To address questions relating to communities of practice and production techniques used to create pre-imperial and imperial style pottery, this study analyzes 300 sherds from Angamuco. Sherds were collected from pre-imperial and imperial habitation and public contexts (Table 1). In addition, 30 samples of natural, local clay are analyzed. All sherd samples were retrieved during stratified excavations that have been chronometrically dated (Cohen Reference Cohen2021). Discussion of the clay sampling both on and around the Angamuco landscape, and ceramic attribute analysis on a larger sample of 5,861 sherds, is available elsewhere (Cohen Reference Cohen2016; Cohen et al. Reference Cohen, Galaty and Fisher2018).
Table 1. Number and location of samples.
1 The total number of NAA samples consists of 300 sherds and 30 geological samples (n = 330).
2 The total number of petrographic samples consists of 36 sherds and nine geological samples (n = 45). The petrographic samples were subsampled from the NAA sample based on the interpretation of geochemical composition groups.
Given that artifacts from Angamuco had not been systematically analyzed before this study, the sampling strategy was designed to lay a foundation for understanding the site's chronology and ceramic artifacts. Ceramic specimens were selected from all seven excavation areas (A–G in Figure 1/Table 1) and all strata that included identifiable forms and/or decorated examples (Cohen Reference Cohen2016:109–112; Cohen et al. Reference Cohen, Galaty and Fisher2018; Cohen et al. Reference Cohen, Pierce and Fisher2019). A nonrandom sampling strategy was based on a hierarchy of variables to maximize contexts that crosscut different sociofunctional excavation areas and decorative and formal classes.
From the attribute sample, 300 excavated sherds were selected for NAA at the University of Missouri Research Reactor (MURR). For petrographic analysis, an additional 36 sherds were intentionally selected that were assigned to NAA compositional groups with the greatest Mahalanobis membership probabilities (>78 percent, but most >90 percent) for each group. This was to ensure that the samples were most representative of that geochemical group. Furthermore, this sampling strategy allows for the correlation of the ceramic groupings produced by several methods, and it incorporates a wider petrographic sample, enabling additional relatively rare types to be assigned to petrographic groups.
Thirty clay samples were collected from excavation contexts or within proximity to the site, such as from probable reservoirs to the southeast and southwest of Angamuco. The clays were made into briquettes and then submitted for NAA. Nine clay samples were selected for petrographic analysis that could be linked with NAA groups. Some samples were unassigned and used to evaluate naturally occurring minerals in clays at the site and nearby, and to determine whether the clays could be matched to any of the excavated ceramics.
Geochemical data overview
Sherds and clays were submitted to MURR and analyzed according to the standard protocol (Glascock Reference Glascock and Neff1992). The MURR analyses produced elemental concentration values for 33 elements in most of the samples. Cluster analysis, principal components analysis, and discriminant analysis using the MURR Statistical Routines and SPSS for multivariate statistics detected patterns in the elemental concentrations. The NAA analysis identified four compositional groups within the Angamuco sherd and clay sample. These groups were initially reported as Groups A, B, C, and D (Cohen et al. Reference Cohen, Pierce and Fisher2019: Figure 3), but they were reanalyzed in a regional study, combined with existing geochemical groups, and renamed systematically as Pottery Group A (PGA), Pottery Group 6 (PG6), Pottery Group 1 (PG1), and Pottery Group 7 (PG7), respectively (Cohen et al. Reference Cohen, Hirshman, Pierce and Ferguson2023). Some specimens could not be assigned (n = 42 sherds, n = 20 clay); further discussion is limited to the total assigned samples (n = 268).
Figure 3. Photomicrographs in cross-polarized light showing the main fabric groups (fg). The differences in the groups are largely based on basalt—whether basalt inclusions are present and whether there are grain-size differences—and on the presence or absence of ash. For example, fg 5 includes small basalt inclusions and ash; fg 6 shows even finer basalt inclusions; the basalts in fg 7a and 7b are small and rounded or coarse; fg 8 basalts are coarse; fg 9 includes rare fine basalt; and fg 11 is coarse vehicular basalt sand. The image of fg 10 shows that it does not include basalt or ash but instead a coarse quartz sand added as a tempering agent. Modified from Cohen et al. Reference Cohen, Galaty and Fisher2018:Figure 3. Photomicrographs by Michael Galaty.
Of the four geochemical groups identified in the Angamuco sample, two are linked with previously established groups in the region, and two are thus far unique to the site. Group PGA (n = 122) is the largest of the groups and found in imperial Area C (n = 35), but also in earlier Areas F (n = 33) and A (n = 27). PGA corresponds with a previously established potting recipe with a broad procurement zone in the southern part of the lake basin (Hirshman and Ferguson Reference Hirshman and Ferguson2012). PG1 (n = 41) is the smallest of the groups and is most common in pre-imperial Areas F (n = 17) and A (n = 11). PG1 also shows a comparable composition to a previously established group from a broad procurement zone in the western and northern parts of the lake basin. Geochemical group PG6 (n = 52) is different from all previous MURR compositional groups and from the Angamuco clay samples. Of the 52 samples assigned to this group, the majority (n = 28) were from imperial Area C, whereas smaller numbers appeared in Areas A (n = 6), B (n = 6), D (n = 3), E (n = 5), F (n = 3), and G (n = 1). Finally, PG7 (n = 43) is the only group that overlaps with the clay samples, suggesting that this group represents a local source (Cohen et al. Reference Cohen, Pierce and Fisher2019; Cohen et al. Reference Cohen, Hirshman, Pierce and Ferguson2023). The 10 matching clay samples were collected from shallow depressions around the site that were probably used to hold water. PG7 samples were recovered in pre-imperial and imperial contexts from excavation Areas A, C, and F (n = 12 each), and in the Areas B (n = 3), D, and E (n = 2 each).
Petrographic data overview
The 36 petrographic subsamples were carefully studied under plane and cross-polarized transmitted light to evaluate their mineralogy and clay matrix characteristics. Clay and ceramic groups were constructed based on qualitative analysis of the thin sections, following the methods outlined by Whitbread (Reference Whitbread1995) that draw from the field of soil micromorphology.
The petrographic analysis identified seven main ceramic fabric groups (fg) and four clay groupings (Figure 3). All the sherds analyzed contained one or more possible tempers, including basalt sands (97 percent) and/or ash (75 percent). Possible tempers include coarse or fine basalt sands of various types (i.e., vesicular, feldspathic [ortho- vs. plagioclase], presence/absence of hypersthene, etc.), ash (composed of clear and/or colored glass), and perhaps coarse quartz sand. The differences between the ceramic fabric groupings are thus most visible in the kinds of basalt inclusions present and in their grain-size differences, as well as the presence/absence and amount of ash. For example, fg 5 was defined based on the presence of small basalt inclusions and ash, whereas fg 6 basalt inclusions are finer than fg 5, with less ash and fewer plagioclase inclusions. The basalts in fg 7 are either small and rounded or coarse, and fg 8 temper consists of relatively coarse basalt sands. Fg 9 includes a rare fine basalt and is not ashy, whereas fg 11 tempers are a coarse vesicular basalt sand that is also not ashy. Fg 10 is the only fabric without basalt or ash, and it includes a coarse quartz sand, possibly added as a temper.
Combining the geochemical and petrographic datasets
When examining the NAA and petrographic data in combination, nearly all of the fabric groups appear in the imperial Late Postclassic contexts, but there are several distinct temper sources exploited for pottery production during that time period that are linked to NAA PG7 and local Angamuco clays (fg 6, 7a, 9, and 10—or 44 percent) (Table 2). Each of these “imperial” tempers provides insight into technological changes or conservativism. For example, in fg 6, the vessels may have been smoothed (Figure 3). They are not reduced, but the groundmasses of some are oxidized (bright red) at the surface. They contain various basalt inclusions, all of which are small and rounded, and a mix of clear and colored volcanic ash. Fg 7a includes mainly reduced and some oxidized surfaces. Samples in this subgroup are nonporous. Small, rounded basalt inclusions are dominant and typically double spaced. The basalt is black opaque with small-to-medium plagioclase inclusions and was probably added as a temper. There is some volcanic ash with clear glass. Fg 9 is reduced and nonporous; samples include very rare, fine basalt sand; and they are not ashy. Mineral inclusions are rare, as are opaques (which, when present, are small and black) and TCFs. Inclusion distribution is open spaced. These samples are very fine compared to all others. Groundmass is inactive, grey in XPL, and tan in PPL. Although it does not appear to include fragments of tuff, the groundmass does perhaps look tuffaceous (i.e., grey in XPL). The clay may be derived from a tuff parent material. Finally, fg 10 is nonporous and very fine, and it contains no basalt and no ash. Mineral inclusions are very rare, suggesting that some processing may have occurred. The sample contains very coarse, quartz sand, which was probably added as a temper. This is unlike the other samples, does not match an NAA group, and could therefore be a long-distance import.
Table 2. Links between NAA and petrographic groups.
1 NAA groups PG6 and PG7—along with fabric groups 6, 7a, and 7b—are from local materials due to links to local clay samples.
2 Late Postclassic only.
3 Classic and Middle Postclassic only.
There are some imperial context samples that can be linked to NAA groups PGA and PG1, both of which are hypothesized to originate in the western and southern parts of the Lake Pátzcuaro Basin (Hirshman and Ferguson Reference Hirshman and Ferguson2012). Only PG1 matches a fabric group, Group 5, which can be associated with samples that date to the Classic through Late Postclassic periods. In other words, clay selection and paste preparation are not unique characteristics of the imperial style ceramics in this sample.
A comparison of the ceramic groups resulting from the NAA and petrographic studies reveals that they are far from identical, but that they do show some correspondence (Table 2). For example, local fg 6, 7, and 11 can be matched with NAA PG7, which in turn matches local clays. Nonetheless, the results also show incongruities, such as the match between fg 5 and NAA PG6 and PG1. In addition, none of the NAA groups match fg 9 and 10.
Nearly all of the fabric groups appear in the imperial Late Postclassic contexts (Areas A, B, C, D), and fg 6, 7a, 9, and 10 (44 percent) were identified exclusively in these imperial areas. Fg 6 and 7a can be matched with local clay sources. These ceramic fabrics could be considered temporal markers for the Late Postclassic period. In contrast, groups 5, 7b, 8, and 11 were identified in excavation contexts that range from the Classic through Late Postclassic periods. Two of these groupings, 7b and 11, are likely derived from local clays. None of the NAA groups were associated exclusively with the Early and Middle Postclassic period areas of Angamuco (Areas E, F, G), although fg 11 was only associated with the Classic and Middle Postclassic period contexts and not the Late Postclassic, which could suggest that the fabric was a temporal marker. The fabric groups can be linked to locally derived clays in all time periods. Overall, these patterns indicate that there was greater variation in ceramic fabric recipes in the later time periods, but that many of the recipes were used throughout the Angamuco occupation.
Discussion
This study focuses on clay and paste recipes as two facets of ceramic production that can provide information on the technological choices of potters within their communities of practice. There are several clay and temper sources exploited throughout the Angamuco occupations that show continuity over time, yet there are also distinct sources used for pottery production during the imperial Late Postclassic period. During each time period, several NAA and fabric groups are present in the sample. One of these, fg 5, was probably produced by taking lake basin clays and adding small basalt inclusions and ash. There is some variation in the amount of ash added, which could reflect individual choice when sifting or levigating out natural inclusions or adding temper. Similarly, during the Late Postclassic imperial period, when potters produced fineware vessels, they relied on local Angamuco clays but used finer basalt inclusions and occasionally some ash.
The persistence in the NAA and fabric groups from the Classic through Late Postclassic periods indicates that despite widespread environmental and societal changes over hundreds of years and a shift in the ceramic forms and styles being produced, the same resources and production techniques were being used. This reflects not only the technological conservatism of the ceramic production and learning, but also the success of the producers in applying the same techniques and raw materials. Although there was a social shift that motivated potters to produce imperial-style vessels, it was not so extensive that they changed the raw materials they selected or the techniques that they used to prepare their clay. Related regional analysis of standardization of form and decoration reinforce the notion that shifts in pottery were confined predominantly to form and surface treatment (Hirshman Reference Hirshman2008; Hirshman et al. Reference Hirshman, Lovis and Pollard2010).
Also noteworthy is that not only was local production continuous throughout Angamuco's occupation, but exchange of vessels or raw materials within the lake basin persisted. Some of the sources of clay and temper used to produce Angamuco vessels can be linked to other parts of the Lake Pátzcuaro Basin based on NAA groupings and previous clay sampling (Hirshman and Ferguson Reference Hirshman and Ferguson2012). This suggests that potters would have travelled to other sources or traded with other potters for raw materials, and/or that vessels were produced elsewhere and traded through exchange networks. Significantly, these fabric types and paste recipes continued throughout the Angamuco occupation.
It is important to mention that this study incorporates data from both paste recipes derived via NAA and from added tempers visible in petrographic samples. This is relevant because ethnographic work with potters has shown that any interpretation of paste continuity over time should take into consideration the thresholds of energy that potters use to travel for resources and locally available raw materials (Arnold Reference Arnold2000). Today—and in the past—most potters do not travel more than 1 km to obtain clays and tempers, and few likely travel more than 7 km (see also Levine et al. Reference Levine, Fargher, Cecil and Forde2015). This means that potters often use local sources of clays and tempers, and that this distance impacts the variability of pastes and interpretations of the variability (discussions in, for example, Arnold et al. Reference Arnold, Neff and Bishop1991; Arnold et al. Reference Arnold, Neff and Glascock2000). By including the petrographic data and geological samples, this study takes into account potential internal variability of the clays. Discussion of the regional NAA dataset and associated raw material sources elsewhere in the Lake Pátzcuaro area (Cohen et al. Reference Cohen, Hirshman, Pierce and Ferguson2023; Hirshman and Ferguson Reference Hirshman and Ferguson2012) supports the interpretation of (1) local (to Angamuco) sources of raw materials and (2) links to paste recipes elsewhere in the lake basin.
When considered with ethnographic accounts in western Mexico, the likelihood that communities of potters were passing down knowledge of raw material sources and paste recipes is unsurprising. In her review of 80 years of ethnographic literature in the Purépecha region of Michoacán, Hirshman (Reference Hirshman2020) emphasizes the resilience of household, kin-based production organization throughout periods of sociopolitical and economic changes in the twentieth century. As a household craft, pottery production persisted through demographic shifts resulting from migrant wage labor to Mexico City and the US, government artisanal programs, and modern construction activities. George Foster's work (Reference Foster1948, Reference Foster1955, Reference Foster and Matson1965:47) pointed out that mid-twentieth-century Lake Pátzcuaro–area potting communities were highly conservative in terms of their technological techniques over time and their broader economic structure. Eduardo Williams (Reference Williams2018) found similar patterns nearby in the town of Huancito, and he interpreted this conservatism as avoidance of economic catastrophe. In sum, the clear continuity in raw material use, paste recipes, and/or trade relationships indicates that although vessel forms and styles changed over time, the potting communities producing and exchanging these ceramic goods were resistant to wider societal changes, including state and empire formation.
Conclusion
This research on ceramic technologies reinforces the importance of examining production techniques such as temper choice and paste recipes alongside larger sociopolitical shifts. As discussed here at the site of Angamuco, Mexico, despite changes in pottery form and styles, long-lived production techniques persisted for centuries. The self-sufficient potting communities of practice were conservative in their production methods, directly passing down knowledge and techniques over the centuries.
Similar methods can be applied to other archaeological investigations of social change. The study of ceramic technologies can provide information on the location of raw materials, the methods of ceramic production, and the extent of exchange networks. This would strengthen understanding of potential continuities or changes in communities of producers, and how they were impacted by large-scale social changes such as empire formation and colonial contact.
Acknowledgments
Thank you to the Consejo de Arqueología of the Instituto Nacional de Antropología e Historia, who permitted the Angamuco fieldwork and the export of artifacts to the US for analysis. I am grateful to INAH Michoacán, the Fontezuelas ejido community, the Tzintzuntzan heritage community council, and the LORE-LPB research team for help during the fieldwork. Ceramic analysis initially occurred at El Centro de Estudios Mexicanos e Centroamericanos (CEMCA) in Mexico City. Mike Galaty conducted the petrographic analysis, and Daniel Pierce led the geochemical characterization. Amy Hirshman and Helen Pollard have provided key insights into Purépecha ceramics; Marion Forest, Elsa Jadot, Karine Lefebvre, and Andrea Torvinen have offered suggestions on pottery identification and interpretation. Earlier versions of this article were presented at the Society for American Archaeology (Ceramic Petrographers in the Americas) and the American Anthropological Association (Ceramic Ecology) conferences. I am grateful for the comments of three anonymous reviewers to improve this manuscript.
Competing interests
The author declares no competing interests.
Data availability statement
Ceramic data are available at the University of Missouri Research Reactor and through the author.
Funding statement
This research was supported by the National Science Foundation (BCS-1344333), an NSF-subsidized grant to MURR (BCS-1415403), the Dumbarton Oaks Research Library, and the University of Washington.
Open access
