A Universal Theory of Pottery Production: Irving Rouse, Attributes, Modes, and Ethnography
240
A Universal Theory of Pottery Production: Irving Rouse, Attributes, Modes, and Ethnography
240eBook
Available on Compatible NOOK devices, the free NOOK App and in My Digital Library.
Related collections and offers
Overview
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.
Product Details
| ISBN-13: | 9780817389444 |
|---|---|
| Publisher: | University of Alabama Press |
| Publication date: | 05/30/2016 |
| Series: | Caribbean Archaeology and Ethnohistory |
| Sold by: | Barnes & Noble |
| Format: | eBook |
| Pages: | 240 |
| File size: | 6 MB |
About the Author
Read an Excerpt
A Universal Theory of Pottery Production
Irving Rouse, Attributes, Modes, and Ethnography
By Richard A. Krause
The University of Alabama Press
Copyright © 2016 University of Alabama PressAll rights reserved.
ISBN: 978-0-8173-8944-4
CHAPTER 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 selects the properties that are to be attributes according to an implicit or explicit theory of description. During attribution most archaeologists reference Western standards of object description. This practice implies that artifacts are objects and may be adequately described as such. Differently put, archaeologists have customarily considered descriptions of artifacts (or groups of artifacts) valid if they adequately reflect Western canons of object description. This practice confuses specimens with artifacts. While all artifacts may be specimens, not all specimens are artifacts. Those standards of description applicable to specimens in general may not be precise enough, or explicit enough, to adequately describe that proper subset of specimens that are artifacts. Unfortunately, if one proceeds in the customary manner — that is, using standards of description applicable to specimens in general to choose which of the many properties of things are significant — and renders as a description of artifacts the sum of these properties, then (1) there is no assurance that the description of artifacts or groups of artifacts from one archaeological site will productively apply to artifacts or groups of artifacts from another and (2) there is no assurance that modifications of group membership will not require a modification of group description.
There have been attempts to resolve the ambiguity inherent in using Western standards of object description through finer-grained (i.e., more detailed) observation and measurement. It has become commonplace to seek the aid of computers in this task. It is true that a computer will produce fuller descriptions as it is fed more and more measurements and observations. If properly programmed, the computer will also correlate these measurements and observations to produce statements of association. If desired, strongly cohering measurements and observations may be identified as significant. But what is gained? To be precise, one is given a statement, or set of statements, identifying those previously taken measurements and observations that are most suitable for efficiently and explicitly remeasuring and reobserving the specimens at hand.
One may assume that the previously discussed observations and measurements will apply to yet-undiscovered specimens. This assumption may not be warranted. In Read's elegant and comprehensive treatment of artifact classification he (2007) discusses this issue at length. Even if future discoveries are similar, there is no in vacuo assurance that the original or the expanded sample is, or will be, significant. To rephrase the issue, would the three blind men of parable fame have a better understanding of the elephant if each could take more measurements on his particular part of the beast? No! Then, too, would applying the same procedures to yet another elephant provide a more determinant picture? No! It is for these reasons that we must seek an alternative grounded in an explicit theory of artifact description. I begin with the term artifact.
Let us consider an artifact as any culturally salient entity intentionally modified by man. For a similar point of view see Clarke 1968:145; Fagan 1978:32; and Read 2007:185. I realize it is customary to use the expression made or modified by man when referring to artifacts (Hester 1976:26; Hole and Heizer 1969:44; Jennings 1974:373; Knudson 1978:482; Sheerer and Ashmore 1979:560; Thomas 1979:456). To include the term made is redundant. All entities made are modified. Hence, those entities made are a special case, a proper subset, of the set of all entities modified. By modified I mean transformed by the application of external force. By an entity I mean any definably bounded, expectable stable substance which occupies space (for a discussion of this point see Harré 1961:42–54). For archaeological inquiry it is axiomatic that no two substances can occupy precisely the same space at exactly the same time (Binford 1958:354). I take the term man as commonly understood in anthropological discourse, that is, as any member of the genus Homo. I would feel comfortable identifying the Australopithecines as men if they are included in the genus Homo. My understanding of intentionally is problematic. It will require additional attention.
Inferring intentionality is the part of artifact identification that requires recourse to regularities in human behavior. To begin I presume that all intentional behavior is purposeful. But it would be incorrect to suppose that all purposeful behavior is the product of conscious, rational action. It would also be incorrect to assume that purposeful behavior is understood by actor and analyst in precisely the same way. In archaeology we apprehend what people did, not what they wanted to do, or thought they did (see White and Thomas 1956:275–308). Nevertheless, if the proposed synonymy between intentionality and purpose is accepted, we may consider it a fact that all purposeful human behavior is patterned.
There are good reasons for identifying purposeful behavior as patterned (Deetz 1968:31–41; Spaulding 1960:60–83). Purposeful behavior is a proper subset of the set of all human behaviors. Now if all human behavior were random, predictability would be impossible. There is, however, a modicum of predictability in the conduct of human affairs (albeit just how much may be debated) indicating that some behavior is not random. We may safely assume that nonrandom behavior is patterned, although the import we may assign to such patterning is a matter to be decided empirically. For the purposes of a given inquiry we may identify instances of patterned behavior as members of that proper subset of nonrandom behavior we identify as purposeful. In fact, we commonly construe patterned behavior as purposeful and assume that it is based in the enactment of those socially contrived and transmitted rules and specifications for proper performance we call culture (Childe 1950:2; Clarke 1968:19–20; Malinowski 1936:440–449; Taylor 1948:102). We do, nonetheless, need an analytical construct that will allow us to move from a general sense of the import of patterned behavior to a specific and archaeologically appropriate claim (or set of claims) about the relationship between patterned human acts and their observable consequences. It was this consideration that led Rouse (1939) to create the mode.
Rouse (1939:313–325) considered the mode to be any standard, custom, or belief to which the artisan conformed when producing, modifying, or using artifacts. Rouse is using customs, standards, and beliefs in the cultural anthropological sense, that is, as socially shared and transmitted rules of and for culturally appropriate behavior. But customs, standards, and beliefs cannot be directly observed. If modes are to be synonymous with customs and so on, they cannot be directly observed either (Taylor 1948:101). It is, however, common ethnographic practice to infer the existence of specific customs, standards, or beliefs from observed and (in some respects) measured patterns of learned-transmitted behavior (Beattie 1964:16–33; Firth 1963:1–40; Gluckman 1968:31–45; Lienhardt 1966:156–163; Malinowski 1961:1–25; Pelto and Pelto 1978:ix–xiii and 1–16; Radcliffe-Brown 1958:166–177). Rouse (1939) claims that it is also possible to infer the past existence of archaeologically relevant customs, standards, and so on from their tangible, observable, and measurable consequences as attributes of artifacts. Modes, therefore, are abstractions created by inference from suitably patterned attributes. But judging the adequacy of attribute patterning may be a worrisome problem. Its solution will require recourse to moderately complex epistemological issues.
While sequences of patterned behavior produce those archaeological manifestations we identify as the products of purposeful action (Deetz 1967:2, 1968), these actions are processes in time and space. These processes can, however, be modeled as a sequence of noncommutative acts (see Nadel 1957). By a process I mean a set of interrelated events that produce a reasonably uniform result. By noncommutative I mean those acts whose relatedness is such that variation in their order of occurrence will alter the result achieved (Krause and Thorne 1971). In sociological inquiry sequences of behavior achieve analytic import when the events within them occur in a fixed order (see Nadel 1957). In archaeology, analytically important event sequences must be determined from an analysis of the potentials and limitations inhering in the interplay between humanly possible acts and the knowledge, tools, implements, and materials available to an artisan. Thus, purposeful human actions may be experienced as sequences of noncommutative events, but those artifact properties that are the products of purposeful human action are only visible as complexes of static attributes (Clarke 1968:20; Schiffer 1976; Wobst 1978:303–309). The two are by no means identical. The intellectually challenging task of transforming static complexes of artifact attributes into models of the dynamic human acts that produced them is crucial. At this point I introduce criteria that will aid in organizing the knowledge necessary for this task. The first is the criterion of redundancy.
In theory there may be as many different ways of acting as there are actors and enactments. I may posit a world in which chaos reigns. I have, however, dismissed this view as incompatible with my experience. I am not concerned with all possibilities. I am interested in ways of acting that are patterned, that is, ways of acting that are governed by rules and, thus, are rendered determinate. One of the first archaeologically relevant things that can be said of such ways of acting is that they must be finite and less numerous than the number of possible actors or enactments (for a similar point of view see Dunnell 1971:123–124; Hester 1976:78). I, therefore, suspect that purposeful ways of acting are repetitive and that significant artifact attributes will also be repetitive. That behaviors or attributes are repetitive means that they are the denotata of a set or class of behaviors or attributes. As such each must have properties in common with others in the set or class to which they belong. It is precisely these properties I take as constituting the necessary and sufficient conditions (the significata) for behavior or attribute identification. This may be symbolically stated as follows:
(1.1) A1 = A2 = φA1 = φA2
where A1 and A2 indicate two instances of the same attribute or behavior and stands for the significata appropriate to each.
If attributes are repetitive, they may be the results of purposeful actions. Yet to assign significance to either behavior or attributes on the basis of repetition alone is risky. Some repetitious human behavior is not customary and not all repetitive artifact attributes can be identified as the products of human action. Some repetitive human actions are genetically determined; some are idiosyncratic. Some artifact attributes are the consequences of natural forces; some are the consequences of the raw materials from which the pieces in question were made. While I may expect that repetitive human acts and repetitive artifact attributes can be understood as members of sets or classes of acts or attributes, it would be unreasonable to assume that all such sets or classes are significant. This fact indicates the need for a second criterion. I shall call it the condition of relative invariance.
The condition of relative invariance is also a consequence of the rule of dependence of purposeful behavior. I may, for example, suppose that rules are centralizing ideas that human beings apply to the perceivable variability in their environment (see Goodenough 1957:167–173). By applying these ideas the continuum of sight, sound, smell, and feel is classified and rendered conceptually manageable. No two sights, sounds, smells, or feels are precisely the same. Nonetheless, human beings, for purposes of organizing objectively different things, identify some of their properties as identical, ignoring for the moment those properties that render each different (Simpson 1961:3–5). It is through this procedure that humans strive to produce order in their social and natural environments (see Kluckhohn and Kelley 1945:26). This perception of order must, of course, organize those ideas that lead to the acts I wish to identify as purposeful.
When translated into action purposeful procedures enable an artisan to produce or use artifacts. Nevertheless, even if artisans are attempting to produce or use their wares by the uniform application of customary standards of thought and action, they will achieve a degree of variability forced upon them by a host of forces — the nature of raw materials, the wear on tools, the inability to exactly replicate sequences of behavior, differences in understanding and motivation to name but a few (Clarke 1968:149–152; White and Thomas 1956:275–308). If I properly measure and graph this variability, I expect a Gaussian distribution. The expected variability will be unimodal because repeated attempts to apply a traditional standard or procedure will constrain the randomizing effects of nonuniform and highly complex external forces (Chang 1968:4; Taylor 1948:119). The essential variability in purposeful human actions should therefore be reflected in some of the attributes displayed by artifacts.
(Continues...)
Excerpted from A Universal Theory of Pottery Production by Richard A. Krause. Copyright © 2016 University of Alabama Press. Excerpted by permission of The University of Alabama Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.