Ceramic Theory and Cultural Process after 25 Years

NVCIZLJHGKFDYAOUTRE Classics Review Ceramic heory and Cultural Process after 25 Years Dean E. Arnold A number of years ago I was talking with colleagues at the Field Museum in Chicago, and one of them introduced me to one of the retired curators by saying: “Dean has written a book about ceramic ecology.” “hat’s a contradiction isn’t it?” replied the former curator. “On the one hand, you refer to static objects, and on the other, you are talking about a dynamic phenomenon.” “hat’s the point,” replied my colleague. “Dean is trying to understand the dynamic relationships of ceramics to culture and the environment.” “Precisely,” I responded. “he book concerns the relationships of ceramics, society and the environment. It is not about the ceramics themselves.” his anecdote expresses both the core of why I wrote the book, and a one-sentence summary of its contents. Ceramic heory and Cultural Process was an attempt to provide cross-cultural evidence for relationships between ceramic production, on the one hand, and the environment and the rest of culture, on the other. I was particularly concerned about articulating the relationship of ceramic production to the environment, not as a deterministic, uni-causal explanation, but as an attempt to restore a neglected perspective to a craft that has signiicant environmental links that transcend the obvious: “Raw materials are necessary to make pots.” Interpretations of ancient artifacts are not inherent in the objects themselves, but rather are based upon the interpreter’s understanding of the relationships between artifacts and behavior. hese relationships may be implicit in the archaeologist’s mind and come from archaeological tradition and experience, but they may also have their source in ethnoarchaeology, ethnography, and archaeological theory. In any event, interpretation is founded upon the implicit or explicit relationships between artifacts and culture whether archaeologists Ethnoarchaeology, Vol. 3, No. 1 (Spring 2011), pp. 63–98. Copyright © 2011 Left Coast Press, Inc. All rights reserved. 63 64 Classics Review realize it or not. Ceramic heory and Cultural Process was written to detail some of those relationships cross-culturally. I am honored to have the editors of Ethnoarchaeology select Ceramic heory and Cultural Process as a classic in ethnoarchaeology. It is a privilege to be asked to write some relections on the book, how it came to be, and its inluence. I am very grateful for the opportunity. I have also written a chapter in the Anna Shepard volume that relects on diferent aspects of the book (Arnold 1991). hese relections, however, will explain why I wrote the book, a brief ethnography of its printing, some proxy measures of its inluence, and then will detail some of the changes to my thinking about the book’s contents in the twenty-ive years since it was published. Beginnings Ceramic heory and Cultural Process was born out of frustration, and the ideas largely came out of my own experience. I had studied pottery making in Mexico, Peru, and Guatemala, but I found that my research in one area had not helped me understand pottery production very much in another. I started working in Ticul, Yucatan, Mexico in 1965 and originally focused on the Maya categories of iring. I eventually learned how to ire in the Maya way, but in the process, I learned much about the indigenous knowledge of clays, tempers, irewood, and the sequences of vessel forming and iring. To obtain this information, I originally utilized an elicitation technique known as “ethnoscience” (Black 1963; Black and Metzger 1965; Frake 1964; Metzger and Williams 1963a, 1963b, 1966). Now considered to be a part of cognitive anthropology (D’Andrade 1995), this technique uses a ield language to discover its cultural categories in order to reveal their meaning and structure. Although this approach was very productive, I didn’t realize how culturally relative it was until much later. Although I had collected a sizable amount of data about the indigenous knowledge of potters in Yucatan during 1965 and 1966, most of it was in the form of texts and question and response frames consistent with the ethnoscience technique. he late Duane Metzger, who originally took me (and three other students) into the ield, largely abandoned us in Ticul. his seeming neglect, whether deliberate or not, frustrated and angered us, but in retrospect, it allowed me the freedom to be independent and go in any direction that I wanted. Metzger suggested that I study pottery iring, but apart from that advice, I was on my own. I was so afraid of failure that I collected massive amounts of data. Classics Review 65 During the course of my research, I discovered that Ticul potters recognized a category of their raw materials that corresponded to the clay mineral palygorskite (then called attapulgite), one of the components of Maya Blue. At that time, no one knew that palygorskite existed in the Maya area, much less Yucatan. I produced a master’s thesis demonstrating this correspondence between the cultural category and the mineralogical category (Arnold 1967a), and the anthropology department at Illinois published it as a research report (Arnold 1967b). here was, however, no Mayanist there to be my advisor and supervise my dissertation. Duane Metzger, under whom I had studied ethnoscience and who taken me into the ield, had departed Illinois in 1965 for a position at the University of California at Irvine, and the late Don Lathrap became my advisor. Don was a South American archaeologist and soon insisted that I learn to do archaeology. So, I enrolled in Art Rohn’s ield school at Yellow Jacket, Colorado in the summer of 1966. Another part of this immersion in archeology was taking Don’s South American archaeology course. I never went into the ield with Don, but he encouraged me immeasurably to continue to do ethnography that was relevant to archaeology even as I was being immersed in archaeology. Lathrap told me to go to Peru and work with Tom Zuidema, an Andean ethnohistorian, who had just been hired by the department. So, in February of 1967, I was on my way to Peru to study pottery making in the Ayacucho Valley under the supervision of Zuidema. Originally, I wanted to examine paste, temper, and clay variability between communities of potters, but for various reasons (see Arnold 1993:xxii–xxiii), I ended up in the village of Quinua, just a few kilometers up slope from the ancient city of Wari, the capital of a pre-Inca empire that ruled much of Peru between 600–1000 A. D. (Williams 2001). he close proximity of Wari and Quinua suggested that the modern ceramic production might have roots in its ancient neighbor. So, instead of following through with my original research design, I ended up studying modern pottery production in Quinua. Unfortunately, I discovered that none of the potters in the community were making pottery. I had arrived in the midst of the rainy season (mid-February). Travel was diicult because of the rain and mud. Rain fell almost every day. here was little sunshine; fog and mist often blanketed the area during the day. Another anthropology student, William P. Mitchell (e. g., 1976, 1990, 2006) was also working in Quinua at the time, and he ofered me a corner of his room for my sleeping bag and air mattress. Fortunately, Mitchell’s host and principal informant was a potter. Unlike most potters, who were also agriculturalists and made pots only in the dry season, Mitchell’s informant worked in the 66 Classics Review government-run artisan school for most of the year, but during the summer vacation (the rainy season of January-March), he made some pottery in his house in order to make extra money. He could not ire it, however, because the dampness kept it from drying completely. Visits to households of other potters did not yield much data until late March and early April and even then, few potters were practicing their craft. Potters were also peasant farmers and needed to work in their ields during the rainy season. hey did not begin making pottery until the harvest was completed in late June when I was scheduled to leave the ield. he early months of my research in Quinua thus were a time of great frustration. Few observations of pottery making were made. For years afterward I felt that much of that ieldwork was a failure because of my inability to gather abundant data on pottery manufacture (Arnold 1993:xxii). After I returned to the University of Illinois in Urbana in mid 1967, I had to come up with a dissertation, and I was in a bind. I had plenty of data for a unique dissertation about the indigenous knowledge of the Maya potter, but I had to write one about my rather failed (or so I assumed) attempt to study pottery making in Quinua. My observations of pottery production in Peru were very limited, but I collected some data about the sources of raw materials that I visited, and about iring when the weather improved. I also collected some samples of raw materials for analysis. Because the data on ceramic production was so scarce, I focused on the designs on vessels that I saw in the weekly market in Quinua, the market in Ayacucho, and among the few potters that were working at the time. None of these data were substantial enough for a dissertation except for the design data. My background in linguistics suggested that the sequence of pottery production was linear like language production, and thus could be described using a linguistic model. his approach followed directly from my work with a Tzeltal informant in Urbana in the fall of 1964. Using the ethnoscience technique, I elicited a folk taxonomy and componential analysis of some contemporary Chiapas pottery and discovered that the Tzeltal folk categorization of vessel shapes and their signiicant components were hierarchically organized. Would this model work with my Quinua data? I had noticed that the design structure on Quinua pottery was consistent across observations (with a few exceptions) with diferent amounts of variability in diferent zones of decoration. Furthermore, the location of these zones was relatively ixed across vessels of the same shape. his design structure could be expressed in several ways, but I chose to describe it as a series of decision trees that relected the sequential choices that Classics Review 67 potters made. hese decisions were linked together hierarchically with the most general choices of design occurring at the top of the diagram, and the more speciic decisions at the lower levels of the tree. Further, these decision trees appeared to be implicit in the way potters painted their vessels. Such an approach to design was not really an etic approach using cross-cultural units of analysis like symmetry analysis (Shepard 1948, 1956), but it was speciic and unique to the community of Quinua (see Arnold 1993:140–196). he designs were not strictly “emic” either because I did not have linguistic terms for the design structure or the designs. Nevertheless, I believed that the design zones, their contents and the design decisions were a series of culturally-relative, but behavioral “emes”. So, I systematically presented my data as decision trees and entitled my dissertation as “he Emics of Pottery Design in Quinua, Peru” (Arnold 1970), publishing the trees in summarized form 14 years later (Arnold 1984). In retrospect, this presentation was an early approach to what has come to be known as “technological choice” (Lemmonier 1978, 1986, 1992, 1993), but in the late 1960s, there was, to my knowledge, no literature on that subject. As I came to realize later, this “choice” or “decision-based” approach to design, although an important descriptive methodology, was too culturally-relative to be used to develop a cross-cultural theory of ceramics. In September of 1969, I started teaching at Penn State as an Assistant Professor. During that irst year, I became friends with a colleague at Penn State’s nuclear reactor (the late William Jester), and he encouraged me to cooperate on a project. At the time, the late Bill Sanders and Joe Michels were in charge of an archaeological survey and excavation in and around the Maya site of Kaminaljuyu in the Valley of Guatemala. Since I had worked with raw materials from Yucatan and Peru, I decided to do an ethnographic project in the Valley of Guatemala testing the assumptions of neutron activation analysis using contemporary pottery materials (Arnold et al 1991). Sanders and Michels liked my idea, and helped by providing project infrastructure in the ield. I received funding from the Penn State College of Liberal Arts in the form of a junior faculty seed grant, and in the summer of 1970, I was in Guatemala, observing pottery making and collecting samples. I focused intensively on four communities (see Arnold 1978a, 1978b) in the Valley of Guatemala (Chinautla, Sacojito, Durazno, and Mixco), and did cursory observations in two others (Sacoj and La Cienaga). I spent considerable time talking to potters, collecting raw material samples, mapping house locations (in 68 Classics Review Sacojito), and observing production when it occurred, which, it happened, was during the rainy season (Arnold 1978a). I also collected some vessels and pottery making tools for the forthcoming museum in Penn State’s Anthropology Department. Again, my knowledge of the craft from Yucatan and Peru did not seem to contribute much to my understanding of pottery making in Guatemala. It was true that my research in each area focused on diferent problems, but except for the broadest generalizations, there seemed to be no commonalities among the six Guatemala communities, and between these communities and those that I studied in Mexico and Peru. his problem vexed me. After all, common processes were responsible for making pottery everywhere, but this similarity seemed trivial and supericial. I had studied Anna O. Shepard’s classic work (1956) on ceramic technology, but except for the common processes in the production of ceramics, there seemed to be no commonalities. Why was this? My study of pottery making communities in Yucatan, Peru, and Guatemala thus disillusioned me because the data from them were too culturally speciic. Cognitive anthropology was extremely fascinating, but from an archeological perspective and from the perspective of theory development, it seemed largely unsatisfying and irrelevant. Indeed, it was diicult to relate my data to pottery production elsewhere because of its culturally particularistic nature. In particular, I had used decision trees for my Peru data, and for some of my Guatemala data, but generalizations about production for cross-cultural comparison seemed impossible. As I came to realize later, some approaches to human culture are culturallyrelative by nature and some utilize cross-cultural units of analysis. Ethnoscience and a structural (decision trees) approach to ceramic design (and technological choice) are methods that can be used cross-culturally, but they do not provide cross-cultural data of the kind that Kenneth Pike (1967) and Marvin Harris (1968) called etic data. Although largely clear in retrospect, it was confusing and diicult at the time, and I refrained from publishing much of my data because the theories and paradigms weren’t there to describe and interpret my results in a way that would beneit archaeology. I did, however, manage to write an article showing the contrast of the emic and etic approaches that was based upon my master’s thesis (Arnold 1971). Years after my ield work in the Ayacucho Valley, I began to evaluate my experiences more objectively. I was preparing some of Quinua material for publi- Classics Review 69 cation (Arnold 1972a, 1972b) and began to relect on my data and the frustrating experiences of trying to study pottery making during the rainy season. With the experiences of observing the craft in Yucatan, Peru, and the Valley of Guatemala, I thought that I had enough data to make some generalizations. I began with a paper that compared pottery production among three linguistic islands of a Maya language called Pokomam. One island of communities (Chinautla, Sacojito, Durazno, Sacoj) made much pottery, another made no pottery (Palín), and a third (Mixco) made a little. All had a similar linguistic and culture historical origin, but why was there a diference in the amount of ceramic production among the communities? Aren’t ceramics supposed to relect culture history? hrough the encouragement of Bill Sanders and examination of the soil maps, I saw that the agricultural potential of each community varied inversely with the amount of pottery production. his led me to write a paper for the 1971 American Anthropological Association meetings on this subject, and it was published in 1978 (Arnold 1978a). Similar insights occurred as I moved beyond my own feelings of failure about my Quinua work, and contemplated the meaning of the suspension of making pottery in the rainy season there. Much to my surprise, I discovered that seasonal production was itself a very important observation. My experience graphically demonstrated to me that weather and climate profoundly afected ceramic production depending on the amount of rainfall and the days with cloudiness, rain, fog, and high humidity. Further, the scheduling of subsistence activities (like agriculture) could preclude ceramic production among potters who were also farmers. Weather and climate patterns were thus important constraining factors which prevented the development of full-time ceramic specialization in Quinua and were probably important variables in inhibiting full-time production in other areas of the Andes with a heavy rainy season. hese insights changed my view of ceramic production as I saw that what was important was not just production itself, but rather how it was tied to weather, climate, and the seasonality of agriculture. Seeing these relationships and the events that stimulated them were pivotal in my thinking at the time and led to writing an article that appeared in Current Anthropology on the role of environmental factors in ceramic production (Arnold 1975a). Preparing this article and responding to the comments were foundational for development of the ideas and thinking that led to writing Ceramic heory and Cultural Process (Arnold 1985). he published comments that followed my Current Anthropology (Arnold 70 Classics Review 1975a) article generally frustrated me, and I realized that archaeologists knew little about ceramic production and its cultural and environmental context. I wrote two responses to the comments, one that was published in the same issue as the article (Arnold 1975a), and the other in the following issue (Arnold 1975b). In writing the second response, I realized that archaeologists needed a summary statement of the relationship of pottery production to the environment. hey didn’t seem to understand my point in the article. At the time, Andean archaeologists traditionally believed that ceramic style was the primary interpretive tool in understanding the Andean past, and the environment had no role in ceramic production except providing the raw materials. (Ceramic raw materials and archaeological pastes were not investigated in the Andes until the last 20–30 years.) his assumption was borne out by an anonymous referee who reviewed the paper for Science before it was submitted to Current Anthropology. he referee latly said that the role of the environment in the Andean ceramic production already had proven to be unimportant, if not insigniicant. Ethnographers and artists who wrote about contemporary pottery production in the Andes further bolstered this assumption, often unknowingly. Most ethnographic and archaeological work there occurred in the dry season when North Americans students and scholars have their summer vacation, and travel in the Andes is easy. When I was a Fulbright Lecturer at the University of Cuzco in 1972–1973, I learned that ceramic production in that region also occurred during these same months, and potters marketed their pots in the yearly fairs during the months of August to October when they exchanged their vessels for their contents of maize or potatoes. I gained other insights about the relationship of pottery production and weather when I was in Cuzco. First, I kept a log of the weather and rainfall to better understand its role in pottery production. Second, I made a few observations of a brick production area during my daily treks to an Inca site to the south of Cuzco, where I assisted local archaeologists of the National Institute of Culture in their excavation. Production was situated on a large deposit of clay, and workers dug clay and formed bricks and tiles and dried them throughout the dry season. Once the rainy season arrived, however, the production stopped. Why hadn’t anyone noticed the role of the environment in Andean pottery production before? Ethnographers and archaeologists described ceramic production when it was occurring, or simply noted that it was done in the dry season when potters could market their products. Consequently, studying the Classics Review 71 craft when it was not occurring seemed counter-intuitive, if not naive (as I was). Further, archaeologists generally do not have experience with potters and pottery making contexts, a point made many years ago by Anna O. Shepard. his problem concerned me so deeply that I realized that archaeologists needed more than one article in Current Anthropology to convince them of the importance of the environment in pottery making communities. As a consequence, my second response to the comments in Current Anthropology (Arnold 1975b) became the précis for Ceramic heory and Cultural Process. About that time, my friend from graduate school, Scott Raymond, was teaching in the Department of Archaeology at the University of Calgary and suggested that I submit an abstract for their annual Chac Mool conference. I decided to expand my brief response in Current Anthropology into a paper, and presented my paper at the conference in 1974. he paper was published in the conference volume (Arnold 1976) and became the foundation for a manuscript that ultimately emerged as Ceramic heory and Cultural Process. Scott heard my paper in Calgary and suggested that I might want to consider a cybernetic approach using feedback as an organizing feature. So, I did some reading on cybernetics, feedback, and Norbert Weiner, and incorporated those perspectives into the manuscript. hose suggestions were the basis for making feedback the organizing principle of Ceramic heory and Cultural Process. Once I learned about the nature of feedback, I remembered my experience of iring pottery in Ticul. When I irst studied the iring process in 1965, I learned all of the Maya categories for the parts of the kiln, the stages of the iring process, and the kinds of irewood used to achieve the desired afects in each stage. But, then I asked myself the question: Did the emic data of my informant’s native categories (his “ideal behavior”) match up with the way that he actually ired pottery? I had already observed many iring episodes, but I wanted to see if the indigenous knowledge that I had elicited and learned was suicient to actually ire a kiln-load of pots. I asked my informant if I could ire his pots, and after some initial hesitation, he consented. I eventually ired ive kiln loads of pottery by myself, and learned that knowledge of the categories that I had elicited was insuicient to ire efectively. Visual feedback from actually bodily engaging the iring process, I learned, was critical for the potter’s success. Feedback afects the choices that the potter makes about his iring behavior by indicating when to add irewood and, according to the choices that he has already made, inish the process. After the potter starts 72 Classics Review the ire, he waits until it burns low before he adds more irewood, augmenting the amount of fuel with each addition. He follows that pattern until the pottery becomes totally black, and the fuel placed away from the ire, but below the pottery, spontaneously bursts into lame. hat information from visual feedback triggers another set of behaviors in which the potter adds additional irewood more quickly to burn of the soot from the blackened vessels. his behavioral pattern then continues, and after a few such additions of irewood, the potter looks into the kiln. If the pottery glows red, iring is inished. If it does not, he continues to add irewood, letting the ire die down each time before he checks the inside of the kiln. In brief, understanding the cognitive categories are necessary to ire pottery, but the actual embodied practice of iring itself involved much more than just the cognitive knowledge of it. Feedback from the iring process was critical to the potter, and he used the information to make decisions about his iring behavior. Plastic Pots I began Ceramic heory and Cultural Process with the lament that archeologists viewed ceramics as totally plastic with only culture imprinting itself on the clay product. In other words, it appeared that the environment and the media itself had limited efect, if any, on ceramic production and its evolution. Another way of expressing this assumption was that the organization of the craft responded to the organization of the larger society with no inluence from the environment, or from the actual production process itself. At irst, my relections upon my experience in studying pottery making in Mexico, Peru, and Guatemala seemed to conirm this assumption because of the culturally relative nature of my data. Upon deeper relection, however, I found that although pottery was indeed plastic and did relect cultural patterns, there were physical and climatic constraints to pottery production depending on the raw material, the weather, and the availability of covered space. Because of my unique experience in studying pottery making in three areas of the Americas, Ceramic heory and Cultural Process was written to counter the assumption that pottery, because it is a plastic medium, uniquely relects the culture of the potter with no intervening environmental variables, or variables from the behavioral sequence of the process. I was also frustrated with the literature on ceramic production that I had read. It seemed that archaeological interpretations of pottery bore little or no Classics Review 73 resemblance to the production in all the communities that I had visited and studied in Yucatan, Peru, and Guatemala. Consequently, I wanted to try and restore a neglected perspective to the craft. One point of the book was that archaeologists had neglected the ecological perspective in ceramic studies because of their own theoretical and paradigmatic biases. Fundamental to this issue, and another perspective that drove me to write the book, was the question: Are there common processual analogies for ceramics that are common to the present or the past that can help the archaeologist interpret ancient ceramics? Or, is the present so diferent from the past that archaeologists have to rely upon their theoretical imagination, and second-hand data to support it in order to understand the ancient society? At one level, the present is very diferent from the past, but on the other hand, the past is only knowable as analogies from the present. Some archeologists, however, seemed to have an implicit epistemology in which the past is not knowable from the present even though they are using their own minds with their own inherent biases to interpret it. Alternatively, there are those archaeologists who believe that the past speaks for itself, and all that one has to do is use modern scientiic technology to unlock or decode the data trapped in the ceramics. A Brief Historical Ethnography of Publication he printing history of Ceramic heory and Cultural Process provides a fascinating account of a diachronic ethnography of publishing technology, its changes, and its organization in the last 35 years. In some ways, this section is an ethnography of the history of production and production organization of the book’s publication; it is an ethnoarchaeological description of an ethnoarchaeological book. he thread of the publication history of Ceramic heory and Cultural Process began in an unusual way. In 1975, clay mineralogist, Bruce Bohor, and I published an article Archaeology magazine that documented the large palygorskite mine in the town of Sacalum in Yucatan (Arnold and Bohor 1975). Since palygorskite was one of the components of Maya Blue and the mine was large, we believed that the mine was an ancient source of the palygorskite used in Maya Blue. Following publication, I received two letters from editors of archaeology books asking if I was interested in writing a book. One of these letters came from an editor at the Cambridge University Press. he Press was interested in expanding their list in archaeology at the time and was looking for manuscripts. Since 74 Classics Review I had already been developing my ideas for Ceramic heory and Cultural Process, I considered submitting a proposal for it. Meanwhile, I started working on a more “etic” approach to the Quinua design data and used the units of symmetry to analyze and describe it (Shepard 1948, 1956; Washburn and Crowe 1988). I prepared a paper for a session about design at the Society for American Archeology meetings in 1980. In the evening after the session, I ended up at a meal with the chair of the session and Dorothy Washburn. he chair tried to convince me that my data revealed deep seated Andean cognitive and social structures, but I was unconvinced. Eventually, however, I recanted and realized that the most common patterns in the designs did indeed express deep structural principles of Quinua society found in its irrigation organization and social structure. As a result of this conversation, Dorothy asked me to write a chapter for her book he Cognitive Structure of Art that would be published by the Cambridge University Press (Arnold 1983). In 1981, I went to the UK to give a paper at a conference at the University of Leicester, and I traveled to Cambridge to talk with the archaeology editor (Robin Derricourt) about my book proposal. He was cautious and hesitant, saying that he would have to see the reviews of my paper in the Washburn volume. Meanwhile, I had already begun working on the book and was writing it as an electronic manuscript. here were no portable computers at the time, and personal computers such as the Osborne, the Kaypro, and Compaq did not yet exist. Further, operating systems, unlike today, were not compatible with one another. he only program available for editing at the time was a mainframe line editor in which every change speciied a line number, the number of characters to jump, and then a command to delete a speciied number of characters, or insert characters in parentheses. Writing, revising, and editing the typescript was a diicult and time-consuming task. I could only work for a little more than an hour at a time before I would go stark raving mad because it was so tedious. I scheduled my editing to no more than once a day to preserve my sanity, and still get the job done. I inished a draft of the manuscript later in early 1982, and submitted it for review. he review process seemed quite daunting because Cambridge had multiple screens for evaluating manuscripts. First, there are two or more anonymous readers, and if they are favorable, the series editors, Colin Renfrew and Jeremy Sablof, and the archaeology editor of the Press evaluated it. he archaeology editor often has training (if not an advanced degree) in the ield. Once approved Classics Review 75 by them, the manuscript must be approved by the Syndics of the Press. he Syndics consist of scholars from various disciplines at the University of Cambridge that meet together at least once a term (once during Michaelmas, Lent, and Easter terms) in the wood paneled board room in a 19th century Neo-Gothic Pitt Building on Trumpington Street in central Cambridge. hey review the readers’ comments and the recommendations presented by the respective editors. At their meeting in early April of 1982, the Syndics accepted Ceramic heory and Cultural Process for publication. Well after the Syndics’ meeting, the title for the book was a concern. he book was originally titled “Ceramic heory and Cultural Process: A Systems Approach”, but the Syndics objected to that title saying that it was not really a systems approach. When Robin Derricourt, the archaeology editor at the time, corresponded with me about the problem saying: “I rather like Ceramic heory and Cultural Process on its own.” I agreed. he title was perfect! It was short, descriptive and catchy. Even today, the succinct appropriateness of the title still astounds me. Because I had a heavy teaching load, and a young family (daughters aged ive and two), preparing the inal manuscript for publication was delayed. Not only was it necessary to prepare the camera-ready igures, and the lengthy tables, but there seemed to be endless errors that needed correcting. As the inal typescript was nearing completion in late 1982, I ofered to provide an electronic manuscript to the Press. I hoped that an electronic manuscript would make the book less expensive to the buyer. Robin Derricourt thought this was a good idea, but he told me that electronic manuscripts were very experimental at that time. As it turned out, he said that Ceramic heory and Cultural Process was one of the irst electronic manuscripts published by the Press. Further, there were issues of compatibility of the electronic iles from our mainframe at Wheaton with the computer that formatted the electronic ile for printing, and a requirement that I insert the printing codes in the electronic text. Meanwhile, the editor who had been responsible for the development of the book, Robin Derricourt, had left the archaeology editorship to head up the Press’ oice in South Africa, and by September 29 of 1982, a new archaeology editor, Katherine Owen, replaced him. Eventually, I inished editing and added the printing codes using the aforementioned line editor. he resulting electronic manuscript was transferred to ¾ inch magnetic tape, and was sent to the Press’ New York oice on January 7, 1983, arriving in Cambridge on February 3. 76 Classics Review Ceramic heory and Cultural Process was originally scheduled for publication in 1984, but there was an incredible amount of details to attend to with production supervisors, designers, and others. Meanwhile, Kate Owen left the press in March to go to Hong Kong with her husband. I received the proofs in May of 1984, but in late June of that year, I was scheduled to leave Wheaton to do research in Yucatán, Mexico, funded by an American Republics Research Grant under the Fulbright Program. My parents were visiting at the time, and I recruited my father and a team of students to work on the index, and we completed it by the time I left for Mérida with my wife and two daughters. he large size of the index created some concern for the editor, but it was accepted, edited, and prepared for printing. By September of 1984, the now late Peter Richards replaced Kate Owen as archaeology editor. Mail between Cambridge and Mérida was slow, with many last minute details a concern. I wanted people on the cover, but Peter believed that such a cover would limit readership outside of the New World. Some scholars, he thought, would dismiss the book as too regionally-focused based upon the people’s clothing. Pots, he said, had a more universal appeal. So, I agreed that the cover should be an image of pots rather than people, and a photo of water storage vessels from the manuscript was selected. he cover designer chose brown as a cover color because of its similarity to the color of pottery. Printing Technology My interest in pottery and printing technology comes out of a long-term interest in extractive and production technology and its organization. I was raised in a small town in South Dakota, and my house was 2.5 blocks from a large quarry of red quartzite. he quarry was always fascinating, and when I was a boy, I surreptitiously explored its edges, former quarries to the south and west of town, and other unique landscapes resulting from the efects of glaciation and melting on the beds of quartzite. Long trains of hopper and gondola cars full of the quartzite were always a source of fascination as they rolled past my house on the way to construction sites throughout the Midwest. When my family went on vacations to the American West, I asked my parents if we could visit mines and their associated industries. hey usually obliged and I still remember tours of the Anaconda Copper Company’s smelter near Butte, Montana, the gold smelting operation in Cripple Creek, Colorado, the Classics Review 77 zinc smelter near Kellogg, Idaho, and the giant open pit copper mine near Salt Lake City. his fascination and curiosity with technology also carried over into other kinds of technology. No visit to a play or musical is complete without a brief visit to the electronic console at the back of the theater, and a conversation, no matter how brief, with its operator about the organization of the production. When my wife and I attended the musical “Mama Mia” in London in 2008, for example, we sat in the sixth row, and since my father had been a high school band director, I spent much of the time watching the conductor. He seemed to be in control of the production and I wondered how the musicians deep in the orchestra pit could see him, and how the singers could sing together so well while they were looking at the audience, rather than the conductor. After asking the conductor, the answer was simple: he faced a tiny television camera with monitors placed deep in the pit and on the front of the balcony. Singers actually were looking at the monitors, not the audience. he printing of my book was no exception to my life-long fascination with production technology and organization. When I was in high school, a friend’s father owned the local newspaper, and I enjoyed visiting him while he was working at his father’s shop, learning about linotype machines, setting type, and the ofset and letterpress printing processes. Although many of the publications of the Cambridge University Press are contracted to other printers, Ceramic heory and Cultural Process was not and was printed by its printing division. Since I was on a six-month sabbatical in Cambridge during the time that the book was in press, I wanted to see the Press’ printing facility, particularly since my book was in the later stages of production. So, in February of 1985, I called Peter Richards, the archaeology editor, and asked if I could tour the Press’ printing plant. He thought it was a strange request, but it was consistent with my curiosity about technology and its organization. Although the Pitt building in downtown Cambridge houses some of the operations of the Press, the publishing division occupies a relatively new brick building of immense size (relative to other buildings in Cambridge) near the railway station on Shaftesbury Road. he printing division occupies a separate building across the street. I went to see Peter the next day, and he indulged my request, calling the printing division to inquire. He told me to go across the street to the lobby of the printing division, and someone would meet me there. When I arrived, a manager was waiting, and asked me why I wanted to visit the printing division. He explained that 78 Classics Review it was printing tests, and security was a concern. I explained that I had no interest in the tests and we could avoid those sections of the plant. Rather, I said, I had a book “in press” there and wanted to see the steps of publication. With that bit of information, my host changed his attitude, and became more accommodating. He made some phone calls to ind the location of the book, and subsequently led me into the plant. After seeing the “paste up” room and the large press room, he took me into an adjoining room where negatives about two meters square hung from the ceiling. He asked the foreman there about the book, and the foreman pointed out a group of eight such negatives hanging on one side of the room. I examined them carefully and sure enough, there was my book. I was thrilled! he negatives would be burned onto ofset plates and then printed on the giant ofset presses in the adjoining press room. On the way back to the entrance of the plant, we stopped in a room that housed a linotype machine with an operator clicking away on its keyboard. Every few seconds, a large arm would take the line of tiny brass molds (called matrices) of each line of type and lift them up into a slot below a reservoir of molten metal. A few seconds later, a solid line of type-metal (an alloy of lead, tin and antimony) dropped into a column of previously formed lines of type (Encyclopedia Britannica 2011a, 2011b). Nearby, we entered a small room of movable type of various fonts and sizes. For a moment, my host opened and closed drawers retrieving a few letters, and then handed me a set that spelled out “Dr. Arnold”. When I gestured to return the type he said, “No, you can keep them,” explaining that the movable type and the linotype were only used for some mathematical books and would soon be retired. After thanking my host and guide for his time, I left the building clutching my name in type. I still have the movable type, but have since glued the letters together. Every time I come across it, I recall my fascinating tour of the printing facility of the Cambridge University Press when my irst book was printed. My irst exposure to the published book came as a surprise, and was mediated by my wife, June. At the time, we were living at Clare Hall, one of the Cambridge Colleges, and family matters and my writing had occupied my mind. I had not yet received my copy of the book even though the Press was just across town. I had temporarily forgotten about it even though it was about to be published. One day my wife and I took the opportunity to have lunch together in downtown Cambridge. When we met, she said cryptically with a furtively sly smile, “Come with me. I have something to show you.” Classics Review 79 She led me over to the historic bookshop at the corner of Trinity and Market streets at the top of King’s Parade where books have been sold since 1581 (Cambridge University Press 2009). Once there she took me to the window on Market Street, and pointed to its contents. “here’s your book,” she said. Ofered as a new release from the Cambridge University Press, I was excited at the sight of my irst book, Ceramic heory and Cultural Process, in the window. Although the book was available for the 1985 annual meeting of the Society for American Archaeology, I did not attend because I was in Cambridge. he next fall, however, I attended the meeting of the American Anthropological Association and stopped by the Cambridge booth to chat with Peter Richards. Like any author of a newly published book, I asked about the sales of the book. I didn’t see the book in the booth, nor in the hands of any of the browsers, so I asked him: “Where is my book?” “It’s right here,” as he gestured over to a rack on the table. “It’s gone!” He exclaimed, “Somebody stole your book!” I was shocked and incredulous. “How could anyone steal a book from the publisher’s booth at a professional meeting?” After a brief awkward silence, Peter quipped: “Hmm . . . the ultimate compliment! Someone wanted it so bad that they stole it!” I went back to the booth several times during the meeting to see if the book had been returned, misplaced, or was indeed missing. It never reappeared, and Peter assured me that it had been stolen. Publication History: he Inluence of the Book I am not the best person to ask about the inluence of my book since I know its strengths and weaknesses. It is very diicult to know the inluence, or lack of it, of one’s own book by other than very subjective criteria. Further, my own lack of objectivity about it may both overemphasize its weaknesses and underestimate its strengths. Some proxy measures, however, provide some objective indication of its inluence: its publication history, its citation frequency, and references to it on the internet. One such measure is the number of copies sold. he original 1985 hardback edition consisted of 996 copies. Approximately 800 of these were sold with the 80 Classics Review remainder as “frees” consisting of review copies and those given for payment for reviewing manuscripts. he hardback went out of print three years after its publication, and the Press made the decision to reprint the book as a paperback. Because a number of errors occurred in the hardback edition, I wanted to make some changes in the new printing. he text was corrected and the 1988 reprint was simply called the “First Paperback Edition”. Because the content was slightly diferent, the Press changed the cover color to orange to distinguish it from the original hardback edition. he book was reprinted again in 1989 and went out of print in 1997 when it was transferred to digital printing and published as an on-demand reprint. here were two separate, but subsequent printings at this time, each with a diferent blue cover with no image—only the name of the book and its author. In 2003, the book was redesigned using a copy of the orange cover of the 1988 paperback and was issued again as an on-demand reprint. By end of 2009, a total of 4041 copies of the book had been sold, a rather surprising number for such a narrow, technical, and specialized book. Booksellers worldwide sell the book. A simple Google web search (using Arnold “Ceramic heory and Cultural Process”) results in about 36,000 “hits” that have varied by several thousand over time for reasons that are an enigma to me. I have seen numbers as high as 100,000, and as low as 10,000 with a mode of roughly 30,000–40,000. Some of these are references in library catalogues, and some are simply from the list in the “front matter” of every book in the Cambridge New Studies in Archaeology series. Other references come from reports and papers published on the internet, course syllabi, and course reading lists. Many of the citations, however, appear to be book sellers and include on-line stores in Britain, France, Germany, India, South Africa, Japan, and Korea, among others. Book sellers often sell both new and used copies, and may also rent them. So, it is unclear from these results how many copies have been sold, resold, and then sold again through the used-book market. It is also unclear whether book dealers that advertise the book have sold many copies, just have one copy in stock, or list it because they can get it on order quickly as an on-demand reprint. A second proxy measure of its inluence is the number of times that the book has been cited. One such measure is the citations listed in Google Scholar. Unfortunately, Google does not reveal the criteria for selecting the sources of the publications that it used for its citation data, but when one clicks the number of citations, the database appears to be scholarly articles that cite the book. Never- Classics Review 81 theless, a search on Google Scholar indicated that Ceramic heory and Cultural Process had been cited 451 times (using Arnold “Ceramic heory and Cultural Process”). Some citations come from my own books and articles, but many are international publications from outside of the USA. According to Google Scholar, the book is my most frequently cited work. A search on Google Books (using Arnold “Ceramic heory and Cultural Process”) yielded a range from 1000 to 1440 at diferent times. Most of these references come from books, but some journal articles also appear. Although a few of the citations are my own references to the book, examining the detail of some of search results revealed that the work was cited in books and articles that covered virtually every area of the world. Besides English, I saw titles in Spanish, French, German, Portuguese, Czech, Greek, Italian and a few that I could not identify. I asked one of the reference librarians at the Wheaton College Library about the databases that Google used for its searches, and he replied: As for Google books, it draws on the bibliographic resources of Worldcat (OCLC) such that you’ll ind a record for most anything. However, Google is still very much in the process of scanning the texts themselves. So it’s fair to assume that the actual number of references to your book is a little higher than Google Books reports. Google Scholar, likewise, can’t be viewed as comprehensive. Google Scholar doesn’t provide a complete list of the journals it indexes, but it is the broadest one-stop search of journal literature available. (Gregory Morrison, personal communication, December, 16, 2010) I have resisted the temptation to search for books by other authors in order to ascertain the meaning of the number of citations. Readers can do their own searches of their favorite books for comparison, but I decided to use other works of mine for that purpose. Judging by Google Scholar, many of my articles have never been cited, and my other book, Ecology and Ceramic Production in an Andean Community does not come close to the number of citations for Ceramic heory. It sold 800 copies and a search on Google Web produced 25,100 results, 428 references on Google Books, and 40 on Google Scholar. his book was more descriptive, area-focused, and much less theoretical than Ceramic heory. Citation rates, however, can be misleading and research on citation frequencies have revealed that they conform to “power laws” that can relect the 82 Classics Review difusion of new ideas as well as the copying of references. hey may also relect scholars’ social networks that share a common Ph.D. advisor or a particular school of thought. However, I have no cohort of fellow graduate students that cite my work. Because I teach in an undergraduate program, I do not produce Ph.D.s in archaeology or ethnoarchaeology. Although my students are very bright, capable, and motivated, they do not choose to study ethnoarchaeology at the post-graduate level, but rather medicine, law, teaching, social work, hird World development, Public Heath, teaching English as a second language, and other programs that prepare them for cross-cultural service to humanity. Citation frequencies, however, are not necessarily indicative of the inluence of the book. First, some publications cite the work, but with little reason. Further, journals such as American Antiquity and Latin American Antiquity require citations of “all the relevant literature”, but sometimes this kind of citation seems post-hoc and appears to reveal little of the inluence of the book on the author of the article. Second, I have seen other works that seemed to build on my ideas in the book, or are elaborations of them, but do not cite the work. Others utilize the arguments in the book and add some signiicant points or insights, but when this author’s work is used in subsequent publications, there are no references to the original source. My sense is that this happens frequently, and thus it is hard to evaluate how important and inluential Ceramic heory and Cultural Process is after 25 years because the generations of scholarly literature may build upon it, but its authors may not know the source of the ideas. As for the more subjective views of inluence of the work, time permits only a very brief review of oral and published information. Down through the years, colleagues and/or students in the US, UK, Japan, the Philippines, and Israel have told me about their appreciation for the book. I have seen the use of the book as a text, and on supplementary reading lists for courses on ethnoarchaeology and ceramics. During a trip to a conference in the UK in 2004, Russian archaeologist, Yuri Tsetlin, told me that the book had been translated into Russian and then placed in his lab in Moscow for his colleagues and students to consult. Israeli archaeologist David Adan Bayewitz used the approach in the book to reconstruct the cultural and economic context of the Galilee during the Roman Period (Adan Bayewitz 1993). New Testament scholar, however, John Dominic Crossan (1996: 226–229), challenged his reconstruction and tried to use my data to show just the opposite. David, however, was unconvinced (David Adan Bayewitz, personal communication). Classics Review 83 Relections on the Content of the Book As scholars have interacted with the content of Ceramic heory, I have also relected on many of the points made in the work. Not surprisingly, I have come to modify some of the ideas. I cannot take the time or the space to engage every work that uses the book, but I will try and make a few points that hopefully will enrich it for those who ind it useful. A span of 25 years since any book was published inds new ideas and new information that date some of the content of the original work. Ceramic heory and Cultural Process is no exception. One change consists of the discovery of more ancient kilns in Mesoamerica and on the coast of Peru. When I wrote Ceramic heory, there was limited evidence for such kilns. In some respects, however, this new information adds little to the work except for one very important point: the discoveries of kilns in many parts of Mesoamerica reinforce the point that kilns, in part, are a signiicant adaptation to a pattern of adverse weather and climate that would signiicantly disrupt pottery production. Feedback Of all of the points in the book, I have probably spent the most time thinking about the notion of feedback. Ceramic heory and Cultural Process argued that there are certain fundamental feedback relationships (I called them “feedback mechanisms” in the book) between ceramic production and the environment that are isomorphic from society to society and are thus universal in all human societies that make pottery. hese isomorphic relationships have their foundation in the physical and chemical characteristics of the clays themselves. Humans must respond to these characteristics in similar ways if they are to make pottery. Feedback consists of the information lowing from the natural and social environment to the human agent through the senses such as the ears, the eyes, the nose, and the skin. his kind of information is probably one of the most fundamental elements of human epistemology, and humans use it to make choices in behavior. Language, of course, represents the most obvious channel of information from feedback, and consists of the morphemes of speech that symbolize semantic categories. Morphemes and their syntax provided the source of most human knowledge is gained, learned, and stored. Language consists of what anthropologists call emic data. his kind of data 84 Classics Review consists of the categories and structure that the natives themselves recognize. Without some behavioral veriication or reference, however, the relevance of emic data to the material and behavioral world of the archaeological record is limited. Anthropologists thus refer to this kind of information as “ideal behavior,” that which people say that they do rather than what they actually do (“real behavior”). For the ethnoarchaeologist, discovering the patterns in a society’s behavioral and material world that are relevant to archaeology require a methodology beyond learning language categories to uncover what actually happens in a society; language data must be veriied with observations of actual behavior. Anthropologists solve this dilemma with the classical anthropological methodology of “participant-observation,” by learning to see the world as the natives see it, but also seeing it from a point of view that is outside of the culture using crosscultural etic categories provided by anthropology. Emic and etic perspectives represent two diferent kinds of epistemologies that are complementary. If one relies on verbal data, one’s ability to generalize cross-culturally will be limited. Furthermore, emic data can’t be used to invalidate etic data, or contradict observations of behavior that have cross-cultural validity. hey only complement such data with culturally-speciic information. Consequently, one signiicant problem in ethnoarchaeology is the culturally relative nature of emic categories. hey must be related to etic units of observation if they are to have any cross-cultural validity. So, etic units of observation are the most relevant to archaeology because they may be material categories that can be used cross-culturally. his is precisely the problem I faced in writing Ceramic heory and Cultural Process; I needed to ind data and their relationships that transcended cultural boundaries, and that could be used and applied in all cultures. My data from Mexico, Peru, and Guatemala, although interesting and important, were too culturally limited because they were too culturally relative, and had limited crosscultural applicability. Rather, in Ceramic heory and Cultural Process, I relied upon etic data to build the feedback mechanisms for pottery production. Feedback implicitly recognizes that humans make choices, and are the agents of continuity and change, but they act upon information that comes from feedback. his information may be conscious and deliberate, but it may also be gathered unconsciously. he potter, for example, receives feedback via the visual, tactile, and aural channels when he engages raw materials, the production process, weather, his Classics Review 85 motor habit patterns (or habitus), and the social and physical environment. his information provides input to make decisions based upon the potter’s own knowledge and experience. As I read about feedback loops in the biological and climatic sciences (Clement et al. 2009; Kerr 2009), I have become more convinced that the information that lows into human brains through the senses can indeed have generalizing power, not as deterministic causes, but as mechanisms that provide lows of information to potters to help them make choices. hey serve as cross-cultural isomorphous processes that are one way to describe and explain human behavior, much like evolutionary archaeologists use the mechanisms of evolution to explain culture change through time. In my enthusiasm to use the feedback mechanism to explain relationships in Ceramic heory, I now believe that at least some of my arguments in which I argue that feedback was responsible for the origin of pottery production were overstated. As others have shown, emic factors also inluence the origin of pottery. Rather, feedback mechanisms operate in the ceramic production process that constrain or encourage production. hey provide information to potters, but they do not necessarily cause the original development of the craft, but only provided information to potters to continue or suspend production over time as a mechanism of selection, or lack thereof. Distance to Resources he distance to resources model described in Ceramic heory and Cultural Process appears to be the most widely used, and perhaps the most inluential part of the book. In the years since its publication, however, I have made modiications to the model (Arnold 2005, 2006), suggesting that the model represents crude probabilities. Many ethnoarchaeologists have found that the model its in their own data in Ethiopia (Arthur 2006), Crete (Day 2004), and other parts of Mexico (Druc 2000). Further, other distance data from Syria (Tsetlin 1998) and Colombia (Duncan 1996:49–52) are also consistent with the model. Finally, some archaeologists have found the model useful in identifying local versus nonlocal ancient pottery (Morris 1994a, 1994b, 1995, 2000, 2001), and many more examples can be gleaned from the literature. Probably the most important elaboration of the model since Ceramic heory was Heidke’s (Heidke et al. 2007) work in reining it, and extracting more inter- 86 Classics Review pretations from it. He expanded the data set of ethnographic distances to clay and temper sources, and provided a statistical summary from which he derived four conclusions. First, he found that the batches of ethnographic distance measurements separately assembled by Arnold and by Heidke were nearly identical. Second, although the ranges of the New World and Old World data were very similar, Old World potters tended to travel a bit farther to collect their clay. hird, potters who traveled by foot to collect their clay traveled no more than 3.3 km and that draft animals or a wheeled vehicle (truck, donkey, cart, or wagon) increased that distance. Fourth, no clear relationship existed between the distance potters go to obtain their clay, and the use and type of temper. Potters who used naturally tempered clay go no more than 5.0 km to collect it. Finally, potters in the Southwest traveled up to 6.9 km to collect their clay although the median distance is 1 km (Heidke 2007:149). Some ethnoarchaeologists appear to believe that they can invalidate the distance model by showing that distance is not a criteria that potters actually use to obtain their resources. As I have argued earlier, emic data do not falsify the model. Emic data concerning the selection of ceramic raw materials enrich our understanding of culturally-relative factors that inluence the source locations, but they do not contradict etic data (Arnold 1971, 2000, 2008). Over the decades of my research, I have found that distance generally is not an overt emic criteria that potters used to obtain ceramic resources (see Arnold 2000). Rather, potters will say that apart from the characteristics of the material itself, tradition, religion, land tenure, and availability are criteria for procuring clay or temper from a particular resource (see Arnold 1971, 1972a, 1972b, 1993, 2000, 2008:153–220). hese emic explanations provide a holistic explanation of behavior that is a part of any good anthropological explanation. But, as I have said before, emic explanations are not generalizable because they are culturally-relative. his realization was one of the reasons that I wrote Ceramic heory. Distance data are not emic data. Rather, they are etic data based upon actual behavior. Distance to resources appears to operate in a way that is diferent from standard emic explanations and appears to operate outside of potters’ awareness. Potters may provide distance to resources if asked, but for whatever reason that the potters use to explain why they use a particular source, the actual distances between the potter and his clay and temper resource does reveal a pattern, and this pattern can be expressed as a power law (log-log). When the distances to clays (Figure 1) and tempers (Figure 2) are plotted 87 Classics Review Distance to Clay Sources as a Power Law 50 45 43 Number of Communities 40 35 30 25 23 20 y = 29.227x-1.05 R² = 0.7954 15 12 10 5 5 7 7 3 1 0 0 2 4 6 8 2 2 2 1 1 1 1 3 1 1 1 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 Distance to Clay Source (in km) Figure 1. A power law curve of distance to clay sources and frequency (number of communities) with the data used in Ceramic Theory and Cultural Process (N = 117). The correlation of the curve with the data is almost identical to that for the curve for temper. Distance to Temper Sources as a Power Law 18 17 16 Number of Communities 14 12 10 8 7 6 y = 10.433x-0.975 R² = 0.8053 4 3 2 2 2 1 1 1 1 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Distance to Temper Source (in km) Figure 2. A power law curve of distance to temper sources and frequency (number of communities) with the original data used in Ceramic Theory and Cultural Process (N = 35). The correlation of the curve with the data is almost identical to that for the curve for clay. 88 Classics Review against their number, and a best it trend line is added, a power law curve best its the data with a very high correlation (R2 $ 0.79). his relationship suggests that the number of communities and the distances to clay and temper resources are a kind of scale-free, self-organizing system found in a wide range of phenomena (Bentley and Maschner 2001; Bentley et al. 2004; Bentley and Shennan 2003). Why is this? Other examples of power laws suggest that some relationships exist among the data that accounts for the curve. Power laws of naming patterns, citation rates, and some ceramic styles suggest that a relationship exists between the data points within the database used. Consequently, copying (or difusion) is postulated as an explanation. In the case of distance to resources, however, postulating cultural difusion as an explanation of the curve seems counter-intuitive because the distances to ceramic resources come from all over the world. Rather, the one common factor in all of these data is the energy expended by the human body to obtain raw materials and suggests that there are limits to that energy in order to make potters’ craft economically viable. A person can only carry 40–50 kg of clay and temper so far on a regular basis. Now, with the expansion of the model by Heidke et al. (2007), draft animals, carts, boats, and modern transportation are essentially “energy extenders” that extend the energy of human carriers by using some other form of transportation. hese diferent modes of transport also extend the distance by either transferring the energy largely away from human carriers, by making transport more eicient by obtaining more raw materials per trip, or by making more trips and transporting a lesser amount per trip as a consequence of some other activity. he distance to resources is thus not necessarily a direct cause for the beginning of pottery making, but consists of subtle feedback that both stimulates ongoing ceramic production for those potters near their resources, and selects against those potters that have to go more than 4–5 km to those resources. Feedback, in this case, selects for energy limits for obtaining ceramic production over time. hose communities within the high frequency distances in the model are selected for, whereas those with lower frequency are selected against. In this sense, it is a processual analogy, and has selective force in the evolution of ceramic production. If energy extenders are introduced, however, the distance to the resources may increase. An example of this evolutionary approach is illustrated by an example that I noted several years ago at the poster session at the annual meeting of the Society Classics Review 89 for American Archaeology. For several years, an archaeologist presented a poster that showed that potters in an abandoned village on one of the Hopi mesas traveled more than seven kilometers to obtain their clay. One of the implicit points of the poster was that the distance exceeded my model, and it was thus presumably invalid because it did not it his data. All seemed to be very convincing until one examined the poster more carefully. First, the potters passed their clay source on their way back to their village from their ields. hey could easily have collected clay without making a special trip. Such procurement behavior is also a kind of “energy extender” because the potter expends energy as a consequence of some other activity and therefore can travel a greater distance to get resources, but travel there more often as a consequence that activity. Second, the occupation of the village was short-lived, and one could easily postulate that even obtaining their clay as consequence of subsistence activities was at least part of the reason that deselected it from making pottery, and it moved to another location. he Ceramic Petrology Group of the UK explored the topic: “What does it mean to be local?” at their meeting at the University of Southampton in 2000. Morris (2001) used the distance model as a way to deine “local” at the conference. Tomber (2001), however, believed that 7 km was too small a distance for the Roman Period because resources came to the production site from 20 km away, and the Romans used roads and probably carts to transport their clay. his distance actually does conform to the model and simply illustrates the role of carts as energy extenders in the distances that human carriers would use. Originally, the model was designed only to apply to human carriers on foot, not to a culture with wheeled vehicles, as existed in Roman times. here is, of course, no substitute for clear evidence of a source of raw materials by comparing pottery made in a location with local raw materials using compositional analysis, but in the archaeological record, such deinitive associations are seldom as clear as one would like them to be. My intention in the presenting the distance model was simply to provide some empirical evidence of what constituted “local” production, and to eliminate improbable hypotheses. Although there is occasional evidence that raw materials are imported into a community from some distance, the model and the evidence upon which it is based shows that this option, although possible, is very improbable, and if it exists, it will not persist very long, particularly if production intensiies. One of the perplexing problems addressed in the book was the distance to sources of volcanic ash temper in ancient Maya ceramics in the Maya lowlands. 90 Classics Review Was it locally obtained, or was it imported? No sources of such ash exist today in the lowlands and there are no volcanoes there. Anabel Ford is studying this problem, and she believes that volcanic ash in pottery comes from local sources, but she has encountered some skeptics who still believe the ash was imported from the highlands. It is more probable, however, that Maya potters were scooping up ash from ash falls and mixing it with their clay to make pottery. he ashtempered pottery ended when the ash falls stopped and the potters ran out of the ash that they stored in their houses. As my distance model shows, the import of clays and tempers is improbable from distances greater than 4–5 km without some kind of energy extender such as carts, boats, or modern transportation. Solutions to this problem might use a diferent approach. In our work with palygorskite and Maya Blue, clay mineralogist, B. F. Bohor found that palygorskite is likely derived from volcanic ash. he evidence for this were minerals such as beta-quartz, euhedral zircons, magnetite, and sanidine in the palygorskite deposits that we sampled. Since volcanic ash falling on land weathers and turns into soil after a few years, it is possible that some of the harder minerals unique to the ash are more resistant to weathering than the glass fraction, and might remain in the soils of the lowlands. he mineralogical analyses of soils there might reveal that such minerals still exist in the soil and could have only come from volcanic ash. Such a inding would support the hypothesis that the lowlands were blanketed with ash from highland volcanoes and that ancient Maya potters used these sources to temper their pottery. Seasonality Weather and climate also have a feedback relationship with pottery making—either limiting or stimulating the development of the craft (Arnold 1985:61–98). he combination of the characteristics of the raw materials and weather and climate may place constraints on production. In order to make pottery, the potter not only needs raw materials, but also needs speciic environmental conditions to facilitate the drying of fuel and pottery and to ire the vessels without damage. Consequently, ceramic production is sensitive to rainfall, temperature, and the amount of sunshine, and potters cannot make ceramics in a climate with a distinct rainy season without delays in production, and damage to drying and iring pottery. his relationship can enable an archaeologist to infer the seasonality, presence, absence, intensity, and scale of ceramic production that occurred in the Classics Review 91 past. Weather and climate constraints can prevent ceramic production from being economically viable and prevent it from being anything but a seasonallypracticed craft. Does the potter have a choice of making pottery during inclement weather? He certainly does. He can choose to ignore the rainfall and cold and take his chances with breakage and sagging. But, in order to avoid damage or breakage to pottery at such a time, he has only four choices to adjust to the problems created by weather: 1) keep production low enough so that vessels can dry in the open space in houses (e. g. Arthur 2006:42–44); 2) build additional structures (larger houses, workshops, and drying sheds) to keep out the rain, wind, and cold to protect drying pottery and provide a dry setting for iring; 3) work as a wage laborer in a production unit that has these facilities; or 4) schedule pottery production during dry sunny weather, and practice some other activity (such as agriculture, a service, or another craft) until the weather becomes more favorable. his latter choice is nicely expressed by Hirth (2009) in his use of new concepts to account for such phenomena: “intermittent crafting” and “multi-crafting” that are a “risk management strategy” by which potters managed their subsistence risk. Scheduling In order for pottery production and subsistence tasks to be compatible, they must be scheduled so that they will not interfere with one another. Without structures with interior space to dry vessels, tasks must be allocated to a dry period in order to avoid damage to the pottery and so that they do not conlict with agricultural tasks. One way to avoid this conlict is to allocate pottery making and agricultural activities to diferent genders, schedule such tasks at diferent times during the agricultural year (D. Arnold 1985:99–108), or at diferent times during the day. In Yucatan, for example, activities of swidden agriculture such as cutting the forest, burning, planting, cultivating, and harvesting can be scheduled so that they can complement, rather than compete with, pottery making activities. A potter can work in his swidden plot during the early morning when fog and moisture may damage newly formed vessels and slow drying, and return to making pottery in the late morning when sunshine and heat are required to dry clay, dry pottery, and ire. he recognition of the universality of seasonality and scheduling in ceramic production and its role in the limitation of full-time ceramic production is now 92 Classics Review at least 35 years old (Arnold 1975a). he efect of weather on the seasonality (and thus the intensity) of pottery making, however, has been slow in coming to theories of craft production in archaeology. Most recently, Hirth (2009) has taken an approach which expands these notions in describing all crafts. Hirth invites the readers to reevaluate production intensity by laying out three alternative concepts to the part-time/full-time distinction. he irst consists of what Hirth calls “intermittent crafting” in which craftsmen only practice their trade for a portion of the yearly cycle. he second concept, “multi-crafting”, involves the practice of several crafts by members of a household, either at the same time, or at diferent times. Hirth’s third concept views craft production as a risk-management strategy in which a household diversiies its subsistence strategy, practicing several crafts (and perhaps agriculture) to insure adequate returns for its sustenance, thus reducing the risks that occur with any one activity. hese concepts are totally consistent with data presented in my Current Anthropology article (Arnold 1975a) of 35 years ago, in Ceramic heory and Cultural Process, and my most recent work (Arnold 2008). Man/Land Relationships As I have pointed out previously in this relection, the data for feedback mechanisms were largely based upon my work in Mexico, Peru, and Guatemala, but supported by other examples drawn from the literature. he marginal agricultural land surrounding communities of potters near Quinua, Peru, and in the Valley of Guatemala are graphic examples of the beneit of pottery production when land is too poor for much agriculture. he reasons for this marginality are many: steep slopes and highly eroded top soil, extensive erosive cutting of gullies, lack of access to irrigation water, and lack of rainfall on the lower slopes (Arnold 1975a). Such marginality, although devastating for much agriculture, provides abundant ceramic resources. How important is agriculture for craft specialists? Agriculture may be critical to potters because by raising their own food, potters can bufer the downswings in the demand of pottery. Agriculture and craft production thus complement one another especially if one regards craft production as a risk-management strategy (Hirth 2009). Since Ceramic heory was published, I have been pleasantly surprised to discover other communities that also it this generalization. One, a town called Classics Review 93 Porvenir in Honduras (Mouat and Arnold 1988), has extremely poor agricultural land to the degree that the men in the community cannot farm locally and must go elsewhere to make a living. Women are thus left at home with little or no inancial support. To provide for their families, they have turned to making pottery (Mouat and Arnold 1988). Similarly, on a trip south of Tegucigalpa in the summer of 1988, I noticed that the highway passed through a relatively lat plateau riddled with large outcrops of granite. Inhabitants had planted some maize ields in some areas, but the plants were small and puny compared to maize in areas with better soils. As we proceeded across the plateau, we came upon roadside stands selling pottery, and it appeared that the inhabitants in this rather marginal agricultural area also made pottery and sold the vessels to passers-by on the highway. In spite of the abundant and graphic examples of potters living on nonexistent or marginal agricultural land, documenting this relationship may be elusive in the archeological record. Douglass (2002) tested this relationship in an archeological survey in the Naco Valley of northwestern Honduras by hypothesizing that the land around pottery making sites would be poor agriculturally. he hypothesis failed. Does this mean that agricultural land and pottery production have no relationship? Not necessarily. Douglass’ testing caused me to rethink the relationship between pottery production and poor or marginal agricultural land. As I have suggested above in the discussion of feedback, the relationship between agricultural land and pottery making provides feedback that selects for or against pottery making (and perhaps for crafts in general as it does in Quinua, Peru). It may not be possible, however, to have a clear assessment of the quality of land used by potters in the archaeological record. First, in any given archaeological site, it is impossible to know if the potters cultivated any land at all, and if they did, where it was located. Secondly, if one does wish to test such a hypothesis, then it is totally reasonable to use the land around a settlement of potters, as Douglass did. If there was no poor agricultural land there as Douglass found, then potters may not have cultivated any land at all, or they may have been multi-crafting, and making pottery intermittently because of the rainy season, and combined it with practicing another craft activity during the inclement weather (e.g., Feinman 1999). As interesting as the relationship is between potters and agricultural land, and its obvious ecological advantages as a source of ceramic raw materials, the relationship between pottery making and agricultural land can be much more complicated than it appears to be. 94 Classics Review Conclusion Ceramic heory and Cultural Process consisted of an approach that sociologists call “Grounded heory”, an inductive approach to theory building. In this case, it was built upon my own experience with pottery production in Latin America. At the time, few archaeologists (e.g., Owen Rye; see Rye 1981; Rye and Evans 1976) had much experience studying the production of pottery in so many diferent communities in such diverse locations on two continents. Although there are a number of tweaks to my interpretations in the book, and richer data is available now, I am encouraged that increased ethnoarchaeological research on ceramics demonstrates the validity of many of the points that I made. Most recently, the work of Ken Hirth and the authors in his edited volume (Hirth 2009) relects many of the concerns and issues that I developed in Ceramic heory and Cultural Process. It is encouraging to see the development of the theory of craft production become more closely aligned with ethnoarchaeological data, particularly that of ethnographic ceramic production. As for the development of a ceramic theory, archaeologists could still beneit by working with one craft at a time, and then inductively building a theory of the development and evolution of all crafts. Nevertheless, there is still much to be learned from those syntheses that deal with all crafts. Acknowledgments I am very grateful to Gary Feinman, Ryan Williams, Daniel Masters, and my wife, June, who read this paper and made many valuable comments that improved it signiicantly. References Cited Adan-Bayewitz, David. 1993. Common pottery in Roman Galilee: A study of local trade. Bar-Ilan Studies in Near Eastern Languages and Culture. Ramat-Gan, Israel: Bar-Ilan University Press. Arnold, Dean E. 1967a. Maya Blue: A new perspective. MA thesis, University of Illinois, Urbana. _____. 1967b. Sak lu’um in Maya culture: and its possible relationship to Maya Blue. University of Illinois, Department of Anthropology Research Reports No. 2, Urbana, Illinois. Classics Review 95 _____. 1970. he emics of pottery design from Quinua, Peru. PhD diss., University of Illinois, Urbana. _____. 1971 Ethnomineralogy of Ticul, Yucatan potters: Etics and emics. American Antiquity 36(1):20-40. _____. 1972a. Native pottery making in Quinua, Peru. Anthropos 67:858-872. _____. 1972b. Mineralogical analyses of ceramic materials from Quinua, Department of Ayacucho, Peru. Archaeometry 14(1):93-101. _____. 1975a. Ceramic ecology in the Ayacucho Basin, Peru: Implications for prehistory. Current Anthropology 16(2):183-205. _____. 1975b. Reply to Haaland and Browman. Current Anthropology 16(4):637-640. _____. 1976. Ecological variables and ceramic production: Towards a general model. In Primitive art and technology, eds. James S. Raymond, B. Loveseth, C. Arnold, and G. Reardon, 92-108. Archaeological Association, Department of Archaeology, University of Calgary. _____. 1978a. he ethnography of pottery making in the Valley of Guatemala. In he ceramics of Kaminaljuyu, Guatemala, ed. Ronald K. Wetherington, 327-400. University Park: Pennsylvania State University Press. _____. 1978b. Ceramic variability, environment and culture history among the Pokom in the Valley of Guatemala. In Spatial organization of culture, ed. Ian Hodder, 39-59. London: Duckworth. _____. 1983. Design structure and community organization in Quinua, Peru. In Structure and cognition in art, ed. Dorothy Washburn, 56-73. Cambridge: Cambridge University Press. _____. 1984. Social interaction and ceramic design: Community-wide correlates in Quinua, Peru. In Pots and potters: Current approaches in ceramic archaeology, ed. Prudence M. Rice, 133-161. Monograph XXIV. Los Angeles: Institute of Archaeology, University of California. _____. 1985. Ceramic theory and cultural process. Cambridge: Cambridge University Press. _____. 1991. Ethnoarchaeology and investigations of ceramic production and exchange: Can we go beyond cautionary tales? In he Legacy of Anna O. Shepard, ed. Ronald L. Bishop and Frederick W. Lange, 321–345. Boulder: University Press of Colorado. _____. 1993. Ecology of ceramic production in an Andean community. Cambridge: Cambridge University Press. (On-demand digital reprint, 2003) _____. 2000. Does the standardization of ceramic pastes really mean specialization? Journal of Archaeological Method and heory 7(4):333–375. _____. 2005. Linking society with the compositional analyses of pottery: A model from comparative ethnography. In Pottery manufacturing processes: Reconstitution and interpretation, eds. Alexandre Livingstone Smith, Dominique Bosquet, and Rémi Martineau, Section 2, 15–21. Oxford, UK: BAR International Series 1349. 96 Classics Review _____. 2006. he threshold model for ceramic resources: A reinement. In Ceramic studies: Papers on the social and cultural signiicance of ceramics in Europe and Eurasia from prehistoric to historic times, ed. Dragos Gheorghiu, 3–9. Oxford, UK: BAR International Series 1553. _____. 2008. Social change and the evolution of ceramic production and distribution in a Maya community. Boulder: University Press of Colorado. Arnold, Dean E., and Bruce F. Bohor. 1975. Attapulgite and Maya Blue: An ancient mine comes to light. Archaeology 28(January):22-29. Arnold, Dean E., Hector Nef, and Ronald L. Bishop. 1991. Compositional analysis and “sources” of pottery: An ethnoarchaeological approach. American Anthropologist 93(1):70–90. Arthur, John W. 2006. Living with pottery: Ethnoarchaeology among the Gamo of southwest Ethiopia. Salt Lake City: University of Utah Press. Bentley, R. Alexander, and Herbert D. G. Maschner. 2001. Stylistic change as a selforganized critical phenomenon: An archaeological study in complexity. Journal of Archaeological Method and heory 8(1):35–66. Bentley, R. Alexander, Matthew W. Hahn, and Stephen J. Shennan. 2004. Random drift and culture change. Proceedings of the Royal Society B: Biological Sciences 271(1547):1443–1450. Bentley, R. Alexander, and Stephen J. Shennan. 2003. Cultural transmission and stochastic network growth. American Antiquity 68(3):1–27. Black, Mary. 1963. On formal ethnographic procedures. American Anthropologist 65(6):1347-1351. Black, Mary, and Duane Metzger. 1965. Ethnographic description and the study of law. American Anthropologist 67(6):141-165. Cambridge University Press. 2009. Cambridge university press bookshop: History of the bookshop. Accessed March 12, 2011. http://www.combridge.org/uk/bookshop/history.htm. Clement, Amy C., Robert Bergman, and Joel R. Norris. 2009. Observational and model evidence for positive low-level cloud feedback. Science 325(5939):460–464. Crossan, John Dominic. 1998. he birth of Christianity: Discovering what happened in the years immediately after the execution of Jesus. New York: Harper Collins. Crowe, Donald W., and Dorothy K. Washburn. 1987. Flow charts as an aid to the symmetry classiication of patterned design. In Material anthropology: contemporary approaches to material culture, eds. Barrie Reynolds and Margaret A. Stott, 69-101. New York: University Press of America. D’Andrade, Roy. 1995. he development of cognitive anthropology. Cambridge: Cambridge University Press. Day, Peter M. 2004. Marriage and mobility: Traditions and the dynamics of the pottery system in Twentieth Century East Crete. In Pseira VIII: he archeological survey of Pseira Island, Part I, eds. Philip P. Betancourt, Costis Davaras, and Richard Hope Simpson, 105–162. Philadelphia: INSTAP Academic Press. Classics Review 97 Douglass, John G. 2002. Hinterland households: Rural agrarian household diversity in northwest Honduras. Boulder: University Press of Colorado. Druc, Isabelle C. 2000. Ceramic production in San Marcos Acteopan, Puebla, Mexico. Ancient Mesoamerica 11:77–89. Duncan, Roger J. 1996. he ceramics of Ráquira, Colombia: Gender, work, and economic change. Gainesville: University Press of Florida. Encyclopedia Britannica. 2011a. Linotype. Encyclopedia Britannica. Accessed Feb. 22, 2011. http://www.britannica.com/EBchecked/topic/342596/Linotype. Encyclopedia Britannica. 2011b. Type metal. Encyclopedia Britannica. Accessed Feb. 22, 2011. http://www.britannica.com/EBchecked/topic/611718/type-metal. Feinman, Gary. 1999. Rethinking our assumptions: Economic specialization at the household scale in Ancient Ejutla, Oaxaca, Mexico. In Pottery and people: A dynamic interaction, eds. James M. Skibo and Gary M. Feinman, 81–98. Salt Lake City: University of Utah Press. Frake, Charles. 1964. Notes on queries in ethnography. American Anthropologist 66(3):132-145. Harris, Marvin. 1968. he rise of anthropological theory. New York: homas Y. Crowell Co. Heidke, James M., Susan Leary, Sarah A. Herr, and Mark D. Elson. 2007. Alameda brown ware and San Francisco grey ware technology and economics. In Sunset crater archaeology: Ceramic technology, distribution, and use, eds. Scott Van Keuren, Mark D. Elson, and Sarah A. Herr, 145–183. Tuscon, AZ: Center for Desert Archaeology. Hirth, Kenneth. 2009. Housework and domestic craft production: An intoduction. Special issue, Archaeological Papers of the American Anthropological Association 19:1–12. Hirth, Kenneth, ed. 2009. Housework: Craft production and domestic economy in ancient Mesoamerica Special issue, Archaeological Papers of the American Anthropological Association 19. Kerr, Richar A. 2009. Clouds appear to be big, bad player in global warming. Science 325(5939):376. Lemonnier, Pierre. 1986. he study of material culture today: Toward an anthropology of technical systems. Journal of Anthropological Archaeology 5(2):147–186. _____. 1992. Elements for an anthropology of technology. Ann Arbor: Anthropological Papers, Museum of Anthropology, University of Michigan No. 88. _____. 1993. Technological choices: Transformation in material cultures since the Neolithic. New York: Routledge. Metzger, Duane, and Gerald E. Williams. 1963a. Tenejapa medicine 1: he curer. Southwestern Journal of Anthropology 19(2):216-236. _____. 1963b. A formal ethnographic analysis of Tenejapa Ladino weddings. American Anthropologist 65(5):1076–1101. _____. 1966. Some procedures and results in the study of native categories: Tzeltal “Firewood”. American Anthropologist 68(2):389–407. 98 Classics Review Mitchell, William P. 1976. Irrigation and community in the central Peruvian highlands. American Anthropologist 78(1):25-44. _____. 1991. Peasants on the edge: Crop, cult, and crisis in the Andes. Austin: University of Texas Press. _____. 2006. Voices from the global margin: Confronting poverty and inventing new lives in the Andes. Austin: University of Texas Press. Morris, Elaine L. 1994a. he organization of pottery production and distribution in Iron Age Wessex. In he Iron Age in Wessex: Recent work, eds. Andrew P. Fitzpatrick and Elaine L. Morris, 26–29. Salisbury, UK: Trust for Wessex Archaeology Ltd. _____. 1994b. Production and distribution of pottery and salt in Iron Age Britain: A review. Proceedings of the Prehistoric Society 60:371–393. _____. 1995. Study 10: Pottery production and resource locations; An examination of the Danebury Collection. In Danebury: An Iron Age hillfort in Hampshire. Volume 6: A hillfort community in perspective, ed. Barry Cunlife, 239–245. CBA Research Report 102. York, UK: Council for British Archeology. _____. 2000. Islands, trade and pottery. he Old Potter’s Almanack 8(1):3–4. _____. 2001. Abstracts: Later Prehistoric pottery review. he Old Potter’s Almanack 9(1):2–4. Mouat, Laurette, and Dean E. Arnold. 1989. Ceramic ecology and pottery production in El Porvenir Honduras. In A pot for all reasons: Ceramic ecology revisited (A Special Publication of Cerámica de Cultura Maya), eds. Charles C. Kolb and Louana Lackey, 239–261. Philadelphia, PA: Laboratory of Anthropology, Department of Anthropology, Temple University. Pike, Kenneth L. 1967. Language in relation to a uniied theory of the structure of human behavior. he Hague: Mouton. Rye, Owen S. 1981. Pottery technology: Principles and reconstruction. Washington, D. C.: Taraxacum. Rye, Owen S., and Cliford Evans. 1976. Traditional pottery techniques of Pakistan: Field and laboratory studies. Volume 21. Washington, D. C.: Smithsonian Institution Press. Shepard, Anna O. 1948. he symmetry of abstract design with special reference to ceramic decoration. Contributions to American Anthropology and History 47:211–292. Shepard, Anna O. 1956. Ceramics for the archaeologist. Publication 609. Washington, D.C.: Carnegie Institution of Washington. Tomber, Roberta. 2001. Local production during the Roman Period. he Old Potter’s Almanack 9(1):4. Tzetlin, Yuri B. 1998. Some notes on modern pottery production in Syria. he Old Potter’s Almanack: Joint Newsletter of the Prehistoric Ceramics Research Group and the Ceramic Petrology Group 6(3):1–3. Washburn, Dorothy, and Donald Crowe. 1988. Symmetries of culture. Seattle: University of Washington Press. Williams, Patrick Ryan. 2001. Cerro Baúl: A Wari center on the Tiwanaku frontier. Latin American Antiquity 12(1):67-83.