In this chapter we want to identify some of the most important issues that have emerged in the course of the Typological Debate, and to discuss them in relation to our own point of view. Some of these are genuine issues that have been around as long as classification itself, and are certainly incapable of resolution. Others from our perspective are non-issues that have arisen through conceptual misunderstandings, or as a result of the intrusion of inappropriate theoretical considerations. The issues that are discussed here are obviously not all of equal importance, nor is the list comprehensive. In general we have selected for discussion those issues that are particularly relevant to our own interests.

Natural vs. artificial classification

This issue dominated the Typological Debate during its configurational/functional phase, but its origins are much older. In fact, it is surely the oldest and the most persistent of all classificatory controversies, tracing back in a certain sense to the age-old debate between “fixists,” who believe in a settled order of things, and “fluxists,” who do not. This question troubled both the Greek and the Chinese philosophers of ancient times (cf. Foucault 1973: 127–8).

The debate over natural vs. artificial classification was evidently flourishing in 1874, when W. S. Jevons published his path-breaking Principles of Science.

We will now consider some corollaries of the philosophy of conceptual instrumentalism that was outlined in Chapter 1. To begin with, it involves a wholly pragmatic view of the role that definitions play in scientific typologies. This in turn has implications for more general philosophical issues, and in particular for the presently much-debated question of whether types are “real.” We will end the chapter by commenting briefly on that issue, and also on certain similarities and differences between typologies and systems of measurement. Here as elsewhere our philosophical comments must be fairly brief, and cannot fully explore the issues that we raise, much less resolve them. Our main concern is to show the interconnection between our views on archaeological typology, and on more general philosophical problems.

Definiteness and definitions

The philosophy of conceptual instrumentalism that was set forth in Chapter 1 holds that typologies and other systems of deliberately constructed concepts in the sciences are instruments that are designed for specific purposes, and they are to be evaluated in terms of their effectiveness in serving those purposes. It is implicit in this philosophy that definiteness in the concepts is not an end in itself, and should only be insisted upon to the extent that it serves the purposes for which the instrument was created. This holds true of any feature in a typology that can vary in definiteness, including type concepts, and the types themselves.

In previous chapters we have considered a number of factors, theoretical as well as practical, that influence the development of classificatory systems. We are now ready to discuss the different kinds of classifications, typologies, and taxonomies that archaeologists actually make and use, and the reasons why they differ one from another. We will begin by re-emphasizing certain basic structural distinctions that affect many different kinds of classifications. We will then go on to consider some specific kinds of classifications that archaeologists make for different purposes. Finally we will discuss some additional, practical considerations that are unrelated to purpose, but that may result in variation from one classification to another.

Basic structural distinctions

Classification and typology. We indicated in Chapter 4 that, according to our usage, a classification is any contrasting set of categories into which a specific body of material may be subdivided, or partitioned. A typology is a particular kind of classification; one designed for the sorting of entities into categories that are absolutely mutually exclusive. The practical importance of this distinction will appear when we discuss the differences between artifact classifications, which are nearly always typologies, and “culture” classifications, which are not.

Typology and taxonomy. A basic typology has only one level of abstraction or generalization; all of its member types are considered to be equivalent in this respect (see Chapter 7).

I (identified in subsequent chapters as EWA) received a BS in electrical engineering and a PhD in philosophy, with a minor in mathematics, at Stanford University. I have been a Fellow of the Behavioral Models Project at Columbia University, Instructor at Wesleyan University, and Professor of Philosophy at the University of California, Berkeley. Aside from philosophy of science I have published papers on mathematical psychology, decision theory, theory of probability, and probabilistic aspects of logic, including a book, The Logic of Conditionals, an Application of Probability to Deductive Logic (EWA 1975). In philosophy of science I have studied and written on induction and confirmation, causation, scientific concept formation, theory of measurement, and the foundations of physical topology and geometry, on which I am presently writing a book.

A word should be said about my philosophical development in relation to the present work. My initial orientation was toward the highly abstract and formalistic approach of the logical positivists (empiricists) and their allies, such as Rudolf Carnap, Carl Hempel, and Karl Popper. This approach dominated the philosophy of science in the 1940s and 1950s, and my early work, especially on measurement, was in that tradition. However, my faith in positivism was shaken by the criticisms of Thomas Kuhn (1962), though I did not accept the Kuhnian historicist position (the “dialectic of scientific paradigms”) that came to succeed positivism as the dominant trend in the philosophy of science in the 1960s and 1970s.

Critics have been heard to complain that for archaeologists typologies are a kind of fetish, and that classification may become an end in itself (Kluckhohn 1939: 338; Bennett 1943: 208; Hill and Evans 1972: 231, 267; Vierra 1982: 164; Hayden 1984: 81). This view seems to imply that some typologies serve no useful purpose. Usefulness, however, is very much in the eye of the beholder. As individuals we constantly encounter typologies that are of no use to us in our own work; for example, WYA has never found a bead typology that adequately describes the beads from his own excavations in Nubia. It is evident nevertheless that the many existing Nubian and Egyptian bead typologies (e.g. Reisner 1923b: 106–27; Steindorff 1935: 46–50; Emery 1938: pls. 43–4) served the practical needs of their makers, in that they permitted the description of a large and diverse body of material within a limited number of pages (cf. Everitt 1974:4).

If a typology appears to serve no obvious purpose, it probably means only that the typologist neglected to specify what his purpose was. This seems to be a fairly common failing (cf. Gardin 1980: 81). It may also be that the typology-maker was himself not consciously aware of his purpose, something that is perhaps also common (cf. Klejn 1982: 51–4). But we doubt that there are any typologies for which the maker had no purpose at all, either consciously or unconsciously.

We observed in the last chapter that the immediate reason for sorting entities is usually, though not always, to permit some kind of enumerative or statistical comparison between types. This in itself is an intermediate step toward the more fundamental goal of determining relationships between types. However, as we saw in Chapter 7, such relationships cannot be expressed within the structure of basic (one-level) typologies, because of the principles of equivalence of types and equidistance of types. In order to express relationships it is necessary to subject types to some kind of secondary ordering which is independent of the type concepts themselves (cf. Dunnell 1971b: 83).

The two most common procedures in secondary ordering are those of taxonomy and sedation. The former is a process of hierarchic clustering, and the latter of linear ordering. Both taxonomic ordering and seriation are common in archaeological classification, but we have delayed consideration of them until this point because, in the great majority of cases, they are incidental rather than essential features of the classificatory system.

Taxonomy

The term taxonomy, like other words relating to classification, has no generally accepted, precise definition. For many systematists, especially in the biological sciences, it is synonymous with classification itself (e.g. Mayr 1942: 3–17; Sokal and Sneath 1963; see also Brew 1946: 44–66). Our usage, however, is more restricted.

In spite of its massiveness, I have insisted that Ceramic Industries of Medieval Nubia (WYA 1986a) is not and cannot be the “last word” on Nubian pottery; it is merely the latest and most detailed in a continuing series of progress reports. The ongoing excavations at Qasr Ibrim are sure to contribute new information, and possibly also new wares, and hopefully future excavations in the Sudan will do the same. It is still also possible that some wares will be eliminated, since the significance of some of my finer distinctions remains suspect. I will here briefly recapitulate the scheme as it exists at the moment, without in any way suggesting that it has attained its final form.

The Nubian Typology in 1989

The Nubian Pottery Classification for the period A.D. 200–1600 today contains 101 wares, grouped into the same seven families and sixteen ware groups that were mentioned in the last chapter. The scheme bears a close but not total resemblance to the familiar type-variety system employed in the American Southwest and Mesoamerica (Wheat, Gifford, and Wasley 1958; Smith, Willey, and Gifford 1960; Sabloff and Smith 1969). The complete listing of wares, ware groups, and families is given in Table 5. Each ware has been formally described in terms of seventy variable features, as listed in Table 6 (WYA 1986a: 411–597).

For better or worse, this book represents something of an experiment. It is, so far as we know, one of the first attempts to achieve a genuine, two-way dialogue between archaeology and philosophy – the respective fields of its two authors. (For a predecessor – not concerned specifically with classification – see Kelley and Hanen 1988.)

It has been a common practice in recent years for archaeologists to borrow concepts and analytical tools from the philosophy of science, and apply them to their own field of endeavor (cf. Fritz and Plog 1970; Martin 1971; Watson, LeBlanc, and Redman 1971; Hill and Evans 1972; Kelley and Hanen 1988). However, this usage has often been both selective and uncritical. That is, the archaeologists have borrowed certain tools while ignoring others, and they have frequently overlooked the fact that the concepts they borrowed were the subject of controversy within the field of philosophy itself (cf. Kelley and Hanen 1988; WYA n.d. 2). Moreover, and more importantly, communication was strictly in one direction: from philosophy to archaeology.

In the present book we have tried to initiate a more genuine dialogue. We believe that, if the archaeologist has much to learn from the philosopher of science, the reverse is necessarily also true. In addition to applying philosophical concepts to the practice of archaeology, therefore, we have tried at various points to show how the practices of the field archaeologist reflect upon current controversies and issues in the philosophy of science.

The relationship between production activities and material residue is neither simple nor self-evident. Linking the two requires research conducted in a contemporary setting. These connections cannot be established directly from the archaeological evidence; those material patterns are the very phenomena we are attempting to decipher. Analyses of production decisions, and their material implications, must be established through research that controls for both behavior and the material record.

At the same time, archaeologists cannot simply rely on ethnographic anecdotes and interpretations through analogy. As noted previously, there is a distinction between learning by experience and learning from experience. The discussion in Chapter 5 suggested that inordinate emphasis has gone toward the former. This section considers how pottery making within Los Tuxtlas may assist with the latter.

This chapter explores ways to link variability in the organization of ceramic production with the material patterns of manufacturing activities. The emphasis on both the material technology and behavioral components of production allows the archaeologist to begin considering the potential contribution of these factors to the archaeological consequences of production. This perspective represents a first step toward developing a middle-range theory of ceramic manufacture.

This discussion makes the point that the material technology associated with a given production system can be as much a function of internal conditions as external variability. In addition to such considerations as the natural environment and/or market demand, the potter must also select production techniques that are suited to the specific characteristics of the context of manufacture.

As we saw in Chapter 4, any classification is basically a set of categories. In scientific classifications, the key to formulating such categories is selectivity. That is, scientific classifications are always based on a consideration of some features but not of others, a selection of certain variables and attributes from a wider field of possibilities. What is involved, as Foucault (1973: 132–3) observes, is a deliberate narrowing of the scientist's field of vision, so that certain kinds or domains of information are systematically excluded; for example, color in the case of most stone tool types. It was the introduction of this factor of selectivity which made possible the pioneering classifications of Linnaeus (1735) and other natural scientists of the eighteenth century, whereas their predecessors had aimed at all-inclusive and non-selective descriptions of each biological species that in effect made classification impossible (cf. Foucault 1973: 125–65).

In the last chapter we saw how different purposes affect the choice of features to be considered and not to be considered in making typologies (cf. also Whallon 1982). Here we want to take a closer look at the features themselves. We will categorize them under three headings: invariants, variables, and attributes. Throughout this chapter it is important to bear in mind our distinction between type concepts and type members (see Chapter 3), and to realize that we are talking about the attributes of types (i.e. concepts), and not necessarily about the attributes of individual objects.

The two previous chapters were concerned, respectively, with the classificatory and with the descriptive procedures involved in the making of the Nubian Typology. It remains here to consider the recording procedures that are involved in its use; that is, in the practical application of the typology to the study of actual ceramic material. While the classificatory and descriptive procedures are, as we saw, rather complex and detailed, we will observe here that the recording procedures are correspondingly simple.

The word “correspondingly” is used advisedly, for there is a negative relationship between descriptive and recording complexity. That is, the more detailed and specific is the information encoded in type descriptions, the less such information need be recorded for individual specimens. The mere assignment of a sherd or vessel to Ware W14 (see Appendix B) is enough to describe its method of construction, fabric, surface finish, and colors, as well as to indicate the range of possible vessel forms and decorative designs.

The material to be studied

The archaeologist typically finds pottery in two quite different forms: as sherds (fragments) and as complete vessels. The differences between these two kinds of ceramic remains are more than just a matter of completeness; they involve also the kinds of information that can be derived from the two sources. Whole pots are usually found in some place where they were intended to be, while sherds usually are not. They have come to their final resting places through successive accidents of breakage, discard, and possibly redeposition.

In Chapter 3 we discussed the physical, mental, and representational elements that go into the making of a type. Of these it is the mental elements that are the most problematical, and also, surprisingly, the least considered in the literature on typologies. As Lakoff (1984: 22) observes, “the classic theory of categories is not a cognitive theory. It is a theory about things as they are in the world, not about how the mind makes sense of the world.” He goes on to remark: “I do not know why the … theory of categories should be an Aristotelian one. But it is an issue that cognitive scientists need to address” (ibid.: 26).

The mental dimensions of typehood will be our concern in the next two chapters. In the present chapter we will discuss the relationship between types and other, related mental constructs: concepts, categories, and classes. Our purpose here is not to develop a theory of knowledge or of classification, but simply to establish working definitions for a great many terms that we will employ later in the book. Much of what we say here is not critical to the arguments that we will present later in regard to the use of types; our purpose is simply to make our meanings clear. We must stress once again that our definitions do not necessarily coincide with those of other authors (see Chapter 1).

Types and scientific thought

Science is defined more by its purposes (ends) than by its methods (means).

In the last chapter we applied certain general considerations abut theory and its use to the understanding of how archaeological typologies are developed and used. Here we will reverse the procedure, bringing our findings about archaeological typologies to bear on general issues of scientific theory and method. In particular we will consider their relevance to the currently controversial issue of scientific progress. This issue arises because the “official” or orthodox epistemology of most of the sciences is a vague empiricism, and it is associated with a special conception of scientific progress against which Thomas Kuhn (1962; 1970) and his followers have launched a powerful attack. The empiricist paradigm will be explained in more detail in the next section, following which we will bring some of our own typological considerations to bear in criticizing Kuhn's critique, as well as in commenting on some related issues, and most particularly on the “problem of pseudoproblems” (Carnap 1927). In a short chapter we can hardly offer definitive resolutions to these very general and very fundamental questions, but we feel that insights drawn from our study of archaeological typologies are suggestive enough to warrant remark. We will start with a sketch of “orthodox empiricism” and the conceptions of scientific progress, and of scientific concept formation, that are implicit in it.

Actualistic studies can be notoriously difficult to apply archaeologically. In most cases, the need for middle-range research evolved out of ambiguity in archaeological material that had already been recovered. Since the results of middle-range research have usually followed on the heels of archaeological fieldwork, it is hardly surprising to find that potentially pertinent archaeological data may not have been collected, or the collection procedures were not congruent with the findings of actualistic studies.

This fact, however, should not dampen the applicability of middle-range research to archaeological material. In addition to identifying new methods for conceptualizing and retrieving information, actualistic studies must be sufficiently flexible to expand the information potential of a pre-existing archaeological data base. The possible information inherent in archaeological remains already recovered cannot be overlooked.

This chapter applies some of the spatial concepts developed in this study to a recently documented Mesoamerican ceramic production system. To be useful, a concern with the spatial organization of pottery making should contribute at two levels of analysis. First, this focus should be able to differentiate between various production entities and supply information on the character of ceramic manufacture. Second, the principles of organizational structure should enable the archaeologists to monitor variability within a given production mode. As noted previously, a static approach to organization is one of the main weaknesses in the conventional model of pottery production (Rice 1984a:233; van der Leeuw 1984:720).