A Universal Theory of Pottery Production: Irving Rouse, Attributes, Modes, and Ethnography

By Richard A. Krause

In A Universal Theory of Pottery Production, award-winning archaeologist Richard A. Krause presents an ethnographic account of pottery production based on archaeological evidence.
 
Krause posits that the careful study of an archaeological site’s ceramics can be used to formulate a step-and-stage theory of pottery production for the area. Krause’s work suggests that by comparing the results of inquiries conducted at different sites and for different times, archaeologists may be able to create a general ethnographic theory of pottery production.
 
Krause demonstrates this process through a comprehensive analysis of potsherds from the highly stratified Puerto Rican site of Paso del Indio. He first provides a comprehensive explanation of the archaeological concepts of attribute, mode, feature, association, site, analysis, and classification. Using these seven concepts, he categorizes the production and decorative techniques in the Paso del Indio site. Krause then applies the concept of “focal form vessels” to the site’s largest fragments to test his step-and-stage theory of production against the evidence they provide. Finally, he assigns the ceramics at Paso del Indio to previously discussed potting traditions.
 
Unlike other books on the subject that use statistical methods to frame basic archaeological concepts, Krause approaches these topics from the perspective of epistemology and the explicatory practices of empirical science. In A Universal Theory of Pottery Production Krause offers much of interest to North American, Caribbean, and South American archaeologists interested in the manufacture, decoration, and classification of prehistoric pottery, as well as for archaeologists interested in archaeological theory.

• Language: English
• Publisher: University Alabama Press
• Release date: May 30, 2016
• ISBN: 9780817389444

Book preview

A Universal Theory of Pottery Production

CARIBBEAN ARCHAEOLOGY AND ETHNOHISTORY

L. Antonio Curet, Series Editor

A Universal Theory of Pottery Production

IRVING ROUSE, ATTRIBUTES, MODES, AND ETHNOGRAPHY

RICHARD A. KRAUSE

The University of Alabama Press

Tuscaloosa

The University of Alabama Press

Tuscaloosa, Alabama 35487-0380

uapress.ua.edu

Copyright © 2016 by the University of Alabama Press

All rights reserved.

Inquiries about reproducing material from this work should be addressed to the University of Alabama Press.

Typeface: Caslon

Manufactured in the United States of America

Cover image: Taino bat; drawing by Richard A. Krause

Cover design: Michele Myatt Quinn

∞

The paper on which this book is printed meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48–1984.

Library of Congress Cataloging-in-Publication Data

Names: Krause, Richard A., 1938–

Title: A universal theory of pottery production : Irving Rouse, attributes, modes, and ethnography / Richard A. Krause.

Description: Tuscaloosa : The University of Alabama Press, 2016. | Series: Caribbean archaeology and ethnohistory | Includes bibliographical references and index.

Identifiers: LCCN 2015032326| ISBN 9780817318987 (cloth : alkaline paper) | ISBN 9780817389444 (e-book)

Subjects: LCSH: Indian pottery—Caribbean Area—Analysis. | Pottery, Prehistoric—Caribbean Area—Analysis. | Rouse, Irving, 1913–2006. | Attribute (Philosophy) | Antiquities—Classification. | Ethnoarchaeology—Philosophy. | Indian pottery—Puerto Rico—Analysis. | Pottery, Prehistoric—Puerto Rico—Analysis. | Indians of the West Indies—Puerto Rico—Antiquities. | Puerto Rico—Antiquities.

Classification: LCC F1619.3.P6 K73 2016 | DDC 972.95/01—dc23

LC record available at http://lccn.loc.gov/2015032326

Contents

List of Illustrations

Introduction

1. An Interactional Theory of Artifact Description

2. A Theory of Ceramic Production: The Focal Form

3. A Theory of Production Steps and Stages

4. The Classification of Artifact Complexes

5. Background for the Study of the Ceramic Sample from Paso Del Indio

6. The Paso Del Indio Sample Size, Morphology, and Manufacture

7. Modes of Appendation

8. Decoration, Drying, and Firing

9. Summary and Discussion

Works Cited

Index

Illustrations

Figures

2.1. Morphological landmarks for vessels with radial symmetry

2.2. Morphological landmarks for vessels with bilateral symmetry

2.3. Shoulderless vessel morphology

2.4. Base and bottom morphology

2.5. Rounded shoulder morphology

2.6. Angular shoulder morphology

2.7. Rim morphology

2.8. Lip morphology

2.9. Decorating tools

2.10. Stamping and engraving tools

2.11. Fabric and finger impressing

2.12. Modified surfaces

3.1. Mass modeling from the bottom up

3.2. Coiling from the bottom up

3.3. Mass-modeling from the shoulder

3.4. Handle manufacture

3.5. Bonfire kiln stack

5.1. Map of Puerto Rico

6.1. Profile views of extra large shouldered vessels

6.2. Profile views of large shouldered vessels

6.3. Profile views of medium-sized shouldered vessels

6.4. Profile views of small shouldered vessels

6.5. Profile views of extra large shoulderless vessels

6.6. Profile views of two large and four medium-sized shoulderless vessels

6.7. Profile views of small shoulderless vessels

6.8. Coils and coiling scars

6.9. Profile views of slab-modeled vessels

6.10. Bottom and base views of slab-modeled vessels

6.11. Vessel width to height ratios by grouped layer

6.12. Bases and bottom shape and construction

6.13. Shoulder morphology by grouped layer

6.14. Lip morphology by grouped layers

6.15. Rim morphology by grouped layers

6.16. Flanged lips

7.1. D- and Ω-shaped handles

7.2. Handle distribution

7.3. Lugs

7.4. Handle frequency by grouped layers at Paso del Indio

7.5. Monkey, human, and humanoid adornos

7.6. Bat adornos

7.7. Bat and bat-like adornos

7.8. Bat and humanoid adornos

7.9. Animal head adornos

7.10. Animal and animal-like adornos

7.11. Tail, limb, and fin adornos

7.12. Frequency of adorno kind

7.13. Adorno frequency by grouped layers

8.1. Pink and red paint

8.2. Area painted by grouped layers

8.3. Pink- and red-painted vessel parts

8.4. Tooled decoration by layers

8.5. Tooled designs 1

8.6. Tooled designs 2

Tables

6.1. Vessel Width to Height Ratios by Grouped Layers

6.2. Shoulder Morphology by Grouped Layers

6.3. Distribution of Direct, Round, and Flat Lips by Grouped Layers

6.4. Distribution of Direct, Everted, and Inverted Lips by Grouped Layers

7.1. Handle Distribution by Grouped Layers

8.1. Painted Sherds by Grouped Layers

8.2. Tooled Decorative Elements

8.3. Tooled Decoration by Grouped Layers

8.4. Ratio of Painted to Tooled Decoration by Grouped Layers

Introduction

This dialogue began as a thought experiment. It was designed to produce an archaeologically derived ethnographic account of pottery production. To achieve this end the meaning assigned the archaeological concepts of attribute, mode, feature, association, site, analysis, and classification were explicated. Then the analyst (1) focused upon the evidence left by production and decorative techniques, (2) created a focal form vessel(s) from the sample’s largest fragments, and (3) organized the evidence by production steps and stages. Organizing by steps and stages was important because it led to a theory of pottery manufacture. Thus if the steps and stages used by a site’s potters are derived from an analysis of a sample of their products (i.e., the modes of manufacture indicated by potsherds) they constitute a theory of how the pottery was made. This theory may then be used to assess transformations in the procedures of pottery production and examine the sample’s place in a potting tradition. In a final step the samples placed in a potting tradition may be productively classified.

The ideas that guided this dialogue were born in the early 1960s in conversations with Irving Rouse and Harold Conklin. They were refined through course work with Irving Rouse, Harold Conklin, and Floyd Lounsbury. Those that survived were forged in the heat of argument with Preston Holder, Raymond Wood, and Donald Lehmer. Still, they remained ideas in need of a body of archaeological materials to which they might be applied. In 1990 this opportunity was supplied by Carlos Solis Magaña, who offered me funding, laboratory space, and the assistance of Elvis Bablonia for work on the ceramics from the site of Paso del Indio in Puerto Rico.

Over the years Holder interrogated me by posing questions that I struggled with. When he asked me, What is archaeology? I responded, following Rouse, Archaeology is the science of artifacts. He retorted, Tell me, boy, just what science is and what an artifact is. When I cited White to the effect that science is what scientists do, he found my answer totally unacceptable and advised me to go back to the drawing board and try again. He was even more skeptical of my textbook-inspired claim that an artifact was anything made or modified by man. Do you really mean to tell me that to practice your science of archaeology I must put soil erosion, air, and water pollution in the same category as pot sherds and projectile points? When I pointed out that similar phenomena had recently been identified as eco-facts he was less than impressed. We have now cluttered the literature with an ill-defined concept and the moon with footprints and other stuff. Is the moon therefore an eco-fact? I think not! There’s something missing here, boy, and I urge you to find it.

When Holder asked me to define a feature I responded, A feature is an artifact too large or too fragile to be moved to the laboratory for analysis. Again he scoffed, Boy, you’re putting the archaeologist ahead of the Indian. Review the term’s use as an archaeological concept. I think you will find that the concept has been used in different ways at different times. The use you reference is a response to the exigencies of archaeological salvage programs conducted by the River Basin Surveys. Can’t you imagine an analytically more useful approach? When I responded to Holder’s question about archaeological sites as places where artifacts were found, he considered my description inadequate. When I asked him in what way he smirked, Think relationships, boy, think relationships.

When I told Holder I was intrigued by Rouse’s (1939) mode, he asked me if the mode was not just another and superfluous way to describe an attribute. Boy, what is an attribute anyway? How does an attribute come into being? I replied by quoting Rouse, Attributes are the properties of artifacts. All modes are attributes but not all attributes are modes. Holder replied by telling me that I had better be able to tell my colleagues what the precise difference might be.

When I described the Sumpter site ceramics as broad-shouldered jars with constricted necks and out-flaring rims, Holder frowned deeply, then responded, If they’re jars why don’t they resemble the ball jars used to can fruit and vegetables? To respond I read a dictionary statement: "n {MF jarre fr. OProv. Jarra, fr. Ar Jarrah earthen water vessel] (1592) 1: a wide-mouthed container made typically of earthenware or glass. Holder was delighted. He scoffed, About the only property the Sumpter ceramics and ball jars have in common is that they are containers, eh, boy. Why describe both as jars? You can do better. If you think more about the issue of containers the properties you want for an explicit description will become apparent and you will be able to answer my question. Once you answer this question I presume you will be more precise in your description of ceramic morphology. As I struggled to produce a more precise description of ceramic morphology, I referred to a specimen as a neck sherd. Holder scoffed: Where does the neck begin? Where does the neck end? Where does the rim begin? When I showed him what I took to be the neck on a second specimen from the same excavation, he laughed. You have showed me different starting points from the same pot. Why? How will we be able to describe or measure the neck if we can’t find the starting or ending point? What will such measurements mean to future inquiry? Imprecision will do more to confuse than clarify."

Holder also sneered at my attempts to classify artifact complexes. He considered them arbitrary efforts to organize the data of prehistory. He challenged me to do better. Must attempts to classify artifact complexes be purely arbitrary?, he smirked as only he could. Can’t we find alternatives that are not grounded in arbitrary criteria? My response was, Many previous attempts to classify artifact complexes have misused the taxonomic principles and systems available to us. I have addressed this issue in several previous publications (Krause 1994, 1998; Jenkins and Krause 1986, 2009). I will try again here.

While I was attending classes and participating in graduate seminars at Yale, American archaeology was experiencing a dramatic transformation. This transformation was first called the new archaeology. As it grew in popularity the new archaeology became processual archaeology. Insofar as processual archaeologists challenged the basic analytic practices of their predecessors, they urged attention to statistical procedures in attribute pattern discrimination and the use of systems theory to interpret the classificatory superstructure produced by cultural historians.

Processualists viewed culture as a system composed of various interrelated subsystems with human behavior as a point of subsystem articulation. Variation in human behavior was considered both a product of subsystem restructuring and the means for establishing systemic harmony at a different level or plane. Culture change was thus construed as achieved through variation in one or more of the subsystems that grew, displaced, or reinforced others as systemic balance was challenged by disruptive social, economic, political, or environmental forces. A prime analytic desideratum was the isolation of each subsystem and its study as a variable within the matrix of forces to which it was exposed. The ultimate goal was to construct archaeologically testable models of the factors producing variation in prehistoric human behavior. The originality of the systemic approach was questioned and the credibility of its aims and goals was debated (Kushner 1970). It, nevertheless, generated a considerable amount of intellectual enthusiasm among a generation of archaeologists and had a large body of adherents. Yet most of them used the basic units of analysis, classification, and sequencing practices developed for the study of culture history.

David Clarke introduced the only serious challenge to Rouse’s interpretation of attribute. Clarke’s hierarchical arrangement of culture, assemblage, culture group, and techno-complex rested upon his interpretation of attribute and artifact type. He did not understand his cultural historical predecessors’ use of these units making his attempt to redefine them unworkable. Clarke defined an attribute as a logically irreducible character of two or more states, acting as an independent variable within a specific artifact system (Clarke 1968:186). If by multistate Clarke meant only presence or absence then his attribute was logically irreducible but was not properly a variable. If Clarke’s character was both variable and repetitive within the population of occurrence, then it was the member of a class of related occurrences. This attribute class was a Boolean-class product. Any such class can be logically reduced to the necessary and sufficient conditions for class membership. Clarke’s (1968:197) definition of an artifact type as an homogeneous population of artifacts which share a consistently recurrent range of attribute states within a given polythetic set was vexing in its ambiguity. Both the range of attribute states and the membership in polythetic sets required a subjective judgment (Read 2007:134–135). Neither Clarke’s construal of attribute nor his definition of artifact type was widely used.

The processualist use of systems theory offered the prospect of a truly anthropological archaeology. Flannery (1967), Binford (1962), Deetz (1968), and Longacre (1970) produced compelling interpretations of archaeological materials that fused ethnographic with archaeological concepts to give social relevance to the cultural historical construal of attribute, mode, artifact, artifact type, feature, association, and site. Thompson and Longacre (1966:270) succinctly state the prevailing processualist view of archaeological sites: All of the material remains in an archaeological site are highly patterned or structured directly as a result of the ways in which the extinct society was organized and the patterned ways in which the people behaved. Krause and Thorne (1971:245–257) were the first to challenge this view. They urged greater attention to determining precisely which materials in an archaeological site were patterned as a consequence of people’s behavior and which were not. Schiffer (1972:156–165) first expanded Krause and Thorne’s challenge then formalized it to create an approach he called behavioral archaeology (Schiffer 1976). The systems approach advocated by the new archaeologists and Schiffer’s observations about archaeological site patterning, when combined, led to the widespread implementation of new techniques of site sampling and excavation and the invention and use of more sophisticated artifact recovery techniques. They did not, however, produce a serious modification of cultural historical principles of analysis and classification.

As processualism matured, its limitations, especially a propensity to interpret social transformations as a consequence of technologically and environmentally stimulated modifications of subsistence practices, became more and more apparent. As Rouse puts it, Where are the people in processual interpretations? See, for example, Rouse’s 1965 publication The Place of Peoples in Prehistoric Research. Holder opined that prehistoric lifestyles must have provided more than a perpetual quest for efficient food production. He was not alone in this thought.

Processualism generated a beyond subsistence (Duke and Wilson, eds. 1995) postprocessual response to the processual interpretation of human behavior. Postprocessualists stress the role of artifacts in the transformation of past human behavior (Hodder 1985). For the postprocessualist, artifacts are more than reflections of human behavior. Artifacts are viewed by postprocessualists as a consequence of human agency and as participants in the transformation of human understanding and action. Postprocessualists distinguish between structures and structuring, between being and becoming (Krause 1995:307–352). Artifacts through human agency participate in structuring social life. Artifacts, if properly interpreted, can be used to create models of the being and the becoming of ancient societies. Postprocessualists produce intriguing images of past societies. Yet they and their contemporary rivals, the neo-Darwinian evolutionists, still used the basic analytic, classificatory, and sequencing constructs developed for the study of culture history.

A neo-Darwinian challenge to processualism attempted to modify the interpretation of attributes and artifacts but did not seriously modify their definition. For the neo-Darwinians attributes and artifacts are important as participants in and indicators of evolution. To produce a fit between static archaeological evidence and evolutionary dynamism attributes and artifacts are defined as evidence for, and the consequence of, adaptation. Read (2007:282–283) argues that by this move the neo-Darwinists reintroduced a formerly discredited biological analogy to archaeological enquiry. The biological construal of adaptation has serious consequences. A biological trait is considered adaptive if it conveys a reproductive advantage. With slowly reproducing species this reproductive advantage can only be measured after the fact. Hence the traits that produced it are introduced as adaptive by virtue of their presumed import in ex-post-facto, just-so stories. The reintroduced biological analogy, even if the archaeologist identifies the artifact traits he thinks are adaptive and even if artifact trait creation or replacement is relatively rapid, makes this biological problem into an archaeological problem. Simply put, the archaeological propensity to produce just-so stories is reinforced by dependence upon Darwinian theory. Without significant and appropriate modification the cultural historical construal of attribute and artifact are not compatible with Darwinian theory. In 1980 Holder, shortly before his death, read Dunnell’s (1978) article in American Antiquity and noted that the new evolutionists had yet to solve the problem of defining basic units of artifact production and/or use that provide a reproductive advantage. Until they do, he opined, Their approach will generate more heat than light. In this he was prescient.

Holder died before landscape archaeology, as Phillip Duke (2008:278) puts it, became the flavor of the month for the last decade or so. Landscape archaeologists view their subject as the long-term study of the intersection of land, place, and people. They construe the landscape as a cultural construction whose interpretation requires reference to the data supplied by knowledge of an area’s topographic, hydrologic, climatic, faunal, and floral composition as well as settlement patterns and site structures. In short, they view landscape archaeology as a holistic study of the way an area’s inhabitants created strategies for negotiating their physical surroundings—a view that requires multiple scales of analysis to implement. In doing so they stress the impact of place, the importance of scale, the artificiality of boundaries, the informative potential of the longe durée, and the landscape as reflector and indicator of ethnicity and class. Yet they still use the concepts of attribute, mode, artifact, artifact type, feature, and site developed for the study of culture history (Krause 2010:249–252).

Several of my colleagues in the cohort following my own venture the opinion that culture history is theory deficient, unimaginative, and uninformative. I hope I can dispel this view as I discuss Irving Rouse’s basic ideas and create from them a universal theory of hand potting. In sum, since most contemporary archaeologists, whether they are processualists, postprocessualists, neo-Darwinian evolutionists, or landscape archaeologists, build their various approaches upon the analytic and classificatory foundation provided by culture history I consider Rouse’s analytic ideas and practices relevant to contemporary archaeological inquiry. In the chapters 1 through 4 I address the issue of attribute, mode, artifact, artifact type, feature, association, and site as necessary elements of archaeological theory building and answer the questions my mentors have asked me about them. For those with little or no interest in the answers please skip to chapter 6 with the description of the Paso del Indio site and its ceramics. For those who choose to read these chapters, I hope the answers do not vex them.

1

An Interactional Theory of Artifact Description

In class, and in private discussions, Irving Rouse identified archaeology as the science of artifacts. If Rouse’s science of artifacts is to be empirical, it must have three major objectives: (1) to isolate the phenomena of import, (2) to describe the phenomena of import, and (3) to establish the means by which these phenomena can be systematically anticipated and explained. These objectives are complementary and interrelated. As commonly understood the explanatory and predictive principles of a science lie in its theories. Further, it is the role of theory to identify and describe the regularities to which individual cases conform. It is precisely these regularities that make it possible to predict (i.e., anticipate an occurrence of the phenomenon in question). Thus, if observed or anticipated regularities are described, they are statements of theory. From this perspective an adequate description is an exercise in concept formation. In archaeology, as in other empirical sciences, concept formation and theory building proceed together.

If archaeology is to be the science of artifacts, the term artifact designates a concept with logical and epistemological priority. It warrants a reasonably full and precise explication. But first I must address the tacit confusion between specimens and data. The two are frequently treated as synonyms. This practice engenders questionable results. In archaeology, specimens are usually (although not exclusively) objects. The specimens of greatest import are those identified as artifacts or artifact by-products. These specimens may be understood as data sources. They are not, in and of themselves, data. To construe them as data confuses the object of an inquiry with the results of that inquiry. Data are not specimens. They are the recorded consequences of observations taken upon specimens.

At first glance the distinction between specimens and data seems trivial. Upon greater reflection it makes possible several important points. First, any specimen identified as an artifact or artifact by-product (or any group of specimens so identified) may provide multitudes of data now and in the future. The number and kinds of observations that can be made about any object or set of objects are infinite. The observations taken on those specimens identified as artifacts are a select sample of all possible observations. Second, when an observer selects a particular property as integral to an artifact or artifact by-product it may be termed an attribution. Its consequence is an attribute. But attribution may be accomplished from different perspectives. The attributes that result depend upon the perspective taken during attribution. Thus attributes are descriptions of object properties. They are the consequence of an interaction between analyst and object. The analyst observes and