SIX

The Social Evolution of Potters’ Households in Ticul, Yucatán, Mexico, 1965–1997

DEAN E. ARNOLD

INTRODUCTION

This chapter provides two answers to the question, What do households do? First, their members participate in craft activities, and second, the social composition of these craft households evolves and changes through time. One type of craft activity that households practice is pottery making, but what happens to potters’ households through time? How do they change? To answer these questions, I will trace the changes in the composition and location of potters’ households in Ticul, Yucatán, Mexico, between 1965 and 1997. By describing these changes, I hope to provide some insight into the processes that affect the organization of potters’ households through time, and how those processes might be applicable to the study of ancient households.

Ticul is one of the major cities in southern Yucatán and, like many cities in Latin America, it has experienced a surge in population during the last forty years. In 1960, its population was 14,789 (Salas 1967:50), and in 1965, the production of pottery, hats, and shoes were major cottage industries. By 1997, a sign outside the city proclaimed that the city had 22,900 inhabitants, an increase of 155 percent of the 1960 population. Hat making had disappeared, the production of shoes had greatly diminished, but the production of pottery had flourished and grown.

THE NATURE OF HOUSEHOLDS

Households should be understood from two different perspectives. First, they need to be understood in terms of the way in which they utilize space (see, e.g., the chapters by Ciolek-Torrello, Douglass and Heckman, and Snow, this volume). This perspective involves the way in which the household space is subdivided and allocated to the activities that are performed within it. Do these areas consist primarily of general, multipurpose space, or do they consist of a set of specialized areas reserved for specific activities? All households utilize both types of space, but the amount devoted to each type varies across time and from culture to culture. The space used for food preparation and cooking, for example, often consists of an area devoted exclusively to those activities and includes a hearth and specific locations for the storage of food, water, and service ware.

Craft activities, however, provide a challenge to households because they often require additional space for the storage of tools, raw materials, and completed and partially completed craft products. Furthermore, weather conditions may require that some craft activities (such as forming and drying pots) take place inside in a protected environment (see Arnold 1985:61–98). As craft activities become more important to a household’s economic well-being and replace subsistence activities located away from the household, more time must be devoted to craft production. This increased amount of time often requires more space for craft activities and this increased space competes with that needed for other, more general household activities.

The second perspective necessary to understand households consists of the way that they are organized with a focus on the relationships of the people within them. While the ways in which households utilize space is a critical dimension for craft production and one that is important to archaeologists, this chapter focuses on only one craft activity (making pottery) and the households that perform that activity. This paper thus utilizes the potter’s household as the unit of analysis and then describes what happens to the number, location, and composition of such households through time.

While the focus of this chapter is the social composition, location, and continuity of households, the utilization of household space cannot be ignored. The main issue concerns the amount of space devoted to craft activities that may increase to such an extent that the term “household” becomes an inadequate way to describe the locus of production without some change in terminology, such as “household production,” “individual specialization,” “workshop,” “household workshop,” “dispersed workshop,” or “workshop industry” (Costin 1991; Peacock 1982; Van der Leeuw 1976). The use of the term “household,” for example, to describe the locus of ceramic production over the thirty-two years of my research in Ticul is inadequate. Households expand and change, and production (or some aspects of it) may move outside of them. One can not simply call the production locus a “household” over such a period because so many changes have occurred to the locus of production since 1965. From a material and spatial perspective, production organization in 1997 looks very different than it did thirty-two years previously.

I struggled with this issue as I tried to categorize and quantify my data. Initially, the term “household” seemed to be a useful way to describe the social unit of ceramic production. As the craft evolved and production facilities expanded, however, a few production facilities moved outside the household into what could be best described as workshops and small factories. These facilities utilized task segmentation with specialists in activities such as mining raw materials and procuring firewood and within production units assigned to forming, firing, and painting (Arnold 2008). This observation led me to use the term “workshop” for such units. But then it became clear that the line between “households” and “workshops” was fuzzy depending on the criteria used for each category. As I saw household after household evolve into “workshop-like” production facilities between 1965 and 1997, it was obvious that trying to separate the two was unproductive and unnecessary. Regardless of the way that they had changed, expanded, and evolved, however, all production units that appeared to be “workshops” were still household-based. In fact, using any criterion to define the production units in Ticul as “households,” “workshops,” or anything else was useless because they were changing so fast and at any one moment of time, there were several different “types” or “modes” of production present in the community, depending on how they were defined.

As I struggled with the description and categorization problem, it was evident by 1997 that none of the production units were “households” of the kind I had observed thirty-two years earlier, even though many had developed from those same households. The only way around this problem was to use a different descriptive terminology. I thus decided to simply describe the locus of ceramic production as a “production unit” and did not worry about categories that archaeologists had tried to use (e.g., Costin 1991; Peacock 1982; Van der Leuuw 1976). This change freed me from the fuzziness of categorization but allowed me to describe how the locus of production units that began as the potters’ “households” in 1965 had changed.

This categorization problem revealed a larger, rather obvious issue about the relationship of social categories and cultural evolution. Preoccupation with categorization may obscure fundamental processes of change because evolution varies according to how one defines the social “types” involved.¹ The Ticul data thus demonstrate that even through the tiny span of thirty-two years (from an archaeological perspective), describing social evolution simply in terms of changing fixed categories invented and imposed by the investigator may provide few insights about underlying evolutionary processes. This awareness thus challenges one to describe change in deeper and more profound ways.

Costin (1991) also struggled with the categorization problem, and her response was the development of a set of parameters that underlie the organization of craft specialization. These parameters are probably the most important outcome of the specialization literature in the last twenty years and provide a useful way to think about specialization, but they give specialized meanings to words already in common use and can easily create confusion, potentially obfuscating the great significance of her work. Using Costin’s parameters conceptually (rather than labels for those parameters) thus helps to uncover the underlying processes of the evolution of craft production and craft specialization. This chapter explores Costin’s (1991:15–16) parameter of “scale,” which consists of the composition of the production unit and includes the number of individuals working in the unit and the way in which it recruits labor for production.

MAYA HOUSEHOLDS

A traditional Yucatec Maya household usually consists of a lot with a house more or less in the center of the property. Such a lot may also be occupied by an extended family consisting of more than one nuclear family related by descent. Each family, however, may have its own house. In cities and towns, a stone fence defines the limits of the property and the house may occur next to the street so that access to the lot behind is possible only through the house. The open space around the house (or in the back of the lot) usually contains ornamental and economically important plants² and trees, a pole-and-thatch structure for cooking and eating, and perhaps other small structures such as a shaded basin (pila) for washing clothes, a storage structure for maize (if the family grows its own maize), and occasionally other structures. Craft activities, such as making pottery, may occur inside the house; outside in the shade, weather permitting; or sometimes in one of the small structures on the houselot.

Spatial configurations of traditional Maya households have changed greatly since 1965. In Ticul, Maya houses in the center of houselots have been replaced by Spanish-style houses with flat, cement roofs that are adjacent to one another along the street. As families grow, households segment into smaller nuclear families, and if the children and their families remain on the land of their parents, the land might be subdivided to accommodate new houses on the property. Such adjoining houses may be spatially distinct, but the rear portion of the lot may have a common patio for all of the houses on the subdivided property.

When the potter’s craft becomes economically more important for a family, however, the use of space in the household changes. One of the most dramatic changes in potters’ households between 1965 and 1997 came with the kind and amount of space used for production. First, the use of general space for craft production has evolved into highly specialized space. In the late 1960s, almost all production occurred in multipurpose household space that was used for a variety of activities, not just ceramic production. By 1984, the space for production became more specialized with specific interior spaces set aside for the storage of raw materials, mixing, forming, and drying vessels. Except for painting and the sale of pottery, which has grown into space outside of households, most of the actual ceramic production has remained in the houselot and expanded with the construction of new facilities.³ The result was that the amount of space for ceramic production (its “spatial footprint”) has increased dramatically.

SOCIAL AND CULTURAL CHANGES, 1965–1997

Traditionally, Ticul pottery production was primarily oriented to the local Yucatec population with vessels fabricated for utilitarian, service, and ritual purposes. In 1965, much of the craft was oriented around the production of vessels for water transport and storage, which were sold throughout much of the northern Yucatán peninsula. Some potters also produced ceramic coin-banks that had been introduced in the 1940s.

Beginning in the late 1960s, a number of social changes had a significant impact on the craft. First, piped water was installed in the cities and towns on the peninsula. When this task was complete in the early 1970s, the demand for water-storage and water-carrying jars plummeted. About 1975, construction began on the tourist resort of Cancún, located 300 kilometers away, and eventually, the demand for pottery shifted. This demand was driven by two consumer values. The first consisted of the value of potted vegetation for interior spaces. Once it was discovered that Ticul vessels made desirable receptacles for large plants, consumers wanted these large vessels to display potted vegetation indoors and on patios and porches. The greatest demand for these vessels came from the Cancún hotels, which reportedly changed their pots every one to two years, but such vessels were also used by hotels in the capital city of Merida and by urban dwellers in these cities and elsewhere.

The second consumer value that emerged was the demand for small vessels painted with copies of ancient Maya designs. Such vessels became symbols of a visit to the land of the ancient Maya and were small enough to fit into a suitcase when tourists returned home. The demand for these vessels was facilitated by a government-sponsored workshop that was established in Ticul in the early 1970s to teach potters to make copies of prehispanic vessels with ancient Maya polychrome designs.

The demand created by these values was facilitated by an extensive transportation infrastructure. As recently as 1951, highways were limited and the extensive railroad network was the principal form of transportation throughout much of the northern Yucatán peninsula (Thompson 1958). By 1965, highways had expanded, but many roads to the interior were still unimproved and were nothing more than vehicle tracks that required four-wheel drive or a high wheel base to negotiate them. By the late 1960s, however, the quality of the roads was improving, and by the 1990s, most roads to the interior of the peninsula were asphalt. Meanwhile, work had begun on a limited access toll road eastward from a point sixty-eight kilometers east of Merida during the 1980s, and by 1994, the drive from Merida to Cancún required only a few hours. With the increase in highway infrastructure, access to the tourist markets on the east coast was easier and relatively rapid in comparison to travel there in the 1960s.

As a result of these changes, ceramic production in Ticul also changed. First, raw material procurement intensified and became professionalized with clay and temper mining carried out by full-time specialists (Arnold 2008:153–220). Consequently, the principal clay source that had been used for centuries became exhausted⁴ and clay procurement shifted to sources in the State of Campeche (see Arnold 2000, 2008:154–170; Arnold et al. 1999). In the late 1980s, a new source of temper was discovered, and by 1997, temper procurement had largely shifted to the new source (Arnold 2008:193–204).

As already implied, vessel shapes also changed between 1965 and 1997. After the early 1970s, production of traditional shapes used for utilitarian and service purposes were largely (but not completely) abandoned. Ritual vessels, however, such as food bowls, incense burners, candle holders, and whistles, continued to be produced on a seasonal basis for the Day of the Dead ceremonies. Some coin-banks continued to be made for local consumption. Nevertheless, production largely shifted to making plant pots and suitcase vessels. While some of these shapes were miniature copies of traditional⁵ vessels, most consisted of totally new shapes that included copies of ancient Maya vessels that were produced in prehispanic times.

Vessel decoration also changed. The traditional decoration, which consisted of red and tan firing slips, continued to be used on plant pots. Using oil-based paint to decorate banks was introduced in the 1940s and this practice continued through 1997. The most dramatic change in decoration, however, was the expansion of oil-based painting to include a variety of vessels and the use of polychrome designs copied from books of Maya art and archaeology.

The pattern of the consumption of Ticul pottery changed as well. Although the local Maya population still purchased pottery for use in Day of the Dead rituals, the consuming population changed from the Yucatec Maya to one that was largely tourist-related. Second, the physical location of the market also changed from the State of Yucatán to the resort of Cancún in the State of Quintana Roo on the east coast of the peninsula. Third, the size of the market and consuming population also changed. In the late 1960s, consumption was generally limited to the Yucatec Maya demand for water-carrying and water-storage jars, coin-banks, and vessels for seasonal rituals. By the 1990s, the demand for plant pots from hotels and urban dwellers and for vessels from tourists seemed insatiable and subject only to the vicissitudes of the ebb and flow of the tourist industry.

Finally, the patterns of distribution have changed. In the late 1960s, potters (or members of their families) sold their pottery directly to consumers. With the development of the tourist market at Cancún, however, most potters could not transport their pottery there using public transportation and were forced to sell their pottery to those who had vehicles. These potters thus lost control of the market and had to sell their wares to middlemen/brokers who transported it to Cancún or other consumer markets.

HISTORY OF RESEARCH

This chapter is a small part of a more comprehensive study of change in contemporary pottery production in Yucatán that occurred over ten visits from 1965 to 1997. These visits have varied in duration from one week to six months (Arnold 1967, 1971, 1987, 1989a, 1989b, 1991, 1997, 1998, 1999, 2000, 2008; Arnold and Bohor 1975, 1976, 1977; Arnold et al. 1999; Arnold, Neff, and Glascock 2000; Arnold and Nieves 1992; Ralph and Arnold 1988). The goals, data, and outcomes of each visit varied, but almost every visit included collecting data about potters, where they lived, where they produced pottery, and the persons who helped them.

One advantage of long-term research in the same location is the high validity that is possible to achieve with such a strategy. Because I have returned to the same community repeatedly over thirty-two years, I have become well-acquainted with potters, their relatives, their residence locations, and their production. Consequently, during each visit, I built upon the rapport and knowledge of previous visits and can assess, even with a brief visit, the veracity of informants’ statements and the validity of short-term and long-term changes. The result is that I can easily detect both deliberate and involuntary deception and verify and cross-check data with other informants. I can see changes from previous visits, and because I know most of the potters in the community, I can easily determine who is making pottery and where they are making it, often from observation alone. The following narrative briefly describes the history of this research, but it is abbreviated to focus on the collection of the data relevant for this chapter.

When I first went to Yucatán in 1965, I spent six months in Ticul and carried out a survey of potters. A sketch map was made locating each household, and data collection included the names of the potters present, how long each potter had been making pottery, the type of pottery made, and where it was sold. A return trip to Ticul in 1966 (Arnold 1967, 1971) included a survey of each potter’s household with its location on a map. Data from two households were obtained from others since visitation was not possible. A sample of 93 percent of the pottery-making households was obtained at that time.

The research visits of 1967 and 1968 focused on raw materials, but these visits also involved a compilation of a list of active potters. In 1968, the preparation of genealogical diagrams of potters’ families was begun. Two brief visits in 1970 involved eliciting a list of active potters. The lists from 1967, 1968, and 1970, however, are biased toward my informants’ families, their lineal relatives, and others known by them since they were not based on visits to households other than those of my principal informants.⁶ Subsequent visits and narratives of the history of each potter’s household have shown these lists to be accurate but incomplete, with approximately 10 to 15 percent of the potters in the community missing from these years’ lists.

I returned to Ticul in 1984 for six months and visited almost all potters’ households, elicited extensive genealogies of potters’ families from a few informants, and filled in the gaps in the diagrams with data from others (see Arnold 1987, 1989a). I returned again in 1988 and 1994. Many potters’ households were visited, but visits focused on the extended family with which I had most familiarity, even though a list of all active potters was elicited. In 1988, the location of each potter’s household was identified by brief inspection and each was placed on a map of the community produced in 1984. Again, narratives of household histories revealed that these lists included approximately 90 percent of the potters in the community.

In 1997, an attempt was made to systematically visit all of the potters’ households. A list of potters and other workers in each household was compiled. Those households that were not visited were identified as making pottery by brief inspection, but most of these had been visited before, and their owners and personnel were personally known to me and verified by my informants. In addition, questions (particularly about kin relationships) from previous visits were answered and verified.

THE ELECTRONIC DATABASES

The data from the ten research visits were assembled into four electronic databases, of which only two are relevant to this chapter.

The Production Unit Database

This database was compiled from data collected during each of my ten trips to Yucatán between 1965 and 1997. Each potter during this period has a record (N = 302) and each record contains a set of fields for each visit that includes the type of potter (whether owner of the production unit, worker, or relative of the owner), the production location, the type of production unit,⁷ the address of the unit, the type of pottery produced, and the other potters working in it. Three other fields recorded the presence or absence of a pottery store associated with the unit and, if present, its location and address. After fieldwork in 1997, fields such as “helpers” and “painters” were added for that year because some production units were becoming increasingly specialized with workers who were not potters.⁸ Additional summary fields were also added to provide numerical comparison between the different visits. In total, approximately eighty-eight fields exist for each record in the database.⁹ The production unit database makes it possible to track individual potters and production units through the thirty-two years of my research. These data, among other aspects of production, show changes in the composition, the number of potters, and the locations of the production units over time.

The Genealogical Database

Genealogies of the entire population of potters in Ticul constitute a second database. The primary purpose in developing this database was to graphically represent the relationships among potters across all of the generations and provide links among the seventy-two kin diagrams elicited in 1984. These links were complex and in some cases were not evident from the paper diagrams. So the data from these diagrams were entered into a commercial genealogy program (Parsons’ Family Origins program) and then updated with field notes from subsequent visits and microfilmed marriage records from the Ticul church (see below).

The resulting database consists of 1,024 individuals, 287 nuclear families (e.g., Mo, Fa, and children), and 659 events. The “events” group of fields consists of birth, marriage, and death dates; residence; place of origin; occupation; and other information that does not occur elsewhere in the database. The data for the “events” fields, however, are uneven because some individuals have many data entries in the “events” fields while others (such as those deceased for more than one generation) have no entries in these fields.

The electronic version of the genealogies provided a quick and easy way to answer specific genealogical questions from the production unit database. Each time a question arose about the relationship of one potter to another, this database would be searched for the answer. The electronic version of the genealogical data can thus be accessed and presented in a variety of ways including ancestors, descendants, family groups, and standard family trees.

During more than thirty years of research in Ticul, data gathered from participant-observation seemed to fit with my genealogical data. There were, however, some ambiguities that seemed “to slip through the cracks,” and I was anxious to resolve them by independent means. This task was accomplished by consulting church records for marriages from Ticul; these records were borrowed from the genealogical library in Salt Lake City and available through the Family History Center in Naperville, Illinois.

Table 6.1. Basic data on potters and production units from 1965 to 1997

Much to my delight, I discovered my genealogical data proved to be amazingly accurate, and the church records succeeded in resolving ambiguities that had puzzled me for almost fifteen years. The data also aided me in completing some missing details about potters, such as their birth and marriage dates.

RESULTS

Several demographic trends are evident among Ticul potters from 1965 to 1997 (Table 6.1). First, the number of potters has grown. Second, although the number of female potters has fluctuated, the percentage of female potters working in the craft in 1997 decreased to 46 percent of the 1965–1966 levels. Third, the number of production units has increased dramatically.

Although the mean number of potters per production unit has fluctuated over the years, this mean increased slightly between 1965–1966 and 1997 (see Table 6.1). When the median size of the production units is considered, the data indicate that the median size has also fluctuated but suggest that most production units are very small. This small size is further demonstrated by the distribution of the number of potters per production unit in 1965–1966, 1984, and 1997 (Figure 6.1). These data reveal that while a few production units have gotten larger since 1965, the size for most units has remained small with most units consisting of one to three potters throughout the period.

Since it was clear from the outset of this study in 1965 that the composition of production units consisted largely of kin who were household members, the first task was to identify the kin relationships of the potters in the production units. Potters in each unit were classified into a kin type according to their relationship to unit’s owner (who is usually male) and placed in the database.¹⁰ These types were then grouped into four larger, more encompassing categories commonly used in kinship analysis: affinal, lineal, collateral, and non-kin. “Nonkin” consisted of three types of persons. First, it included potters working at the tourist hotels at the archaeological site of Uxmal in “attached workshops” (see Brumfiel and Earle 1987) or “retainer workshops” (Costin 1991:9).¹¹ A second category included the wage laborers who worked as potters in production units in Ticul, and a third category consisted of potters that had ritual compadrazgo (or fictive kinship) relationships¹² with production-unit owners.

Figure 6.1. Number of potters per production unit in Ticul in 1965–1966, 1984, and 1997; these three periods of observation had the most complete survey data

The first way of describing the kin composition of the production units over time was to chart the numbers of kin types grouped by the major categories (lineal, collateral, affinal, and non-kin) according to each observation year (Figure 6.2). In this analysis, the numbers of different kin types were counted by observation year and then plotted. The result reveals some variability in the numbers of types, but over the years there has been an increase in the number of types of collateral kin working in the units.

When the number of potters in each of these categories was expressed as a percent of the total population and then plotted over the entire thirty-two-year period of the study, some dramatic changes were evident (Figure 6.3). Considerable fluctuation occurred in the frequency of all the categories, but there were two clear trends. First, there was an overall decreasing trend in the percent of affinal relatives in production units, and second, there was an increase in the percent of non-kin in these units.

Figure 6.2. Number of different kin categories of potters from 1965 to 1997; types are defined by their relationship to the production-unit owner

When the potters attached to the workshops at the tourist hotels at Uxmal (from 1965 to 1970) are removed from the “non-kin” category and the number of wage laborers are expressed as a percent of the total population, the percent of potters who were wage laborers increased dramatically between 1965 and 1997 (Table 6.2). This change was first evident in 1984, when owners of some production units were hiring many non-relatives who were potters or were training non-potters to make pottery as wage laborers. This change corresponds to the increased demand for plant pots for resort hotels and vessels for tourists.

Kin classification can also be described by generation (Figure 6.4). Using the kin relationships that identified potters by their relationship to the owner, these relationships were classified by generation. In these cases, the reference individual (“ego” in kinship reckoning) was the owner of the production unit and is in the “0 generation.” A father or mother thus would be in the “+1 generation” while a grandchild would be in the “–2 generation.” Similarly, a brother would be in the “0 generation” while a FaBr would be in the “+1 generation.” The plots of these classifications revealed no clear trends except perhaps the use of late adolescent and adult grandchildren (the “–2 generation”) in the production units since 1988.

Since the owners’ kin have constituted the majority of potters in production units since 1965, it is useful to ascertain which kin type has been the most common. When the most frequent types are expressed as a percent of the total population of potters and then graphed by year, the data reveal that son, wife, daughter, and mother (in rank order) have been the predominant types of potters during this thirty-two-year period (Figure 6.5). Again, considerable variation in these percentages has occurred over the years, but the clearest trend has been a decrease in the percentage of wives working as potters.¹³

Figure 6.3. Percent of total potters by kin type; kin types are identified by their relationship to the production-unit owner and then grouped into the categories used here

Table 6.2. The number and percent of non-relative wage laborers in production units in Ticul, 1965–1997

So far, this discussion provides a measure of the aggregate of changes in production units between 1965 and 1997. It does not, however, provide a picture of continuity (or lack of such continuity) of individual production units through time. To assess this continuity, the evolution of each unit was traced through time and its trajectory was placed into one of four categories (Figure 6.6). The first category (“same”) consisted of those units whose production locus did not change at all. The second category (“segmented”) included those units that had internally segmented into differentiated nuclear families but who lived in the same houselot. All such nuclear families consisted of children of a potter in the previous generation. The third category (“fissioned”) were households that had been previously part of another but had fissioned because the parents’ unit was too small to accommodate the space needs for production by an additional family. A fourth category (“continuing”) included those units that had been making pottery during at least one previous survey and were missed because of methodological bias or had temporarily stopped making pottery at one time in the past and had begun making pottery again. This category also included individuals who were making pottery during a previous survey but had moved because a financial crisis had caused them to sell their household land.

Figure 6.4. Frequency of potters classified by generation of the kin type that defines their relationship to the owners of production units from 1965 to 1997

Figure 6.5. Most common kin of potters working in production units from 1965 to 1997; types are defined by their relationship to the production-unit owner

Figure 6.6. Changes in the locations in production units in Ticul from 1965–1966 to 1997

When all production units were tracked, compared, counted, and graphed, several patterns emerged. First, comparing the 1984 and 1997 surveys with 1965–1966 data, it is clear that great household continuity occurs throughout the thirty-two-year period (see Figure 6.6). In 1984, 35.5 percent of the households were the same as those in 1965–1966, and an additional 29 percent were units derived from existing production units through segmentation and fissioning. Similarly, 26.7 percent of the production units in 1997 were the same as those in 1965, and another 33.3 percent were derived from those units through segmentation and fissioning.

The development of “new” units is also evident (see Figure 6.6). In 1984, 36 percent of the production units were “new” units without any continuity from units in 1965–1966. By 1997, the number of totally new production units had increased in absolute numbers and in the percentage (40 percent) of the total. These “new” units are the result of entrepreneurs who had established production facilities to take advantage of the tourist market. They were usually painters (not potters) and hired potters as wage laborers (see Table 6.1).

Figure 6.7. Changes in the location of production units from 1970 to 1997 based on movement from their location in the previous survey; the 1970 data are compared with the 1965–1966 data

When one looks at the changes in location relative to each successive period of observation and the graph of each “change” category compared to the previous visit, a slightly different picture emerges (Figure 6.7). There was a dramatic jump in new units in 1984, but there was still a strong continuity of production units from the previous survey and a diminishing number of “new” production units since 1984. This pattern again marks the change in the orientation to the tourist market that began in the late 1970s.

Finally, another way to look at the continuity of the population of potters is to look at the overall kin-relatedness of production units through time (Figure 6.8). “Kin-relatedness” was defined as the kinship relationship of two production-unit owners traceable across no more than one marriage.¹⁴ By “traceable,” I mean that the kin relationships had to be traceable in my electronic genealogical database or paper kin diagrams. If the owners of two production units met these criteria, they were considered to be “kin-related.” The number of units that met these criteria was then counted, grouped, and graphed by year. For simplicity, only the data from the first two field seasons in 1965–1966 and the last field visit (1997) are presented here (see Figure 6.8) because they show the amount of kinrelatedness among production units at the beginning and end of this study. These data show a very strong kin continuity of the production units in Ticul, both across the entire population of potters and through the entire period of this study.

Figure 6.8. The amount of kin relatedness among production units in Ticul in 1965–1966 and 1997, based on traceable kin relationships

SUMMARY AND CONCLUSIONS

In conclusion, what does study of the production units in Ticul across the thirty-two years of research tell us? First, the strong continuity in the location of production units through time is the result of the use of the nuclear family (So, Da, Mo, Wi) as production personnel and indicates the co-occurrence of most (but not all) production units with households. The small size of most units is also consistent with this finding. Second, the stability and change in the location of most units are the results of processes of household continuity resulting from patrilineal inheritance of household land, segmentation, and fissioning. This finding is consistent with a previous study that used data collected in 1984 (Arnold 1989a).

It is now time to revisit the notion of “household” that was temporarily replaced at the beginning of this chapter in favor of a more neutral term “production unit.” By examining the results of this study in light of the term “household,” it is clear that while some production units are getting larger and adding wage laborers as workers, most, if not all, production units are still household-based, use household personnel, and remain or change locations based on household processes such as inheritance of household land, segmentation, or fissioning. When fissioning occurs, the residence location is still relatively close to the source of the fissioning household as the result of patrilineal inheritance of household land and its subdivision. By 1997, residence was virilocal and was not simply the result of post-nuptial residence “rules” but rather a result of the desire of parents to have their children live relatively near them. To ensure this proximity, a father would buy land nearby for his sons (most frequently) and his daughters (less frequently).

Probably the most remarkable aspect of these data is that in spite of massive social and cultural changes and the changes in raw materials, procurement practices, vessel shapes, decoration, demand, and market that has occurred since the late 1960s, pottery production in Ticul is still household-based and is still largely perpetuated through household processes of procreation, socialization of the young, land inheritance, post-nuptial residence, and other processes that affect household composition. Since the household and its continuity is critical for the perpetuation of the society through procreation and socialization of new members of the society, it is not surprising that given a demand for crafts produced within the household, the technology of their production is perpetuated by the same processes as those that perpetuate the household. The Ticul data collected over the thirty-two-year period of this study suggest that the production population is probably the most conservative aspect of ceramic production in Ticul. This continuity persists precisely because production is household-based and tied into household processes rather than macro-level processes outside of the household.

Household continuity in production thus transcends the life of individuals. By 1997, 32 percent (27 of 85) of the potters living in 1965–1966 had died, and at least 18 percent (16 of 85) more had left the craft for one reason or another. Nevertheless, 60 percent of households in 1997 had continued from 1965–1966 because of the inheritance of household land, internal segmentation of families, and fissioning. The continuity for ceramic production thus appears to be even more conservative than population continuity, even in light of massive social, technological, and cultural changes.

What has occurred in Ticul during this thirty-two-year snapshot is the beginning of the evolution of a more complex organization of ceramic production. In 1965, the craft was tied to generalized household space (Figure 6.9), but by 1997, production space had become more specialized but was still tied to the household (Figure 6.10). More important, the evolution of the changes in personnel from households to larger workshops has seen the increase in the use of wage laborers. During the period of this study, there was much variability in size, composition, and location of production units. Through it all, however, household organization has not disappeared but continues to form the basis of the composition of production units and is responsible for their persistence because households provide the social context in which the craft is learned and perpetuated. Even though the amount of production space has increased and the craft has become increasingly specialized, production units that no longer look like households are still household-based because their social organization is still largely based on the kin relationships in the household and change by the same processes that maintain and transform households through time. Household pottery production thus has great potential to permit an increase in the amount of production space and in the amount of task segmentation without fundamental changes to household social organization.

Figure 6.9. The back of Lorenzo Pech’s house and his production unit in 1965, looking south through the house to the street; in this image, his father, Lucas Pech, is making water-storage vessels in a thatched structure at the rear of his house

Figure 6.10. Lorenzo’s house (in the distance) and the workshop behind it in 1997, showing the space used for the quantities of drying and fired pottery; this photo was taken in a position identical to that of Figure 6.9 but twenty meters further north toward the rear of the workshop

How do these data square with Costin’s description of “scale” and its usefulness as a parameter for defining and organizing craft specialization? In a phrase, rather well. Costin argues that the two extremes of production are small, individual or family-based production units, while at the other extreme are “wage labor forces of the industrial west, where employment is contractual in nature and based on skill and variability” (Costin 1991:15). While this dichotomy is a truism in many respects, it is also true that when family-based production units meet the industrialized West, as seen in Ticul, a mix of largely small household-based production units exist alongside much larger highly specialized units. Both are largely household-based. In responding to Torrance (1986), that there is a necessary link between the nature of the demand and the scale of production, Costin argues that both independent and attached facilities can be large or small (Costin 1991). And so it is in Ticul. With the high demand of the tourist-related market, there is a mix of small and large production units with a predominance of small units. What is important about the Ticul data, however, is that large units are still family- and household-based, but these large units also include wage laborers who often are members of the extended family, individuals related to production-unit owners by fictive kin relations, as well as non-relatives. These changes correspond precisely to the observation that Costin made from Kleinberg’s (1979) work with village pottery production in Japan: “As production units grow in size, new labor is recruited first among distant, fictive, and adoptive kin. With further growth, nonrelated individuals are added to the workforce” (Costin 1991:15). In Ticul, the composition of production units has reflected most of these categories between 1965 and 1997, but over time, the use of more distant relatives and non-relatives as supplements to the nuclear family has increased. The addition of personnel is much more complex than this, however, because over time, the types of collateral relatives and the percentage of the population of non-relatives used in production units have increased, but the percentages of affinal kin, women, and wives as potters have decreased.

While Costin’s notion of “scale” is an important descriptor of the organization of craft production, the Ticul data suggest that its usefulness as an index of the evolution of craft specialization is limited because of the great variability of the size and composition of the production units even within a short span of thirty-two years. Not only is there variability in the type of craft production units in Ticul, but great variability also exists in their scale with the principles of kinship and household recruitment still being the predominant means of labor recruitment. Does this mean that pottery production in Ticul has not evolved and become more specialized? Of course not. Rather, a better indicator of the evolution of Ticul pottery production is the increasing use of specialized household space for production and the resulting increase in the size of its total spatial footprint in each production unit (compare Figures 6.9 and 6.10).

The production of Lorenzo Pech provides a dramatic example of the change in production space. In 1966, his production space consisted of a house and an attached porch (see Figure 6.9). Thirty-one years later, Lorenzo expanded his house, built a large production facility behind it, and established a second production unit along the highway in another part of town (see Figure 6.10). Although his production space increased more than any other production unit in Ticul, such changes were reflected in a lesser degree in all of the other production units in the community. Production space in all units changed from generalized space to specialized space, and the amount of specialized space increased.

The evolution of specialized space from generalized space for ceramic production and the increase in the amount of specialized space are changes most visible to archaeologists. These changes reflect an increased amount of pottery produced and increasing task segmentation. They also co-occur with the development of increased homogeneity of ceramic vessels, but for other reasons (Arnold 2008:265–272). Elaboration of these changes, however, is the subject of a future monograph currently in preparation.

Finally, one of the most interesting concluding observations about these data concerns the adaptive nature of households. The Ticul data show that households are dynamic and changing entities that are elastic in the way in which they can organize production socially and serve as its spatial locus in light of changing social and economic conditions. Households are not just an abstract type of unit on a unilinear scale of modes of production organization but rather are changing, dynamic entities capable of adapting to increased amounts of production with a larger production scale and intensity. In this sense, one can see how the use of Costin’s notion of “production scale” for describing production organization is a better tool to describe evolutionary changes in production than the use of abstract types. Her descriptors also provide more insight into the evolution of households that is not possible with the use of finely graded abstract social types.

With Ticul pottery production, as with the households of ancient Ejutla of the Oaxaca Valley (Feinman 1999), production intensity should be decoupled from production scale just as Feinman suggested. The household, as Feinman (1999) has found archaeologically, is capable of far more production intensity than was previously thought possible. Households thus are not just elastic in scale but are also capable of producing a greatly increased amount of pottery without being coerced or organized by social and economic entities outside of the household.

Acknowledgments. This paper was originally presented at the 24th Annual Meeting of the Midwest Conference of Mesoamerican Archaeology and Ethno-history, University of Iowa, Iowa City, IA, March 24, 2001, and at the Annual Meeting of the Society for American Archaeology, New Orleans, LA, April 21, 2001. Fieldwork in 1997 and release time to prepare the data for publication were supported by the Wenner-Gren Foundation for Anthropological Research (grant no. 6163), the National Endowment for the Humanities (grant no. RK 20191-95), and the Wheaton College Alumni Association. Other research in Yucatán was funded by the Wheaton College Alumni Association, an American Republics Research Grant (Fulbright Program), the Wheaton College Norris Aldeen Fund, a Vice President Research Initiation Grant (Pennsylvania State University), the University of Illinois Research Board (with B. F. Bohor), the Ford Foundation Center for Comparative Studies (University of Illinois), the Department of Anthropology (University of Illinois), and the Interuniversity Project for Behavioral Science Training in Yucatán. The tables and bar charts were prepared by Heidi Biddle and Christy Reed.

NOTES

1. This problem is one reason why I tried, as much as possible, to explicitly avoid typologies in Ceramic Theory and Cultural Process (Arnold 1985).

2. Some trees that I have observed in potters’ households are avocado, orange, perennial chili pepper, and the tree gourd (Cresentia cajeta), used for making gourd bowls and scrapers/shaping tools for forming pottery.

3. In one case, production has expanded into a location outside the household, but generally most such activities have remained within houselots, although the households themselves have changed greatly.

4. A visit to the interior of the clay mine at Hacienda Yo’ K’at in 1968 revealed that this clay source was used in the Terminal Classic period (Arnold and Bohor 1977). Since cross-cultural models of resource distances indicate that clay sources are seldom more than seven kilometers from production centers (Arnold 1981, 1985), it appears likely that pottery production took place in, or near, Ticul during the Terminal Classic period. The most probable location for this production is the archaeological site of San Francisco just north of Ticul.

5. “Traditional” vessels are those vessels described by Thompson (1958) that were used for cooking, carrying and storing water, soaking maize kernels, and serving water and food.

6. The methodological bias toward lineal relatives in these lists can be seen in Figures6.2 and 6.3. In light of surveys during other years, it is clear that those potters who are affinal and collateral relatives of production-unit owners did not occur in the lists of those years.

7. This category was initially “household” or “workshop,” but this distinction turned out to be useless. See the previous discussion in the text for an explanation.

8. This separation into “potters” and “painters” had already occurred in 1984, but since I was not interested in painting at that time, I did not collect systematic and quantitative data about the painters in the production units.

9. Some fields have few data points (such as those for the 1967 and 1968 visits), whereas other fields (such as those for the 1965, 1966, 1984, 1988, 1994, and 1997 visits) have many data points. Since the 1965 and 1966 visits were only six months apart and the data were complementary, the data from these visits were combined into a new set of fields called the “1965–1966” fields.

10. Up until recently, only males could legally inherit land upon the death of a father, unless the land was legally deeded to a daughter before her father’s death.

11. From 1965 to 1968, potters worked at only one Uxmal hotel, but a second hotel added potters in 1970. At both hotels, the shapes, decoration, and the number of vessels produced were controlled by the management of the hotel. Here, “attached workshops” means “attached specialization” (Brumfiel and Earle 1987), in contrast to the “independent specialization” that was characteristic of Ticul production at this time. Although the term “attached specialists” refers to ancient production, the category does fit the production in these hotels, in contrast to “independent specialization” in Ticul itself.

12. Co-parent (compadres), godparent, and godchild relationships.

13. While most of the production-unit owners were males, some units were owned and operated by females. In most cases, these women owners were unmarried, widowed, or divorced.

14. This definition does not include relationships acquired through serial marital unions. Rather, it simply indicates that relationships that could be traced through a marriage of affinal relatives were not considered to be related by kinship. So, affinal relatives of a production-unit owner were considered to be kin-related and a potter married to an affinal relative was considered to be kin-related. A traceable relationship through the marriage of an affinal relative other than one’s spouse, however, was not considered to be kin-related.

REFERENCES CITED