2006 Can There Be A Philosophy of Archaeology?

CAN THERE BE A PHILOSOPHY OF ARCHAEOLOGY? Blank Page CAN THERE BE A PHILOSOPHY OF ARCHAEOLOGY? PROCESSUAL ARCHAEOLOGY AND THE PHILOSOPHY OF SCIENCE BY WILLIAM HARVEY KRIEGER iii Copyright Page Contents Figures and Photographs vii Preface ix Acknowledgments xi Introduction 1 1 Logical Positivism and Scientific Explanation In the Beginning: Explanation and the Vienna Circle The Nature of Hempelian Explanation Explanation and the Social Sciences 5 5 14 21 2 The New Archaeology The Origins of New Archaeology Archaeology: The Need for Change The Aims, Features, and Methods of New Archaeology 31 33 37 39 3 The New Archaeology’s New Archaeologists Introducing the New Archaeologists The New Archaeology in Practice: Problems and Issues Philosophical Accounts of Archaeological Explanation 47 47 62 66 4 Philosophy and Archaeology: Post-Positivism What the New Archaeology Could and Could Not Do Laws and the Philosophy of Science Disunified Science and the Future of Archaeology 69 69 74 95 vi Contents 5 Philosophical Problems: Archaeological Responses What Can Archaeology Do for Philosophers of Science? Realism, Antirealism, and Archaeological Entities Science, Values, and Archaeology A Call for Re-evaluation 99 99 99 110 118 6 Future Studies: Archaeological Explanation and the Philosophy of Science Where Do We Go From Here? Theory, Geography, and the (Dis)unity of Science Moving Forward 121 121 121 129 Bibliography 133 Index 143 About the Author 147 Figures and Photographs Fig. 1. Philosophical/Archaeological Timeline (Not To Scale) Fig. 2. The Form of Explanations, According to Hempel and Oppenheim Fig. 3. Death’s Head, Cherub, Urn and Willow Fig. 4. Philistine Bird or Eating Crow? Fig. 5. Pot or Rock? Fig. 6. A Hypothetical Tel Fig. 7. (Solomonic) 6 Chamber City Gates (not to scale) Fig. 8. The Logical Positivistic Explanatory Model of Science Fig. 9. A Modified Version of Wylie’s Explanatory Model of Science Fig. 10. Puako Petroglyph Archaeological District, Hawai’i Fig. 11. Tell el-Far’ah (South), viewed from wadi Besor viii 15 71 90 100 111 112 116 117 124 125 Philosophy of Science American Archaeology Syro-Palestinian Archaeology 1890 Petrie: Tel el Hesi (brings stratigraphic archaeology and pottery seriation to the region) Vienna Circle: first meeting 1922 Hempel: “The Function of General Laws in History” 1942 Hempel and Oppenheim: “Studies in the Logic of Explanation” (introduces covering law models of explanation) 1948 Taylor: “A Study of Archaeology” (early call for a more scientific approach to archaeology) 1952 Kenyon: Tel Jericho (introduces Wheeler-Kenyon excavation method and radiocarbon dating) 1962 Binford: “Archaeology as Anthropology” (introduces his new archeology) 1964 Wright: Tel Gezer (1964-1965) (testing ground for new archaeology in Israel) Hempel: “Deductive-Nomological vs Statistical Explanation” and “Explanation in Science and in History” Scriven: “Explanations, Predictions, and Laws” (systematic critique of Hempelian Explanation) Hempel: Aspects of Scientific Explanation Kuhn: The Structure of Scientific Revolutions 1965 1967 Binford: “Smudge Pits and Hide Smoking” 1968 Hill: Broken K Pueblo Binford: “Post-Pleistocene Adaptations” (introduces covering laws to archaeology) 1969 Munson: “Comments on Binford’s ‘Smudge Pits and Hide Smoking’” 1970 1972 Binford: “Archaeological Reasoning and Smudge Pits – Revisited” 1981 Gitin and Dothan: Tel Miqne (1981-1996) (classic new archaeology in Israel) Salmon: Explanation and Archaeology 1982 Cartwright: How the Laws of Physics Lie 1983 van Fraassen: Laws and Symmetry 1989 Fig. 1. Philosophical/Archaeological Timeline (Not To Scale) Preface When I was a graduate student, I began my work as a field volunteer at Tel Miqne/Ekron. During my years and training at that site, I spent many a hot afternoon discussing my interests with my dig directors Sy Gitin and Trude Dothan. During one such conversation, Sy asked me what I would write my first book on. I told him that I wanted to write on philosophical explanation and archaeological theory. He leaned back in his chair, and said: “ Do you really think there is enough there for a book?” In truth, archaeologists, from the most theoretically self aware to the classical “dirt archaeologists” engage in issues surrounding philosophical explanation. The rationale behind digging at a particular site, the utility of certain objects as sources of data, the interpretation of those data, and the goals of the archaeological enterprise are all governed by a range of theoretical decisions, some more explicit than others. Of course, this was not news to Sy (I am pretty sure that he just asked the question to have a bit of fun with me), but his question is one that I have heard before and since, as archaeologists have had a tumultuous relationship with philosophers of science In the 1960s, archaeologists en masse were voicing dissatisfaction with the archaeological status quo. Archaeologists reached out to philosophy as a way to change their discipline from a sub discipline of history to a scientific enterprise. Rather than record static facts as historians, archaeologists wanted to study fluid processes as scientists. As Hempelian explanation, where an event is explained when it is subsumed under a law or law-like statement, showed promise as a way to recast archaeology in this manner, it was chosen as the theoretical base for what became known as processual, or “new archaeology.” Unfortunately, the new archaeology ran into a number of serious theoretical and methodological problems. Archaeology lacked the regularities that were thought to exist in sciences such as physics, so archaeological laws were difficult, if not impossible to define. In addition, equifinality, or under-determinism, made it impossible for archaeologists to choose between equally subsumable, albeit contradictory explanations. These issues were not limited to archaeology. Philosophers of science were already aware of the problems plaguing Hempelian explanation, and were searching for answers for to the continuing issues surrounding scientific explanation. In the end, although as a result of this first meeting, some archaeologists still shy away from anything too theoretical. Despite the feelings of those archaeologists who felt stung by their first contact with philosophy, there has been a lot of continued study and activity in subsequent years. A number of theoretically minded archaeologists and archaeologically inclined philosophers of science have been asking the questions brought up by the first marriage of archaeology and philosophy. Clearly, although that first marriage was not to be, it was the beginning of a beautiful friendship. In a manner mirroring other field sciences, archaeology, has kept deep connections with philosophy, much to the benefit of both fields. This book can be broken up into two major sections. The first three chapters are designed to provide a brief history of this exciting period, both from an archaeological and a philosophical perspective. The objective here is not to write a x Preface definitive anthropological study of this period. This has already been done by a number of authors. The goal here is to present a nuanced view of the many factors resulting in the first meeting of philosophers and the new archaeological scientists. As it is important to note the successes and failures of both the logical positivist movement, and its processual archaeological counterpart, each will receive a separate treatment. After a survey of the problems left in the wake of the new archaeology, the second half of the book asks and begins to answer questions about the future of archaeological and philosophical explanation. In short, the goal of this book is to show that, due to the questions both groups continue to ask about the nature of science, this is an incredibly exciting time to be a theoretically minded archaeologist or an archaeologically minded philosopher of science. Throughout the book, there are a series of philosophical and archaeological case studies designed to tease out these themes and show how archaeologists and philosophers of science have, either consciously, or unconsciously, worked to bridge the gap between the two disciplines. By reengaging with philosophy, archaeologists have shown that the failure of a particular type of theoretical model (here, new archaeology) should not be seen as a failure of philosophical archaeology. In fact, for archaeologists, ongoing questions about the future of the discipline mirror general debates going on about the future of philosophical explanation. The same sorts of questions that caused the fall of the new archaeology, forcing archaeologists to reexamine their scientific foundations are questions that every philosopher of science struggles with to date. At the same time, a growing number of philosophers of science have seen that they have much to gain from archaeology. As a trans-disciplinary science, archaeology puts an interesting spin on standard philosophical problems, as these case studies will show. For instance, archaeological practice can be seen as questioning the centrality of laws to scientific explanation, and raises questions about the unity of science. In addition, claims that scientists have to be realists in regard to theoretical entities may need to be rethought, due to antirealist tendencies found in some archaeological contexts. Finally, archaeology can be used to move beyond strictly “archaeological” or “philosophical” problems, and to explore inter-field issues that are the hallmarks of the field sciences. These final chapters use an archaeological perspective to focus on the relationship between science and values, and to raise questions about the relationship between the theory and practice of science. In summary, my hope is to encourage new theoreticians to enter the fray. Despite the growing cooperation between philosophers of science and archaeologists, there is still a lot of wheel reinventing going on in both disciplines. This work will provide concrete examples of how a study of problems central to archaeology can provide philosophers of science a new way to approach philosophical issues. At the same time, these examples should prove to even the most theoretically shy “dirt archaeologists” that a study of contemporary philosophy of science will strengthen the epistemological foundations of their relatively young science. Acknowledgments I would like to thank all of the people who helped me get this book to press. However, given the limitations of space, (not to mention my memory), this list will be incomplete. So, if I missed you, please accept my apologies, and feel free to add your name here__________________. (If you type it in Times New Roman, no one will even know the difference). The impetus for this book was my doctoral thesis. The goal of that project was to ask questions about the history and philosophy of archaeology, and in the course of that project, my committee suggested that the answers (or the road to those answers) would make an interesting book. I agreed (as I hope you will) and so I would like to thank Brian Keeley, Tammi Schneider, Steadman Upham, and John Vickers for their insights and advice in the formative stages of this work, and in the cases of Tammi and Brian, for their continuing help in the archaeological and philosophical development of this book throughout. Once this book got underway, it required a large amount of time and attention. I would like to acknowledge my home institutions, Tell el-Farah South and California State Polytechnic University, Pomona, for giving me the institutional, and at times financial support, needed to complete this project. In addition, much of this material was worked out in conversations with my colleagues and students (both in the classroom and in the dirt). In particular, I would like to thank Jeffrey Chadwick, a colleague and friend who has offered his advice and opinions over many a steak dinner off-site in Israel, and who has allowed me to use his drawing of a Hypothetical Tel in this book. For all of those who have helped me in this process, I have tried to cite you where appropriate, but in case I have forgotten to do so recently, I thank you for your help. Chapters 5 and 6 of the book are direct results of conference activity and other acts of scholarly exchange. In particular, I would like to thank the NEH, and organizers Peter Machamer and Sandra Mitchell for allowing me to participate in their 2003 summer institute in Science and Values, hosted by the University of Pittsburgh’s HPS program. Not only did the institute give me the opportunity to present and discuss my thoughts on the effect of the sci/values debate on archaeological explanation, it provided me with a network of friends and colleagues who have been sounding boards for my continued work in the field. In particular, I would like to acknowledge Alison Wylie, whose work in the field continues to be an inspiration to me. Her comments and feedback, both at the NEH and other conferences, and more recently in providing a wealth of suggestions (both structural and substantive) for the book, gave me the input necessary to make this work something that I am truly happy to present to you. Finally, I would like to thank the staff at Lexington Books. Not only did you make this book literally possible, your support throughout the process is much appreciated. Again, my thanks, Introduction The title of this book brings to the fore a couple of questions: 1. Is there (or has there ever been) a philosophy of archaeology? 2. What is the relationship between the philosophy of science and the sciences themselves, or in this case, archaeology as a science? 3. What do philosophers of science and archaeologists have to gain from continued (and expanded) contact and discourse? The first three chapters address the first question, and answer it affirmatively. There was a time when archaeologists and philosophers of science would agree that there was a philosophy of archaeology. In the early twentieth century, when archaeology as a discipline became fully self-aware, archaeologists recognized that if their field were to be more than a descriptive enterprise, it would need to look to the goals and nature of science, and determine whether archaeology could be made answerable to these types of criteria. Explanation, confirmation, and prediction are attributes of science that archaeologists found attractive, and they sought to integrate Hempelian philosophy of science, the prevalent model of the day, into their discipline. This philosophy of archaeology was called “new archaeology,” for a number of reasons that will become clear below and in the following chapters. However, it is important to note that Hempel’s models were as groundbreaking for philosophers of science as they would be some 20 years later, when they were incorporated into archaeology. Hempel’s models, introduced in the 1940s and expanded to include the social sciences in the 1960s, were seen as a considerable advancement for philosophers of science. For Hempel, an explanation is a special type of argument in which an event is subsumed under a law or law-like statement. When an event’s prior conditions and the appropriate laws are known, the event is said to be explained. Hempel’s models, applied first to deterministic laws and later to statistical laws, gave a method by which explanation and prediction would be possible. Archaeologists, including Lewis Binford, were interested in Hempel’s models as a means to set archaeology on a new and sound foundation. In 1962, Binford’s “Archaeology as Anthropology,” and much of his subsequent work, showed that a Hempelian style of archaeology was not just a theoretical goal, but that it could be put into practice.1 For Hempelian explanation to work for archaeologists, archaeological data would need to be subsumable under some sort of general laws. If this could be done, archaeologists could explain past events by interpreting excavated remains. At the end of an excavation, the data and archaeological laws would be combined to reconstruct and explain history. This process would transform archaeology from a descriptive historical enterprise into a science. Instead of merely cataloguing dates and places, archaeology would focus on the processes of change, and archaeologists would thus begin to understand what really went on in the past. Along with this theoretical shift came a name change. Archaeologists would no longer be called culture historians. Instead, scientific archaeology would be practiced by new archaeologists. 2 Introduction Archaeology’s quest to become a science led to changes in the practice of archaeology that were as radical as were the theoretical changes. These changes included the creation of relative chronologies (from ceramic and architectural analysis), absolute chronologies (from dendrochronology, and from carbon-14 and neutron activation dating), and the collection of entirely new sets or types of data (including micro- and macro-floral and faunal remains, coprolites, and both land- and ice-cores). New archaeologists believed that this methodology would provide the seed data needed to start proposing potential laws. However, shortly after archaeological theorists turned to Hempel’s models, a number of problems surfaced. At first, it was thought that these problems might be the fault of archaeology, as Hempelian explanations could not produce laws archaeologists could use in the field. While Hempelian general laws might explain why water boils at 373° K, it does not explain why a particular pot sherd appears in a given context, much less the connection between that sherd and the person who used it. In addition, equifinality (the archaeological term for underdeterminism) was as much a problem to these new archaeologists as it was to their predecessors. These problems were not limited to archaeology. Philosophers of science were fully aware that Hempel’s theory of explanation, however groundbreaking, was controversial at best. In fact, by the time archaeologists such as Binford (1962) appropriated Hempel’s models, philosophers had already recognized as crucial the very problems that archaeologists would eventually come to face, and many had moved beyond Hempelian explanation. Archaeologists, generally unaware of the philosophical literature, were left in the dark, and in the end, saw the edifice built on Hempel’s model start to crumble around them. The abandonment of the new archaeology marked the end of a unified “philosophy of archaeology.” Initially, many archaeologists believed that philosophy in general, and not just Hempelian models specifically, failed to make good on its promise to provide a model of explanation and prediction that could transform archaeology from a cataloging enterprise to a modern scientific discipline. Philosophers of science, who have a clear sense of the changing and open-ended nature of philosophical models, retorted that archaeologists were naïve to adopt the Hempelian model with the hope that it would solve all of their problems. After the “fall” of new archaeology (also known as processual archaeology, due to its focus on changing cultural processes), a number of archaeological schools developed, each believing itself to be the successor to new archaeology. To date, these divergent, post-processual schools continue to battle for dominance, fighting each other, the remnants of the processual school, and a group calling itself “dirt archaeologists,” who believe they only need hold to the strict empiricism required for fieldwork. Interestingly, all of these competing theoretical movements have accepted the methodological changes brought on by new archaeology, while dismissing any relevant theoretical connections. Post-processualists use the data gathered by processualists in order to disprove the latters’ theories, and dirt archaeologists have claimed methodological and laboratory advances as their own as well. However, regardless of changes in the way that archaeology was done in the Introduction 3 field, the theoretical side of the first cooperative archaeological and philosophical effort failed to satisfy either archaeologists or philosophers of science. In failing, however, both fields have made significant progress. Archaeologists have a much more sophisticated understanding of the problems surrounding explanation than they did as culture historians, and philosophers of science have learned quite a bit about scientific practice since Hempel’s time. In the past couple of decades, a number of philosophers of science and theoretically minded archaeologists have (until recently, quietly) continued to ask questions about the status of archaeology as a science, and it is this work that leads to the second and third questions posed above. It is the view of this project not only archaeologists, but also philosophers of science have quite a bit to gain from a heightened dialogue, a topic explored in the second half of this work. This trans-disciplinary project will contribute to both philosophy of science and to archaeology, seeking to show the benefits of opening the communication between fields to the larger archaeological and philosophical communities. A number of archaeologists, stung by the failures of the new archaeology, do not realize that issues they are considering are fundamentally philosophical issues. Sharing between the two fields, will be nothing but positive for archaeologists who want tools to help put the discipline back on the road forward, wherever it may lead them. At the same time, some philosophers of science do not see the benefits of studying and testing theories on this relatively new science. Philosophers of science have been working with scientific models for decades, and have questions about a variety of issues, including the relevance of laws to explanation, the unity of science, and scientific realism and antirealism. Archaeology, as a relatively young science, is struggling with these same issues, and philosophers of science, en masse, would have a lot to gain by studying archaeology as it tries to find its place among the sciences. Archaeology shares the methodology of the natural sciences and the interpretive problems of the social sciences. More philosophers should use this transitional science as a laboratory, helping archaeologists work through issues relevant to philosophy, and at the same time gaining insight into how science (or at least this science) develops and changes over time. Since the answers to question two and the resultant contemporary worries that make up question three will continue to shape both the philosophy of science and the philosophies of archaeology, this work will conclude with the investigation of a few examples, meant to point the way forward. Chapter 5 will be a study of entity realism/antirealism, a philosophical problem with interesting archaeological implications. Chapter 6 lays out possible directions for future research, showing how problems illustrated in the preceding chapters might find their solutions, and pointing toward other interesting problems that archaeologists and philosophers of science are just beginning to recognize as significant. The final goal of this book is to show the benefits of new cooperative relationship between philosophers and archaeologists, a relationship that promises to provide for an exciting future for both fields. 4 Introduction Notes 1. Actually, there is some question as to the development of Binford’s program and its connection to Hempel’s views on explanation. Recently, a correspondence between Hempel and Binford has surfaced. As this material is catalogued, it is my hope that we will have a better understanding of the formation of Binford’s new archaeology. Chapter 1 Logical Positivism and Scientific Explanation The Vienna Circle (“Der Wiener Kreis” in the original German) was the name given by Neurath in 1929 to a group of scientists, mathematicians, and philosophers who, through their association, came to share a general conception of science. Although differences in goals, methodology, political factors, and the death or dispersal of some of its members would eventually lead to the dissolution of this group, its influence continues to be felt by philosophers of science, and by scientists in general. A number of features proved especially relevant to the scientific outlook of the Vienna Circle, including a belief in science as an unified enterprise, a desire to eliminate metaphysics, and a need for a new definition of philosophy. As the pursuit of these goals has had a significant impact on archaeological explanation, an understanding of these theses will be needed in order to understand their effects on contemporary archaeology.2 This chapter will provide philosophical background material about this rich period, focusing on the goals of the Vienna Circle and on Hempel’s reconstruction (called logical empiricism). In addition, it will introduce and explore the covering law models of explanation, relating this form of explanation to both the physical sciences, and the social sciences, including both history and archaeology. The role of history here is important for a two reasons. First, Hempel focused on the study of history, using it as the template for the introduction of the social sciences to his covering law models. Second, there are many methodological similarities between history and archaeology. The chapter will conclude with a discussion of problems with Hempel’s models. Some critiques are endemic to logical empiricism and to Hempel’s models of explanation, and others are specific to the application of those models to transitional sciences like archaeology. Only once this discussion is done will a proper explanation of archaeology be possible, and this will be the task of the following chapters. In the Beginning: Explanation and the Vienna Circle In a pamphlet entitled Wissenschaftliche Weltauffassung (The Scientific Conception of the World), the members of the Vienna Circle3 stated that they had discovered the proper goals and methods of scientific explanation: We have characterized the scientific world-conception essentially by two features. First, it is empiricist and positivist: there is knowledge only from experience, which rests on what is immediately given. This sets the limits for the content of legitimate science. Second, the scientific world-conception is marked by application of a certain method, namely logical analysis. The aim of scientific effort is to reach the goal, unified science, by applying logical analysis to the empirical material. (Der Wiener Kreis, 331, emphasis in the original) 6 Chapter 1 Science, according to the Circle, would be based on experience (as opposed to a priori generalizations, guesswork, witchcraft, and theology), and further, it would progress by means of deductive reasoning, with the goal of eventually unifying science. By reducing science to these key components, the Circle believed they could revitalize and clean up scientific discourse. Their project would lead to both the elimination of pseudo-scientific nonsense, and to the simplification and eventual unification of all science. The elimination of nonsense was exactly what the Vienna Circle had in mind when it set out to finally dispense with metaphysics. In setting epistemology up as the study of empirical facts by means of logic, the goal was to gain strict verification of the objects of knowledge. The Circle’s members did this by redefining the objects of knowledge in terms of sense events and rules of logic. This limited worldview was seen by the Circle as a strength of the program, as they were prepared to answer with certainty any question that could be parsed into terms representing experienced events and logical propositions: Questions regarding the “validity and limits of knowledge” disappear. Everything is knowable which can be expressed, and this is the total subject matter concerning which meaningful questions can be raised. There are consequently no questions which are in principle unanswerable, no problems which are in principle insoluble. (Schlick 1959, 55-56) For the logical positivists and their sires, the logical empiricists, everything is knowable (at least in principle) so long as the question can be asked in an appropriate manner. However, this methodology is more than a matter of logical etiquette, as strict verification, a stated initial goal of the Vienna Circle, was possible only under these conditions. Verification and Confirmation A statement is verified if shown to be true, verifiable if it could be shown to be true, and, similarly, falsifiable if it could be shown to be false. The verification criterion of meaning claims that only verifiable and falsifiable statements are (empirically) meaningful. The criteria for verification originally limited philosophers to either directly observable data or a priori truths. As this focus on verification would mean that much of science would become meaningless (as many truths of science are not based on direct observation), the verification principle was modified by Ayer: I require of an empirical hypothesis, not indeed that it should be conclusively verifiable, but that some possible sense experience should be relevant to the determination of its truth or falsehood. If a putative proposition fails to satisfy this principle, and is not a tautology, then I hold that it is metaphysical, and that, being metaphysical, it is neither true nor false but literally senseless. (Ayer 1936, 31) Logical Positivism and Scientific Explanation 7 According to Ayer, anything not verifiable, at least in principle, would not be considered an instance of knowledge. As a result, metaphysics, which the logical positivists took to mean the vast majority of philosophical endeavors up to this point, could finally be dismissed. As metaphysical statements purport to deal with essences and other insensible qualities, the logical positivists determined that the whole of metaphysics was no more than a series of pseudosentences and linguistic errors: What have been considered such up to now are not genuine questions, but meaningless sequences of words. To be sure, they look like questions from the outside, since they seem to satisfy the customary rules of grammar, but in truth they consist of empty sounds, because they transgress the profound inner rules of logical syntax discovered by the new analysis. (Schlick 1959, 56) For the logical positivists, if a question could not be relieved of its metaphysical baggage, it was not a question at all. Questions such as “Does G-D exist?” long considered to be problematic for a number of reasons, were now shown not merely to be confused, but to be literally meaning-less. These sentences use words that appear to have meaning, but upon close inspection, they are vacuous. Here (according to the logical positivists, taking their cue from Kant), the word “exists” provides no useful (or meaningful) information to the reader. Were someone to ask whether G-D were blue, that would be a meaningful question, but the substitution of meaningless words (such as “exists”) in place of proper predicates (such as “is blue”) makes for a sentence that sounds proper, but to the logical positivists, is merely confused. Unfortunately, there are a number of problems with verification. As Carnap (and others) pointed out, verification is for the most part unattainable: If verification is understood as a complete and definitive establishment of truth then a universal sentence, e.g. a so-called law of physics or biology, can never be verified, a fact which has often been remarked. Even if each single instance of the law were supposed to be verifiable, the number of instances to which the law refers—e.g. the space-time points—is infinite and therefore can never be exhausted by our observations which are always finite in number. We cannot verify the law, but we can test it by testing its single instances i.e. the particular sentences which we derive from the law and from other sentences established previously. If in the continued series of such testing experiments no negative instance is found but the number of positive instances increases then our confidence in the law will grow step by step. Thus, instead of verification, we may speak here of gradually increasing confirmation of the law. (Carnap 1936, 30, emphasis in the original) Since a law cannot ever be completely verified, Carnap stopped speaking of verification, and instead focused on confirmation. A law would be confirmed (gradually) by means of testing. The more instances that conformed to the law, the more that law would be confirmed. As confirmation seemed a better and more attainable goal than verification (one of the goals of logical positivism), 8 Chapter 1 the logical empiricists used confirmation to set themselves apart from the logical positivists. As opposed to verification, Carnap believed that confirmation was possible, and that it would work by means of an incremental approach: “To be sure, in many cases we reach a practically sufficient certainty after a small number of positive instances, and then we stop experimenting. But there is always the theoretical possibility of continuing to series of test-observations. Therefore here also no complete verification is possible but only a process of gradually increasing confirmation” (Ibid., 31, emphasis in the original). Unfortunately for Carnap, confirmation has problems of its own. Despite Carnap’s assertion that incremental confirmation is attainable, logic would seem to prove otherwise. According to logicians, any system based on incremental confirmation commits a basic fallacy, called affirming the consequent. For instance, if a scientist wants to show that a certain substance is acidic (A) by seeing if it turns litmus paper red (R), the experiment would take the following form: If it is acidic, then it will turn litmus paper red. The paper was turned red. It is acidic. If A then R or R A The conclusion, seemingly confirmed by the experiment, does not follow, as is plain by simply substituting in other terms (which would be a perfectly legal move for a valid argument). If the experimenter wanted to know if a room was engulfed in fire (A), given the presence of smoke (R), the experiment would be as follows: If there’s fire, then there’s smoke. There’s smoke. There’s fire. If A then R or R A As there are many situations (such as dry ice and steam) where smoke exists without fire, logic shows that this form of argument is invalid, and therefore fails to provide confirmation. According to Popper (1953), good hypothesis testing can be used to disconfirm hypotheses. However, there is nothing in the logical form of explanations that provides any degree of confirmation. So, according to Popper, either confirmation is logically impossible, or there is something to confirmation that is not captured by logic. As either option would be unacceptable to a logical empiricist, this problem continues to be a problem for scientists interested in confirmation. Unification and Reduction Setting aside the problems with verification and confirmation, which were well known at the time and continued to be the subject of debate by logical posi- Logical Positivism and Scientific Explanation 9 tivists and logical empiricists, the Vienna Circle’s program hinged on a desire for the unification of science and further (and more controversially), on the reduction of all science into a single enterprise or discipline. For the logical positivists, the quest for all scientific truth amounts to a singular attitude, or way of thinking about science: The scientific world conception is characterized not so much by theses of its own, but rather by its basic attitude, its points of view and direction of research. The goal ahead is unified science. The endeavour is to link and harmonise the achievements of individual investigators in their various fields of science. From this aim follows the emphasis on collective efforts, and also the emphasis on what can be grasped intersubjectively. (Der Wiener Kreis, 327-8, emphasis in the original) A principal aim of the Vienna Circle was to unify the sciences. So the goals, processes, and conclusions of all science should be examined, and brought into line with the verificationist principles stated above. Once that was achieved, science would move forward as a unified program. The second thesis, reduction, was seen to follow from unification by the members of the Vienna Circle, in much the same way that Frege, Russell, and others attempted a reduction of mathematics to logic: When the mathematicians completed the reduction of the calculus to arithmetic, the philosophers took over… Frege and Russell argued that arithmetic itself was reducible to logic… Whitehead and Russell were the first to see in this reductionist episode a model for epistemology. The idea was to do with all our knowledge claims about the empirical world what logicism had tried to do with mathematics... (Coffa 1991, 211) Although there were still a number of problems in the reduction of all mathematics to logic, the Vienna Circle saw a real opportunity to reduce science in a similar way down to its purest, most basic form. Stripped of metaphysics and brought down to bare experience and logic, science would finally be able to speak in terms of certainty about its objects. Translated into scientific language, this reduction would amount to preserving only a set of axioms (a sort of minimal instruction set) and a collection of statements that serve to connect these axioms to the empirical world: A system of axioms, cut loose from all empirical application, can at first be regarded as a system of implicit definitions; that is to say, the concepts that appear in the axioms are fixed, or as it were defined, not from their content but only from their mutual relations through the axioms. Such a system of axioms attains a meaning for reality only by the addition of further definitions, namely the ‘coordinating definitions’, which state what objects of reality are to be regarded as members of the system of axioms. (Der Wiener Kreis, 333-4) The original idea was to reduce science to a combination of logic and expressions of bare experience. Coordinating definitions moved these basic units 10 Chapter 1 from the realm of experience to that of language. Carnap, Schlick, and Neurath, in their famous controversy about knowledge and experience spoke in terms of “protocol statements.” These, for Carnap, were linguistic elements (not direct descriptions of experiences), as statements could be used in arguments. Regardless of the actual content of protocol statements, however, the logical positivists agreed that the knowledge content of these statements reduces to logic.4 Given that everything except experience (be it mediated or not) and logic had been dismissed as metaphysical, the activity of science would be greatly simplified. There would be only two sources of data for the entire scientific enterprise. Any change, growth, or further reduction could only come by manipulating one or both of these: “The development of empirical science, which is to represent reality by means of as uniform and simple a set of concepts and judgments as possible, can now proceed in one of two ways, as history shows. The changes imposed by new experience can be made either in the axioms or in the coordinating definitions” (Ibid., 334). This focus on scientific development highlights two of the largest problems addressed by logical positivists, decisions between potential axiom sets,5 and questions surrounding the form and content of coordinating definitions. In addition, even were it possible to use language to gain epistemological access to experience, there are questions about our ability to relate this information to others. Since experiences are personal, then it is difficult to imagine how people could objectively communicate (use assertions) to represent these intersubjective events. While internal debate crippled early efforts to move toward reduction, these issues would be taken up by both the later logical positivists and by their successors, the logical empiricists such as Hempel (as we will discuss below), and by others who used these issues to deny the unificationist project altogether.6 Regardless of these disputes, the members of the Vienna Circle agreed on the object of scientific reduction. Just as the purified form of mathematics was found in logic, the pure science would turn out to be the most basic of the sciences, physics: The methodological problem of the application of axiom systems to reality may in principle arise for any branch of science. That these investigations have thus far been fruitful almost solely for physics, however, can be understood from the present stage of historical development of science: in regard to precision and refinement of concepts, physics is far ahead of the other branches of science. (Ibid). Physics, as the most basic of the sciences, was taken to have made the most progress toward truth. The fundamental nature of physics (not to mention that the Vienna Circle was predominately made up of physicists and philosophers of physics) made it the most likely candidate for the ultimate science. As physics was observed by the group to be the most fundamental, and the most scientific of the sciences (due to its focus on experiments and universal laws) it was the logical choice to serve as a model for unified science: “For there is but one science, and wherever there is scientific investigation it proceeds ultimately ac- Logical Positivism and Scientific Explanation 11 cording to the same methods; only we see everything with the greatest of clarity in the case of physics, because it is the most advanced, neatest, most scientific of all the sciences” (Hahn 1959, 147). Once so defined, physics, and by extension all science, would be thought of in terms of the perceived strengths and focus of that fundamental science. The Vienna Circle’s principles of reduction were not meant to be limited to the natural sciences, presumed to be reducible to physics. Scientific unity and reduction would also extend to the whole continuum of scientific activity. Oppenheim and Putnam (1958) went so far as to categorize the sciences from the most complex to the most basic, to better facilitate reduction: 6………………….Social groups 5………………….(Multicellular) living things 4………………….Cells 3………………….Molecules 2………………….Atoms 1………………….Elementary particles Any whole which possesses a decomposition into parts all of which are on a given level, will be counted as also belonging to that level. Thus each level includes all higher levels. However, the highest level to which a thing belongs will be considered the “proper” level of that thing. This inclusion relation among our levels reflects the fact that scientific laws which apply to the things of a given level and to all combinations of those things also apply to all things of higher level. Thus a physicist, when he speaks about “all physical objects,” is also speaking about living things—but not qua living things. (Oppenheim and Putnam 1958, 409) Although Oppenheim and Putnam were well aware that there were problems to overcome before reduction to the most fundamental level could occur, they believed their work to have shown the value of such a program: We hope to have shown that…a tentative acceptance of this belief, and acceptance of it as a working hypothesis, justified, and that the hypothesis is credible, partly on methodological grounds (e.g., the simplicity of the hypothesis, as opposed to the bifurcation that trivial suppositions create in the conceptual system of science), and partly because there is really a large mass of direct and indirect evidence in its favor. (Ibid., 421, emphasis in the original) Given that their thesis was indeed both justified and credible, Oppenheim and Putnam believed that the partial reductions of contemporary science would eventually give way to complete reduction to the most basic level, once the sciences were fully understood. The members of the Vienna Circle, as well as Oppenheim and Putnam, believed that the only difference between sciences such as physics and chemistry and others such as history and economics was that the latter were less well developed: “As we have specially considered with respect to physics and mathematics, every branch of science is led to recognize that, sooner or later in its development, it must conduct an epistemological examination of its foundations, 12 Chapter 1 a logical analysis of its concepts. So too with the social sciences, and in the first place with history and economics” (Der Wiener Kreis, 337). To the logical positivists, there was no substantive difference between the goals and results of the individual sciences. Physics and chemistry had been focal points of study for hundreds of years. In that time, they had moved from the realm of superstition and magic to their current status as sciences. Once the social sciences matured, went through enough years of study, and analyzed enough data, they would be in a position to make a similar transition. History, as the social science that Hempel believed had the closest connections to the real sciences, was thought to be the next emergent field of scientific study.7 According to the strict reductionists, at a certain point, first the natural sciences and then the social sciences would lose their individual character and would take their places as a part of the larger unified scientific enterprise. Just as the physical sciences were thought to reduce to physics (as physics was thought to be the fundamental science) so would the social sciences lose their artificial boundaries, and similarly lead to more foundational principles and, finally, be reduced to physics itself. In subsequent years, a number of philosophers, including Jerry Fodor (1974) and Alan Garfinkel (1981), came to believe that reductionism should be dropped as a goal. Fodor questions the reductionist implication that there is a direct correspondence relationship between higher and lower level phenomena. For example, Fodor asks whether psychology could be reduced to neurological principles (which, by the way, would correspond to a reduction from level 5 to level 4, according to Oppenheim and Putnam): “Suppose the functional organization of the nervous system cross-cuts its neurological organization. Then the existence of psychology depends not on the fact that neurons are so depressingly small, but rather on the fact that neurology does not posit the kinds that psychology requires” (Fodor 1974, 439). According to Fodor, each science focuses on specific issues. The belief that the categories that are significant to physics will then correspond to those of the other sciences has not been proven, and to Fodor, is probably untenable: If science is to be unified, then all [of the disparate sciences’] taxonomies must apply to the same things. If physics is to be basic science, then each of these things had better be a physical thing. But it is not further required that the taxonomies which the special sciences employ must themselves reduce to the taxonomy of physics. It is not required, and it is probably not true. (Ibid., 440, emphasis in the original) Fodor’s view, that explanations in the special sciences may not be captured by reduction is taken further by Garfinkel (1981). Garfinkel believes that explanations proceed at many levels, and further, that lower level explanations are unable to explain higher level phenomena, just as higher level explanations are not suited to cover lower level phenomena. Garfinkel (1981, 451) asks whether human action can be reduced to neurophysiology: Logical Positivism and Scientific Explanation 13 Here reductionism says that the action of raising my arm “is just” the physical movement of the arm (the underlying state), together with its microexplanation (the neurophysiological causes of the movement of the arm). But the problem with such an identification is that we do not want an explanation of why my arm moved in exactly that way… My arm did not have to move in exactly that trajectory for it to have been the same action…If for example, the explanation of why my arm moved is that I was shaking hands with someone to whom I was being introduced, this explanation gives us what the underling neurophysiology does not… (Garfinkel 1981, 451-452, emphasis in the original) Although Garfinkel and Fodor give reasons to believe that strict reductionism should not be a goal of scientific inquiry, the weaker thesis of unification (especially those versions focusing on the methods of scientific inquiry) is still seen as important to some philosophers of science. Indeed, some of Wesley Salmon’s more recent work (1990) suggests the possibility of rapprochement between different types and levels of explanation, a point that will be discussed again in chapters 4, while unification on other levels is critiqued in chapter 6. Philosophy and the Vienna Circle As the Vienna Circle was made up of some of the greatest scientific and philosophical minds of its day, one might assume that as science continued to flourish, so would philosophy. However, according to the views of the logical positivists, philosophy would be changed radically. In fact, after the unification and reduction of all science, there would be no place for a field or science of philosophy: “Every science, (in so far as we take this word to refer to the content and not to the human arrangements for arriving at it) is a system of cognitions, that is, of true experiential statements. And the totality of sciences, including the statements of daily life, is the system of cognitions. There is, in addition to it, no domain of “philosophical” truths. Philosophy is not a system of statements; it is not a science” (Schlick 1959, 56). As a result, the logical positivists believed that philosophy was to be separated into two components. The first, consisting of the majority of philosophy, would be considered metaphysics, and would be discarded as meaningless. The remainder of philosophy was logic, and this activity alone would have a place in the new science: “We see in philosophy not a system of cognitions, but a system of acts; philosophy is that activity through which the meaning of statements is revealed or determined. By means of philosophy statements are explained, by means of science they are verified. The latter is concerned with the truth of statements, the former with what they actually mean” (Ibid., emphasis in the original). Philosophy (or logic, to be more precise) was the means to understand statements, and was considered an aid to science, but not a science in and of itself. Wittgenstein, an occasional friend of the Vienna Circle, believed that philosophy (as opposed to logic) was essentially nonsense, and was useful only as a 14 Chapter 1 tool. Once used, it was to be thrown away, not retained as an independent source of knowledge.8 Like Wittgenstein, the members of the Vienna Circle believed that philosophy’s only remaining merit would be as a tool: “I maintain that there is nothing in the nature of philosophy to warrant the existence of conflicting philosophical ‘schools’” (Ayer 1936, 32). According to the Vienna Circle, the ascendancy of logic meant that philosophy would survive only as a method. Philosophy would not have anything to add beyond its use as a tool. The Nature of Hempelian Explanation Hempel took seriously his commitment to the fundamentals of the Vienna Circle, as well as his understanding of the shortfalls of the original program. While he carried on many of the tenets of his teachers, he (and other logical empiricists) had some ideas that set him apart from the logical positivists. For example, Hempel was a realist in terms of theoretical entities (unobservables which are the apparent referents of theoretical terms). Logical positivists, as a whole, would be considered antirealists, as the existence of entities such as electrons is neither empirically verifiable nor logically necessary. Hempel, however, believed that there was a way to be a realist without falling into a metaphysical trap. Hempel saw these entities as being there to be found in principle, if not in practice, and his willingness to think in these terms set him apart from his mentors. Hempel’s belief was “…that a sentence makes a cognitively meaningful assertion, and thus can be said to be either true or false, only if it is either (1) analytic or self-contradictory or (2) capable, at least in principle, of experiential test” (Hempel 1959, 108).9 The Covering Law Models Hempel focused specifically on explanation, or answering the question “why did ‘X’ happen?” He did this by creating of a series of explanatory devices which he called “covering law models.” Although the models differed in a number of respects (see below), they had in common a number of basic features. Hempel and Oppenheim, in their famous and influential paper, “Studies in the Logic of Explanation,” define an explanation as being made up of two elements, called the explanandum and the explanans: By the explanandum, we understand the sentence describing the phenomenon to be explained (not that phenomenon itself); by the explanans, the class of those sentences which are adduced to account for the phenomenon…the explanans falls into two subclasses; one of these contain certain sentences C1, C2, …, Ck which state specific antecedent conditions; the other is a set of sentences L1, L2, …, Lr which represent general laws. (Hempel and Oppenheim 1948, 247) Logical Positivism and Scientific Explanation 15 Hempel and Oppenheim believed that the explanandum sentence is properly explained (by means of explanans sentences) by properly understanding the situation at hand and the laws that apply in that instance. Hempel believed that all scientific explanations have in common a focus on general laws. For this reason all of Hempel’s models have been designated as covering law models, as these events are explained when they are understood as being a product of, or covered by, those laws. Many people feel that an explanation does its job if it leads people to understand (psychologically) how or why an event occurred. For Hempel, this induced feeling of satisfaction was not sufficient to make something an explanation. Covering law explanations were defined in terms of sentences about an event to be explained (the explanandum), and certain facts about its occurrence and relevant laws for the given situation (the explanans). Only in the presence of those conditions would explanation be possible. When Hempel first proposed his covering law models, detractors (such as Mandelbaum 1959 and Dray 1957) objected that his models seemed to fly in the face of how explanations were done.10 Hempel did not dispute this claim. In fact, he agreed that rather than focusing on how science was done, scientists should focus on how, ideally, science should be done: “these models are not meant to describe how working scientists actually formulate their explanatory accounts. Their purpose is rather to indicate in reasonably precise terms the logical structure and the rationale of various ways in which empirical science answers explanation-seeking why-questions. The construction therefore involves some measure of abstraction and of logical schematization” (Hempel, 1965, 412). Hempel was aware that science was rarely, if ever, done in a manner that explicitly goes through the necessary steps to explain phenomena fully. However, he believed that his models were a somewhat abstract means to reach that goal. That initial roadmap, crafting scientific explanations as arguments, was what Hempel and Oppenheim (1948, 249) originally posited in the following explanatory scheme: C1, C2,…,Ck L1, L2,…,Lr Statements of antecedent conditions General laws Explanans Description of the empirical phenomenon to be explained Explanandum Logical Deduction E Fig. 2. The Form of Explanations, According to Hempel and Oppenheim 16 Chapter 1 This initial model would later be expanded by Hempel into a number of different types, but the idea that explanations are deductively derived from an understanding of initial conditions and general laws would be the central feature in all of them. Although Hempel’s models had in common a reliance on covering laws, the nature of these laws could be either deterministic or statistical, and his explanations could likewise either be certain or probable. The three explanation types are known as the Deductive-Nomological (D-N) model, the DeductiveStatistical (D-S) model, and the Inductive-Statistical (I-S) model. Given the common reliance on covering laws, there are significant differences between these models. In fact, the evolution of these models was dialectical. D-S models were developed in response to the shift from verificationism to the weaker criteria depending on confirmation, and both the D-S and I-S models were posited after critiques of the original D-N model. By focusing on Hempel’s models, it will be possible both to understand the changing criteria Hempel thought crucial for proper explanations and to see how these models fared, given such criteria. The Deductive-Nomological Model In the Aspects of Scientific Explanation, Hempel (1965, 337) explained the D-N Model: “If we imagine the various explicit or tacit explanatory assumptions to be fully stated, then the explanation may be conceived as a deductive argument of the form” C, C2, …, Ck (D-N) Explanans S L1, L2, …, Lr E Explanandum-sentence The D-N model states that an event (E) is explained, given a set of explanatory sentences (S), formed by initial conditions (C) and appropriate general laws (L). This model would serve as the paradigm for Hempelian explanation. Here, so long as the appropriate laws and circumstances could be understood, the hope was that events could be fully explained. For example, a scientist might wonder why, from within a rowboat, the parts of the oars that are underwater appear to be bent upwards, Hempel and Oppenheim believe that this explanation is really just a matter of deductive reasoning, based on general laws “mainly the law of refraction and the law that water is an optically denser medium than air” (Hempel and Oppenheim 1948, 246), and on certain known facts “especially the facts that part of the oar is in the water, part in the air, and that the oar is practically a straight piece of wood” (Ibid). To Hempel and Oppenheim, when scientists ask why an event occurs, they could just as easily ask about which general laws and which initial conditions resulted in the event in question. Essentially, asking for the former is the same as working through the latter. These explanations can be used on a number of levels. They can tell us why specific events (such as bent oars) occur, or they can be asked in a form that focuses on why general laws work (such as why light is refracted in such and such a manner). In either case, the reliance on general Logical Positivism and Scientific Explanation 17 deterministic laws and understood initial conditions will lead the scientist to the proper explanation. The Deductive-Statistical Model Among the early criticisms of Hempel’s models was the claim that D-N explanations were oversimplifications of the ways that real explanations work. Specifically, Hempel’s detractors believed that few explanations depend solely on the types of general laws mentioned above. As a response to these critiques, Hempel introduced the D-S covering law model. The D-S model was similar to the aforementioned D-N model, with the exception that this model involved at least one covering law that was statistical in nature. In addition, the force of a DS explanation is that it allows us to infer probabilities from relative frequencies. For instance, an explanation of the odds that a fair coin, having been heads for a number of tosses, will be heads on the next toss, would be something like the following: The tosses are thoroughly independent and equiprobable. The r/f of heads is ≈ 1/2. The probability of heads on the next toss is ≈ 1/2. Like the D-N model, D-S explananda would provide deterministic explanations. However, these explanations took into account the fact that many laws used in science are statistical. As a result, D-S explanations are deductive arguments, where one (or more) of the laws involved is statistical. This explanation form was welcome news to social scientists, many of whom believed that laws governing their practices are irreducibly statistical. For archaeologists, explanations that use carbon-14 as a means to assign dates to objects could easily be seen as examples of the D-S model. Radiocarbon dating is the analysis of the concentration of C14, a radioactive isotope of carbon, in artifacts. C14 is created in the upper atmosphere, is incorporated by respiration into all organic matter (plants and animals) throughout the organism’s life, and decays (becoming nitrogen-14) at a constant rate, with a half-life of 5,730 years. By knowing the proportion of C14 to carbon-12 (C12, the stable, non-radioactive isotope) in an artifact, and the rate of decay of C14, scientists believed that they could determine the age of that artifact. The belief that knowledge of the decay rate of C14 can be used to determine a specific sample’s age is dependant on radioactive decay being seen as relatively constant. This belief is predicated on a D-S explanation, where the probability that radioactive decay will remain relatively constant can be explained in the following manner: Consider, for example, the hypothesis that for the atoms of every radioactive substance there is a characteristic probability of disintegrating during a given unit time interval…This complex statistical hypothesis explains, by deductive implication…the following: Suppose the decay of individual atoms of some radioactive substance is recorded by means of the scintillations produced upon a sensitive screen by the alpha particles emitted by the disintegrating atoms. Then 18 Chapter 1 the time intervals separating successive scintillations will vary considerably in length, but intervals of different lengths will occur with different statistical probabilities. Specifically, if the mean time interval between successive scintillations is s seconds, then the probability for two successive scintillations to be separated by more than n•s seconds is (1/e)n, where e is the base of the natural logarithms. (Hempel 1965, 380-381) Given what is known about the decay rates of particular radioactive elements like C14 (which can be represented by statistical laws), D-S explanations show why there is reason to assume a general uniformity in that rate of decay. Here, a D-S explanation of radioactive decay would give an experimenter the confidence necessary to use C14 as a dating method.11 The Inductive-Statistical Model I-S models are unlike D-N and D-S explanations in that, given the appropriate information and laws, there are no expectations that these explananda would result in any deterministic or probabilistic laws being instantiated. In place of such a law, I-S explanations may result in something resembling a likely outcome, but the event to be predicted or explained does not follow logically from the explaining laws and sentences expressing the antecedent conditions. For a boxing match the strength and endurance of each fighter is known in advance. On this particular night, there is a mismatch. According to reputable boxing experts, one fighter (A) is clearly a better fighter than his opponent (B), meaning that the probability (P) is high that fighter A will win. Given the preceding, the question is: in this particular boxing match, where A is fighting B, will A win? P(W, A*B) ≈ 1 A fights B It is highly likely that A wins (an assertion) This model would be far more flexible than either the D-N or D-S models, since the explanation is not encumbered by the deductive form, and only claims probability, as opposed to certainty. Unlike in the case of its fellows, however, there is no reason to believe, even in principle, that it would be possible to achieve the kind of rigorous explanation expected of covering law explanations. Even if all relevant conditions were known and all laws were invoked, the best possible result would be a good bet as to the outcome. Of course, covering law predictions would be totally out of the question. Hempel did not believe that I-S explanations would replace the other models. Many explanations would fit into one of the deterministic structures, and I-S explanations would fill the gap left when D-N and D-S explanations could not be used. Explanations fitting into both the D-S and I-S models will be explored in chapters 2 and 3, but, one can see that there is a significant difference between D-S inference and I-S inference by means of a couple of quick examples. In the first (D-S) case, the archaeologist would assume a law or generalization. The Logical Positivism and Scientific Explanation 19 conclusion establishes that there is a deductive consequence of the statistical law. An example of this relationship would be the following: Ceteris paribus, the relative frequency of late Bronze age artifacts in a particular region is p. An artifact was found in this region. The likelihood that this artifact is from the Late Bronze age is p. So long as the probability set in the first statement is correct, an archaeologist could assume that a certain artifact found on site would have probability p of being from the late Bronze age. In the second (I-S) case, no deductive laws are employed. Thus, there is no guarantee (either in terms of certainty or of statistical probability) of the given conclusion. An example would look something like the following: Certain stone tools that are used today by a group in a particular region function as hide scrapers. Ancient artifacts, similar in form and/or perceived function to those mentioned, are found in the same region. It is likely that those artifacts, like their contemporary analogues, are hide scrapers. Note that this conclusion about the function of this type of artifact is not in any way certain, although the addition of premises can be used to strengthen the conclusion. For instance, if an artifact’s provenience (physical context when found) seems to indicate a hide-scraping function (as opposed to a cooking, or other possible function), this confirmation would add strength to the given conclusion. However, by design, I-S arguments cannot provide the deductive certainty that is associated with D-N and D-S models above. Despite the limitations of I-S explanations, they were seen by Hempel as fulfilling a need. I-S explanations were found to be more flexible than their deductive counterparts, and therefore could be used in situations where D-N and D-S explanations could not. Where deductive models of explanations could not be posited (either in fact or in principle), I-S explanations would provide scientists reasons to continue their work, with some confidence of their results. General Problems With Hempel’s Models12 Explanations and Relevance Critics claimed that it would be impossible to know when a potential explanation would be considered complete. In some simple cases, the relevant facts and laws might be called up, and an explanation could be offered. Of course, some believed that outside of a laboratory (or a few oversimplified examples) it would be impossible to decide which circumstances or laws were relevant to a given event. Given this indeterminacy, there could be no way to deliver a complete explanation, free of irrelevant facts, auxiliary hypotheses, and the like.13 In 20 Chapter 1 complex, real situations, it can be difficult to decide how much (or which) information is relevant. In fact, information that seemed immaterial might end up being crucial to an explanation, while law-like statements or facts that seemed relevant may turn out to be irrelevant. A question that a D-N model (or any covering law model) theorist would be unable to answer is whether there is any reason to believe that future events will resemble seemingly similar past and present events. If the future does not resemble the present, then can be no method to guarantee the veracity of any of Hempel’s candidates for general laws. This problem did not begin with Hempel. Ever since Hume, scientists have been aware that their belief that predictions made about future will hold as they have in the past is based on a measure of faith and not on certainty. A number of yesterday’s laws have been discredited today, and any laws posited by Hempel could similarly be found lacking in the future, potentially disrupting any attempt at deductive explanation. Hempel, recognizing the problem, spends a good amount of time trying to delineate laws from statements that are merely true by happenstance. Hempel claims that, given the potential for error in assigning statements as “laws,” scientists could substitute law-like generalizations or law candidates, principles that have the form of laws but which may not end up being true. Scientists can believe in law candidates without needing the same commitment to their absolute reliability as they would if they called these generalizations “laws.” Another serious criticism of covering law explanations has to do with the relevance of information to an explanation. There is no requirement in Hempelian D-N explanations that covering laws must be relevant. These explanations required only that the inference be valid, and this circumstance can lead to some very counterintuitive explanations. For instance, Salmon (1982, 106) shows how a perfectly valid explanation can be constructed for the fact that John Jones is not pregnant, by noting that he takes birth control pills. This example can be reparsed as follows: Conditions: Laws: Explained: Mr. John Jones takes birth control pills regularly. Ceteris paribus, people who take birth control pills do not get pregnant. John Jones is not pregnant. There is only one problem with this otherwise perfectly good explanation: John is male. This fact need not be a part of the aforementioned explanation in order for it to be valid, but without that relevant piece of information, the proffered explanation does not explain in any real sense of the word. Explanation and Prediction Hempelian explanation was seen as especially enticing to scientists because a Hempelian explanation of past events was seen as symmetrical to a prediction of similar future events: Logical Positivism and Scientific Explanation 21 Let us note here that the same formal analysis…applies to scientific prediction as well as to explanation. The difference between the two is of a pragmatic character. If E is given, i.e. if we know that the phenomenon described by E has occurred, and a suitable set of statements C1, C2, …, Ck, L1, L2, …, Lr is provided afterwards, we speak of an explanation of the phenomenon in question. If the latter statements are given and E is derived prior to the occurrence of the phenomenon it describes, we speak of a prediction. It may be said, therefore, that an explanation of a particular event is not fully adequate unless its explanans, if taken account of in time, could have served as a basis for predicting the event in question. (Hempel and Oppenheim 1948, 249) Covering law explanations, properly filled out, are in a pragmatic sense only temporally dissimilar to predictions. So long as the explanation is complete, the same laws that explained why a certain event occurred will provide the scientist the same materials needed to predict future occurrences of that event. For example, to an astronomer: E.g., those initial conditions and general laws which the astronomer would adduce in explanation of a certain eclipse of the sun are such that they might also have served as a sufficient basis for a forecast of the eclipse before it took place. (Hempel 1965, 234-5) The astronomer, having more or less complete information about the factors relevant to explain astronomical phenomena, has by extension the same materials needed to tell the world when next to expect this event. Given the potential for success in this example, it would seem that complete explanations are the goal of science, and that this goal is attained by scientists on a regular basis. However, in reality, complete explanations are not so easily found. Hempel recognized that despite the apparent symmetry between explanations and predictions, in practice, “only rarely, if ever, are explanations stated so completely as to exhibit this predictive character” (Ibid., 235). Even in the case of most D-N or D-S explanations, there is little chance that a scientist would be able to sort out all of the relevant facts and understand the appropriate covering laws in order to offer complete explanations. In the case of I-S explanations, the gap was, even in principle, insurmountable, as these explanations are not even deductive (certainty-generating) in nature. Explanation and the Social Sciences The Controversy Over Historical Explanation Despite general problems surrounding covering law explanations, a number of social scientists believed that their fields could become more scientific by adopting Hempel’s models. Historians, and social scientists in general, continue to question whether their fields could be considered scientific disciplines (in the 22 Chapter 1 same manner as were the natural sciences). Both history and archaeology are seen as being torn between two factions. One group of historians sees their field as it has been traditionally viewed, as a narrative account of the past, while another sees the future as being in line with scientific progress. Archaeologists have been experiencing exactly the same debate. The most radical of postprocessualists such as Shanks and Tilley (1987) claim that archaeology should leave the sciences and return to an historical approach. They are facing off with the neo-processualists and other archaeological scientists such as Binford (1972), and philosophers of science Salmon (1982, 1990, 1991), and Wylie (2002), who insist that archaeology is a science and should be treated as such. This tension has much to do with the proper placement of both of these fields in the academic community. Although Hempel considered history to be a science, it is difficult to draw a clear line between history and literature (which is clearly not meant to be scientific). Similarly, archaeology, recast as scientific by Binford, has traditionally been housed in a variety of academic departments, including anthropology, history, classics, oriental studies, and art. As a result, in order to make a case for an archaeological science, it will first be necessary to understand Hempel’s approach in relation to history, his paradigmatic social science. Unified Theories Although some historians believed that history was best served by moving in another direction than the physical sciences, Hempel believed that historical explanation could be cast in covering law form: “Our characterization of scientific explanation is so far based on a study of cases taken from the physical sciences. But the general principles thus obtained apply also outside this area” (Hempel 1965, 251). Hempel thought that history was the most evolved of the social sciences. As a result, he believed that he could use historical explanations as an ideal model. If he could prove his case with history, he could show that his explanatory models fit the social sciences as well as they fit the physical sciences, such as physics. In “The Function of General Laws in History,” Hempel attempted to mirror the work he did when investigating physics. In order to head off charges that history is not a science, Hempel underscored the difference between working scientists and the theory behind what they do: It is a rather widely held opinion that history, in contradistinction to the so-called physical sciences, is concerned with the description of particular events of the past rather than with the search for general laws which might govern those events. As a characterization of the type of problem in which some historians are mainly interested, this view probably can not be denied; as a statement of the theoretical function of general laws in scientific historical research, it is certainly unacceptable. (Ibid., 231) Just as the explanations of physicists and chemists do not naturally resemble DN explanations, neither do historical explanations. However, general laws are behind both types of explanation, and “general laws have quite analogous func- Logical Positivism and Scientific Explanation 23 tions in history and in the natural sciences, and they form an indispensable instrument of historical research” (Ibid). Despite claims to the contrary, Hempel believed not only that regularities relevant to history exist, but also that they are important tools, making the explanation of the past as possible as it is in the more mature sciences: “Historical explanation, too, aims at showing that the event in question was not a “matter of chance,” but was to be expected in view of certain antecedent or simultaneous conditions. The expectation referred to is not prophecy or divination, but rational scientific anticipation which rests on the assumption of general laws” (Ibid., 235). The goals, Hempel claimed, are the same in history as they are in the physical sciences. In truth, the only real difference between explanations in history and in physics is in the vagueness14 found in the former, and in the degree of clarity found in the latter. Covering law explanations were more successful in physics because experimental results were explained by reference to well known physical laws. “Most explanations offered in history or sociology, however, fail to include an explicit statement of the general regularities they presuppose” (Ibid., 236). Regularities were not only missing from historical explanations, their very existence in principle was suspect. In addition, any candidate for a general law was seen as either so specific or so general as to be completely useless. Few, if any general laws (especially deterministic laws) could be found that were useful to historians, making Hempel’s claims about the existence and utility of such laws hard to accept. Historians were unable to explain their field adequately, much less give complete explanations or generate predictions about analogous events in the past or possible events in the future. Hempel was aware that the lack of clarity was shared between history and the natural sciences: “But there is no difference, in this respect, between history and the natural sciences: both can give an account of their subject-matter only in terms of general concepts, and history can ‘grasp the unique individuality’ of its objects of study no more and no less than can physics or chemistry” (Ibid., 233). In his earlier writings, Hempel agreed with his mentors, the logical positivists, that the mature sciences were naturally predisposed to scientific analysis. However, by 1965 Hempel had time to realize that complete explanations, predictions, and the like were not going to be forthcoming in physics, much less in history. At best, the resulting explanations would be explanation sketches, partial views of the historical events in question: “What the explanatory analysis of historical events offer is, then, in most cases not an explanation in one of the senses indicated above, but something that might be called an explanation sketch” (Ibid., 238, emphasis in the original). These sketches were supposed to show what a full explanation would be, were the details available. The presumption was that, as the science of history evolved, general laws would be forthcoming, and explanation sketches would turn into true explanations: “A scientifically acceptable explanation sketch needs to be filled out by more specific statements; but it points into the direction where these statements are to be found” (Ibid.). Physics had started out as speculation and had matured to the point of being able to explain happenings in the natural world. It was assumed that history would similarly mature and grow, given this new direction. Explanation 24 Chapter 1 sketches would point historians in the right direction. So, armed with hypothetical sketches of what historical explanations should look like, Hempel believed that historians would then be able to wait for enough data to be collected for laws to fall out. Unfortunately, they are still waiting. Responses to Hempel: From Unity to Disunity Responses to Hempel’s call for unified explanation were varied. Some, like Patrick Gardiner, believed that Hempel’s models were on the right track, though perhaps overambitious. He believed that philosophers of science were overzealous in their belief that history could ever compare with the natural sciences: One is inclined to say that generalizations of the kind in question provide indications, and rough ones at that, of the sorts of factors which, under certain circumstances, we expect to find correlated with other sorts of factors; but that they leave open to historical investigation and analysis the task of eliciting the specific nature of those factors on a particular occasion, and the precise manner in which the factors are causally connected with one another. (Gardiner 1952, 93). As history is a less exact science than is physics, Gardiner believes that good, historical explanations would have the form of the explanation sketches that Hempel believed were the starting point for historical explanation. Gardiner believed that historians have an obligation to investigate the mental causes of historical events (meaning that historians needed somehow to understand the mindsets of the historical agents in order to explain their actions): “The apparent antimony created arises out of the view that the world is made up, on the one hand, of ‘dead matter,’ and, on the other, of ‘mind’; in human beings it is thought that these are miraculously intermixed. A human being is ‘matter’ with ‘mind’ attached…Whether we regard human beings in their ‘physical’ aspects or in their ‘mental’ aspects depends on our interests” (Ibid., 138). To Gardiner, the only way to understand the causes of events, and thereby to explain them, would be to bridge the gap between the physical and mental worlds inhabited by the people who made that history come to pass. In The Function of General Laws in History, Hempel argues that the view that historians can somehow connect themselves with the mental states of ancient peoples is untenable: Stated in crude terms, the idea underlying this function is the following: The historian tries to realize how he himself would act under the given conditions, and under the particular motivations of his heroes; he tentatively generalizes his findings into a general rule and uses the latter as an explanatory principle in accounting for the actions of the persons involved. Now, this procedure may be heuristically helpful; but it does not guarantee the soundness of the historical explanation to which it leads. (Hempel 1965, 240) If it serves some purpose to delve into the thoughts and dreams of historical characters, Hempel has no quarrel with this use of mental imagery. However, he urges that it would be a serious error to confuse (albeit occasionally helpful) Logical Positivism and Scientific Explanation 25 fantasies with reality. Covering laws cannot be based on picturesque guesses as to the mental states of people long dead.15 Gardiner’s defense of his position lies in his belief that present day scientists can delve into a long dead person’s motivations by appealing to reason: We all know (it would be an abuse of language to say that we ‘assume’) that people often act rationally, that in many situations they can be counted upon to give us good reasons for that which they are doing and to do it in a certain kind of way. As rational beings ourselves, accustomed to choose between various methods of attaining certain results, it is not surprising that we are able, readily and often without hesitation, to understand why, in similar situations, other people have acted as they have, and to imagine the sort of reasons they would have given for the actions they performed, had they been asked. (Gardiner 1952, 135) From Hempel’s point of view, claims that it is possible know the mental states of others, that these others are (for the most part) to be considered rational, and that analogous reasoning can somehow be employed in order to truly understand their motivations are fatally flawed. As there would be no way to verify, or to show any real connection to the modern results by matching them to long vanished mental states, this is an explanatory non-starter. Regardless of the final form that scientific explanations do take, appeals to mental events do not seem likely to lead in a direction acceptable to scientists.16 Here, the claim is not that Hempel was against empathy. If having a feeling about a past culture leads to new thoughts about that group, Hempel would have no problem using those feelings to generate hypotheses. Testing hypotheses by appealing to ancient rationality or claiming some sort of psychic empathy, however, would be unacceptable. The Forces of Disunity William Dray went beyond Gardiner’s belief that covering laws were just a poor fit for historical explanation. Dray believed that Hempel’s models did a disservice to explanation, in that they limited explanation in ways that were both unwarranted and unnatural. For one thing, Dray found that explanation sketches, by Hempel’s own rules, make their explananda meaningless: If the candidate law ascends too far into generalities it loses its methodological interest; but if it descends from the stratosphere it becomes possible to deny it without withdrawing the explanation. In the face of such a difficulty, covering law theorists [such as Hempel and Gardiner] often employ more cautious substitutes for ‘entailment’, which suggest its advantages without laying their claim open to strict test…If the use of such terms is intended to mean no more than that a covering law is suggested by an explanation—i.e. to admit that there is no tight logical connexion at all—it would be difficult to quarrel with the logician’s claim. But the cost of modifying the covering law claim in this way would surely be rather high; for an explanation can scarcely be said to stand or fall by what it suggests. (Dray 1957, 29, emphasis in the original) 26 Chapter 1 Dray rightly pointed out that by loosening the rules of entailment, Hempel took all of the force out of historical explanations. The explanation sketches that remained were not truly connected to any general laws, and they also could not be used to construct predictions. So, covering law explanations seem superfluous in fields like history. Secondly, Dray believed that Hempel’s focus on covering laws unnecessarily limited the range of what is generally meant when the term “explanation” is used. While standard “why-necessarily” explanations may or may not be best served by covering laws, other types of explanations, called “how-possibly” explanations would not have been served by Hempel’s models at all: I shall maintain, the historian need not show that what is to be explained happened necessarily in the light of the particular events and conditions mentioned in the explanation, and, a fortiori, need not show that it happened necessarily in the light of some covering law or laws. For the demand for explanation is, in some contexts, satisfactorily met if what happened is merely shown to have been possible; there is no need to go on to show that it was necessary as well. (Ibid., 157, emphasis in the original) Although Hempel could (and did) counter that these explanations can, in many cases, be re-parsed into covering law explanations, Dray claimed that such reparsing was counterproductive: “To insist, nevertheless, that no explanation is complete until a lurking covering law has been discovered is surely just to fall into a kind of determinist myopia” (Ibid., 168). Even if it would be possible to change how questions into why questions, the change would not address the issue of whether the latter are somehow better than the former. The Possibility of Historical Laws One final point recognized by Hempel was that historians believed that there was a lack of laws that historians could call their own: “Quite probably, some of the historians who tend to minimize, if not deny, the importance of general laws for history, are prompted by the feeling that only ‘genuinely historical laws’ would be of interest for history” (Hempel 1965, 243). This belief that history needs historical laws, according to Hempel, was a non-issue, as the only reason that historians would feel a need for purely historical laws would be if they believed their science to be autonomous. If historians did not anticipate any contact with other scientists, or believed their science to be somehow different from the other sciences, then finding and developing laws of history would probably be an important goal for historians. Since Hempel’s goal was unification and reduction, however, a lack of historical laws would not be a problem. Those laws, or law-like statements, that history does tend to use may come from a variety of fields, and this accords well with a unified view of science. So the presence or lack of purely historical laws does not show that the covering law model is inapplicable to history.17 Logical Positivism and Scientific Explanation 27 The Controversy Over Archaeological Explanation It could be argued that Hempel’s “The Function of General laws in history” is even more relevant to archaeology than it would be to history, as archaeology is positioned between the natural and social sciences. Archaeologists draw data from physics, chemistry and biology, on the one hand, and from sociology, anthropology, and history on the other. As a result, archaeologists have sought to use the tools of the natural sciences to answer many of the same questions asked by historians. This dual focus leads archaeologists to occupy a unique space. Since the 1960s, archaeologists have wondered whether covering laws are appropriate in archaeology, and specifically whether the archeological explanations generated by Hempel’s models are relevant to the questions archaeologists ask in the field. For instance, explanatory underdeterminacy, a charge that can be directed at Hempelian explanation in general, is especially relevant in an archaeological context. For most archaeologists, having multiple equi-possible explanations for a site’s past is unacceptable.18 For example: Several large stones are found in a specific area atop a tel.19 These stones are not native to the region, and are piled up in a conspicuous manner along one edge of the tel. Given that “end state,” and little additional information, archaeologists could posit a number of “beginning states,” not limited to the following: 1. The stones were a part of an ancient fortification system, destroyed during the conquest of the last phase of occupation. 2. The stones were a part of an ancient fortification system, destroyed by natural processes, such as erosion, long after the end of the last phase of occupation. 3. The stones are a part of the ancient city’s dump. When conquerors, settlers, etc, come upon an abandoned tel, they first level the surface, in order to build upon it. These stones were removed from throughout the tel, and were cast aside in one unused location, in order to facilitate this process. 4. The stones were “robbed” from throughout the tel by modern builders, using the ancient remains as a “quarry” for a nearby village. The remaining stones are either refuse, or as yet untapped resources. 5. Modern farmers, attempting to reclaim the surface of the tel for agricultural purposes, found the stones while tilling the soil, and piled them on the side of their new “field” so that their crops could grow, and so that their equipment would not be damaged by the stones’ presence. This wide range of possible explanations (1-5) is but one vastly oversimplified example of a continual archaeological problem. Equifinal explanations are common in the field, as archaeological remains do not always provide the full range of information (contextual, functional, etc.) that an archaeologist would need to narrow down the contenders. One of the most severe modern critiques of archaeology is that it has no systematic means to work with this problem.20 Note 28 Chapter 1 that equifinality is not solely a feature of Hempelian explanations. This issue has been at the heart of archaeology (and of all other underdetermined sciences) since its start. The hope was, however, that covering law explanations would be able to bring new tools to the table, and to succeed where past attempts had failed. When Hempel introduced covering law explanations, he acted in the tradition of the logical positivists, a group of scientists and philosophers of science who were interested in properly locating and understanding scientific objectivity. His attempts to understand what goes on in scientific reasoning were initially well received, and when philosophers of science began to note problems (discussed above), Hempel expanded his models to encompass broader forms of explanation, and he introduced explanation sketches to deal with underdeterminism. The versatility of Hempel’s explanatory models was caused scientists from other disciplines to interest in the application of this work to their fields. As we will see in chapters 2 and 3, archaeologists were especially interested in using covering laws as the basis of their updated discipline. Before engaging in the archaeological incorporation of logical positivism, however, it will be necessary to provide some relevant archaeological background. The material in chapter 2 will show how archaeologists first came to the realization that they needed a theoretical basis, and it will chart the impact and influence Hempelian models had on archaeology as a discipline. Finally, if covering law models are not taking archaeology in the right direction, then questions surrounding the right direction abound. Archaeologists would then have to decide whether to modify Hempel’s models for archaeology, if any other sort of strict adherence to laws would be able to avoid the problems of Hempel’s particular approach, or if laws would have to be done away with altogether. A range of alternatives also exist, from a return to past methods to a call to find a replacement to Hempel’s strategy, as will become evident in the next few chapters. Notes 2. Note that this is not an exhaustive (or uncontroversial) treatment of the Vienna Circle’s views. For those interested in this topic, Sarkar (1996) has edited a 6 volume work on the Vienna Circle and science. 3. According to this pamphlet (340), the members of the Vienna Circle included Rudolph Carnap, Kurt Gödel, Otto Neurath, and Moritz Schlick. Friends of the Circle included Hans Reichenback, Ludwig Wittgenstein, Albert Einstein, and Bertrand Russell. 4. “All these possible modes of representation—if they otherwise actually express the same knowledge—must have something in common; and what is common to them is their logical form” (Schlick 1930, 55). Logical Positivism and Scientific Explanation 29 5. Struggles to decide between axiom sets led to a huge debate over the nature of truth (correspondence vs. coherence) that was never resolved by the Circle. 6. See Reisch (1994) for an account of Neurath’s attempt at unification (the International Encyclopedia of Unified Science) as well as its collapse. 7. Hempel believed that the only difference between history and its older sister sciences was in their relative stages of development. His covering law models were designed to bring history the rest of the way into the fold. Inadequacies with historical explanation were believed to be due to failures in the practice of history, not in the models themselves. 8. These views are found throughout Wittgenstein’s Tractatus, which was widely read by the Circle. 9. Hempel’s realism will be of special interest to archaeologists, as discussed in chapter 5. 10. This critique was also made of new archaeology, as many working archaeologists objected that the new archaeology bore little resemblance to the way that archaeology was done. 11. It has since became clear that a number of factors have caused carbon-14 levels to vary over time and place, forcing scientists to come up with complex calculations and adjustments to bring these dates into line with other, known and reliable, sources of information. See chapters 2 and 4. 12. Problems specific to the application of covering law models to the social sciences will be addressed later. 13. Philosophers of archaeology such as Salmon (1982) have mentioned this problem. Although the covering law model works in simple examples taken from physics, attempts to find covering laws that govern an archaeological excavation (or any real scientific enterprise) continue to elude investigators. 14. This sloppiness is in terms of their lack of predictive power, their incompleteness, and in their apparently indeterministic nature. 15. Collingwood would take this belief fully into absurdity, claiming that the only relevant way to study history was to abandon the “positivistic conception, or rather misconception, of history, as the study of successive events lying in a dead past, events to be understood as the scientist understands natural events, by classifying them and establishing relations between the classes thus defined…So far as historians yield to it, they neglect their proper task of penetrating to the thought of the agents whose act they are studying” (Collingwood, 37). 16. That is, unless one wants to give scientific legitimacy to psychics. Feyerabend (1975) championed this position (see chapter 4). 17. That is not to say that such laws do or do not exist. Hempel was agnostic on this matter. 18. Some archeologists, such as Shanks and Tilley, might disagree. Their views will be examined in chapter 5. 19. A tel (or tell) is an archaeological feature peculiar to (though not always limited to) the ancient near east. It is a mound, created by continual occupation at a site by numerous people over the years. The numerous cycles of growth, destruction, and rebuilding cause its vertical aspect and peculiar shape, leading to its easy identification. 20. This problem will be discussed in more detail in chapters 4-6. Note that equifinality is not a problem created by Hempel’s models. This problem was one of the reasons that archaeologists sought to systematize the field. The fact that Hempel’s models provide no means to solve this problem is significant, however. 30 Chapter 1 Chapter 2 The New Archaeology In “The Function of General Laws in History,” Hempel sought to transform history from a descriptive enterprise into a scientific discipline. Instead of preparing tabular lists of dates and places, scientific historians would focus on understanding why the past happened as it did. Following Hempel’s course, the archaeological community saw an opportunity to leave behind the descriptive focus that had been archaeology’s only raison d’etre, and instead to explain a vibrant past. Traditional archaeologists, or culture historians, believed that their job was to create catalogues of artifacts, to establish a link between a particular set of cultural activities or traits and those artifacts, and to order the artifacts (and cultures) spatially and temporally. The emphasis here was on the reconstruction of events from history. The result of the culture historical approach would be analogous to creating a sort of scrapbook of images. These images (like a set of photographs taken on vacation), could be placed in order, providing the viewer with a full account of the vacationers’ activities. However, anyone who has been on vacation knows that the pictures provide a hollow account of what happened. If some of the pictures did not come out, or if the camera was not handy, there will be gaps in the record. Further, even if the trip is fully documented, the viewer still needs a running commentary from the vacationer in order to understand the photos. The photographer can talk about the reasons for going to a location, the story represented by the single photo, or the context of the scene in view. Without the storyteller, however, the photographs are meaningless. Explanation, as practiced by scientific, or new-archaeologists, would have an entirely new focus: “The new archaeology in America is tending to be more concerned with culture process and less concerned with the descriptive content of prehistoric cultures” (Caldwell 1959, 304).21 New archaeology focused on dynamic cultural processes, as opposed to culture history (new archaeology was also known as processual archaeology), would focus on a more lifelike reconstruction of the past. Here, the study of history is a dynamic process, with many interlocking and interactive parts. An archaeological reconstruction for a new archaeologist would be more analogous to a full video recording of a vacation, taken from a removed, omniscient perspective. Not only would the visual evidence be described, it would also be put into context. Dynamic processes could be observed interacting with the subjects under study and, in the end, the archaeologist would know not only what happened, but why it happened as well. Culture process, or processual archaeology would soon be associated primarily with the application of Hempelian philosophy of science to archaeology by theoretical archaeologists such as Lewis Binford, John Fritz and Fred Plog, but the move from culture history to culture process did not come from a vacuum. A sort of restlessness or knowledge that the field had stagnated and needed new direction had been building for years. Decades before archaeologists such 32 Chapter 2 as Binford made his case for Hempelian processual archaeology, others saw the future of archaeology as being tied in some way to scientific methodology. In 1917 Clark Wissler, having read of Nelson’s excavations of Chaco Canyon, claimed that Nelson’s methods were groundbreaking in their scope and attention to detail. Wissler, forecasting the recasting of archaeology as a science (tied to anthropology), complimented Nelson’s work: “Such are the results of the real, or new archaeology, as a part of the science of anthropology” (Wissler 1917, 100 emphasis added). Walter Taylor, a graduate student in the 1940s, also believed that the archaeological status quo was going nowhere. However, Taylor saw the future of archaeology as being more than a matter of increased precision. He believed that archaeologists needed to adopt a very different perspective: “what is necessary is that we compare, not individual items either separately or in groups, but rather cultural contexts and/or broad cultural complexes as wholes” (Taylor 1948, 169). Taylor complained that archaeologists kept their attention on singular entities, at the expense of large trends. The future, Taylor believed, would involve these larger complexes and he called the study of these trends a “conjunctive approach.” Although Taylor stated outright that the conjunctive approach could not be captured in any single set of methods, he believed that there were certain features that would set his approach apart from that of his fellows. Specifically, he saw the need for this approach in the interpretive phase of an excavation: “It is at this point that the cultural conjunctives, the relationships between the archaeological data of a society or cultural entity, play their important part. Often this interpretation is neither easy nor the means obvious, but if archaeology intends to attain the level of context, if it professes to become either anthropology or historiography, this is an essential step in the process” (Ibid., 193). The goals of processual archaeology would be realized in Taylor’s lifetime. In his 1972 article, “Old Wine and New Skins: a Contemporary Parable,” Taylor claimed that in the course of a quarter century his conjunctive approach had become the archaeological mainstream. The search for scientific method, the focus on anthropology, and the interactive approach that Taylor spoke of would become the central foci of new archaeology. The focus of this chapter will be on a number of issues that led archaeologists such as Taylor to demand a theoretical shift in focus from description to explanation. A desire for change would lead archaeologists to adopt Hempelian covering law models of explanation, and this choice (as much as the initial desire for change) would shape the discipline in fundamental ways, both theoretical and methodological. New archaeology, the union of philosophical explanation and archaeological practice, would have both positive and negative effects on archaeology as a science, effects that will be the focus of the remainder of this project. However, among the most significant changes was a shift in the perceived goals of archaeology. In the past, when archaeologists focused on better ways to “do” archaeology, they believed that the practice of archaeology could be improved by précising existing methods. However, when the new archaeologists reacted to what they saw as the shortcomings of culture history they questioned the goals of The New Archaeology 33 their enterprise, and sought to change their discipline on a fundamental level. These changes would affect archaeological methodology, but those effects would be artifacts of the real enterprise; the examination of what it means to practice archaeology. New archaeology marks, in fact, the transformation of archaeology from an activity whose goal was to document history to a discipline devoted to explaining that history. This transformation, its motives, its workings, and its effects are the subjects of this chapter. New archaeology would, in the 1960s and 1970s (and in some circles, to this day), dominate archeological theorizing and have pervasive effects on archeological practice in the Americas, Europe, and the Middle East. So, prior to exploring new archaeology, this movement will be put into its historical context. To that end, a brief sketch of its origins and history will be the first priority. The first section of this chapter will be devoted to some background into traditional (or culture historical) archaeology, and The second section will present an exploration of issues leading archaeologists to become discontented with traditional archaeology, or culture history, leading them to search for a new archaeology. The remainder of the chapter will focus on new archaeology itself, examining its promise and its effects on the discipline. The Origins of New Archaeology Pre-Theoretical Archaeology Archaeology as a field has a rich history that dates back to at least the third century, C.E. Helena, the Roman Emperor Constantine’s mother, was one of the first “archaeologists” on record. She spent a good portion of her life searching for religious sites throughout the ancient Near East. In contrast to archaeology’s long and rather colorful pedigree, the history of scientific archaeology, archaeology as an explanatory and theoretically self aware discipline, is rather short. Prior to archaeology’s theoretical awakening in the late 1800s, the only major changes to archaeology were in the development of excavation techniques and tools of the trade (see Schnapp 1997 for an excellent introduction to this topic). Scientifically, archaeology is still a relatively young discipline. During this pre-theoretical period (roughly lasting until the end of the nineteenth century), archaeologists for the most part came in two varieties. An archaeologist could be an antiquarian, collecting and cataloguing significant finds, ranging from coins and trinkets to whole buildings (at times moved piece by piece to a foreign museum). These academic archaeologists used their finds and excavations to increase the academic community’s knowledge of a site’s occupation or ethnicity (although, according to faunal specialists such as Maher (1999), the ability of an archaeologist to determine ethnicity from archaeological finds has recently been called into question). The other variety of archaeologist was decidedly nonacademic. Rather than cataloguing finds, tomb raiders plundered sites to increase their wealth or status, or made a career out of trafficking 34 Chapter 2 in artifacts collected in this manner. The line between these types of archaeologists was in many cases blurry, as many academic archaeologists had private collections of their own, and even those academics who preferred not to pilfer artifacts themselves would routinely use grave robbers and thieves to enhance their collections or provide them with specimens for study. Snead remarks that these two types of archaeologists were integrally connected: “The relationships between eastern ‘museum man’ and western relic-hunter, between scientific aspiration and economic motivation, between professional ambition and popular demand were at the heart of the relationship between Southwest archaeology and American society as the end of the nineteenth century drew near” (Snead 2001, 4). Whatever their respective motivations, the relationship between these types of archaeologists resulted in a field discipline focused on the collection of significant artifactual finds. Back in the laboratory or university, the collective efforts of these two types of archaeologists gave the researcher the tools to establish a given site’s chronology. Further, with enough data, the archaeologist hoped to be able to infer traits about people living in each temporally separable cultural unit. For instance, if a faunal specialist can determine the age of a domesticated animal at its time of death, this information could be used to determine the primary use of that animal. Animals killed at relatively young ages would have been used for meat, while older animals (which are less tender) would have been used for other purposes, such as milk production or labor. If the bones were found in a particular datable archaeological context, the archaeologist believed it possible to associate that feature (a particular type of animal use) with the cultural unit living at that time. By separating groups under study into cultural units and by focusing on learning details about each extinct culture, the archaeologist hoped to learn the site’s culture history. As a result, the archaeology of the period became known as culture history. Culture History Culture history archaeology focused on gathering as much information as possible about a definable area (such as a city) in order to learn about its inhabitants at particular periods of time. As an example, the best Syro-Palestinian22 archaeological excavations of the nineteenth century were large scale operations. These teams were made up of a small staff of professionals, including cartographers, sketch artists (and later, photographers) and a large contingent of unskilled (usually local) laborers. Romanticized images of neatly dressed archaeologists drinking tea as the workers moved huge mounds of dirt come from this period. The local workers were paid extra for finds (to mitigate theft), and all finds were brought to the archaeologist for authentication. Site reports catalogued artifacts found in particular regions (although those regions could cover multiple sites), and special attention was paid to particular architectural features (such as tombs and palaces) and particular finds (those that would be considered valuable to collectors and museums). The New Archaeology 35 Archaeologists such as Sir William Flinders Petrie, while continuing on with many features of this old method, came up with some innovations that he believed applied specifically to Near Eastern Archaeology. Although Petrie’s innovations and their specific ties to regional geography will be the subject of chapter 6, these changes included a belief that sites in this region should be excavated systematically, one layer at a time. This approach, called the stratigraphic method, would provide a way to separate an area into roughly contemporaneous materials that could then be grouped together. By paying close attention to changes in soil composition and artifact styles, Petrie hoped that a team could collect all of the material belonging to one phase of occupation before starting on another. Petrie’s hope was that this initial grouping activity would be the starting point for being able to use these stylistic changes to date the phases of occupation at a site, and then for an entire region.23 Artifact (and phase) dating was initially done by setting up relative chronologies (using ceramic, architectural, or other stylistic seriation) and then comparing these artifacts to others that could be given absolute dates (coins, textual references and dendrochronology). In the twentieth century, advances in science would add to this list other absolute dating techniques. Carbon-14 dating techniques, for example, were touted as making possible precise dating of artifacts without reference to relative dating techniques. Since carbon-14 levels would be determined in a lab, the resulting dates were supposed to be more accurate than previous methods. In addition, the act of moving the dating process to the lab also led archaeologists to believe that the results would be free of any biases that could be related to the particulars of their discovery. Interestingly, relative dating methods, instead of being replaced by the more “scientific” carbon-14, are exactly the tools now being used to correct carbon-14 errors (mentioned in chapter 1). As it is now known that carbon-14 levels are not constant, the older methods, including seriation (relative dating), and dendrochronology, are being used to calibrate carbon-14 dating, making it more accurate.24 In addition to these techniques, laboratory advances allowed for the analysis of new types of data including micro floral and faunal analysis, which were making their way from botanical and forensic circles into the archaeological mainstream. By analyzing the macrofaunal, microfaunal and microfloral remains at a site, an archaeologist could reconstruct what foods the site inhabitants ate, could determine whether they had domesticated animals and plants at a particular period, and could be used to hypothesize about how the animals were used. Major advances in the field, including critical re-evaluations of dig strategies (such as the change from the use of local unskilled laborers to the use of skilled field hands, either locally trained or imported) centered on methods of maximizing the strengths of stratigraphic analysis of archaeological sites. The stratigraphic method which began as a part of culture history would be refined and carried into new archaeology. Despite the efforts of archaeologists such as Wheeler and Kenyon (1952), whose work modernized stratigraphic analysis (to date, variants of the Wheeler-Kenyon method are used throughout the archaeological world), this method took a comparatively long amount of time to gain acceptance in the United States. While stratigraphic methods were accepted in 36 Chapter 2 Europe and the Mediterranean by the 1860s, stratigraphy was not recognized as significant in the United States until much later. This difference in approach was due to a large extent to unique features differentiating the old world from the new. While old world sites were generally occupied for long periods of time, new world sites had short periods of occupation and little to no stratigraphy to show for it. As a result: “many American sites, especially those of Eastern North America, did not lend themselves readily to stratigraphic digging” (Willey and Sabloff, 89). Sites with long enough occupations to leave layered remains were eventually found in the American Southwest and Mesoamerica in the early 1900s, and only then did stratigraphic analysis take hold. An interesting artifact of the change to stratigraphic methodology and laboratory aided analysis was a dramatic change in the concept of archaeological significance. Although some finds considered valuable or significant in the past (such as whole vessels, statues, precious metals) continued to be valued, archaeologists began to spend more and more time studying materials that antiquarians and the like would not even have considered to be of interest (ceramic and architectural fragments, faunal and floral remains, etc). In addition, since the holdings of thieves and illegal antiquities dealers necessarily would lack any stratigraphic context (as admitting that they knew the original context would implicate the dealer in any illegal activities associated with the artifacts’ removal), many once valuable trinkets suddenly lost their archaeological value. (See Schnapp 1997 for more on this changing process). In order to further link archeological value with archeological context, many in the archaeological establishment have recently sought to ostracize those who would still deal in these black market antiquities. Although, despite these efforts, personal greed and financial desperation (as well as whole countries who still refuse to return those antiquities taken from foreign countries centuries ago) still make for a thriving black market, changes in archaeological methods have had an impact on what is or is not to be considered archaeologically significant, as well as socially acceptable. As, by the late 1800s, the only advances in archaeology seemed to be coming either from the laboratory or from the field, culture historical archaeology had moved into what Kuhn (1970) would have called a pure problem-solving mode: that is to say, no new theoretical ground was being covered. Archaeologists were merely tightening up the techniques and methods upon which they had already agreed (for instance, trying to incorporate more ceramic sub-forms into a typology to increase precision in setting ceramic chronologies). Although archaeologists were excited about this increase in precision, many believed that this must be a first step in a new direction, and not just the final tune up of the current model: “It is well known that the fortunes of archeology have been greatly improved by new technical aids such as radiocarbon dating. A more important but far less celebrated advance is represented, I think, by a shift of interest in recent years toward problems of far greater generality than pertain to any single excavated prehistoric site” (Caldwell 1959, 303). It is clear in retrospect that the old model, culture history, had gone as far as it could go. A group of forward thinking archaeologists, newly exposed to the aforementioned methodological and technical advances in the laboratory and in strat- The New Archaeology 37 igraphic analysis, began to see the limits of the traditional approach and started thinking about archaeology’s future. Specifically, culture historical archaeology’s major limitation was in the conclusions an archaeologist could draw at the end of analysis. Culture history employed an inductive approach to science, one that precluded any information gathered about a site from being used to explain or predict why or how the same things might occur at other sites. The inductive approach involved gathering data about a site, interpreting that data in order to describe cultural traits, and then trying to make inferences about the existence and meaning of those traits in other contexts, sites, etc. This project did not involve an explanation of these traits and there was obviously no predictive element. In addition, there was no real attention to explicit hypothesis formation or research design, and the results were not subject to any sort of rigorous testing. As archaeologists became more interested in understanding what went on in the past, instead of merely documenting it, they turned to science, which they saw as the means to this new end. Archaeology: The Need for Change The push for change came at a time when archaeology had essentially become stagnant. The traditional views of culture historians had for some time held archaeology to its standard task of data collection and chronology setting. In the beginning, before archaeologists agreed even in principle that there needed to be a consistent general research design, even this task was very difficult, as data collection implies the existence of data and in the absence of consensus the definition of archaeological data is going to be inconsistent. However, as field techniques were scrutinized and stratigraphy was accepted as the best means to set relative chronologies, the goal of understanding a site’s chronology became more realistic. Pottery seriation, the idea of separating pot styles into types and then documenting changes to those types over time, added additional data for setting relative chronologies, and this led to the recognition that similar pottery forms at different locations represented the possibility of connecting chronologies between sites. For culture historians, inter-site comparison was predicated on the idea of the individual site as the primary and self-contained unit. An archaeologist first studied changes in the architectural, ceramic, and artifactual record to establish a site chronology. Only after understanding the site and having established its chronology did the archaeologist look to the outside. Site-centered archaeology is in opposition to a regional approach which would become one of the hallmarks of processual archaeology. The regional approach takes the unit of study as a region, as opposed to a particular site within that region. Instead of believing that a single group lived at each site, a regional approach might assume that a group lived in a region and that differences found at particular sites within that region might be more a matter of idiosyncrasy or accident than of cultural design. As regions are a more fluid concept than sites, this change from the site based to the regional approach would do more than just add to the amount of 38 Chapter 2 data studied by an archaeologist. Changing patterns of habitation, shifting alliances, and trade are all interactive systems, so the shift to the regional approach necessitated a corresponding shift in theorizing about how best to understand antiquity. The recognition that regions were significant, either as an extension from the site or as the primary focus of activity, would be considered a major theoretical advance for archaeologists. This shift in focus was seen as the first real step in changing archaeology from its treasure-hunting phase to its current status as a scientific discipline. Although this process of establishing site chronologies became more and more exact over time, with the advent of better relative and absolute dating techniques and more finely developed field work, archaeologists were well aware that this could only be a first step, if archaeologists wanted to be more than librarians who got dirt under their nails. The descriptions offered by pre1960s archaeology were, in spite of improved techniques, still oversimplified and far from the sort of scientific knowledge to which archaeologists aspired: In the early 1960s, traditional archaeology was challenged on the ground that it had failed to provide any substantive understanding of the cultural past. Major criticisms of older archaeology were that it had remained heavily descriptive, that it accepted too much as unknowable, and that when interpretations were offered, they were likely to be generalizations extrapolated from a particular data base in what was called a narrow inductivist mode. (Kelley and Hanen 1988, 1) Given a best case scenario, and even if a site could be fully analyzed, a traditional archaeologist could never move beyond an oversimplified list of changes in pottery design, population, architectural styles, and the like. There could be no explanation as to why or how things occurred. In the absence of explanations, even the validity of the lists made by culture historians could be questioned. Every time an archaeologist puts trowel to soil, decisions are made that affect the types of data that are collected. When archaeologists categorize their findings, there is always the question of whether these delineations have any resemblance to those that existed in the past. There was considerable debate about this matter: Were archeological categories natural and invariant over time or were they transient artifacts of contemporary methodology? Archaeologists thought that only after they understood why changes occurred would they be able to understand which changes were real and which were accidents of the recording process. The answer to these and other questions was thought to be found in a theoretical shift away from the historical approach of culture history, toward a more scientific way to do archaeology. As archaeologists looked through the scientific and philosophical literature to find a way to make archaeology more scientific, they found in Hempel a philosopher of science who was expanding the boundaries of science in their direction. The New Archaeology 39 The Aims, Features, and Methods of New Archaeology In “The Function of General Laws in History,” Hempel argued that scientific covering law explanation need not be confined to physics and the other natural sciences. He led the way for a number of historians and, later, archaeologists to think about the possibility that there might be covering laws that would govern the social sciences. According to archaeologist Paul Martin: “Since 1960, my goals and interest have been modified by a trend that is now widespread in American science—a trend that involves a shift from emphasis on particularisms to an imaginative era in which anthropologists and archaeologists build a cultural-materialist research strategy that can deal with the questions of causality and origins and laws” (Martin 1972, 7) Hempel was one of the few philosophers of science who concentrated his efforts on the expansion of philosophical explanation beyond the natural sciences, mapping out a method to transform history (and by extension, the other social sciences) into “real” sciences; thus, a number of forward thinking archaeologists found themselves drawn to his work. After reading Hempel’s “The Function of General Laws in History,” archaeologists, including Fred Plog, saw an opportunity to remake archaeology as a science in a manner similar to history. As archaeology was, until the 1960s, thought to be an offshoot of history it made sense that the method for changing one would be the method for changing the other. The goal for archaeology, as for history, was to give archaeologists access to the powerful methodological apparatus employed by physicists and the other scientists. For the new archeologists, archaeology would no longer be run from the trenches. Antiquarians dug artifacts up and then haphazardly (and subjectively) tried to come to conclusions based on the information garnered from those pieces. Archaeology was to be a science, guided by laws and utilizing its particular artifactual data to test hypotheses: “Archaeology is a science that uses data from the past to test hypotheses concerning past cultural processes. A discipline that recovers data from the past and attempts to do something with it is antiquarianism, not archeology” (Plog 1974, 4). This new archaeology was not to be compared with past changes and improvements. Unlike past achievements that merely gave archaeologists a new set of tools, archaeologists such as Binford and Plog believed that new archaeology represented a radical break from the past. Archaeologists as antiquarians used to recover and manipulate data. Plog believed that new archaeologists as scientists would focus instead on hypothesis testing. Although data collection would always be a part of archaeology, its role as the final goal of archaeology had come to an end. Archaeologists as scientists would study the dynamic processes of cultural development and change instead of making lists. They would favor explanation over description, expect replicability and offer predictability. They would focus on deduction and objectivity over induction and subjectivity. 40 Chapter 2 Binford’s New Archaeology Acquiring these abilities became the goal of the new archaeologists. Lewis Binford’s article, “Archaeology as Anthropology,” moved archaeology one step closer to its goals by divorcing archaeology from the humanities and history, and wedding it to the sciences and anthropology. Archaeology’s divorce from history in favor of science is ironic, by the way, as the point of Hempel’s “Function of General Laws in History” was to show history to be a science. Since archaeology was using the same roadmap, it would seem that making a break from history should be unnecessary. Putting aside for a moment Binford’s decision to recast archaeology as a science, his decision to collapse archaeology into anthropology would prove dangerous to archaeology’s very existence. As the unification and reduction of science was a goal of logical positivism, the reductionist platform seemed wholly appropriate at the time. In fact, by claiming that archaeology was just a type of anthropology, archaeology was almost swallowed up by anthropology. To date, a number of anthropology departments in the United States consider archaeology to be just a source of data for anthropological study. However, as scientific reduction is no longer seen as the correct direction for science (as discussed in chapter 1), and as anthropology has itself since fragmented into a variety of competing schools, there seems to be no reason for archaeologists to so ally themselves. As archaeology uses data from a wide variety of disciplines including anthropology, chemistry, history, geology, and literature, a good argument can be made for archaeology to leave anthropology behind and move on as a self-contained scientific discipline. That is not to say that philosophers are uninterested in the unification of science in some sense of the term, but the strict unification and reduction of science to physics advocated by the logical positivists is neither viable nor desired. In any case, Binford believed that, as an anthropological science, archaeology (and anthropology) would have two new goals: “Initially, it must be asked, “what are the aims of anthropology?” Most will agree that the integrated field is striving to explicate and explain the total range of physical and cultural similarities and differences characteristic of the entire spatial-temporal span of man’s existence” (Binford 1962, 217, emphasis in the original). Binford believed that archaeology (as an arm of anthropology) had two foci, the description of ancient cultural differences and similarities, and the explanation of how ancient peoples lived and changed over time. As for the first criterion, description (or explication), Binford saw that as well represented in the archeological community. By pursuing the goals it had before the 1960s, archaeology as a discipline had made great advances in the documentation of ancient civilizations: “Archaeology has certainly made major contributions as far as explication is concerned. Our current knowledge of the diversity which characterizes the range of extinct cultural systems is far superior to the limited knowledge available fifty years ago” (Ibid., emphasis in the original). Advances in data collection and analysis (as well as recording) made it possible for archaeologists to catalog and understand everything from architec- The New Archaeology 41 tural styles to idiosyncrasies in ceramic firing. So long as they could somehow show that these idiosyncrasies were real, and not artifacts of the archaeological process, archaeologists believed they could use those quirks to differentiate peoples both spatially and temporally. As for the second goal, explanation, archaeologists were not so selfcongratulatory: “So little work has been done in American archaeology on the explanatory level that it is difficult to find a name for it” (Willey and Phillips 1958, 5). The problem here was not only that archaeologists were not explaining their work, but more importantly, that they had no idea of what an archeological explanation was. Cultural historical archaeology was a purely descriptive exercise with results that could at best be used to show the viewer a static moment in time. The new archaeologists believed that this was a wrongheaded and simplistic approach. “To treat the site as if it were occupied throughout a phase is to treat it as if all rooms were built in a day, occupied continuously, and abandoned in a day” (Plog 1973, 193). According to the new archaeologists, the only way to understand a civilization would be to understand the fluid systems (and the fluid interactions between those systems) making up that civilization. Binford thought of archaeological explanation as more like studying the factors affecting natural coastline formation than as looking at still photograph of a day at the beach, and he understood culture as perpetually going through processual change: The meaning which explanation has within a scientific frame of reference is simply the demonstration of a constant articulation of variables within the system. Processual change in one variable can then be shown to relate in a predictable and quantifiable way to changes in other variables, the latter changing in turn relative to changes in the structure of the system as a whole. (Binford 1962, 217, emphasis in the original) While archaeologists in the past focused on the changes in one feature over time, Binford wanted archaeologists to study the system as a whole, not just the pieces making it up. New, or processual archaeology was a reflection on this new focus on dynamic systems and processes as opposed to static moments of time. A number of schools were born during the rise of processual history, and many of these were interested in the dynamics of systems. Although there are divisions within this group, the positions have in common a belief that the focus of archaeological (and anthropological by extension) research should be toward understanding how systems interact with one another. While General Systems theorists focused on the bigger picture and on the interconnectedness of all systems (see Willey and Sabloff 1974 for an introduction to this topic), others chose single systems as primary; single systems theorists included those who preferred the ecological approach and the proponents of social archaeology (see Renfrew and Bahn 1991 for a discussion of these approaches), choosing factors that they took to be significant as the basis of other interactions. In any case, the shift from static descriptions to fluid systems was seen as significant to the new archaeologists. 42 Chapter 2 Archaeological Systems The systems approach to explanation was very complex. A multitude of factors, known and unknown, all combine to form a system: “Culture is not a univariate phenomenon, nor is its functioning to be understood or measured in terms of a single variable—the spatial-temporal transmission of ideas. On the contrary, culture is multivariate, and its operation is to be understood in terms of many causally relevant variables which may function independently or in varying combinations” (Binford 1965, 205). Given that scientists do not generally know enough about a single natural process such as the weather to be able to explain or reliably predict its future states, it would seem to be practically impossible to give a full explanation of a complex living system, taking into account (or even recognizing) all of the relevant factors making up a community, culture, or the like. Archaeology would add to the aforementioned complexity an additional component. Archaeologists have no way of directly studying their subject matter. Archaeologists, no matter how scientific they are, can only study ancient cultures from a distance. The closest that a new archaeologist can come to his or her quarry is the collection and distribution of those artifacts and ecofacts (ecological residues left behind by human activity) that the dead culture left behind. Making that connection, between artifacts and the groups that used them, would not be easy. Binford set out to discover that connection. He wanted to understand the types of artifacts used and discarded by societies and believed that the key to archaeological explanation was in understanding how artifacts functioned and in uncovering the relationship between the artifacts and the cultural systems of which they were a part. Binford (1962) separated artifacts into three types, technomic (related to technology and work), socio-technic (related to social interaction), and ideo-technic (related to religious and other idiosyncratic factors), as well as stylistic attributes. Binford claimed that explanations of the past require archaeologists to understand more than the functions of artifacts. An understanding of past systems will only be possible when scientists are able to connect those artifacts to the technomic, socio-technic, and ideo-technic institutions of which they were once a part. Once archaeologists have understood how the different types of artifacts function, and how particular collections, or assemblages of these artifacts interact with one another within a given society, they will be in a position, given a certain assemblage, to understand the original society, or system. To this end, archaeologists looked to either ethnographic links between past and present or to computer based and statistical models as the means to make these connections. Binford believed that this correlative data would be found by means of ethnographic research and that archaeologists, as anthropologists, needed to understand fully the components making up living culture systems in order to be able to use artifacts to understand these relationships in dead civilizations: “[According to Binford,] the processual archaeology of the 1960s represented a shift away from explaining homologous similarities in the archaeologi- The New Archaeology 43 cal record to focusing on and explaining analogous similarities between living cultures and those of the past” (Lyman and Dunnell, 230). Archaeologists as historians would never have the proper training or enough of an understanding of culture change to accomplish this task, but Binford believed that archaeologists as anthropologists could develop this ability. When done properly, the result of this activity; the understanding of cultures and the development of archaeological explanations and laws; would come as a matter of course. In fact, Binford has since devoted a large part of his career to the study of living cultures; his hope is to eventually developing the requisite cultural database for archaeological explanation.25 Like new archaeologists, culture historians once believed that, given enough time and research, the data recovered would reach a critical mass and that the data would begin to make sense of their own accord. Early scientists studying artificial intelligence in machines and neural nets had a similar idea. They believed (and some still believe) that once a machine had taken in enough information it would be able to transcend its programming and learn and think on its own.26 A number of approaches, ranging from inputting encyclopedias, game instructions and strategies, news clippings or weather patterns on the net, have been employed to create thinking machines. If there is such a liminal boundary, separating data gathering from understanding, that information limit has still not been reached. Despite this problem, culture historians believed that when they had amassed large enough amounts of data, the past would come alive and speak to them: “Prior to 1960, in common with most of my colleagues, I had emphasized culture traits, trait lists, histories of sites and/or areas—all organized in a time-space dimension. I entertained the idea that the facts would speak for themselves once presented. I was carrying on what Kuhn (1962) calls “normal science” or solving jigsaw puzzles” (Martin 1972, 7). With the move to new archaeology, archaeologists believed that they would now be able to stop waiting for the data to show them the solution. Science, in the form of Hempelian covering laws, would give archaeologists a deductive approach to force the past to give up its secrets. These covering laws were seen as the perfect tonic to cure the ills of pre-1960s archaeology. Where archaeology once ended at the descriptive level, covering laws would allow for archaeological explanation. Archaeological Laws In order for this promise to be fulfilled, archaeologists had but to discover the nature of archaeological (or other relevant general) laws, develop hypotheses, and use deductive scientific methodology to test and hopefully confirm those hypotheses. The resulting explanations would give the archaeological record its voice and would make archaeology a science, just like physics. Physicists explain changes in pressure in a closed system by reference to general laws (with the attending changes in temperature and/or volume of the liquid or gas in that system). As Binford introduced in his 1969 article “Post-Pleistocene Adap- 44 Chapter 2 tations,” and filled out in his subsequent theoretical work and archaeological excavations (see the smudge pits case study in chapter 3) archaeologists would have a similar goal. They would plug in the appropriate data, artifacts, or whatever, and they would then be able to explain, predict, and fully understand culture change. According to John Fritz and Fred Plog, , at worst, the D-N model can be seen as heuristically significant, and at best: “[it] points the way archaeologists must travel if they are to contribute to the corpus of laws of human behavior” (Fritz and Plog 1970, 406). For reasons already discussed in chapter 1, physics was seen as the most advanced of the sciences and was considered the model for all prospective sciences to follow: “Many of their [e.g. the new archaeologists’] procedures are patterned after the physical sciences, and their ultimate goal is to discover laws of human behavior that are analogous to the laws of thermodynamics” (Flannery 1973, 50–1). The symmetry of explanation and prediction in covering law explanations means that, to Hempel, Boyle’s law serves to predict changes in pressure before they occur, given variation in volume or temperature as an antecedent condition, and can just as well explain those changes post hoc by citing those same variations. Given that the laws are universal, ceteris paribus, the same initial conditions will produce the same end state. Archaeologists were very excited by the promise that in creating complete archaeological explanations, general laws would also grant scientists of archaeology the ability to make predictions: “It is a major goal of archaeology to explain the past, but as indicated by Hempel and Oppenheim [1953], all scientific explanations, including those in archeology, help to explain contemporary events and to predict future events” (Watson, LeBlanc, and Redman 1971, 6). Given that certain laws have been established, an archaeologist could predict that, ceteris paribus, any newly found feature that displays the same relevant initial conditions (be they environmental, artifactual, or other) as one already understood must have the same explanation. Archaeology’s Changing Focus One of the most significant byproducts of the quest for archaeological general laws was a shift from the site approach to what has been termed regional or “landscape archaeology.” A culture historian’s limited focus was on the archeological site (or, at best to cataloguing the “movement” of a trait from one site to another). A new archaeologist, interested in explaining culture change, would instead focus on the big picture. In addition to the proposed theoretical benefits (discussed above) of a landscape approach, archaeologists found that this change in focus made sense for a number of methodological reasons. In the American Southwest, some sites were only temporarily (seasonally) occupied; so a report of a particular occupation zone would give little insight into the whole lives of the people living there (just as a study of a summer condo would not give an accurate account of its inhabitant’s life). Other groups lived in multiple sites throughout a region or alternatively engaged in significant amounts of trade. For The New Archaeology 45 these reasons, some archaeologists even grew to doubt the relevance of sitecentered archaeology, saying that it focused attention on urban life or that it focused on royalty or other forms of power structures, ignoring the lifeways of the poor or the majority. In some cases, it was impossible (especially in the Americas) to delineate a site clearly.27 Finally, the needs of the archaeologist outstripped the offerings of single sites. If archaeologists wanted to gather enough information to make universal laws, they needed to gather huge amounts of materials. The site-centered approach was seen as either a limiting factor on data acquisition or as a source of bias, and therefore this approach needed to be replaced. While the move from excavation to landscape archaeology did alleviate some of these problems, surveyors ran into a batch of new issues including: the impossibility of collecting a truly random or properly weighted sample, myths surrounding collection controls, and problems applying statistical analysis to small, bias plagued samples (see Cowgill 1986, 379–393). Despite the fact that these issues were quickly recognized and never resolved, regional surveys and surface surveys, once used to gather data for site excavations, became the object of new archaeology. Armed with a wider range of data, the new archaeologists began to look forward to the promise of Hempelian explanation. In order to come up with covering law explanations, these archaeologists knew that they needed more than data. They had to be able to use that data in conjunction with general laws. The new archaeologists set out to find these laws, and a number of reports demonstrating the promise of new archaeology came out in the late 1960s and early 1970s. These new archaeologists showed (with varying degrees of success) how archaeological explanation could work and also exposed a new range of issues and problems with covering law models of explanation. In order to understand new archaeology in practice, it will be necessary to show these questions in their proper context. As such, an exploration of new archaeology as practiced by new archaeologists will be the focus of the next chapter. Notes 21. See also Meggars 1955. 22. Archaeology in this area is so ideologically charged as to even affect the naming of archaeologists who specialize in the region surrounding ancient Canaan. As there is no politically neutral term for specialists in this region, the author has (reluctantly) chosen to use the name most often cited in the literature, despite its political overtones. 23. Among the best examples of this style of excavation are Petrie’s manual of field archaeology (1904), Petrie’s later excavations (1930), and Macdonald, Starkey, and Harding (1932). 46 Chapter 2 24. As Wylie (2002) points out, the dependence of carbon-14 on relative dating methods means that the resulting claims that carbon-14 dating is a separable (and therefore independent) source of data are thrown into question. 25. There are a number of issues surrounding ethnographic analogy, including some severe problems trying to show any connection between living cultures and their dead analogues, as well as issues surrounding the scientific validity of ethnography in either the discovery or testing phases of scientific explanation. These issues will be taken up later in chapters 3 and 4. 26. For more information on the search for computer intelligence, see Hayes-Roth, Waterman, and Lenat (1983). 27. Although this topic will be taken up again in chapter 6, problems delineating sites are relatively limited to American archaeology. Some severe problems have resulted from the export of American new archaeology as the ‘correct’ way to do archaeology to other geographical locations. Chapter 3 The New Archaeology’s New Archaeologists Introducing the New Archaeologists The best way to see how archaeology changed in response to advances in archaeological theory is to examine the results of these changes on or in the ground. The two analyses chosen (well known pieces by Lewis Binford and James Hill) are paradigmatic examples of the shift in theory and methodology that separated new archaeology from its predecessor, culture history. These studies are not meant to be in any way exhaustive as the only examples of new archaeology in practice. The two selections represent some of the best known scientific archaeological activity; each was chosen to highlight key features of new archaeology as well as some problems endemic to the discipline. Specifically, Binford and Hill conducted their studies with an explicit understanding of scientific methodology. Problems were stated, deductive hypotheses that would explain the problems were offered and tested, and conclusions were drawn based on those hypotheses. The contexts of hypothesis discovery and testing (also known by Hempel as the contexts of discovery and justification) were kept separate by each archaeologist to avoid the appearance of circular reasoning (i.e., that the same evidence used to create the hypothesis would then be used to confirm it), with varying degrees of success. The conclusions reached were based on the results of those tests, not on outside presuppositions or other idiosyncratic judgments. When the tests did not confirm the hypotheses, this information was not hidden. The hypotheses were changed, retested, and new conclusions were drawn in the manner of any other experimental science. Binford and Hill practically demanded that people view their methodology. The expectation was that, as with any science, others duplicating their “experiments” would arrive at the same conclusions. This belief is especially true of Binford, who would appeal directly to the scientific method to judge whether a colleague’s critique was fair and whether it would cause an equifinal result. In addition to their attention to scientific methodology, Binford and Hill also had in common a reliance on ethnographic data. However, they differed radically on the role that ethnographic data played in their investigation. The philosophical rationale for their debate will be discussed below, but it is worth taking a moment to explain the reasons that ethnography has become a central part of new archaeology. Ethnography became a central part of archaeology in America in the early 1900s. The “direct historical approach” to archaeology, used by a variety of archaeologists, has as a central tenet the belief that the past resembles the present: Methodologically, the direct historical approach involves the elementary logic of working from the known to the unknown. First, sites of the historic period are 48 Chapter 3 located. These are preferably, but not necessarily, those of identifiable tribes. Second, the cultural complexes of the sites are determined, Third, sequences are carried backward in time to protohistoric and prehistoric periods and culture. (Steward 1942, 377) This approach to archaeology is based on the premise that cultural evolution can be separated into distinct stages, and that these stages are essentially universal. If groups evolved differently, and if cultural interactions, social groupings, and societal goals differed on an individual basis, then there would be no foundation for this approach. However, proponents of this approach claimed that the data were on their side, and that they could show that there were consistent, universal stages of cultural evolution: Combined anthropological research in Peru, sparked by archaeology, has recently posited on stratigraphic grounds a sequence of roughly sequential culture epochs, including, from early to late, (1) the prehorticultural, (2) the incipient horticultural, (3) the formative (of local civilizations), (4) the florescent, (5) the fusional, and (6) the imperial epochs… Steward (1949) has carried this comparative approach still further, suggesting that a similar culture-era formulation may well fit the major archaeological and early historical periods of the Near East, Egypt, India, and China, as well as what Kroeber has designated “nuclear native America.” (Strong 1953, 393) Of course, the existence of universal epochs was not universally accepted, even among the direct historians. For instance, while Gordon Childe did believe that there were distinctly measurable stages of cultural growth, he believed that local factors and environmental differences made each region (and each region’s stages) unique. Despite Childe’s views, however, most of those archaeologists connected with the direct historical approach believed that gaps in the universal system would be filled in as time progressed, and the source of those data would come to a large extent from ethnographic study. More recently, the large role that ethnography has played in archaeology has been questioned on a number of fronts. For instance, Kent Lightfoot shows that written records generally reflect the biases of those who wrote them. He remarks on this as regards American colonial archaeology: Unfortunately, most colonial accounts were written from the perspective of affluent European men who documented little about the lifeways of lower class laborers and their relations with local native men, women, and children. Ethnohistorical research often provides little or highly selective information on the pluralistic laboring class in colonial settlements. Yet while these people were largely invisible in written documents, the material remains they left behind are recoverable and interpretable by archaeologists. Archaeology is the field of choice for examining the lifeways and interactions of poorly documented peoples in the past. (Lightfoot 1995, 201) Although social critiques of archaeology will be taken up again in chapter 4, it is worth mentioning that the biases of history writers are hard to ignore. In cases such as the one documented above, ethnography would be a hindrance to The New Archaeology’s New Archaeologists 49 archaeological attempts to understand the lifeways of the majority of the people living in colonial America. Later in his paper, Lightfoot talks about the impact of ethnographic data, and of ethnographers themselves, on prehistoric civilizations. In many cases, written records are given more weight than they should. In fact, Lightfoot asserts that archaeologists generally resolve conflicts between ethnographic and archaeological information by favoring written records over scientific data. These biases are the result of a number of factors, but Lightfoot believes the best solution would be to properly train archaeologists to understand the limitations of ethnographic data: “If every student of North American archaeology better understood the biases and limitations of different sources of written records, then many of the most flagrant abuses of direct historic analogy would probably cease, and the privileging of written records over archaeological materials might be curtailed” (Ibid., 206). In the end, Lightfoot stresses that ethnographic data will be (and should be) used, but that archaeologists will only gain knowledge of the past when ethnographic data are used correctly. Binford and Hill (below) each have differing beliefs in the utility of ethnography leading them to very different conclusions about its place in archaeological explanation. Another noteworthy difference between new archaeology and its predecessors is that the scope of Binford and Hill’s excavations was wider than the single site focus of pre-1960s archaeologists. In fact, for both Binford and Hill, very little excavation (in the classic sense of the word) was done. Instead, sites (or assemblages found at multiple sites) were surveyed and put into a wider context. This survey approach was in part successful due to the new archaeologists’ change in focus. Binford and Hill were both seeking explanations, as opposed to descriptions. However, Neither Binford nor Hill would say that descriptions did not occur in the course of their investigations; each archaeologist described well the site (or feature) under study, properly documenting relevant features or applying statistical analyses where relevant. However, the goals of each project were to explain a certain phenomenon present at that site (or in that region). The conclusions of each study would go beyond descriptive questions, such as: “what was there?,” moving to scientifically more interesting questions, like: “why was it there” and “what role did it play in a larger context?” Case Study: Lewis Binford – Smudge Pits and Hide Smoking In 1967, Binford noted the existence of a series of small pits, dated circa C.E. 470–1000, filled with burnt corncobs “from a number of Mississippian sites in the southern Illinois valley, northern Georgia, and eastern Texas area as well as from sites in the lower reaches of the Mississippi [River]” (Binford 1967, 6). He decided to focus his investigation on the explanation of these smudge pits and to demonstrate the proper use of analogy in an archaeological context. Analogy, as discovered by ethnographic study, had been a traditional source of data for anthropology and was being used in new archaeological explanations. Anthropological (or new) archaeologists, like their culture historian predeces- 50 Chapter 3 sors, believed that they could use present-day cultures as a sort of template to study past civilizations. For instance, if an archaeologist found an artifact similar in form to those used by members of a living culture it might be possible to infer the function of the unknown artifact from knowledge about its modern analog. While leaving aside, for the time, the larger question of whether there is any rationale supporting ethnographic data as a justification of analogous behavior between known and unknown groups and especially in archaeological explanation, a question that to be revisited in chapters 4 and 5, Binford believed that ethnography did have a place in covering law explanations. In fact, Binford demanded rigorous testing of the assumptions behind the proposed analogy, (what Binford refers to as middle range theory) in order to use those analogies as a part of his explanada. However, Binford saw that archaeologists were assuming analogical relationships with very little evidence, meaning that this source of data was being applied inappropriately in the field. When Hempel applied covering law models to history, he was very specific as to how a proper explanation should work. He separated hypothesis formation from explanation, or hypothesis testing. While the latter part was necessarily quite rigid (a criterion that many archaeologists such as Hill either did not care for or did not see as important), Hempel had an “anything goes” approach in regard to the former: We have tried to show that in history no less than in any other branch of empirical inquiry, scientific explanation can be achieved only by means of suitable general hypotheses, or by theories, which are bodies of systematically related hypotheses. This thesis is clearly in contrast with the familiar view that genuine explanation in history is obtained by a method which characteristically distinguishes the social from the natural science, namely, the method of empathic understanding: The historian, we are told, imagines himself in the place of the persons involved in the events which he wants to explain; he tries to realize as completely as possible the circumstances under which they acted and the motives which influenced their actions, and by this imaginary self-identification with his heroes, he arrives at an understanding and thus at an adequate explanation of the events with which he is concerned. (Hempel 1965, 239, emphasis in the original) Ethnographic analogy is, for Binford, a means to gain Hempel’s empathic understanding. Although questions remain as to the proper role of ethnography as a source of analogical data, archaeologists continue to use ethnographic data for comparative purposes and as a basis for explication. A wide range of anthropological tools are available to show that an ancient artifact with a form analogous to a modern tool will share that present tool’s function. These techniques do sometimes bear fruit, and Binford did not seek to dismiss their value in this regard. He recognized that ethnography’s only value was that it could be used to generate hypotheses. Although it might feel right to move from the assumption that ethnography can give hypotheses about the function of ancient artifacts to the belief that ethnography can prove that function, this move would be to commit the fallacy of wishful thinking. For Hempel, ethnography could give archaeologists the former, providing them with a certain connection (which he called empathy) to past peoples, but The New Archaeology’s New Archaeologists 51 not the latter. Empathy was not to be confused with the activity of science. Hempel did not disregard this empathy, nor did he specifically dismiss any potential source of hypotheses. He drew the line, however, at claiming that these sources constituted explanations: “This method of empathy is, no doubt, frequently applied by laymen and by experts in history. But it does not in itself constitute an explanation; it rather is essentially a heuristic device; its function is to suggest psychological hypotheses which might serve as explanatory principles in the case under consideration” (Ibid., 239–40). Hempel believed that scientific discoveries can come from analogies, dreams, or falling fruit, but falling fruit is not science, it is inspiration. Binford agreed with Hempel that ethnographic analogy could only be properly used in the context of discovery, and that it was an error for archaeologists to depend on ethnography in the context of justification: Analogy serves to provoke certain types of questions which can, on investigation, lead to the recognition of more comprehensive ranges of order in the archaeological data. In short, we ask questions about the relationships between types of archaeologically observable phenomena that had possibly not been placed in juxtaposition or viewed as orderly. In doing so we can develop a common ‘explanation’ for observed variability in a number of formally independent classes of archeological data, and thereby we can approach more closely the isolation of systematic variables which operated in the past. (Binford 1967, 10) In other words, Binford believed that ethnographic analogy has an important role to play in archaeology. Connecting the known present to the unknown past would narrow the range of inquiry to a manageable level. In fact, Binford was convinced that by refining ethnographic data, the archaeologist could devote more time to hypothesis testing and would be able to do better archaeology. As a result, Binford has spent a large amount of his professional career studying living peoples in order to see better the connections between living people and their artifacts and to be able to apply this information to past cultures as well: “Such studies have as their aim the delineation of behavioral correlates for material items, and the purpose of archaeologists undertaking such research has been to maximize their interpretive powers by increasing their knowledge of living peoples—that is, to make more secure the analogies they draw between lifeways of peoples known archaeologically and those known ethnographically” (Binford 1968:13). Ethnography was one of many tools that the archaeologist had to understand the past. Given the scope of Binford’s research, he clearly believed that this study would give archaeologists more tools with which to investigate the past. Even though Binford believed that ethnography had a part to play in archaeological analysis, he clearly believed that there was more to the latter than simply fitting ethnographic data to past civilizations: “Fitting archaeological remains into ethnographically known patterns of life adds nothing to our knowledge of the past. In fact, such a procedure denies to archaeology the possibility of dealing with forms of cultural adaptation outside the range of varia- 52 Chapter 3 tion known ethnographically” (Ibid., 86). Although ethnography could be of use to archaeology, archaeology would be severely limited if it were limited to what ethnographers have discovered about culture. Archaeology’s goal is to understand extinct civilizations. As such, archaeologists hoped to learn about cultural systems that were similarly unique. In those cases where ethnographic data did exist, Binford did not argue that archaeologists should not infer analogous function upon seeing analogous forms. He believed that in many cases science alone cannot provide enough material to form a hypothesis or to choose which hypothesis to test. For Binford, a bunch of ash filled holes proved to be the sort of instance where ethnography and analogical reasoning would provide the additional material needed to form initial hypotheses: On the basis of (a) the convincing correspondence between the formal attributes of smudge pits as known archaeologically and smudge pits used in smoking hides as known ethnographically, (b) the strong positive analogy between the distribution of smudge pits in which corncobs were used as fuel and the use of corncobs as fuel for smoking hides as documented ethnographically, and (c) the relatively late archeological documentation for the use of smudge pits, which would make continuity between the archaeological and ethnographic periods reasonable, we postulate that the archeologically-known features described were in fact facilities employed in the task of smoking hides by the former occupants of the archaeological sites on which they were found. (Binford 1967, 7–8) As stated above, analogical reasoning (or any other method of producing suitable hypotheses) is very appropriate in that analogy can move beyond science and help narrow the field of potential hypotheses. In some cases, analogy can help to quickly rule out some possible contenders and can help the archaeologist decide which hypotheses to test first. An analogy would not be used to test the hypotheses it suggested; however, analogical reasoning does provide a means to initially separate hypotheses that look promising from the rest of the field. In this case, Binford found a number of pits located throughout the area he was studying and delineated those bearing certain qualities from other otherwise similar pits. Binford did this to set the limits of his inquiry: “In summary, smudge pits are a class of archaeological features sharing (a) small size, (b) contents composed diagnostically of carbonized corncobs, lacking kernels, and (c) contents exhibiting a primary depositional context” (Ibid., 6). Binford then cited (Ibid., 6–7) ethnographic reports of how these pits were used by a wide variety of peoples in the Southeast and Great Lakes regions by ethnographic study of known cultures and proceeded to explain how these smudge pits would not have been able to be used in other contexts (due to their form, size, contents, location, etc.). Binford felt confident that, given the strength of the analogy drawn out in his study, he could present as a hypothesis that the pits he found were smudge pits. Hill was mistaken in his assumption that these initial hypotheses could continue to carry any weight once the testing began. In that regard Binford showed himself to be a good scientist. Instead of becoming attached to his ethnographic The New Archaeology’s New Archaeologists 53 hypotheses, Binford let the results speak for themselves. This hypothesis would then need to be subjected to rigorous testing and would be confirmed by Binford as a good explanation of the function of these pits. Here, Binford believed that he was led to the correct explanation by his use of analogy and his adherence to the “positivistic philosophy of anthropology and archaeology” (Ibid., 10). It is evident, however, that had Binford’s initial hypothesis not been confirmed he would not have faulted his methodology. To Binford, new hypotheses would come from his initial failed hypotheses; a contrary result would have been as useful to Binford as a confirming instance: It should be pointed out that these gains may obtain regardless of whether the original analogy led to a correct postulate. In short, I do not view interpretations, or syntheses of interpretations as an end product of our investigations; on the contrary, we should be seeking generalizations regarding the operation of cultural systems and their evolution – something which has not been described ethnographically nor thus far achieved through the observation and analysis of contemporary events. (Ibid., 10) Binford believed analogy to be a study) one of his hypotheses was form more hypotheses and these stage. Once the hypotheses were would take over: means, not an end. If (in the course of this disconfirmed, he would use the new data to would be tested as well in the justification offered, the justification (or testing) phase, The procedure which should be followed in refuting or increasing the probability of the validity of the proposition would be as follows: (1) Determine if there are any spatial correlates of the activity of smoking hides; in other words, determine if the activity was regularly conducted in any particular location? If so, determine whether or not the smudge pits exhibit such a distribution. (2) Determine if there are any temporal correlates of the activity of smoking hides; in other words, determine if the activity was regularly conducted at any particular period of the annual cycle? If so, determine whether or not the smudge pits exhibit such an association with respect to relevant seasonally variable phenomena. (3) Determine if there are any formal correlates of the activity with respect to other implements or facilities which were employed as parts of a set which also included hide-smoking pots. Was hide smoking normally conducted at the same place and at approximately the same time as the manufacture of clothing from the hides? If so, then there should be demonstrable concomitant variation between the incidence of smudge pits and implements used in clothing manufacture, such as needles. (4) Determine if there are any other activities which employed facilities which shared the same formal attributes as observed in hide smoking pits. If so, then the specific postulate could be refuted, but a more general one could be stated which could then be tested along the dimensions of time, space, and form. (Ibid., 8) 54 Chapter 3 It was this justification procedure that Munson, a critic of Binford attempted to use in order to propose an alternate more general hypothesis to describe the function of the “smudge pits.” In 1969 Patrick Munson looked at the smudge pit data that Binford analyzed in 1967 and believed that Binford had overlooked some pits that were important for a proper analysis of the data. Munson added those pits and reanalyzed the data, offering an alternative explanation for the pits that was seen as a criticism of both Binford’s conclusions and of his methods for reaching them. Although Munson’s criticism was important, and an analysis of that critique follows, it is important to mention that Munson’s critique had nothing to do with Binford’s methodology. In fact, this was a clear vindication of Binford’s methods. Munson’s study mirrored Binford’s. Munson used the same method of hypothesis generation by use of ethnographic analogy and the same method of hypothesis testing, and, like Binford, showed that his conclusion was based strictly on those test results. According to those new results, Munson claimed that the data pointed to an alternate (more generalized) explanation for the existence and function of Binford’s smudge pits: “The arguments which have been marshaled to support the position that the function of these features was that of smudge pits are convincing. However, Binford’s more specific postulate, i.e., that the features were used for smoking hides, is challenged as being too narrow” (Munson 1969, 83). Munson’s alternative hypothesis, which he based on additional examples and ethnographic data (Ibid., 83–4), was that the smudge pits (both Binford’s and Munson’s additions) were used for pottery smudging as well as for hide smoking. To justify this conclusion, Munson explained that corncobs have been seen through ethnographic study to have been used for smudging pottery in ethnographically studied cultures. He believed, therefore, that Binford’s smudge pits were also used for this other activity. Munson paraphrased Binford’s conclusions, substituting “pottery smudging and/or hide smoking” for Binford’s uses of “hide smoking” in the appropriate places (Ibid., 84). As a result, Munson’s critique should be seen as confirmation of Binford’s approach, if not his conclusions. Although Binford’s methods were not under fire by Munson, there is, in fact, a real conceptual problem, not mentioned by Munson, but that Munson’s article brought to the fore: If Binford’s and Munson’s accounts of the smudge pits are equally worthy and if further investigation does not show one to be the better, the result would be to have incompatible explanations of the same phenomenon with no clear way to choose between them. Binford remarked on this in his reply to Munson and pointed out that were this the case, it represented an instance of equifinality. The obvious invitation to relativism entailed by equifinality is, of course antipathetic, to the scientific attitude shared by Munson and Binford. Indeed, Binford had remarked on this in his original 1967 paper and restates it in his reply to Munson: “As far as the truth value of any given proposition is concerned, the presence of equally defensible alternative propositions does not in any way diminish the potential truth value of any of the alternatives. The only method available for testing the truth value of equally valid proposi- The New Archaeology’s New Archaeologists 55 tions is to test them against relevant materials from the archaeological record” (Binford 1972, 53). In a situation where two outcomes would be equally likely, there could, in Binford’s view, be no decision between them based on anything other than scientific empirically verifiable testing. In this example, Munson posited that there could be no substantive test to choose between the two alternatives. If this were truly the case, in the absence of other data, Binford’s conclusion would have been to accept a more general inclusive explanation, one that would include both possible variants, while weeding out others as less likely: “In cases where hypothesis testing of alternative propositions such as hide smoking versus pottery smudging proves inconclusive, one may then be forced, as I previously suggested (Binford, 1967, p 8) to offer a more general and inclusive proposition. In this case the more general inclusive proposition would be that the observed archaeological features were smudge pots as distinct from lighting fires, roasting pits, etc.” (Ibid., 57). In cases where two equally valid explanations existed, the more general explanation would have been the only alternative. However, despite Munson’s assertion, Binford suggested that equifinality was not at issue here. Binford could give a reason to choose between the alternatives. Binford went through Munson’s additions and showed that given a full ethnographic study of the area and materials used for hide and pottery smoking and smudging, although corn cobs could be used for smudging pottery, corn cobs were not the smudging tool of choice of the people being surveyed. They were simply one (and not even necessarily the best) of many smudging punk fuels, and the appearance of corn cobs in every pit would not have occurred were there other, equally useful alternatives available. In addition, Binford argued, only one of Munson’s examples involved pottery being smudged in a pit of the sort that Binford mentions. Pottery smudging, therefore, was an activity that could be done in a variety of ways, using whatever fuels and facilities were on hand. Hide smoking, however, involved a very specific combination of materials, including corn cobs. As a result, Munson’s explanation was not as specific as the facts permitted. Although, according to Binford’s original article, Munson’s generalized hypothesis could not be disproved using testing, Binford did not find Munson’s hypothesis to be as likely as his own. As a result, Munson’s solution ends up not being equiprobable; so equifinality is avoided. Binford’s analysis is a fine example of new scientific archaeology being done correctly and successfully: Precise and interesting hypotheses are formulated and tested against the data in the light of strong ethnographic analogies. The Binford-Munson dialogue addresses just the right questions of fact and conjecture, and in the end a parsimonious conclusion is reached for the right scientific reasons. But, in fact, the retrospective consensus of the archeological community has it that the very strengths of the smudge pit case point out severe limitations of the new archaeology. One reason for the success of Binford’s study was that it focused on a very limited phenomenon; the existence of a certain sort of pit within a limited and well defined area. Binford was able to take a neat, easily defined, and delineated problem, and use scientific testing to confirm his hypotheses. But, small, clearly delineated problems are not supposed to be the 56 Chapter 3 subject of new archaeology. New archaeology in practice seems to be restricted to precisely defined and limited explanations while the point of new archaeology was supposed to be a move to general trends and large issues. As the questions at issue become more complex, the subject cultures multiply and the geography more extensive and less precisely bounded, the hypotheses lose the precision upon which testability depends. However, the reason Hempelian models were sought was because of their power to create and use general archeological laws. Covering law explanation and prediction depends completely on explicit laws. But when the field data exhibit the messy heterogeneity so essential to archeological exploration, laws are either unattainable or so vague as to be vacuous. There just do not seem to be many nontrivial archeological laws that have explanatory power. Case Study: James Hill—Broken K Pueblo James Hill’s focus was on an even smaller area than either Binford or Munson covered in their aforementioned surveys (the Broken K Pueblo site takes up no more than one square mile), but the goal of his 1967 study was much more ambitious. While Binford and Munson desired only to explain the existence of a single feature, Hill wanted to use the data from a pueblo site to explain inheritance patterns and migration tendencies in the northern American Southwest. Hill’s hypotheses would include precise descriptions of male and female roles in life and work throughout this area and would, in addition, offer a comprehensive understanding of regional population shifts. In order to explain these features, Hill chose a site that he deemed representative of the pueblos in the region: “The focus of the analysis is Broken K Pueblo, located eleven miles east of Snowflake, Arizona…This site is a 95-room, single-storied, surface masonry pueblo, dating from about A.D. 1150 to 1280” (Hill 1968, 104). Hill had two objectives when he came to Broken K. His first goal was to use the data obtained from his survey excavation to reconstruct the life ways of the people living at the pueblo during its occupation. He intended to do this in a scientific manner, by gathering data, establishing hypotheses and testing them: This paper attempts to describe a limited aspect of a cultural system that existed in eastern Arizona during the thirteenth century A.D. My primary concern is threefold: (1) to describe the locational patterning of various kinds of cultural features at the site, (2) to offer explanation of the demonstrated patterning in terms of past behavior, and (3) to test the accuracy of these explanatory propositions. This is intended as a case study in the demonstration of “activity-areas” or “activity-structure” within prehistoric communities. (Ibid., 103–4) Hill’s focus on explanation and testing makes it clear that his intent was to go far beyond simple description. Using the tools of culture history, Hill would not have been able to go beyond narrative. However, Hill wanted to explain the structure of this community in precise, general, and empirically testable ways. The New Archaeology’s New Archaeologists 57 His study was presented as an explanation of the life ways of the thirteenth century inhabitants at Broken K. Hill’s second goal was to use his results as a model for others to follow. He wanted to show how a proper scientific excavation and a proper archaeological explanation could work. Although he excavated a site that, like most American sites, was only sparsely and shallowly stratified and whose analysis depended on ethnographic data, Hill believed his theories to be nearly universal: This paper also serves as an example of a general methodological approach that should prove useful in nearly all archaeological studies, regardless of the kinds of archaeological remains being considered (cf. L.R. Binford, 1957 [sic 1967]). I have attempted to illustrate a means by which it is possible to go beyond merely making inferences about past behavior; we can also test these inferences, and gain confidence (or lose confidence) in their validity. This can be done, even in those cases in which direct ethnographic evidence is not available. (Ibid., 104) Hill’s purpose in conducting this study was to identify matrilineal social structure by focusing on mother-daughter transmission of a variety of elements, including the distribution of ceramic design elements and of “pottery-types, firepit-types, storage pits, ‘chopper’-types, and animal bone” (Hill 1970, 58). His plan was first to separate spaces in the site into room types and family units. Next, he would show that different activities occurred in different room types and that these activities were separated by sex. Finally, he argued that stylistic elements of those artifacts associated with female activities (unlike those used by males) were transmitted from parent to child (mother to daughter). Like Binford, Hill believed that discovery and justification needed to be independent of each other. However, Hill chose to use ethnographic data in the testing phase (as opposed to Binford, who used it in the discovery phase). Hill chose to discover his hypotheses by an analysis of function and to justify them by analogy to ethnographic data, collected recently from nearby indigenous groups. According to Binford, and arguably, to Hempel, ethnographic data are suitable as sources of hypotheses but are un-testable. There is no way to give any acceptable level of assurance that the cultures being compared are indeed analogous in the relevant manner. During Hill’s analysis of room types, Hill first claimed to be able to determine the function of each room type by its size as well as ceramic and other artifactual evidence found during the excavation. However, a serious problem with Hill’s architectural analysis stems from the fact that the data given do not seem to bear out Hill’s divisions of rooms in this manner as being anything other than arbitrary. Hill claims, based on size, that two types of rooms exist and that this distribution was not random (Hill 1968, 108). However, Hill then notes on the same page that, based on the actual data, the distribution might be trimodal. Hill dismisses this, claiming that only seven rooms, which he did not consider to be enough to be significant, were of the “extra-large” variety. So, Hill collapsed these extra large rooms into the large category. At the same time Hill concludes that, in addition to small and large rooms, a third category of “special rooms” 58 Chapter 3 exists. As there are only four of these special rooms, his earlier dismissal of seven as statistically insignificant is puzzling at best. In addition, at another point, Hill seems to contradict his previous statement, stating that six of the rooms found were neither large nor small but were somewhere in between (Ibid., 109). For analysis, Hill assigned these rooms as either large or small, depending on characteristics he does not mention. Based on this information, it seems clear that Hill could only have divided the rooms as he did based on ethnographic analogy to current (Hopi and Zuñi) pueblo dwellers who have small rooms, large rooms, and special rooms, each with functions that he attributes to the analogous rooms as Broken K. Hill’s analogy from ethnography would have also forced him to dismiss the existence of extra large rooms (which had no modern analogue), and to find ways to designate of each of the aforementioned six “hybrid” rooms as either large or small. Regardless of the veracity and basis of his initial hypotheses, Hill then moved to the hypothesis testing phase: “After designating the rooms functionally, the next step was to test this proposition of functional equivalence; and this required some additional (and independent) data” (Ibid.,136). Hill used ethnography as the source material for his independent test by comparing his findings to data taken from local Native American pueblos. Hill did this while believing his initial hypothesis to have been purely functional, and held that his use of functional analysis was separated from his use of ethnographic data in the testing phase. Had Hill’s beliefs been true (had his hypothesis been derived from a strictly functional analysis), then Hill could have made the argument that he had kept separate the phases of discovery and justification (a separation that he stated needed to be observed). However, as there was no independent (nonethnographic) way to confirm that the ethnographic data lend credence to Hill’s hypothesis, then he has ethnography justifying the use of ethnography (a case of circular reasoning). In this case, Hill was unable to follow through on his claim that his ethnographic and functional analyses were separate. Ethnography not only played an enormous part in his testing stage, but also in developing many of his hypotheses: from his decision to divide and categorize rooms by size in a manner analogous to contemporary pueblo settlements, (Ibid., 108) to his designation of these rooms’ functions and names along contemporary terms, to the assumption that male and female activities would have been the same in the tenth to fourteenth centuries as they are today, in accordance with ethnographic reports. Hill was so committed to his belief in the explanatory power of ethnographic analogy that he even defended it when the data seemed to be in conflict with his findings. When the archaeological record did not result in findings consistent with Hill’s analysis, Hill explained that examples of disconfirmation in his findings were a way to track culture change. When supposed analogues were disconfirmed, Hill hypothesized that societal change was responsible for these discontinuities. For instance, the differences in room shape between the Broken K and modern pueblos could be explained as idiosyncratic changes in those particular cultures over time. Although this hypothesis may or may not be true, it practical- The New Archaeology’s New Archaeologists 59 ly demands that archaeological data are only useful when ethnographies did not exist or did not apply. Since archaeological data would be ignored whenever they conflicted with ethnographical data, archaeological data would only be used to fill in gaps left by the ethnographic reports: “In addition to providing evidence of culture-change, they provide new information that cannot be obtained from ethnographic materials” (Ibid., 137). In a follow up to his 1968 study, Hill (1970) used ethnographic evidence to move beyond his preliminary findings with the goal of designating certain activities (located in particular room types) as “female” activities (as opposed to “male” activities): “There would seem to be little doubt that ceramic designelements and pottery-types were associated with female activities primarily, as this is the case ethnographically” (Ibid., 62). Cooking, similarly, was designated as a female activity: “Although Hopi and Zuñi women may not actually construct their firepits, it may well be that they have some control over their planning and construction. This is a reasonable idea, since firepits are used primarily for cooking, and cooking is a function of women. This would probably be correct on a worldwide basis too” (Ibid., 62). Since these (and other activities) were designated as female, Hill further explained that if female activities took place in certain rooms (due to the exclusive presence of “womanly” artifacts), he could show that those rooms must have been associated with females. Next, Hill looked at the distribution of what he called female and male stylistic elements. He concluded that a number of unique female design elements on ceramics could be localized into areas of the pueblo. Hill believed that these separate areas represented family units, and that each family’s “female” rooms contained ceramics and fire-pits that were distinct from those in other family units. Therefore, according to Hill, the family units were probably uxorilocal (matrilineal). As there was not enough data to know whether male stylistic elements were also distributed randomly or not, Hill (Ibid, 64) could not rule out a duolocal residence pattern, but he thought that unlikely. Hill set the site’s period of occupation from circa C.E. 1050–1300 and posited that the ceramic traits for each family unit remained essentially unchanged for the duration of the site’s occupation. As people of that time period lived a minimum of sixty five years per family unit, Hill found the relative stability of ceramic traits over the two hundred year period reinforced his own conclusion. “Considering all of the preceding evidence, it seems quite likely that [each unit] constituted a large uxorilocal residence unit” (Ibid., 64).28 In the absence of any archaeological evidence pointing to women as potters and cooks, Hill’s reliance on ethnographic data was fundamental. As such, it can be argued that Hill’s conclusions are as much a product of his modern research choices and his biases as they are a product of his survey. If Hill had found modern ethnographic evidence that men cooked (or if they were involved in making ceramics, fire-pits, etc), then the study might as easily have pointed toward a virilocal (patrilineal) society, given the same data. Alternatively, it might be possible to imagine that each family had either a male or female craftsperson who specialized in ceramics. This person would teach his or her successors, leading to the same stylistic integrity, without allowing for any conclusions to be 60 Chapter 3 drawn about the site’s societal makeup. From the preceding scenarios, it should be clear why too much reliance on ethnography is problematic. As Binford believed that archaeologists are supposed to go beyond ethnographic evidence, this example would serve as proof that ethnography should not be used to test explanations. Archaeologist Steadman Upham sees the problems with ethnography as far more fundamental than did most archaeologists, including Binford or Hill. He rightly points out that the biases involved with ethnography are potentially so disastrous as to make suspect any conclusions reached on their behalf. As he shows in his summary of North American Indian tribes: Cultural systems do not exist in isolation and it is undeniably naïve to suggest that the core of Puebloan organization has remained relatively unaffected by changes that occurred even during the “ethnographic present,” let alone during periods prior to 1880. In fact, a body of literature is now accumulating that indicates it is precisely because the Pueblo were able to change and modify organizational structures that they were able to survive periods of successive domination. (Upham 1987, 267–268) Not only are changes brought upon by well known factors such as investigator bias and recording biases, but Upham points out that rapid change in the face of outside factors, such as invasion, may even be a deliberate, effective adaptive strategy of survival. Given the wide range of problems surrounding ethnographic data, Upham criticizes attempts to use ethnographic analogy as the basis for hypothesis formation, much less in the testing phase: “It is clear, at least in the studies of…Hill that the correlation between ceramic manufacture, mother-daughter transmission of design styles, matrilineality, and matrilocality would not have been suggested if the ethnographic literature on the Pueblos were absent” (Ibid., 268). In all of these cases, and in all uses of ethnographic analogy, Upham posits that no one has been capable of living up to the aforementioned standards set by Binford. Ethnography is rarely used only in the process of hypothesis formation and excluded from hypothesis testing, even when this intention is stated and thought to be met: “Although hypotheses are tested using archaeological data (i.e., independently of the ethnographic record), the evaluation for goodness-offit is accomplished by recourse to the ethnographic record” (Ibid). Regardless of the high hopes of archaeologists, the real dangers associated with the use of ethnographic analogy make its utility suspect at best, and worthless at worst. 29 Ironically, as a result of ethnography’s problems, analogy in the absence of ethnography would seem to actually be safer than the ethnographic variety. Although at face value ethnography would seem to give more hints into an otherwise silent past, the amount of bias associated with ethnography makes any conclusions suspect. Although analogies in general can be subject to the same categories of investigator and ethnocentric biases, the gap between present and past is wide enough to preclude the scientist from being as sure about his or her conclusions. These more tentative analogical ascriptions would thereby be subject The New Archaeology’s New Archaeologists 61 to more rigorous testing, and would prove more valuable as a source of data than would their more modern analogues. Hill’s explanations were not all based on ethnography. Hill used hypotheses from a variety of scientific sources to explain how changes in the environment affected the settlement patterns at the Broken K Pueblo, and the nearby Carter Ranch site (excavated by William Longacre): A number of sociological changes occurring at this time (about AD 1050– 1300) in the Southwest seem to have been promoted by a minor environmental shift. This shift is documented by the nearly simultaneous occurrence of at least the following events: 1. A shift from a relative abundance of arboreal pollen to a relative abundance of nonarboreal pollen. 2. A shift in the width of tree-rings, from wide to narrow (probably reflecting a shortage of effective soil moisture). 3. A widespread cycle of erosion. Evidence suggests that these events were related to one another, and they may reflect conditions inimical to agriculture. (Ibid., 108) Hill’s hypothesis, that environmental factors led to the decline and eventual abandonment of those sites seems to be well confirmed, given the information gathered from the aforementioned tests. Here, Hill’s analysis of a variety of data sets led him, without explicit recourse to ethnography, to show that these environmental changes occurred, and that they were factors relevant to the decision to abandon Carter Ranch and Broken K Pueblo. Although Hill would not be able to give a full Hempelian explanation here, as the laws Hill appeals to (laws connecting effects of general environmental trends on specific events), are sketchy,30 Hill is able to at least offer an “explanation sketch” of this abandonment phenomenon. However, a closer look at Hill’s conclusions reveals that Hill was walking a fine line between creating overly-specific laws (governing just the site) and useless generalizations. Hill never mentioned which laws he was trying either to create or understand in the context of this study. If Hill meant only to refer to the Broken K and Carter Ranch sites during the period in question, he could certainly come up with laws governing the abandonment of those sites. However, any laws that would be exact enough to explain why people specifically left Broken K and Carter Ranch would be too specific to be useful in other contexts. Many potential factors affect peoples’ decisions to stay at a site or move on. The aforementioned factors at Carter Ranch and Broken K may be good reasons to have left a site, but in another context, other factors might prove more relevant. If, on the other hand, Hill wanted to make his explanation universal, the best he would have been able to do would be to say that people abandon sites when they run out of plants, have no fertile soil, and run out of water. Although this law would usually be considered true, it is also trivial. Here, as in Binford’s “Smudge Pits,” the status and relevance of Hempelian Laws to archaeological explanations is in question. Although both Hill and Binford provided paradigmatic case studies of Hempelian archaeological explanations, neither is able to provide the kinds of explanations desired by archaeolo- 62 Chapter 3 gists and promised by logical empiricists. This shortcoming has led to a number of theoretical questions about the role of covering laws in the future of archaeology. The New Archaeology in Practice: Problems and Issues A Lack of Laws As has been noted above and in the previous chapters, the drive to produce general laws, Hempelian explanations, and deductive science, ended in a sense of disappointment. At best, covering laws are scarce for all of science. Archaeologists were very frustrated, and this search for laws was seen by some as an ideal quest, and by others as quixotic fantasy. Regardless of any particular archaeologist’s stance on the status of general laws (an issue that will be addressed again in chapters 4 and 5), the new archaeology produced some of the best and some of the worst archaeology to date. While some archaeologists (including Binford) managed, at least for very well defined, small, and specific problems, to produce the kinds of explanation sketches believed possible by new archaeologists, in most cases, the results were less impressive: In some cases, the statistics used have been so much more powerful than the raw data deserved, that one feels he is watching a grenade launcher being turned on a field mouse. In other cases, it seems that in order to discover a ‘natural law’ in the allotted six weeks of his field season, the investigator was forced to attack a problem of the utmost trivia; this has produced a series of low-level generalizations that some critics have called “Mickey Mouse laws.” (Flannery1973, 51) On rare occasions, archaeologists like Binford could produce scientific explanations of their data. However, for the most part, general laws of archaeology were shown to be too specific to be useful outside of their precise context (as in Binford’s case), so general as to be completely trivial, or nonexistent. This lack of success initially caused archaeologists to believe that the lack of laws were due to problems with archaeology as a science. However, exactly the same issues arise in the other sciences, and archaeological responses to these problems have a great deal in common with current theories of philosophical explanation. Computers were seen by some archaeologists as a means to salvage the covering law model. Since cultural systems are complex and archaeologists are unable to keep all of the requisite data in their heads, the limitless storage capacity of computers were seen here (as in many fields) as the solution. General laws even sound a lot like computer code; computer programs create “worlds” with specified laws. Data are then presented to these worlds, and the results of these interactions are computed as the output. Archaeologists had already become aware that advances in scientific methodology had resulted in a large increase in the types and quantities of data coming in from the field. With the powerful data The New Archaeology’s New Archaeologists 63 processing facilities that computers promised, archaeologists would be able to keep and manipulate all of the different data sets to make predictions, and to better understand interconnected systems. Just as modern meteorologists use general laws of weather and constantly updating satellite images to predict future weather patterns, archaeology could program general laws into simulators, input their data, and analyze the complete record: In the future, it should be possible for archaeologists to feed data about small selected subsystems, which they have studied intensively, along with other relevant archaeological information, into the computer, enabling them to produce a simulation, or a series of alternate simulations, of the entire system. Archaeologists can then go back and test the predictions of the simulations against other subsystems. They can also hold constant various factors and develop new simulations which will show what the systemic effects of different changes would be. Predictions and tests concerning the nature of systemic change through time can then be made. Thus, even with data on whole systems incomplete, archaeologists should be able to test hypotheses about the evolution of such whole systems (Flannery, Lecture at Harvard University, 1971). (Willey and Sabloff 1974, 193) Although computer models have since been adopted by a number of excavations, archaeologists have, for the most part, used them for cataloguing purposes, or to make 3D reconstructions of sites, using GIS software. Predictive modeling has not been successful, probably due in part to the lack of general laws in archaeology. Predictive modeling can only work in situations where both initial conditions and general laws exist. As a result, Flannery’s ideas regarding computer aided prediction are probably a matter of wishful thinking. The lack of archaeological laws and the utter failure of archaeological prediction caused many early supporters of the new archaeology to become disillusioned: “The law-and-order archaeologists receive their nickname from the fact that they not only believe that Carl Hempel rose from the dead on the third day and ascended to heaven—where he sits at the right hand of Binford—but, to use their own words, ‘have made the formulation and testing of laws (their) goal.’ Thus they view the archaeologist’s major responsibility as developing the body of general laws necessary for a covering law explanation” (Flannery 1973, 50). Explanations and Explanation Sketches Archaeologists were especially disillusioned with their inability to create Hempelian explanations. Processual archaeology represented data as a web of interconnected systems. Even with the aid of today’s computers, there would be no way to accurately represent all of these factors correctly. So the complexity of the data means that complete explanations (desired by Hempel) could never be given. At best, Binford’s model would limit the new archaeologist to offering what Hempel would call “explanation sketches” of past civilizations, given that it is possible to learn enough from the artifacts to even tentatively connect them 64 Chapter 3 to systems long since gone. This disillusionment led some archaeologists (such as Flannery) to modify archaeological, covering law explanations, making them less strictly explanatory. Instead of having the goal be the delivery of a complete explanation, the Hempelian explanation sketch was seen as adequate. However, although this goal is more realistic, its attainability has a steep cost. Only a complete explanation would be capable of generating the sorts of laws and explanations desired by the new archaeologists. By “dumbing down” explanations to the levels of sketches, archaeologists completely removed any of the explanatory (not to mention predictive) power they might otherwise have had. If explanation sketches replace explanations as the goal of new archaeology, a serious argument could be made that the relevance of the entire enterprise could be called into question. Is Archaeological Evidence Relevant? Questions about the appropriateness of certain archaeological tools caused other problems for covering law explanations. One example of this relates to the use of texts as evidence. Archaeology has traditionally been divided into prehistoric and historic varieties, with the difference being the existence of readable texts. In the case of prehistory, there would be no conflict between textual and archaeological evidence, but such is not the case in historic period archaeology. For a number of reasons, (including, but not limited to ideological or other biases of the writers) texts, which can span political, mercantile, religious, moral, and historical grounds, can directly contradict archaeological reconstructions. In the case of such conflicts, Hempel would most likely have come down on the side of science, or archaeology, over unconfirmed, untested (and un-testable) texts. However, textual evidence was routinely touted as being superior to archaeological data. The bias of text over artifactual evidence is not a new phenomenon, and the bias continues to date. Although the Hempelian approach would suggest that collected data should be more trustworthy than textual information, many new archaeologists explicitly accepted textual evidence, ignoring or downplaying the significance of the “scientific” data they painstakingly collected. “In the former [prehistoric] case, the synthesis must be based primarily on archaeological evidence and in the latter [historic] case on documentary evidence, with the remains used only to fill in gaps in that evidence or to illustrate it” (Rouse 1973, 26). Hempelian archaeologists have not realized that their epistemological stance places textual evidence on shaky ground. Hempel never specifically addressed texts as a source of scientific data, but as textual data are in many cases unprovable and biased, they should not be considered any more reliable than ethnographic data. In both the case of text and ethnography, the archaeologist is trying to apply context to data stripped of their temporal significance. In the case of ethnography, archaeologists assume that ancient cultures interact and change in the same way as modern groups. For texts, the problem is similar. When ancient texts are taken at face value, there is an assumption that the data in those The New Archaeology’s New Archaeologists 65 texts are similar to modern historical records. Putting aside questions surrounding the objectivity of contemporary historical records, there is reason to believe that ancient texts had different functions than their modern counterparts. For instance, it is evident that ancient historical texts “borrowed” stories from more ancient sources. The biblical flood story, which has counterparts in other (later) religious traditions, can be found in an Akkadian version of the same story (in the Gilgamesh epic), written much earlier. Heroic epics of Greece and Rome contain stories of heroes that were borrowed from earlier stories (about earlier heroes, or about the Gods themselves) for effect. Egyptian stelae depict victories in battles that almost assuredly did not occur. Texts in antiquity had a wide range of uses, and although they may have something to add to the archaeological record, their preference over that of archaeological evidence is strange, especially to Hempelians, who claim to be free of the biases contained in those records. Historians were well aware of the biases associated with documents, be they of gender, empire, politics or theology, and it would have made sense for new archaeologists to follow that lead. Either because of Hempel’s silence on the matter, or possibly as a result of an improper understanding of Hempel’s program, many archaeologists who claim to be new archaeologists still use texts as the foundation of their research. Just as Binford gave Hempelians a means to use ethnography (as a source of hypotheses), a very good case can be made for a similar treatment of texts as texts.31 If textual data are only used to stimulate thought, they would not violate the principles of new archaeology. Ignoring texts altogether would be as great a mistake as depending upon them too heavily. There is information in texts, and that information should be scrutinized and used in the same way as is other theory-laden evidence, such as ethnography. The mistake would be to confuse thoughts about texts as useful sources of information with a belief that texts should be the proof of such information. Just as ethnography was shown by Binford to only be useful in the hypothesis generating phase of scientific inquiry, texts should assume the same role, and should be excluded from the testing phase altogether. Although questions about evidence are common for logical positivists, the specific debate over ancient texts is peculiar to fields such as archaeology and history. From within the hard sciences, logical positivists were (or believed themselves to be) able to clearly distinguish between discovery and justification, and could easily set aside the former when it was time to do the latter. Here, however, the status of texts and ethnography blur this distinction; texts and ethnographic data are commonly used in the discovery phase (see Binford), in the justification phase (see Hill), or in both phases, which is clearly unacceptable to a Hempelian scientist. In the interests of sustaining this separation, these sources of data would need to be limited to the discovery phase (where Binford has proven these data are useful), eliminating their use in the testing phase, regardless of the impact of this decision on the desires of archaeologists. Another problem surrounding texts had to do with the meaning of the terms “prehistorical” and “historical.” A number of cultures were deemed prehistoric by Europeans for a range of reasons having nothing to do with the nature of the 66 Chapter 3 people being studied. At times, archaeologists considered cultures, such as the Mayans, prehistoric if they (the researchers) were unable to read the codices of the indigenous cultures. In other instances, peoples with rich oral histories but no texts were judged prehistoric. In addition, a number of archaeologists deemed peoples with written histories that are considered to be biased, or ideologically tainted, as essentially prehistoric. For instance, a number of Syro-Palestinian archaeologists called minimalists (due to their desire to minimize the biblical account)32 believe that Israelite archaeology should be practiced as if the biblical text did not exist. Their belief, which is predicated on the belief that there is an extreme (and uniquely pervasive) ideological bias of the biblical account, is especially odd, as the minimalists seem to have no problem accepting the authenticity of texts from adjacent regions, such as Assyria and Egypt, while admitting that these texts are full of political biases, including victory steles for battles that were never won. Other archaeologists routinely use texts as primary data in their scientific studies of the past, and have had great success in doing so.33 Philosophical Accounts of Archaeological Explanation The text issue brings up in important schism between archaeologists and philosophers of science. Archaeologists believed that texts were useful tools, explanations should be applicable to single sites and individual problems, and universal laws should result in predictability. Philosophers of science had different ideas about these issues. The role of texts is questionable, the focus is on large trends and general laws, and predictability for the social sciences was found to be unlikely. Problems with new archaeology, including difficulties integrating texts into new archaeological explanations, new archaeology’s inability to eliminate equifinality, and its focus on explanation sketches and general trends, led to thoughts that archaeological explanation might be in some way different than philosophical explanation. Detractors saw the theoretical basis of new archaeology as being taken piecemeal from another discipline, and expected to work in a totally new context: “As is so often the case where concepts are borrowed, the introduction of philosophical concepts into archeology has been selective and incomplete” (Kelley and Hanen 1988, 2). Put simply, if archaeologists are suddenly unable to use any of their tools, are unable to work toward what they perceive as their goals, and are told that their explanations are not explanations, either archaeologists know nothing about their subject matter, or the problem is not one with archaeology, but is instead a problem with the borrowed theory. This issue will be spoken of again in chapter 6. Archaeologists, having been told that they were doing bad science, could easily retort that Hempelian scientists were simply not doing the same thing that archaeologists do when each side provides explanations. Archaeologists can provide examples of successful studies that do not reference either laws or hypotheses (one such study will be examined in the next chapter) and in the end, they were able to produce a satisfactory explanation of the features in question. The New Archaeology’s New Archaeologists 67 The fact that archaeologists can provide explanations without relying on laws has led some to question not only whether covering laws are appropriate, but also whether philosophical explanations in general provide any help to archaeologists on the ground. Philosophers of science have a reply to this critique. Before an archaeologist puts a trowel to the soil, he or she has a plan (conscious or not) in mind. Although laws, hypotheses, and the like may not be explicit, they are there nonetheless: It is sometimes argued that one need not worry about the absence of explicit hypotheses because all research has implicit ones. Kluckhohn (1939) asserted many years ago that no one collects data in the absence of some theoretical orientation that causes him to collect some data and ignore others. But, a fundamental identity usually exists between good research and research that has been done according to a well-planned design. When part of the research process is implicit, explicit consideration of the viability of the research design is unlikely. (Plog 1974, 17) Good research demands that there be some sort of research design, and better research will come from researchers who are conscious of the problems they wish to solve. So, regardless of (or despite) the problems associated with new archaeology, philosophers of science could make a case that archaeologists have a lot to gain by resuming their association with philosophers of science. According to these philosophers of science, archaeologists should be aware that the failures of the new archaeology are not a reflection on philosophy per se. Archaeologists read Hempel’s “The Function of General Laws in History” and believed it to be the answer to their theoretical and methodological problems. They did not realize that philosophical explanation was (and continues to be) in a state of flux: The appeal to philosophy of science, was, initially, incomplete and somewhat shallow, however. For one thing, the work of a few philosophers (notably Carl Hempel, Ernest Nagel, and Karl Popper) tended to dominate the archaeological perspective, and this in itself gave a rather narrow perspective to the discussions. For another thing, the archaeological appeal to philosophy occurred at a time when philosophy of science was itself in a transitional state, but this was not evident at the time to archeologists. A consequence was that archaeologists were not at first aware that what they were borrowing from philosophy of science was under serious question and in the process of being replaced by philosophical views thought to be more adequate. (Kelley and Hanen 1988, 2) Even as the new archaeologists were grafting historical explanation onto archaeological theory, philosophers of science were working on and changing their accepted models of explanation. For instance, Hempel’s statistical models of explanation (the D-S and I-S models) were reactions to criticisms of his original covering law model (the D-N model). Other philosophical models of explanation competed with Hempel’s models, and the field was (and is) in a constant state of change. In fact, the goals of logical empiricism (which became the foun- 68 Chapter 3 dations of new archaeology) are now no longer universally accepted goals of scientific explanation. These issues will be the topic of the next chapter, and once they are understood, chapter 5 will reintroduce the disunified range of contemporary archaeologies to a set of post-Hempelian philosophical schools that will look surprisingly familiar. Notes 28. Hill asserts the ceramic record’s stability, and the unique nature of each family unit’s ceramic traits. Hill’s results have since been disputed by Factor Analysis studies, such as those done by Lischka (1975). 29. For a contemporary defense of the potential for analogical reasoning (despite its real dangers), see Wylie’s “The Reaction against Analogy” (2002). 30. Hempel (1965) 236 gives an example of how this sort of explanation sketch might work in his “dust bowl” example. 31. As texts are written in particular styles, with particular materials, and on potentially datable media, texts qua artifacts are not subject to this problem. It is only in suggesting that texts are accurate representations of the past that this problem presents itself. 32. The minimalists include Philip Davies, Niels Lemche, Thomas Thompson, and John Van Seters. 33. An example of this is Deetz and Dethlefsen’s study, the subject of chapter 4. Chapter 4 Philosophy and Archaeology: Post-Positivism What the New Archaeology Could and Could Not Do The new archaeology (and its new archaeologists) was the focus of chapters 2 and 3. However, the promise, successes, and failures of Hempel’s models did not exhaust archaeological activity during this period of history. Either knowingly or not, a number of archaeologists, such as James Deetz and Edwin Dethlefsen (see below) were producing archaeological explanations that seem to ignore the “advances” and features of new archaeology. The ability to produce non Hempelian, archaeologically significant explanations will framed debates between archaeologists and philosophers of science in the 1970s and 1980s as to whether philosophical explanation was at all relevant to archaeology. While very few believed that the answer to this question should be answered with an emphatic “no,” there is still much work to be done on the development of alternative philosophical approaches that are directly relevant to the needs of archaeologists. Many archaeologists were, and still are, largely unaware of ongoing philosophical debates over the proper form and goals of scientific explanation. As a result, when the new archaeologists found themselves unable to provide covering law explanations (as discussed in the preceding chapters), post-processual archaeologists such as Hodder (1991, 1992) and Shanks and Tilley (1987) claimed that it was time for critical approaches to replace the philosophically naïve new archaeology. However, the post-processualists (for the most part) continued to do scientific archaeology. They collected and analyzed data like scientists, and they came to conclusions based on their data. So, the failure of new archaeology was not a failure of science (and philosophy of science, by extension), but that a particular model of scientific explanation was at fault. In fact, the range of post-processual archaeologies closely resembles the range of philosophies of science studied today. To prove this claim, it will be necessary to show that philosophical explanation should not be strictly equated with Hempelian (or any single) explanation, and that the application of philosophical explanation to archaeology is similarly fluid. Deetz and Dethlefsen’s case study makes no reference to the general laws required by new archaeology, but according to the excavators, their results provide a complete explanation of the material. If explanations can be given without laws, this leads to a general question surrounding the necessity of laws in scientific explanation. The range of answers to this question (although this treatment will not be exhaustive), and the utility of these positions to archaeologists, is the subject of this chapter. This contemporary debate about the relationship between philosophical models of scientific explanation and theoretical approaches to archaeological 70 Chapter 4 explanation will set up the questions asked the final section of this chapter and answered (hopefully in the affirmative) in the rest of the book. These questions are interrelated and are as follows: “are there philosophical problems that should be of special interest to archaeologists?” and “are there archaeological problems that should be of special interest to philosophers of science?”. Case Study: Death’s Head, Cherub, Willow and Urn Like Binford (1967), Deetz and Dethlefsen chose a narrow focus for their study. Instead of analyzing small pits, they set out to explain changes in colonial grave markers over time, and given a well researched and clearly demarcated area, Massachusetts and its surrounding states. Grave markers are about as good a source of data as any artifact can be. They are found in specific areas, they generally have absolute dates inscribed on them, and they have designs or patterns that make it possible to categorize them by type. “For a number of reasons, colonial New England grave markers may be unique in providing the archaeologist with a laboratory situation in which to measure cultural change in time and space and relate such measurements to the main body of archaeological method” (Deetz and Dethlefsen 1967, 2). In this study, Deetz and Dethlefsen noted a certain thematic progression in New England tombstones, from Philosophy and Archaeology: Post - Positivism 71 Death’s Head, To Cherub, to Urn and Willow. Fig. 3. Death’s Head, Cherub, Urn and Willow 72 Chapter 4 Deetz and Dethlefsen sought not only to document this change, but also to attempt an explanation for this seemingly coordinated activity: Enter almost any cemetery in eastern Massachusetts that was in use during the seventeenth and eighteenth centuries. Inspect the stones and the designs carved at their tops, and you will discover that three motifs are present… The earliest of the three is a winged death’s head, with blank eyes and a grinning visage. Earlier versions are quite ornate, but as time passes, they become less elaborate. Sometime during the eighteenth century—the time varies according to location—the grim death’s head designs are replaced, more or less quickly, by winged cherubs. This design also goes through a gradual simplification of form with time. By the late 1700’s or early 1800’s, again depending on where you are observing, the cherubs are replaced by stones decorated with a willow tree overhanging a pedestaled urn. (Ibid., 1) This study of the change from death’s head, to cherub, to urn, was unlike Binford’s project in a number of ways. In this case, there was no need to wonder whether a given object was or was not a tombstone, as opposed to Binford and Munson, who could not even agree on the members making up the set of smudge pits. As a result, the whole set of tombstones could be found and studied. Sampling strategies (and problems) could thereby be totally avoided. As another departure from the studies mentioned in chapter 3, the object of this study was not the identification of function (or at least the primary function) of the tombstones (as this is readily apparent). The object here was an explanation of the reasons for particular stylistic changes in these markers.34 Deetz and Dethlefsen’s hypothesis was that the changes in tombstones were the product of culture change, and specifically in terms of ecclesiastical changes in the area. The orthodox Puritans of the 1800s saw iconography as heretical, and eschewed graven images of the divine, instead focusing on mortality and death: “While the use of cherubs might have verged on heresy, since they are heavenly beings whose portrayal might lead to idolatry, the use of a more mortal and neutral symbol—a death’s head—would have served as a graphic reminder of death and resurrection” (Ibid., 3). As those restrictions became more lax, a shift in emphasis from death and mortality to the afterlife and immortality complimented a relaxation in the restriction against iconography, and the cherubs replaced the death’s head as the preferred style of grave marker: “Given the more liberal views concerning symbolism and personal involvement preached by Jonathan Edwards and others later in the eighteenth century, the idolatrous and heretical aspects of cherubs would have been more fitting to express the sentiment of the period” (Ibid., 3). Finally, the “willow and urn” motif accompanied an intellectual revolution, as opposed to an ecclesiastical one. These markers stood as memorials (and not even necessarily on the spot where the deceased’s remains lie), and were neither references to the end of life or to eternity, as were the death’s head and cherub respectively: The earlier stones are markers, designating the location of the deceased or at least a portion of him. In contrast, “In memory of” is simply a memorial state- Philosophy and Archaeology: Post - Positivism 73 ment, and stones of this later type could logically be erected elsewhere and still make sense. In fact, many of the late urn and willow stones are cenotaphs, erected to commemorate those actually buried elsewhere, as far away as Africa, Batavia, and in one case—in the Kingston, Massachusetts, cemetery— “drowned at sea, lat. 39 degrees N., long. 70 degrees W.” The cultural changes that accompany the shift to urn and willow designs are seen in the rise of less emotional, more intellectual religions, such as Unitarianism and Methodism. (Ibid., 4) Possibly the most troubling aspect of this analysis to a new archaeologist is that no hypothesis testing was involved, and no laws or conditions for the study were ever stated; so although Deetz and Dethlefsen backed up their conclusions with sound research, a Hempelian archaeologist would not consider this article to have explained anything at all. Hempelians might well have dismissed this study as anachronistic, might have demanded that it be redone in properly scientific terms., or might claim that although no laws were explicitly stated, those laws are part of the explanatory apparatus implicitly. However, it appears that this explanation stands well on its own, in that it explains the observed thematic progression of tombstones observed by the investigators. In addition, if Deetz and Dethlefsen are invoking laws, implicit or not, these laws must be very limited in their scope (probably only applicable to their particular study) and these specific laws could hardly count as the general laws desired by logical positivists. Despite the theoretical differences cited above, Deetz and Dethlefsen’s conclusions sound very similar to the sorts of conclusions reached by new archaeologists. Deetz and Dethlefsen used archaeological evidence to explain the reasons for the changes in death markers in a particular region over a specific period of time, just as Binford and Hill used archaeological evidence to explain the role and presence of specific data in their respective regions. The similarities between Hempelian philosophical explanations and Deetz and Dethlefsen’s archaeological explanations raise an important point. If Deetz and Dethlefsen are doing good archaeology without proposing laws and testing hypotheses based on those laws, then an argument can be made against the necessity of covering law explanations. On the other hand, if examples of good archaeology such as this one are to be disregarded as unscientific and as failed explanations simply because they do not appeal to covering laws, then this decision would call into question much of what archaeologists seek to explain. Archaeologists are as interested in the kinds of explanations given by Deetz and Dethlefsen as they are in the kinds of explanations provided by Binford and Hill. Both in this situation and in the discussion of texts and other questionable pieces of scientific evidence in chapter 3, differences between the goals and interests of archaeologists and philosophers of science mirror differences in the structures of archaeological and philosophical explanations. In some ways, archaeologists accepted (or tried to accept) the new direction of logical positivism. As a result, new sources of data suddenly became relevant (as discussed in the preceding chapters), and other sources of data (such as ethnography, and arguably, texts) were relegated to the discovery phase of explanation. Either as a re- 74 Chapter 4 sult of the failure of new archaeology, or as a parallel development, other successful archaeological explanations (such as the sort provided by Deetz and Dethlefsen) showed that there may be other ways to explain the past, ones that do not depend on the structure of new archaeology, but still supply information that archaeologists can use. If these two groups cannot agree on what makes for a good explanation, more work needs to be done to rectify the situation. The relationship between archaeological and philosophical explanations will be revisited in chapters 5 and 6. However, before archaeologists can engage debate with philosophers over theory formation, it is worth taking some time to explore the present range of options open to philosophers of science working in explanation. In fact, philosophy of science is not so monolithic in its support of Hempelian explanation as archaeologists were led to believe. Significant problems with Hempel’s models, both philosophically and in their application to archaeology were explored in the preceding chapters. Now, it is time to ask whether there are alternative models, or at least alternative trends that both philosophers of science and archaeologists can see as providing the groundwork for the future of their fields. As a result, it is time to look a bit harder at science. Hempel’s models of science were accepted by many as they provided a clean method for moving forward. The question is whether this picture is true, or if it is no more than wishful thinking. As early as 1982, Merrilee Salmon pointed out that archaeologists (as well as other scientists) have spent a lot of time examining the features of scientific explanations in fields such as physics. She says that problems are bound to occur when archaeologists try to emulate the features of explanations that seem to work in physics without really understanding which of those features are necessary parts of the scientific enterprise: “It is important for archaeologists to have a fair understanding of the nature of science if they are not to waste their time emulating inappropriate and nonessential aspects of other sciences in their efforts to make their own discipline more scientific” (Salmon 1982, 181). Salmon is not denying that there are essential aspects of science. The difficulty is in deciding which features of a science are necessary for all science. If there are features of scientific explanation that transcend specific sciences, such as physics, the question remains whether those universal explanatory features will have biases toward the hard sciences, or if those features of science that seem to discriminate against archaeological forms of explanation will disappear. Laws and the Philosophy of Science One way to frame this discussion is by asking whether scientific explanations need to reference some form of covering laws. If Hempel is right, then not only are laws crucial for science, but his concepts of laws are crucial to science. However, if Hempel’s conception of science is wrong, this leads to a number of questions about where the problem lies. There is no question that when archaeologists accepted the need for laws for scientific explanation, they radically Philosophy and Archaeology: Post - Positivism 75 changed their discipline. However, whether archaeology should continue to remain wedded to law-driven explanations is another (and open) question altogether. Assuming for the moment, that archaeology should concern itself with laws, questions remain as to the proper character of these laws, and their relationship to the field. This option will be examined in chapter 4. Alternatively, if archaeology should disassociate itself from laws, then archaeological science will need to be based on something else. That option will be explored in chapter 4 as well. Although critiques of logical positivism and logical empiricism continue to be of particular interest to philosophers of science, criticisms of Hempel’s covering law approach began decades ago. Scriven’s 1962 “Explanations, Predictions, and Laws” was a thorough critique of Hempelian explanation. Scriven focused on a number of problems, including some that have already been discussed (see chapter 1) and others attacking Hempel’s intuitions about the nature and limits of explanation. For example, Hempel believed that explanations and predictions were interdefined. In fact, for Hempel (Hempel and Oppenheim 1948, 249), a complete explanation of a phenomenon would provide everything needed for a scientist to make a prediction about its next occurrence. Scriven gives a number of reasons to believe that Hempel’s necessary connection between explanation and prediction is mistaken. For one thing, Scriven provides an occasion when prediction is possible without explanation: “For example, we may discover that whenever cows lie down in the open fields by day, it always rains within a few hours” (Scriven 1962, 176). Although an argument can be made that explanations could come from this sort of statement (although I have never heard an especially persuasive attempt), this prediction, however strong, can not in any way be equated with an explanation. Alternatively, there are times when explanations are entirely possible, but predictions are not: A man in charge of an open-hearth furnace may be suspiciously watching a roil on the surface of the liquid steel, wondering if it is a sign of a “boil” (an occasionally serious destructive reaction) on the furnace lining down below or just do to some normal oxidizing of the additives in the mixture. Suddenly, a catastrophe: the whole charge drops through the furnace lining into the basement. It is now absolutely clear that there was a boil which has eaten through the lining, apart from sabotage (easily disproved by examination) there’s no other possibility. But no prediction is possible to the event, using the data then available. (Ibid., 189) In both of these cases, a Hempelian could claim that the deficit in Scriven’s examples is that the scientist making the prediction or providing the explanation is limited to the data then available. A complete explanation would contain the missing data, and would restore the symmetry between prediction and explanation. However, Scriven’s tactic, to point out difficulties with the relationship between explanation and prediction that Hempel asserts, is effective, and was recognized as problematic to Hempel (resulting in his inductivist models of explanation). 76 Chapter 4 According to Scriven, this line of reasoning points to one of the real problems with Hempelian explanation, the belief that complete explanations are possible. Leaving aside questions as to how it would be possible to collect all of the information necessary to provide a complete Hempelian explanation (already addressed in this work, as well as by Scriven 1962), There are questions surrounding the adequacy of general laws for providing the desired explanations. Answers to these questions can be separated roughly into two positions. On one hand, there are those who believe that failures in explanation have been due to the types of laws proposed by philosophers such as Hempel. For these philosophers, such as Merrilee Salmon (1982, 1990, 2001), Wesley Salmon (1998, 1999), Peter Kosso (2001) and Alison Wylie (2002), the future of explanation lies in either the discovery of archaeological laws that will do what Hempel’s could not, or in: “how to distinguish laws from well-founded empirical generalizations, how to understand various scientific models of confirmation and explanation, and whether these models are useful in archaeology” (Salmon 1991, 239). For some, the problem with explanation is ultimately tied to the reliance on laws themselves. For these philosophers, such as Ronald Giere (1999), Bas van Fraassen (1980), and Nancy Cartwright (1983), explanation is hindered the self-imposed focus of explanation on archaeological (or any scientific) laws. Alternatively, maybe the concept of scientific objectivity itself needs to be changed, according to Thomas Kuhn (1970, 2000), or eliminated altogether, according to Paul Feyerabend (1993). These questions will be the subject of the remainder of this chapter. Laws in Archaeology: Repairing a Damaged Structure Salmon’s Critique and Proposed Revision According to Merrilee Salmon, philosophy and archaeology work together to reconstruct “on a firm evidential basis past cultural, social, and economic systems” (Salmon 1982:1). Both philosophers and archaeologists have in common a desire for explanation and description, and for this reason, archaeological theory has been traditionally thought to be best developed and understood by philosophers of science. In her groundbreaking Explanation and Archaeology, Salmon began by asking whether laws are necessary for explanation in archaeology. She believes that the answer is yes, but that archaeologists were wrong in believing that covering laws were good candidates for archaeological laws. She focused on Hempel’s models, showing their initial promise, and then articulating the major philosophical and archaeological problems, similar to those critiques made by Scriven and those found in the preceding chapters, with the covering law method (Ibid., 104– 105). Salmon did not believe, however, that problems with Hempel’s models meant that archaeologists should abandon their belief that laws are relevant to explanation. In fact, she did not even believe that archaeologists should abandon new archaeology. Salmon thought it was a mistake to assume that new archaeology had to be equated with Hempel’s particular vision of explanation. She saw Philosophy and Archaeology: Post - Positivism 77 the strengths of new archaeology (and logical empiricism) as stemming from its reliance on hypothesis testing, and not as being wedded to Hempel’s particular models: The emphasis on testing, which demands controlled observations and/or experimentation, is a reaction to what in the past may have been an excessive reliance on imaginative reconstruction, appeals to authority, hearsay, and written records of dubious authenticity or semimythical character. It is a serious mistake, however, to equate the admitted importance of hypothesis formulation and testing with the acceptance of [Hempel’s] method, and the rejection of other forms of inductive reasoning. (Ibid., 41) Salmon believed that new archaeology was still alive and well, despite the serious problems archaeologists and philosophers had with Hempel’s D-N model. In fact, she posited that logical empiricism would be taken seriously as a model of science only when scientists stopped depending on Hempel’s models as an indicator of success. Putting aside for a moment the merits of Hempel’s models as being able to explain (largely the subject of the preceding chapters), it has become clear that Hempelian explanations do not necessarily provide good archaeological explanations. Hempel’s models, if properly applied, would eliminate or at least severely limit the utility of certain types of archaeological evidence (such as texts). In addition, covering law explanations are just as unable to differentiate between equifinal archaeological end states as the explanations they sought to replace. These explanations fail in their promise to allow archaeologists to make predictions. Finally, the focus of covering laws is on general trends and not on specifics (the focus of Deetz and Dethlefsen’s study). So it might be the case that covering laws explanations are not relevant to the needs of archaeologists. It would seem reasonable to suggest that archaeological (or any scientific) explanations need be relevant to the scientists in question. Explanatory laws need to utilize the tools and data sets given by a discipline in order to answer those questions that are relevant to working scientists in the field. Archaeologists are generally as concerned with how something occurred as they are with why that sort of thing occurs. Due to the focus on general laws that is integral to Hempelian explanations, a new archaeologist could answer the second question, but they rarely, if ever could answer the former question. Perhaps a new archaeologist could develop covering laws to explain why groups tend to migrate (such as, in times of drought), but it would be a different enterprise to try to discern how it is that a particular group chose to migrate at a particular time, or during a particular drought, as opposed to another. Merrilee Salmon believes that an account of causal or functional explanation would be needed to address both the “why” as well as the “how” questions: Philosophers, when confronted with archaeologists’ concerns about explanation, should see that their formal accounts of the structure of scientific explanation fail to capture some of its most important features. They must pay attention to some highly general but nonformal aspects of explanation (such as its causal features) if their accounts are to be adequate. They should also recognize that 78 Chapter 4 functional explanation plays a central role in archaeology, just as it does in biology. Thus, if their models are to be models of scientific explanation, and not just models of explanation suitable for a narrow class of explanations in the physical sciences, they must deal with functional explanation in a serious manner. (Ibid., 180, emphasis in the original) Formal covering law explanations are rarely on archaeologists’ minds when they seek to reconstruct the past. In fact, according to Salmon, most scientists are comfortable working on an substantive (or intuitive) level and, instead of making a specific model of explanation work, set their sights on establishing a collection of explanatory facts, usually regarding the cause or function of the artifacts in question (a focus that will be addressed below). Explanations that are built upon these facts might fit into any number of philosophical models. Determining which model is the best or most adequate is more substantive than formal, and Hempelian models of explanation are unable to move beyond formal matters to cover these substantive (and incredibly important) issues. Another, related problem concerns how Hempelian explanations were being applied to archaeology. As Hempel’s D-N model was the primary model being discussed in the scientific literature, this was the model that archaeologists sought to apply to their work: Thus at the same time that archaeologists were adopting the rhetoric of D-N explanation, they were attending more closely to the substantive features of archaeological explanations. Moreover, careful attention to the content of archaeologically acceptable explanations revealed the difficulty of accommodating them to the D-N model. Laws were difficult to come by, and if the did exist, they were more likely to be statistical than universal. (Salmon 2001, 236) Merrilee Salmon proposed that a modified statistical-relevance (S-R) model, (based on the work of Wesley Salmon35) could accommodate the needs of archaeologists in a way that D-N explanations could not. To explain this model, it is worth understanding its origin. The original S-R model (the non-causal version) was based on the belief that an explanation was only possible when all of the relevant factors could be accounted for and weighted appropriately. According to the original S-R model: “An explanation of an event…is an assemblage of all factors statistically relevant to its occurrence, along with an assignment of the probability of the events occurrence in the light of those factors” (Ibid., 131, original quote from W. Salmon 1971, 71). Among the problems with the Hempel’s original S-R model is its inability to account for causal factors, which the Salmons believed necessary. Their point is well taken. In the absence of some sort of mitigating factor, a statistical relevance explanation would be incomplete. For example, in a small tribal group, an increase in population could be statistically relevant to that group’s decision to separate into smaller units. Alternatively, the separation of a group into smaller units would be statistically relevant to the group then increasing in size. Although this may not seem like an important distinction statistically, the temporal order of these events is significant to an archaeologist. In the first case, an archaeologist would seek to explain why a population increase caused a group to Philosophy and Archaeology: Post - Positivism 79 split. Factors here might be limited resources, increased social stability, or a political disagreement, leading to separation. In the second case, the aforementioned factors might be totally irrelevant, causing an archaeologist to look elsewhere for an explanation. If an archaeologist cannot use an explanation to account for causal relationships, then archaeology might just as well go back to description. “Which came first” dilemmas help explain the difference between the causes of change and the effects of those changes, and an analysis of these features would be necessary to any science interested in explanations of events that occur over time. As a response to this temporal problem, Wesley Salmon supplemented the S-R model to account for the role of causes in explanation. So the causal S-R model required the following: “Along with the complete set of statistically relevant factors for an event’s occurrence and the proper probability assignment, he now says that there must also be a causal account of any statistical regularity that is used in the explanation” (Ibid., 131, emphasis in the original). The addition of causal language to the S-R model is designed to eliminate archaeological “chicken and egg” disputes, where scientists cannot determine whether cultural change causes environmental change, or whether the reverse is the case. The Salmons claim that there are examples in contemporary archaeological thought of causal S-R accounts. They reference recent attempts (by Renfrew 1971, 1973) to link cultural changes to changes in population size, “establishing (statistical) correlations between increasing population size and certain other archaeological recognizable features” (Ibid., 131). These statistically relevant features would not alone be enough for a complete explanation, as there is no way to tell which feature set the others in motion. Salmon claims that Renfrew (quoting Forge) provides the additional causal reasoning needed to provide a full explanation of the organization or fragmentation of growing populations: Forge says that “Homo sapiens can handle only a certain maximum number of intense face-to-face relationships, successfully distinguishing between each (quoted in Renfrew 1973:115).” According to his view, when populations grow beyond a certain size, compensatory measures, such as introducing structure or fragmentation into smaller groups, need to be taken to ensure continued communication and social integration. (Salmon 1982, 132) The causal S-R model is not without its problems. Salmon discusses a number of real issues that go hand-in-hand with causal explanations. For instance, causes are traditionally seen as having very specific spatial and temporal boundaries. For those archaeologists who believe that all explanations must be causal in nature, apparent interactions from a distance require a lot of work to prove. “When actual contact cannot be supported by evidence, they mount expeditions such as the Kon-Tiki to prove that contact was at least possible” (Ibid., 137). Although most archaeological explanations will not involve building a boat out of balsa wood and sailing it from South America to Polynesia, just to prove that it could have been done in antiquity, many archaeological explanations do involve assertions about changes over considerable distance or time. As such, causal explanations are difficult to prove. 80 Chapter 4 In some cases, there is no real way to determine what caused an event or artifact. In cases when it is difficult to find causes for events, scientists are forced go to great lengths to attempt to manufacture the needed connections. For instance, when Syro-Palestinian archaeologists first saw small stone flecks in the matrix of certain types of pots, they asked why this feature existed, or, what caused its existence. There could have been a number of possible answers to this question. Perhaps all of the local clay had these inclusions, and the potters were unable to remove them, implying that the form was accidental. Alternatively, the potters might have believed the inclusions to have some aesthetic value. Another possibility (and the one that has since been accepted) has to do with function. Just as boiling chips (small pieces of ceramic) are used by chemists to stabilize the boiling point of some reactions, archaeologists posit that ceramics with these inclusions function better as cooking pots (as they have been named) than their “chip-less” ancestors. At times, it is impossible to know what causal factors are in play, especially when an artifact’s function cannot be determined from its form. This indeterminacy has led to an unfortunate (but common) situation where anything with an unknown function is said to be cultic (having religious significance). As articles used in religious ceremonies are almost by definition idiosyncratic, claims that unknown artifacts were used in such contexts are used to explain otherwise unexplainable items. At Tel Miqne/Ekron in the 1990s, a group of archaeologists in Field 1 found a large raised rectangular platform with a round opening cut into it. It was clearly an artifact, due to its size and context, but the group was at a loss to understand its function. They posited that since the form of the object did not necessitate any particular function, it might well have been a bamah (high place, or altar). Given this hypothesis, additional hypotheses were added to provide confirmation for the cultic hypothesis. For instance, the hole could have been filled by an incense burner. As the platform was believed to be a significant find, a soil sample from the installation was sent off to the lab for analysis. When the results came back, the team was a bit embarrassed to find that their cultic stand was actually a late Bronze Age or early Iron Age toilet. Had the site been a little less advanced technologically, that proto-toilet might have wound up mislabeled in a museum; the outhouse area might have been designated a temple complex. In this example, the artifact’s function was ultimately discernable, but this is not always the case. Salmon was well aware that there is no way to guarantee explanations will be possible in every case. Regardless of the difficulties, however, Salmon believes that archaeological explanation is causal in nature: “Sometimes hard work discloses nature’s secrets, sometimes genius is required. Luck surely plays a role, and probably, in some cases we will never find the answers we seek. But regardless of any barriers to the discovery of causes, no satisfactory theory of scientific explanation can dispense with its causal features” (Ibid., 138). In her introduction of the modified S-R model, Salmon did not intend to simply replace Hempelian D-N explanations with her own model. Salmon proposed her work as one of many possible alternatives, believing that multiple Philosophy and Archaeology: Post - Positivism 81 perspectives might be brought to bear under the umbrella of new archaeology. In doing so, Salmon calls into question the idea that there was (or should have been) a single unified new archaeology. “New archaeology was never a homogeneous movement. While all new archaeologists embrace the aims and practices of science, they sometimes disagree about the nature of the scientific method and how to apply it” (Salmon 1991, 239). Although Merrilee Salmon focused her efforts primarily on archaeological explanation, she was well aware that the problems archaeologists face are problems found throughout the scientific community. In fact, the assumption that these explanatory issues exist due to the limitations of the social (and allegedly therefore inferior) sciences is equally false. Salmon points out that many of the features that make archaeological explanations difficult are found in the sciences in general: Some understanding of the development of modern physics, for example, can convince us that no precise method of discovery replaces trial and error, luck, and creative genius in the enterprise of theory construction. Also, because statistical laws are fundamental in modern physics, one cannot reasonably claim that genuine science demands a full set of universal laws. Reflection on the character of the other physical sciences (such as geology, astronomy, and meteorology) enables one to see that limits on the repeatability of events one is studying, on experimentation, and on predictability are less formidable barriers to genuine science than many archaeologists have assumed. (Salmon 1982, 181) Although Salmon does not argue that the natural and social sciences are identical, she believes that the sciences have quite a bit in common, in terms of their goals and methods. More importantly, Salmon urges scientists not to preserve failing models of explanation by claiming that explanatory gaps are the fault of social scientists. Scientists and philosophers of science should instead focus on communicating their needs to one another, and to answering questions that both parties would agree are relevant and interesting. Statistical and Probabilistic Explanation: Many of Salmon’s critiques of Hempelian explanation were focused on problems that the original D-N model had in dealing with statistical and probabilistic laws. Hempel was aware that the D-N model was not suited for these kinds of laws, and subsequently created those statistical models (D-S and I-S) evaluated in previous chapters. Regardless of the success or failure of both Hempel and the Salmons in their use of statistics and probability in explanation, it is clear that these play a major role in any modern explanatory account in science. Three methods of using probability in explanation have different strengths and applications in the interpretation of archaeological data. These methods (classical hypothesis testing, Bayesian methods, and likelihood methods) will be discussed presently.36 Classical hypothesis methods (named for statistitians Neyman and Pearson) are best used before the archaeologist sets foot in the field. One potential goal of 82 Chapter 4 archaeology, explicitly stated in the age of new archaeology, is predictability. Neyman-Pearson methods are designed to use experimental data to test the relative merit of an hypothesis. One could certainly imagine a situation where an archaeologist has an hypothesis about artifact transmission patterns and wants to test it. Neyman-Pearson tests are designed to give grounds to accept or reject the proposed hypothesis, given a number of limiting factors, such as the desire to limit false positives and false negatives (designed to separate hypothesized distributions from those occurring by chance). Once the test has been designed, data are collected, and the hypothesis is either justified or rejected. If the data matches the results of the hypothesis (preset parameters determining when the hypothesis should be rejected), there is reason to believe that the hypothesis is true, and that the model suggested has merit. In fact, archaeologists and archaeometricians have begun to look at the way that classical tests can be utilized in archaeological analysis to test hypotheses. For instance, Schneider and Bentley (2000) posit that statistical models37 may allow archaeologists to establish a relationship between ceramics found on a tel’s surface and the distribution of those ceramics buried at the site. So, Neyman-Pearson methods can provide an archaeologist the means to decide between hypotheses before even taking to the field. Likelihood methods, as opposed to Neyman-Pearson tests, work best after data have been collected. These methods are the quantitative form of inference to the best explanation. They give a precise way of comparing the explanatory power of alternative hypotheses for the same data. The likelihood of a given hypothesis given data is proportional (by means of a constant) to the probability of the data, given the hypothesis. Put formally:, “L(H|R)… is proportional to P(R|H)” (Edwards 1972, 9). Here, instead of focusing on the changing data, given a stable hypothesis, attention is placed instead on finding the best hypothesis, given the data. For example, likelihood models would be useful for archaeologists if they found a particular style of pottery and wanted to know if it is local or imported. The two hypotheses (the local hypothesis and the import hypothesis) would each have a certain probability, and the most likely result would be chosen. The goal is to find the hypothesis that best fits the data, as opposed to the classical hypothesis testing model that asks whether the data confirm the hypothesis. Whereas Neyman-Pearson hypotheses are generated before data are collected, and likelihood is assigned after data are collected, Bayesian methods assign what is known as a posterior probability. Posterior probabilities are used to determine whether a hypothesis is confirmed (has a higher posterior probability than its prior) or disconfirmed (has a lower posterior probability than its prior).38 This value is determined, given the results in a trial, and knowing the likelihood of that result if that hypothesis succeeded and if it failed. Bayes' theorem asserts: P(H1| E) = P(H1) * P(E, given H1) P(H1) * P(E, given H1) + P(H2) * P(E, given H2) where H1 is the hypothesis to be tested, H2 is the alternative and contradictory hypothesis, and E is the observed evidence. Although Bayesian methods are Philosophy and Archaeology: Post - Positivism 83 strongest in cases of repeated exchangeable (i.e. when order is irrelevant) trials, they can still serve in small samples when priors are well-supported, but they work best in the repeated, exchangeable trials. The rationale for this is that every time a trial is run, its posterior probability can serve as a prior for a successive trial. In the long run, or after enough trials, the posterior probabilities will approach their real values, regardless of the original priors set. Although there are not too many repeatable exchangeable trials in archaeology, and there are not many instances when an archaeologist will be confident enough to assign precise priors for a single trial, Bayesian methods are used in sciences that impact archaeology, such as their use in calibration for carbon-14 dating.39 Without delving into a thorough treatment of statistical theory (well beyond the scope of this work), there appears to be no compelling reason to exclude these or any other types of statistical analyses from archaeological theorizing. It is important, however, that the archaeologist understands the limitations of each particular method. Although, given certain conditions, all three methods can be made to resemble one another, doing this eliminates the strengths of each individual method. Properly used, Neyman-Pearson, likelihood, and Bayesian methods can provide information to justify assumptions, and help to explain the presence or significance of archeological data. Functional and Causal Explanation Functional explanations, and the aforementioned causal explanations are natural ways to explain the presence of items in the archaeological record. Functional explanations, for archaeologists, are simply that, explanations based on the function or role of material items. For example, when we explain why a well protected site location was chosen by a group, we assume that the function of a tel was to protect its inhabitants from enemies. When archaeologists seek to explain the tel’s function, they are providing a type of causal account of how a need for protection caused the tel to be built. Obviously, not all causal explanations are functional. When someone decides to purchase a car, they may have function in mind (in that the car will have to get them to work), but the decision to buy an classic car may be based on other desires, such as status. However, when archaeologists explain the past functionally, they do so in the belief that they are providing the causes of past behavior. Explanations based on causal relationships are historically problematic to logical positivists, and empiricists in general. In book 1, part III of the Treatise of Human Nature, Hume demonstrated that causality has no a priori or a posteriori justification, and philosophers of science have to date been unable to reassert knowledge of causality on empirical grounds.40 As a result, logical positivists have to believe that both causal and functional sorts of explanations must either be dismissed as being without foundation, or it must be shown that apparently functional explanations are actually founded on other, presumably more solid grounds, such as those provided by Hempel. Certainly, more general causal theories such as functionalism, a belief that all explanations are functional in character, would be dismissed, as Hempel believes: “in this generality, the claim 84 Chapter 4 is apt to be either meaningless or covertly tautologous or empirically false” (Hempel 1959, 372). Despite the failure of functionalism, Hempel does not suggest that specific functional analyses are without use. He suggests that the functionalist approach can play a supporting role in explanation: “The functionalist mode of approach has proved highly illuminating, suggestive, and fruitful in many contexts. If the advantages it has to offer are to be reaped in full, it seems desirable and indeed necessary to pursue the investigation of specific functional relationships to the point where they can be expressed in term of reasonably precise and objectively testable hypotheses” (Ibid). Hempel considered functional analyses to have the potential for discovering hypotheses that could then become parts of a real (covering law) explanation. In fact, according to Hempel function-based explanations are best seen as heuristic aids to help reveal real (or covering law based) explanations (Ibid., 371–372). Hempel believed that functional explanations were only sources of hypotheses. As such, those that could not be reduced to, or restated as covering law explanations would be incomplete. Hempel argues that functional models do not provide explanations. Given a system, a functional account of an object or item can only account for the presence of a set of functional equivalents, and not of the item itself. Given a proper functional analysis, along lines given by Hempel in this essay, “its explanatory force is rather limited; in particular, it does not provide an explanation of why a particular item i rather than some functional equivalent of it occurs in system s” (Ibid., 368, emphasis in the original). According to Hempel, functional analyses are only able to provide a number of functional equivalents that could each explain the state of a given phenomenon. Functional accounts stop short of being able to explain why one of the equivalent possibilities should be chosen over the others. Of course, Hempelian explanations can be just as problematic as functional explanations (see the earlier discussion of equifinality). In the end, it is not an easy task to decide whether Hempelian equifinal explanations are somehow better or worse than equifunctional explanations. Nancy Cartwright dismisses Hempel’s assertion that functional (or more generally, causal) explanations are incomplete. On her view, covering law explanations are based upon an oversimplified view of science, and causal claims better represent scientific understanding: “I am in favor of causes and opposed to laws. I think that, given the way modern theories of mathematical physics work, it makes sense only to believe their causal clams and not their explanatory laws” (Cartwright 1980, 379). Modern physics, and modern science by extension, is thought to be governed by laws, but according to Cartwright, this is not so: “Many phenomena which have perfectly good scientific explanations are not covered by any laws. No true laws, that is. They are at best covered by ceteris paribus generalizations—generalizations that hold only under special conditions, usually ideal conditions” (Ibid. 1983, 45). When scientists refer to laws of nature, they generally refer to approximations that are false in all but limited examples. As these laws are false, they cannot explain. The function of these oversimplifications is to present a unified Philosophy and Archaeology: Post - Positivism 85 front to science, leading students to the belief that scientific laws are real, objective, measures of reality. However, there is no reason to believe that the regularities that result from ceteris paribus generalizations will be true reflections of the world: “We are lucky that we can organize phenomena at all. There is no reason to think that the principles that best organize will be true, not that the principles that are true will organize much” (Ibid., 53). As Cartwright believes that the world is too complicated to be held together by covering laws, this leads to the question of how scientists really explain. For Cartwright, explanations in physics are not appeals to oversimplified laws: “This is typical of modern physics. ‘Competing’ theoretical treatments—treatments that write down different laws for the same phenomena—are encouraged in physics, but only a single causal story is allowed. Although philosophers generally believe in laws and deny causes, explanatory practice in physics is just the opposite” (Ibid. 1980, 386). If laws are wrong in all but ideal situations, then the real world, which tends to be messy, cannot be governed by laws. However, Cartwright sees scientists providing explanations for phenomena, based on experiments. She gives examples (Ibid., 381–384) of explanations that are based on causal reasoning. In cases where general laws may not apply, scientists focus on finding the causes of their results. For a scientist, causes, and not laws, are the foundation of scientific explanation: “I have sometimes summarized my view about explanation this way: no inference to best explanation; only inference to most likely cause” (Ibid. 1983, 6). Unlike Hempel, who believes that unifying laws are the only means to explain, and Cartwright, who believes only in causal explanation, Wesley Salmon proposes a unifying account of explanation. According to Salmon, causal mechanisms, such as the causal S-R model described above, are one way (but not necessarily the only way) archaeologists explain the past: To explain the location and contents of a particular burial, for instance, it may be necessary to ascertain the age and gender of the individual interred, and to determine the cause of death. In constructing causal explanations it is necessary, in general, to infer or postulate the existence of causal processes that are no longer available for our direct inspection. Moreover, causal explanations often appeal to entities such as atoms, molecules, or bacteria that are not directly observable under any circumstances; theory observation or detection requires some sort of special apparatus. The causal explanation consists, in large part, of exposing hidden mechanisms. (W. Salmon 1998:359, emphasis in the original) These explanations, focusing on single events and particular circumstances, are opposed to covering law types of explanations, which Salmon has recently come to believe have another, complementary role to play: “Unifying explanations involve reference to broad structural features of the world” (Ibid). Although in his earlier work (referenced in Merrilee Salmon 1982), Wesley Salmon used his causal S-R model to critique covering law models of explanation, his recent work has been far more conciliatory. Salmon (1990) talks about causal explanations (or “how” questions) and unifying explanations (or “why” ques- 86 Chapter 4 tions) as both explaining the same phenomena on different levels. Neither of these is the only correct method of explanation. Indeed, he argues that the complexity of science is proof of the complexity of scientific explanation: “When we think seriously about the very concept of scientific understanding, it does not seem plausible to expect a successful characterization of scientific explanation in terms of any simple formal schema or simple linguistic formulation. It is not surprising that there might be the kind of duality I have been discussing” (Ibid., 104). For Salmon, both types of explanations answer questions of interest to scientists. Sometimes an archaeologist will want to know about why peoples migrate. This sort of explanation might lead the investigator to larger trends or generalizations. At other times, an archaeologist may want to know the circumstances surrounding one particular group’s decision to move. This question will best be explained by focusing on causal mechanisms. The data gathered can be used to help form both types of explanations, and Salmon considers both to be important tools of science. Any archaeological explanation based on cause or function will force archaeologists to be especially wary of ethnocentricity. There is little way to determine whether a given explanation explains based on the desires and motivations of the people being studied, or whether the modern scientist’s own desires or motivating factors are unconsciously attributed to the people under study. This problem cannot be overstated, but an awareness of the issue , according to archaeologists such as Binford, can lessen its force. As discussed in chapter 3, Binford confines the use of ethnographic analogy to the discovery phase of his explanations. Binford would never use ethnographic information as a means of testing hypotheses, however, for this might result in conclusions tainted by the ethnographer’s biases. Ethnography is still useful to Binford as a means to generate hypotheses, and even as way to choose between hypotheses to test. Binford believes that any ethnocentric bias that makes it into an hypothesis will cause it to fail in the rigorous testing phase. Wesley Salmon is the first to admit that his work is incomplete. He agrees with detractors that a satisfactory philosophical account of cause has yet to present itself, but he believes that this account will be forthcoming, and that, in the meanwhile, his causal S-R model of explanation will do: “The result, we hope, will be an account of scientific explanation that will be adequate to the needs of anthropologists. In the meantime, it seems reasonable to call the S-R model to the attention of anthropologists concerned with covering law explanations, and to remind them that philosophers have not yet provided definitive answers to all of these problems—but we are trying.” (Salmon 1998, 345, emphasis in the original). An example of causal reasoning can be found in Steven Mithen’s recent work on the origins of the human mind. In The Prehistory of the Mind, Mithen explains the rise of Homo sapiens as being caused by changes in the human mind. Both Homo sapiens and Homo neanderthalensis (Neanderthals) lived on Earth about 100,000 years ago. Assuming that Neanderthals and Homo sapiens are separate species, 41 Mithen asks why it is that Neanderthals—creatures who Philosophy and Archaeology: Post - Positivism 87 were both better adapted physically to the cold climate of the Middle Paleolithic period and who had as big (or bigger) a brain than their fellow Homo sapiens— became extinct; replaced by the smaller, less well physically adapted (more tropically suited) Homo sapiens. Mithen’s answer to this was that the Homo sapiens’ advantage was in the way that their brains processed information. Mithen demonstrates (in chapter 7) that according to archaeological evidence, including complex tool making (such as the Levallois flake tool making method, diagrammed on page 135 of his book), and a change in hunting and living patterns which indicates a complex social structure, Neanderthals were intelligent. Their intelligence, however, was specialized. Mithen claims that Neanderthals had specific cognitive domains (language, social intelligence, technical intelligence, and natural history intelligence) with little integration between them (in a domain called “general intelligence”) (Ibid., 162–163). Homo sapiens, on the other hand, posses what Mithen calls cognitive fluidity. Mithen believes that the Homo sapien mind was able to integrate the previously compartmentalized different types of knowledge (Ibid., 173). This integration resulted in the birth of art, religion, and agriculture, as well as advances in hunting and animal domestication (Ibid., 204). Mithen states at the outset that his explanation of the origins of the human mind is incomplete. During a talk at Claremont Graduate University (2003), he explained that more needs to be said about cognitive fluidity. Archaeological data can provide evidence for change, but these data can only indirectly point toward the cause of that change. Mithen does not claim that cognitive fluidity results from documented, specific changes in neuroanatomy. In fact, there is no direct evidence that any change in the brain occurred. As cranial volume is one of the only measures of ancient brains that exists to the archaeologist, there is no way to test for cognitive fluidity, nor any explanation as to how such a test could be performed. Mithen assumes the archaeological record is best seen as a product of cognitive fluidity (or as Cartwright would say, an inference to the most likely cause). Another assumption that Mithen makes is that the early hominid brain was similar in kind (though not in complexity) to its closest living ancestor, the chimpanzee: “To compensate for [a lack of information about proto-humans] we have considered the behaviour of the chimpanzee, assuming that the chimpanzee mind has a similar architecture to that of the common ancestor of 6 million years ago” (Mithen 1996, 102). This belief that proto-humans would be cognitively similar to primates is an analogy based on their places on the evolutionary tree, and the archaeological record. Given the limits of analogical reasoning, Mithen assumes that creatures with similar origins would have similar mental architecture. It would seem like an easy way to weaken the strength of Mithen’s analogy would be to point to the long temporal gap between primitive Homos and contemporary chimpanzees. Mithen defends his belief in cognitive similarity based on the archaeological record. In the entirety of Neanderthal civilization there is little or no evidence for cognitive growth, at least in the fashion that separates Neanderthals from Homo sapiens. Going back to the example of the Levallois flake method, alt- 88 Chapter 4 hough Neanderthals were able to replicate this complicated method for producing pointed tools, they never innovated on the basic design. According to Mithen, this ability to create technologically advanced tools, coupled with the inability to innovate is evidence for a certain type of specialized mind, and this mind is similar to that of the modern chimpanzee. So, despite the temporal distance, Mithen proposes that Neanderthals’ mental cousins are found in humanity’s closest living relative, the chimpanzee. Analogical Explanation While Binford showed that properly used ethnographic analogy can provide information an archaeologist can use, either to posit function or to suggest hypotheses, Mithen’s work on the origins of cognitive integration shows that analogy in general can serve the same purpose. Setting aside problems specific to ethnographic analogy (already discussed in chapter 3), analogical reasoning is an important element for archaeological explanations. According to Alison Wylie: “the connections between material and behavioral or other cultural variables—the determining structures or relations of structural and functional interdependence—are just what archaeologists cannot observe directly; they are among the features of past cultural contexts that must be reconstructed inferentially” (Wylie 2002, 151). Regardless of real or perceived problems with analogical reasoning, inferences about an artifact’s function can only be made by resorting to analogy. By definition, archaeologists have no opportunity to question the artisans or users of the artifact, and there is no way to directly observe the item being used in its original context. Since archaeologists are temporally distant from their object of study, they have no other option than to base their assumptions on the features and context of those artifacts they find. Without analogy, archaeology would be severely limited. Unfortunately, analogical reasoning can lead to serious error. For example, Hill’s analysis of ceramic traits led him to believe that the Broken K Pueblo family units were matrilocal. His assumption that ceramics were made by women was based entirely on analogy to ethnographically studied Pueblo groups. As has already been shown above. the data gathered from Broken K could have led to a number of plausible alternative conclusions. Another limitation on analogy is archaeologists’ reliance on form leading to function. Although an object with a certain form may be more efficient at a particular function, other factors may be equally (or more) relevant. For instance, chopsticks are not the most efficient means for eating noodles, Victorian dresses with corsets are not the most functionally sound examples of dancing attire, and basketball hoops are too small to efficiently allow for a maximum of baskets. In each of these cases, other factors are considered to be more important than function, and analogical assumptions of form begetting function would fail. Of course, form can limit function for more utilitarian objects, or in cases where form is relevant to function. Presumably, these objects, as opposed to things like ceremonial objects, will be easier to pin down. Context can impose a further restriction, given an object with a certain form. If an archaeologist comes Philosophy and Archaeology: Post - Positivism 89 upon a flat-bladed tool, two analogous possible functions, that of a scraper or a spatula, could be posited. If the flat-bladed tool is found in contexts associated with food preparation, the assignment of the tool as a spatula would be further supported, while in other contexts (such as at a brickyard), the scraper might be more credibly posited. Of course, even context can be deceptive, as people routinely reuse, carry away, and drop things. Items can be given as gifts, changing their function and affecting their context. Further, a range of natural phenomena, ranging from shifting earth and flowing water, to burrowing animals can intentionally or unintentionally move objects from their original locations. Humans can also unintentionally confuse archaeologists by plowing up fields or digging trenches, mixing up otherwise stratified materials. For instance, in 1998, a team of workers at Tell el-Far’ah, South found a series of puzzling artifacts in a presumably clean Middle Bronze Age context. These artifacts included a sardine can, a broken syringe, and a “London” pin. Only after these materials were found was the team aware that a deep trench had been dug in the area during World War I, that these intrusive materials were left from that period, and that the same soil was used to fill the trench later (leading to an almost perfect soil match between the old and new materials). Had those artifacts not been left behind, any other artifacts dumped back into the trench to fill it might have been assigned the wrong context, and would have been explained improperly. A final problem, especially relevant to archaeologists utilizing analogies in their explanations, occurs when form does not dictate function. As an example: can artistic expression be used as a means of learning about past? Archaeologists routinely use artistic representations of peoples to gain knowledge about those peoples. One brief example is the use of the “philistine bird.” One (though by no means the only) piece of archaeological evidence for the existence and origins of a people known biblically as the Philistines was a particular corpus of pottery. Philistine pots were derivatives of a particular style of pottery, a monochromatic style from Mycenae. The local variant, called Mycenaean IIIc:1b, found in Cyprus and Canaan in the Late Bronze age. The philistine style of pottery featured Mycenaean motifs in local clay, and was also distinguishable from its forbears by the use of red and black (bichromatic) paint. Philistine bichrome pottery, which is thought to have made its appearance in the twelfth century BCE is used as a proof that a new, seafaring people appeared on the coastal plane of Canaan at that time. Trude and Moshe Dothan describe the initial identification of this pottery in a pit locus in Ashdod: The lowest levels of the pit contained only the earliest types of Philistine pottery, which would be dated to the period of their initial settlement and expansion in the twelfth and early eleventh centuries B.C. But in the upper levels, from a slightly later period, around 1000 B.C., there was a surprising tranformation of their traditions. From her work at Tell Qasile, Trude had assumed that except for the heavey kraters with the crude black spirals, the distinctiveness of the Philistine ceramic tradition had been complete lost by this time. Our garbage pit revealed an unexpected artistic innovation, a previously unknown class of pottery whose “ridged rims” were covered with red slip, burnished and often decorated with think black and white bands, still bearing some of the traditional Philistine shapes. (Dothan and Dothan, 1992, 134-5) 90 Chapter 4 Additionally, according to much of the archaeological establishment, the pottery, along with other evidence (which is also currently under renewed scrutiny) “shows” that the new inhabitants came from Mycenae, solving the problem of both the presence and the origins of the biblical Philistines. Fig. 4. Philistine Bird or Eating Crow? Philosophy and Archaeology: Post - Positivism 91 The purpose of this discussion is not to prove or disprove the existence of the Philistines, however it is worth mentioning that the example above is from a contemporary T-shirt, and not from a pot from the late Bronze or early Iron age. This begs a question: why should a particular image point to the presence of a particular people? In many contemporary cases, people latch on to fads, either foreign or domestic. Additionally, people who cannot afford or find the desired item will purchase a locally made copy. Why not posit that the “philistine copy” was the Times Square fake Rolex of the twelfth century BCE? Granting the ubiquitous nature of Disney, will future archaeologists posit the presence of “the animal worshiping Disney people” every time they see a set of Mickey Mouse ears? In short, our assumptions about antiquity, as well as our modern habits and desires, have an affect on the connections we make between modernity and antiquity. In the absence of methods to properly tie modernity to antiquity (what Binford calls middle range theory, or what Hodder calls hermeneutics) ethnographic analogy can be lead to almost any conclusion. In the absence of checks and balances that have yet to be established, archaeologists are forced to depend on a tool that is difficult to manage.42 Lawless Archaeology? The many different characterizations of science mentioned above assume the goals of science to be inexorably bound to the concept of laws. This approach, although dominant in the scientific community, is not the only way to do science, which raises an obvious question as to how science could exist without the structure and authority of scientific laws. A number of archaeologists have, either consciously or unconsciously, made careers as scientists while ignoring or attacking the existence of explanations based on general laws.43 Before exploring the contributions of those archaeologists who do not believe that laws are necessary components to scientific inquiry (which will be discussed in the next chapter), it is worth looking at the philosophical foundations of their arguments. Although Scriven was one of the first to provide a systematic critique of Hempel’s vision of scientific explanation, another of Hempel’s contemporaries, Kuhn questioned the idea of scientific objectivity (as the goal of scientific progress) itself. Kuhnian/Feyerabendian Explanation In his Structure of Scientific Revolutions, Kuhn showed that real science is not the sort of idealized program of discovery and justification given by Hempel and other covering law theorists. Kuhn saw established science as a combination of normal science (known as puzzle-solving) and periods of revolution (when crises cause scientists to radically change science and proceed in a new direction). According to Kuhn’s model, science does move forward, but in a way very different than Hempel’s view of scientific progress. For Kuhn, progress is 92 Chapter 4 anything that makes scientists better at answering the questions they have set out to solve. Unlike covering law theorists, who see progress as leading toward objective truth, Kuhn sees progress as just a part of puzzle solving. In fact, after a paradigm shift, science reorients itself and scientists seek to answer new questions. So, progress in the former paradigm would be meaningless in the new paradigm. So science progresses, or moves forward much in the same way that evolution is considered to be progressive.44 However, just as there is no “end” to evolution, there is no reason to believe that this “progress” is moving toward any objective goal of true understanding. Current models of science, just like their ancient analogues, will come up against crises, and these crises may similarly lead to new revolutions. The resultant new science will start over, with new problems to solve, new directions, and new goals. This position would seem to place Kuhn squarely in a relativistic camp. Kuhn both admits and rejects this position: “Though the temptation to describe that position as relativistic is understandable, the description seems to me wrong. Conversely, if the position be relativism, I cannot see that the relativist loses anything needed to account for the nature and development of the sciences” (Kuhn 1970, 206). On one hand, he considers the progress that goes on in puzzle-solving (normal) science as objective, in that it allows scientists to better understand their world (in the context of their paradigm). On the other hand, Kuhn rejects any concept of scientific truth being paradigm independent, or “really there.” In the end, Kuhn remains dissatisfied with the label of “relativist,” but can offer no clear objection to the label. As an interesting note, Kuhn further muddies the waters of scientific objectivity as it regards the social (or human) sciences. In his discussions with Charles Taylor (referenced in Kuhn 2000, 216–223), Kuhn recognizes that the human sciences provide a special challenge to his view of science. In each of the physical sciences, the chaos and disunity of competing pre-paradigmatic schools of science gave way to an unified program of normal science. It is only after this initial unity that science could “progress,” as progress can only exist within the puzzle solving activity found in a paradigm. The unified science might later be punctuated by the occasional chaos and revolution, but those revolutions would end, and the unified project of science would continue. The human sciences, in contrast, lack the unified program of the physical sciences. Consequently, Kuhn claims that the human sciences are locked in the preparadigmatic phase of science, with competing schools fighting for ascendancy. Whether this case of stunted growth is a matter of principle or age is left open. Kuhn posits that some social aspects of social science may eventually become ordered enough for a paradigm to be formed. On the other hand, though, it is possible that an inherent instability in these sciences and their subject matter precludes the necessary basis for paradigmatic science. With the comparative stability of the natural sciences, Kuhn argues that science had a fertile ground from which to spring. However, Kuhn does not think this is so for the social sciences: “Without that stability, the research responsible for the change could not have occurred. But stability of that sort cannot be expected when the unit Philosophy and Archaeology: Post - Positivism 93 under study is a social or political system. No lasting base for normal, puzzlesolving science need be available to those who investigate them; hermeneutic reinterpretation may constantly be involved” (Kuhn 2000, 223). Although this argument is compelling, one problem with it is that Kuhn seems to be suggesting that scientific data remain stable in the face of changing paradigms. However, this directly violates Kuhn’s body of work, as Kuhn believes that data only exist within a paradigm. Regardless of this contradiction, Kuhn believes that all mature science is paradigm driven. As such, there could be no end to science, no truth to approach. If this is science, then the laws of science sought after by Hempel must similarly be no more than features of paradigms. Kuhn clearly believes that scientists pick out those features of science they find interesting or important, and ignore features that other scientists thought equally relevant. If the laws of science are chosen to direct research, there is reason to question the necessity of scientific laws to archaeological explanation. Feyerabend, while critical of Kuhn’s work, opened a door unlocked by Kuhn’s ideas. While Kuhn did not believe in the objective certainty of science, Kuhn saw merit in scientific progress (as problem solving) and genuinely believed in the utility of those answers science provides. Feyerabend, on the other hand, spent a good portion of his career making science out to be a villain, one that must be toppled by any means necessary: Any ideology that breaks the hold a comprehensive system of thought has on the minds of men contributes to the liberation of man. Any ideology that makes man question inherited beliefs is an aid to enlightenment. A truth that reigns without checks and balances is a tyrant who must be overthrown, and any falsehood that can aid us in the overthrow of this tyrant is to be welcomed. (Feyerabend 1974, 55–56) Science was not always the tyrant, according to Feyerabend. During the enlightenment period, science was the underdog and it served to unseat the tyranny of religious thought. In the twentieth (and twenty first) century, however, science has totally replaced religion as the dominant power over people’s lives. Now the slave has become the master, and that master must be brought down by any means necessary. Despite his thoughts on the tyranny of science, Feyerabend never claims that he wants science to disappear. Again, to make the comparison to religion, Feyerabend is as happy with the survival of scientific thought as he is with the survival of religious thought, or of any other. His objection is to the association of science with truth, and its privileged status in the modern world. Just as the state once sanctioned the methods and results of the church, now the state has devoted itself to science. This condition, Feyerabend says, must change: “The most important consequence is that there must be a formal separation between state and science just as there is now a formal separation between state and church” (Ibid., 61, emphasis in the original). Feyerabend suggests a variety of alternative sources of meaning, outside the scientific realm. The point, according to Feyerabend, is to educate people to be contrary to any sense of authority, be it 94 Chapter 4 scientific or otherwise. Scientists (along with crystal healers, Delphian oracles, and the like) may certainly make suggestions, but decisions need be democratically decided by an unbiased body. If truth (whatever that means) is sacrificed by dismantling authoritarian structures, so be it. Feyerabend believes that freedom of thought, though possibly slower than scientific progress, is far more important a virtue.45 Truly Lawless Science For some philosophers, the concept of laws are seen as more a hindrance to explanation than they are worth. In the introduction to Laws and Symmetry (1989), Bas van Fraassen points out that the concept of laws have their origins in theology. Universal laws were seen as expressions of divine simplicity. As G-D is perfect, there must be a set (a small set) of divinely created rules that represent that order. Modern philosophers and scientists extricated the need for a deity from the concept of law, but retained the belief that simplicity and universality were concepts worth pursuing. However, in the absence of an omnipotent force, both van Fraassen and Ronald Giere (1999) ask whether the concept of law has any real standing anymore. Although their arguments differ, both van Fraassen and Giere believe that laws are artifacts of the past that do no real work. Instead, both appeal to particular notions of symmetry. For Giere symmetry means a set of complex, and occasionally conflicting relationships between theories and the world: Imagine the universe as having a definite structure, but exceedingly complex; so complex that no models humans can devise could ever capture more than limited aspects of the total complexity. Nevertheless, some ways of constructing models of the world do provide resources for capturing some aspects of the world more or less well. Other ways may provide resources for capturing other aspects more or less well. Both ways, however, may capture some aspects of reality and thus be candidates for a realistic understanding of the world. (Giere 1999, 79) The world is very complicated, but parts of it can be more or less explained by our models. At the same time, other models may better explain other aspects of the world. These models may conflict with one another, but this is not a problem, as it would be were Giere concerned with laws. Unlike laws, which are supposedly unchanging, there is nothing “real” about a model, other than its ability to help us understand our world. However, Giere does not believe that models are total fictions. He compares them instead to maps: Imagine, for example, four different maps of Manhattan Island: a street map, a subway map, a neighborhood map, and a geological map. Each, I would say, represents the island of Manhattan from a different perspective, appropriate, for example, for a taxi driver, a subway rider, a social worker, and a geologist. Maps exhibit the two primary characteristics I earlier ascribed to perspectives. First, they are always partial. There is no such thing as a complete map. Se- Philosophy and Archaeology: Post - Positivism 95 cond, maps may be maps of something. So maps can be understood realistically. (Ibid., 81, emphasis in the original) These differing perspectives, for Giere, are constructions. As they help us to make sense of underlying features, they are realistic, even if they are not actually real (whatever that might mean). Bas van Fraassen’s approach differs from Giere’s, in that van Fraassen believes that some aspects of theories are in fact real. For van Fraassen, empirical entities (objects that can be experienced, in fact or in principle) are real, and theories about these things refer to real objects. However, when sense perception ends, so does reality. “Science aims to give us theories which are empirically adequate; and acceptance of a theory involves as belief only that it is empirically adequate” (van Fraassen 1980, 268, emphasis in the original). Van Fraassen’s view, called constructive empiricism, only depends on a theory’s ability to explain what is perceived. Theoretical entities, such as electrons, can be parts of theories, but they are only placeholders, useful fictions that explain things we see, such as vapor trails in cloud chambers. Without commenting on the relative strengths and weaknesses of these positions (as this has been done by a number of authors, including Paul Churchland 1982), the antirealist and anti-law tendencies of these philosophers of science is mirrored by similar tendencies in archaeology. Meehan (1968), a political scientist, suggested a systems approach to explanation, relying on symmetry relations similar to the ones discussed above, and archaeologists such as Tuggle, Townsend, and Riley (1972) presented this approach to archaeology. Most archaeologists (and most scientists) have been unwilling to consider the possibility of lawless science. Instead, these archaeologists see inconsistencies or failures in scientific explanation to be a product of their methods, and not as a reflection of the world’s structure. To many archaeologists, it is preferable to deal with endless ceteris paribus conditions than to consider the possibility that science might not be the unified, universally structured enterprise they were brought up to believe in. Disunified Science and the Future of Archaeology The wide range of views stated above about the nature of laws and explanation should not be a source of concern to archaeologists qua scientists. Although it may seem from the preceding discussion that archeology is in an especially dangerous place, the problems reviewed above are endemic to scientific enquiry, and not unique to archaeology. Although none of the positions stated above give a clear view of the future of science, or of archaeology in particular, archaeologists are helping to explore these issues in the philosophy of science by testing the boundaries of science itself. Archaeology is a science that is both theoretical and dirt-driven. In some fields debates about theoretical issues can seem pretty ethereal. In physics, for example, arguments surrounding the existence of theoretical entities (such as 96 Chapter 4 electrons, quarks, and strings) are interesting, but many argue that the resolution of these debates is not so important. So long as electrons do the work they are supposed to, many believe that the very question of whether these entities actually exist is not a pressing matter. For archaeology, this is not the case. Claims that archaeological entities, traces from the past only visible indirectly in the present, do or do not exist end up being pretty fundamental to people in the field. Rough correspondence between theory and truth is more difficult to explain to someone who has spent 20 years digging in a trench. For this reason, archaeology provides the philosopher of science an interesting and new vantage point from which to understand fundamental questions. Chapter 5 will introduce some philosophical terminology to archaeological issues, and explore some ways that archaeology can help put a new spin on some problems relevant to contemporary philosophers of science, as well as historians of science. Chapter 6 will conclude this project by showing how improved communication between archaeologists and philosophers of science can result in a better future practitioners and theoreticians in both fields, and will further explore the questions and themes began the book. Notes 34. Another way to think of this distinction is to consider law based explanations “why” questions, as opposed to “how” questions, referring to the reasons for a particular feature. This distinction will be discussed in in chapter 4. 35. Wesley Salmon later modified his thoughts on explanation. His subsequent thoughts on explanation will be discussed later in this chapter. 36. This section will not provide a comparison or critique of the various uses and limitations of different models of probability. For a thoroughly (and surprisingly) readable introduction to this topic, see Hacking (1965). Deborah Mayo (1996) provides an update to this topic that is geared toward a wide audience (including those with little or no background in statistics). 37. Specifically, they refer to Poisson distributions and Chi-Squared analyses (two types of classical tests). For an early example of Chi-squared analysis applied to archaeology, see Watson, LeBlank, and Redman 1971, 141. 38. Of course, this equation assumes that the prior probability is not 0, for if the prior is 0, the numerator (and therefore the result) will be 0 as well. 39. Steel (2001) presents an interesting account of the use of Bayesian methods to calibrate the difference between a given carbon-14 (or radiocarbon) date and the real (or calendar) age of the sample. 40. Attempts to move beyond Hume’s problem, and find a new foundation for causality are well beyond the scope of this work, but have been attempted. For an example, see Wesley Salmon (1998, essay 14) “Probabilistic Causality,” an attempt to explain a probabilistic notion of causality. Philosophy and Archaeology: Post - Positivism 97 41. As there are a wide range of opinions as to the place of Neanderthals on the evolutionary tree, this assumption is contentious, but lies outside the scope of Mithen’s (and this) book. 42. For an interesting look at just how close Binford’s Middle-Range and Hodder’s Hermeneutic approach may be (despite Binford and Hodder’s intentions), see chapter 3 of Kosso 2001. 43. Including the Syro-Palestinian “minimalist” archaeologists and a number of American post-processualists, such as Hodder (1991, 1992) and Shanks and Tilley (1987). 44. For more on this, see the postscript to Kuhn’s Structure of Scientific Revolutions (1970) and Kuhn (2000). 45. Feyerabend’s dislike of “monolithic” institutions is echoed by Marxist and Feminist critiques of science. Although these responses to science come in many forms, they claim that the interpretation of science has been constrained to fit into a single pattern. This way of doing science determines what is to be researched, considered data or relevant, etc. So, explanation is both self-fulfilling and imperialistic, expanding knowledge by crushing alternative accounts of knowledge. For an introduction to social/Marxist archaeology, read Shanks and Tilley (1987b). For an introduction to Feminist critiques of archaeology, see Gero (1989) and Wylie (1994). 98 Chapter 4 Chapter 5 Philosophical Problems: Archaeological Responses What Can Archaeology Do for Philosophers of Science? Given that even laws may not be as foundational to philosophical and archeological explanation as once supposed, it would be fair to question whether there is a future for the philosophy of archaeology. It is my contention that the range of possibilities afforded by the disparate positions examined in chapter 4 could make a robust future for archaeology possible. Rather than the past monolithic focus on covering laws, the future of archaeological theory rests in a reacquaintance with the resources of contemporary philosophy of science. In this chapter, two such problems will be used to highlight the sorts of issues that philosophers of science can see from a new perspective, given the unique character of archaeological science. In the next section, the relationship between archaeological science and the philosophical “realism/antirealism” debate will be discussed, and following that discussion, the “science and values” debate will be explored from an archaeological perspective. In both cases, it is the author’s contention that archaeological problems can be understood as philosophical problems. Realism, Antirealism, and Archaeological Entities Many of the critiques against logical empiricism have been centered around Hempel’s covering law theories, and many critiques of Hempelian, or new archaeology focus on the application of these theories to archaeology. However, a case can be made to consider Hempel a theoretical entity realist (a question that will be addressed below), a position which would have separated him from many of his fellow logical positivists and empiricists. It could be argued that Hempel’s theoretical entity realism, or at least the perception of him as a realist, even more than the covering law methods of “doing science,” has had a lasting affect on archaeologists, and on philosophers of science in general. Put another way, the question is whether invisible entities are responsible for the visible traces we perceive, and further, what we can know about those entities, if we do grant their existence. For an entity realist, not only are sensed events such as vapor trails in cloud chambers real, so too are theoretical entities that refer to those visible artifacts. Theoretical entities are things that are imperceptible, but which are posited (by realists) to exist nonetheless. This realist position is in contrast to a range of antirealist stances. To an entity antirealist, theoretical entities either have no real 99 100 Chapter 5 content or are at least unknowable; they could be useful fictions that allow us a measure of empirical adequacy or devices that have some other pragmatic or utilitarian purpose, but aside from their potential utility, they are nothing. For physicists, theoretical entities range from the well known, such as electrons, to the less known, such as superstrings. Similarly, archaeological theoretical entities could range from the well known, such as ancient dinosaurs, pots and walls (represented by contemporary fossils, pot sherds, and mudbrick or stone debris), to less certain, such as the claim that a particular area of a city was an industrial zone in the eighth century B.C.E. (based on the relative positions of certain artifacts at a modern site). The question is whether archaeological theoretical entities can be thought of in terms of the realist/antirealist discussion. This parallelism may seem strange, as it is sometimes very easy to reconstruct a dinosaur from fossils, or a pot from potsherds, and an archaeologist would argue that modern remains clearly prove the existence of the ancient animal or pot. However, this has not always been the case. Even today, volunteers new to the field are unable to identify potsherds or discern the difference between these artifacts and flat stones. Fig. 5. Pot or Rock? Archaeological Responses to Philosophical Problems 101 Fossils were once thought to be the remains of fallen angels. It was only comparatively recently that fossils were understood as representing animals that lived millions of years ago. Similarly, biologists now know that illnesses are caused by microscopic entities like bacteria. However, people once believed that illness was caused by imbalances in the body’s humours or was due to demonic possession. At that time, bacteria would have been theoretical entities. Just as we now consider bacteria real and knowable, based on the visible traces these entities leave, some archaeological theoretical entities are less controversially real than others. In fact, some theoretical entities (such as bacteria, visible through light microscopes) can be considered visible in principle, although not in practice (due to arbitrary limits on the focal abilities of the human eye). It might be possible to think of the broken pot as representing a whole pot, similarly invisible in practice but not in principle. A number of arbitrary factors result in the pot’s currently shattered state. Were conditions slightly different (had the pot been slightly better preserved, or had the investigator been more careful in excavating it), the pot might have been visible, as opposed to the collection of sherds currently before the investigator. Other entities, on the other hand, are in principle invisible. Just as superstrings and subatomic particles can not be seen, the hair color of the author’s ancestor from the eighth century B.C.E. is similarly unknowable.46 A possible argument against this characterization of archaeological entities as theoretical entities is that unobserved archaeological entities are somehow different in principle than the entities of theoretical physics. Constructive empiricists have spoken of this supposed difference as one between unobserved entities (such as items invisible due to temporal constraints) and unobservable entities (such as items invisible due to our sensory apparatus. A good response to this charge is given by Paul Churchland when he asserts: “We are all willing to concede the existence of Hume’s problem—the problem of justifying the inference to unobserved entities. But the inference to entities that are downright unobservable appears as a different and additional problem…[however], the appearance is an illusion” (Churchland 1985, 39). As humans are more able to control our spatial position than they are able to control time, people seem to be more comfortable with thought experiments involving trips to Venus than we are with those involving trips to the Early Iron Age. However, as Churchland’s spatially rooted arboreal philosopher, Douglas van Fiirrsen would point out, spatiotemporal realist/antirealist lines drawn on idiosyncratic characteristics (such as visual acuity or bipedal locomotion) would then be no more than a matter of perspective (Ibid.). Claims that scalar unobservables are somehow fundamentally different than temporal unobservables will have to be made more clear if they are to be effective. If archaeological entities can be cast as theoretical entities, then a good case can be made that archaeologists have unknowingly been focusing more and more of their energies on what can or cannot be known, or on the realist/antirealist controversy, the subject of much study in philosophical circles. In fact, many of the contemporary critiques of new archaeology could be seen as 102 Chapter 5 more relevant to new archaeology’s realist underpinnings than they are to the specific models suggested by covering law theorists. For instance, archaeologists today continue to debate the differences between traditional and contemporary archaeologies. This debate is precisely over what each group believes archaeologists can know about the past. As a result, it is worth exploring this parallel development to show how contemporary archaeologists, consciously or not, are heavily invested in the realist/antirealist discussion. It is worth noting that archaeologists would benefit from consulting with philosophers of science on this issue, as an understanding of contemporary issues in the philosophy of science (such as realism/antirealism) would be of great help to archaeologists tying to navigate the theoretical/philosophical issues separating the plethora of post-processual archaeologies.47 Philosophers should note that in some sciences, debates about the status of theoretical entities, however interesting, are not central to their disciplines. This is not true for archaeology. For an archaeologist, theoretical entities are central to scientific explanations, and therefore, having an appropriate answer to realist/antirealist questions becomes crucial. Hempel’s Realism As a reminder (from the preceding chapters), the theoretical backing for the new archaeology was largely based on Lewis Binford’s adaptation of Carl Hempel’s covering law models of explanation. These models differed in a number of respects, but all covering law explanations are based on the belief that there are regularities, in nature, and that there are laws that express or describe them. For Hempel, explanations are “arguments whose conclusion is the explanandum sentence, E, and whose premises-set, the explanans, consists of general laws L1, L2, …, Lr and of other statements, C1, C2, …, Ck, which make assertions about particular facts” (Hempel 1966, 51). Archaeologists who assembled the new archaeological platform, reading Hempel’s work in the 1960s, had reason to believe that Hempel was a realist in terms of theoretical entities. Hempel (1966, 3–6) refers to a famous example from the 1840s which makes a strong case for entity realism. Ignaz Semmelweis, a physician in the Vienna General Hospital, noted that a large number of new mothers in maternity ward 1 died or became very ill, compared with relatively low incidence of illness and death in the mothers in the second maternity ward at the Vienna General Hospital. Semmelweis tested and rejected a series of hypotheses, ranging from cosmic influences and psychological factors to others, including overcrowding, rough examinations, and birthing position to explain the high incidence of childbed fever. None of these withstood scrutiny or controlled testing, so each such hypothesis was in turn discarded. In the end (and by accident), Semmelweis found the answer. One of his colleagues cut his finger on a scalpel after dissecting a cadaver, and developed similar symptoms to the sick women. It was found that the doctors in ward 1 went from dissection labs to examining women after only a superficial cleaning. Doctors in ward 2 did not do Archaeological Responses to Philosophical Problems 103 dissections, and thereby did not use cadaver tainted instruments on their expectant mothers. Although this was years before the development of the germ theory of disease, Semmelweis hypothesized that the disease was due to what he called the transmission of “cadaveric matter” from doctor to patient. Semmelweis had the doctors in ward 1 wash themselves and their instruments in a chemical solution after dissections, and the mortality rate stabilized at the low level found in ward 2. Hempel claimed that Semmelweis had used a covering law explanation, modeled on the type Hempel and Oppenheim introduced in 1948 (and discussed in chapter 1 of this book) to solve his problem. This explanation would look something like the following: Laws: Conditions: Explanandum: cadaveric matter causes fatal disease cadaveric matter is present in a small cut an injured person dies of a small cut For Hempel, the explanation that cadaveric matter causes disease only makes sense if cadaveric matter exists, and if we can have enough knowledge about that matter to understand its role in the explanation. This is what philosophers of science would call a realist position. In his Philosophy of Natural Science, Hempel defends his realist conception of science, arguing against a number of rival (antirealist) theories about the existence of theoretical entities. For example, while antirealists believe that there is a line that can be drawn between visible objects (considered real or knowable) and invisible entities (considered fictitious or unknowable), Hempel believed that any such distinction would be arbitrary, and in the end, indefensible: Presumably [the class of observables] should include all things, properties, and processes whose presence or occurrence can be ascertained by normal human observers “immediately”, without the mediation of special instruments or of interpretive hypotheses or theories…Wires…might count as observables. But we would surely not want to say that a rather fine wire becomes a fictitious entity when weakening eyesight compels us to use glasses to see it. But then, it would be arbitrary to disqualify objects… that no human observer can see without a magnifying glass. By the same token, we will have to admit objects that can be observed only with the aid of a microscope, and so on down to objects that can be observed only by means of Geiger counters, bubble chambers, electron microscopes, and other such devices. Thus, there is a gradual transition from the macroscopic objects of our everyday experience to bacteria, viruses, molecules, atoms, and subatomic particles; and any line drawn to divide them into actual physical objects and fictitious entities would be quite arbitrary. (Hempel 1966, 81–82) Without getting into the relative success or failure of Hempel’s agenda, his position at this point of his work was clearly viewed by archaeologists as realism.48 According to the new archaeologists, in order for Hempel’s models to be called explanatory, not only must observable entities be seen as real, so must the 104 Chapter 5 unobservable entities (like cadaveric matter and subatomic particles) that play a role in those explanations. Realist and Antirealist Archaeology The new archaeologists shared the same epistemological outlook as they saw in Hempel. They were realists in terms of theoretical archaeological entities. Not only were potsherds real, but so were the (invisible in practice) pots they were once pieces of, as well as (in principle) unobservables, such as long abandoned pottery factories where the pots were found. For example, given some initial conditions, say, a large number of oil presses found within a small area, and laws about the potential uses of these features, they posited that the existence of the visible oil presses could be explained only by a reference to a particular and real, albeit invisible activity or object that existed in the past (here, an industrial zone). This realist position would prove to be crucial to the new archaeologists as scientists. The question remains whether temporally distant, and therefore unobservable archaeological entities can be considered properly theoretical. According to Peter Kosso, there is reason to believe that scientific and archaeological unobservables are fundamentally similar: The pastness of the objects of historical and archaeological study is not a significant factor in the process of observation and its contribution to knowledge. Being in the past is neither in principle nor contingently an impediment to an object’s being observed. Studies of the human past are not distinguished from the sciences of the present in terms of their ability to observe, directly or indirectly, the objects of study. (Kosso 2001, 40) For Kosso, a scientist using vapor trails to identify electrons is no different than an archaeologist using archaeological remains to identify past civilizations: In the context of science we expect theories to be justified on the basis of what can be observed, the evidence. The scientist has no direct, observational access to the events and objects of interest, the theorized causes and constituents. She must use the evidence found in the effects of things like germs or the Big Bang to justify the claims about the unobservables (Ibid., 20-1) If the Big Bang, an event that happened billions of years ago, is something that can be studied scientifically, how can the study of something that happened hundreds or thousands of years ago be seen as fundamentally different, and therefore unscientific? The answer to this question cannot have anything to do with the type of evidence used to explain the phenomena in question. Unlike the Big Bang, some past events do not fall into the category of the purely theoretical. For instance, the Holocaust is not at all like this. No one (not even those misguided souls who believe that the Holocaust never happened) holds that the Holocaust is a theoretical entity. However, this may be because Archaeological Responses to Philosophical Problems 105 there are people, documents, and other sources of contextual data providing that requisite independent evidence. There is no analogous evidence that a pile of stones, found on the surface of an unknown (and therefore undocumented) flattopped mound in the middle of the desert was anything, independent of the visual information available to the investigator, who sees those stones and believes that they represent a particular feature in the past. So at least in some sense, archaeological entities are theoretical entities. Furthermore, this might bring up another question: at what point in the past would a building change in status from a distant memory to a theoretical entity? Although this continuum is different than Hempel’s thin wire example (in regard to sight), the question is an interesting one, and could serve as a point of connection between the various sciences. Regardless of the answer to this particular question, or perhaps because of the fact that such questions exist, the realist/antirealist controversy is ripe for archaeological discussion. Scientific Realism and Antirealism It has been claimed that scientists are realists, and that scientists can’t be scientists and still be antirealists. This is one of the most sustained critiques of antirealist positions in science. Over 40 years ago, Maxwell claimed: “that… contentions [that the entities referred to by scientific theories are only convenient fictions] strike me as…incongruous with…scientific and rational attitude and practice” (Maxwell 1962, 3). Ian Hacking agreed, asserting that at the level of scientific practice, “scientific realism [is] unavoidable” (Hacking 1983, 71). According to these philosophers, realism, at least as regards theoretical entities, is necessary for any working scientist. Although this position seems generally to be in accord with scientific practice, in archaeological circles, this contention is plainly false. For instance, a number of post-processual archaeologists would claim to be antirealists with regard to theoretical entities. There are other antirealist scientists, but some (e.g. antirealist theoretical physicists) tend to work with entities that have little impact on their daily lives. Even if superstrings cannot be known, the concept of superstrings could still serve an utilitarian purpose. No harm would come of this explanation, and so the question is simply not earth shattering. For archaeologists, however, this is not the case. If archaeologists can know nothing of the past, then why should they spend their days digging and baking in the sun? Archaeological Realism and Antirealism Given that realism would appear to be central to archaeology, the question is how archaeologists can be antirealists and why they would choose such an apparently counterintuitive position. Archaeologists claiming to be antirealists could do so on one of two levels: they could be doing so from either an ontological or an epistemic perspective. As ontological questions surround existence 106 Chapter 5 while epistemological questions concern knowledge claims, this distinction is an important one. If the archaeological antirealist is working on an ontological level, then the claim is that there literally is no past. If this is the intention, then archaeology as it is known now would be pointless. Of course there might be a new reason to do archaeology, but the field that has been the subject of this project would be dead. As archaeologists universally go into the field and come back with data, it would appear that they see some value in gathering information of the past. Were the past truly a fiction, this would be unnecessary. So the very practice of archeology would imply that the distinction cannot be ontological. If the distinction is epistemic, the antirealist archaeologists would not deny that time is linear, and that there was a past. If this is the case, then although the past exists, it is impossible to know anything about it, much less reconstruct or explain it. An example of this skeptical argument for an archaeologist might revolve around the existence of an ancient wall. The modern evidence for that long since destroyed wall consists of a set of data, or artifacts such as stones, charred mudbricks, and some bronze arrowheads. Given these data, an archaeologist might propose the hypothesis (H1) that the wall was destroyed in battle. However, the modern remains could also be explained by alternate pasts which are incompatible with the archaeologist’s hypothesis. For instance, the modern conditions could also explain (H2) a wall, able to withstand repeated attacks, but destroyed by other causes, such as a natural disaster or (H3) erosion. Further, even the existence of that wall could come into question, as another archaeologist could see that material as (H4) evidence of an ancient refuse heap, something with no architectural significance or military relevance whatsoever. As each of the hypotheses (H1–4) are equiprobable in this particular case, there appears to be no way to choose between them, leading some archaeologists toward skepticism. In fact, most contemporary archaeologists are not skeptics. There are a range of positions concerning the status of archaeological entities. There are staunch realists, represented by archaeologists such as Lewis Binford and William Dever who believe they can have knowledge of a real past, down to individual figures like King David. Other realist positions along the continuum include archaeologically experienced philosophers such as Alison Wylie and theoretically informed archaeologists like Bruce Trigger who see the existence of a real past, but see its reconstruction in terms of more fluid systems of interpretation. Feminist and Marxist archaeologists hint at the existence of a real past, but argue that it has been so masked by engendered or socially constructed biases that it is difficult to recover; making knowledge claims scarce, or at least limited to generalizations. Finally, at the other end of the spectrum are archaeologists who might believe that archaeology should admit that there is no way to objectively study archaeology, and that there is no singular past: “There is and can be no monolithic undifferentiated PAST. Rather, there are multiple and competing pasts made in accordance with ethnic, cultural and gender political orientations” (Shanks and Tilley 1987, 11, emphasis in the original). For these archaeologists, the task of archaeology focuses more on the present, reconstructing the past to serve contemporary (political, gender based, etc) goals. Archaeological Responses to Philosophical Problems 107 So, given that there is a range of archaeological attitudes, both archaeologists and philosophers of science might find it interesting to see which archaeological positions could be reparsed into realist/antirealist terms. Archaeologists could use philosophical models to determine the limits of consistent inquiry. As (mentioned above) ontological antirealism would lie outside any serious study of archaeology, this problem is epistemological. As such, philosophical realists and antirealists could use archeological examples to see how their realist/antirealist positions work when seen from this “new” scientific context. For instance, Bas van Fraassen’s form of epistemological antirealism, called constructive empiricism, is based on an assertion that the point of any proper scientific explanation is to “save the phenomena.” In other words, if electrons are useful to scientists to explain what they see (vapor trails in cloud chambers) then scientists can talk about electrons. However, to van Fraassen, electrons do not really exist, or at least they cannot be thought of as being knowable epistemically. To van Fraassen, “Even if observability has nothing to do with existence… it may still have much to do with the proper epistemic attitude to science” (van Fraassen 1980: 272). Van Fraassen is not claiming that electrons cannot exist, but as far as the pursuit of knowledge goes, electrons (and any other theoretical entities) are merely heuristic devices that allow us to understand what we perceive. The question would be whether an archaeologist could work from within van Fraassen’s framework. For instance, given the “fallen wall” example above, multiple hypotheses (H1–4) are equiprobable. Although these equifinal explanations would result in the same observables (the fallen stones), none of the competing archaeologists would claim that there is no difference between the various explanations. For van Fraassen, the proper epistemic attitude to science is focused on the observable. For an archaeologist, however, the proper focus may be on something altogether different. If a wall fell in battle, that will say something very different about a given city than if it fell naturally. In the former case, the archaeologist might try to determine why the city was significant militarily, while in the latter case, the investigator might seek to answer why the city was abandoned or in decline. One response to this characterization might be to cite the open-endedness of science as the vindication of van Fraassen’s position. If the presently available evidence equally confirms two hypotheses, why not claim that van Fraassen’s scientist would look for more evidence to separate them? The answer to this has to do with the goals of science. For an entity realist, the goal of scientific enquiry is to solve problems like “do electrons exist?” or “can we gain knowledge about entities like electrons.” For a constructive empiricist, the goal of science is different. Van Fraassen can be ontologically agnostic as to the existence of theoretical entities, but epistemologically, the goals of science are clear: “Acceptance [of a theory] involves not only belief but a certain commitment. Even for those of us who are not working scientists, the acceptance involves a commitment to confront any future phenomena by means of the conceptual resources of this theory. It determines the terms in which we shall seek explanations” (van Fraassen 1980, 268). For van Fraassen, empirical adequacy is the 108 Chapter 5 goal of scientific explanation. Once that is achieved, there is no reason to continue. To do so would be to claim that there was something more to entities van Fraassen claims are epistemically vacuous. Traditionally, philosophers of science have created models using physics, and have tried to then apply those models to other sciences. Although this goes beyond the scope of this work, an interesting question about the proper directionality of this relationship presents itself here: If van Fraassen’s model is inadequate for archaeological explanations, should that mean that its adequacy in the other sciences should likewise be requestioned? Clearly, the constructive empiricist position is unacceptable to most archaeologists. Even those post-processualists, ranging from Ian Hodder to Michael Shanks and Christopher Tilley, who believe that knowledge of the past is a social construction, do not deny that explanations can be judged by their abilities to posit the existence of ancient social units, or at least of specific sorts of ancient buildings, given modern remains, regardless of any subsequent interpretation of the meaning of those remains. If a philosophical position denies that this, at least, should be a focus of archaeology, there will not be much support for that approach in the archaeological community. The rejection of van Fraassen’s particular form of antirealism should not be misconstrued as an archaeological rejection of antirealism per se. Hodder (1999) sees the objectification of data, the constraints of objective interpretation, and, it could be argued, realism as the enemy. It is his contention that archaeology will only have utility when archaeologists drop claims that their data and analyses are objective: “It is no longer acceptable to close debate by saying ‘these are the objective data,’ since other interests may wish to make sense of data in different ways” (Hodder 1999, 208). For Hodder, an important point of archaeology is to give groups a voice: “We can encourage the un-networked to use the past to express their rights and identities or to become networked. We can use archaeology as a flexible, open, and participatory project in order to break established patterns of thought and domination” (Ibid.). Given such views, Hodder and other like-minded archaeologists would do well to look at antirealist models of science to fit their own epistemological intuitions. In terms of realist archaeologies, a number of philosophers of science stake out positions that are complementary to archaeological realism. Alison Wylie, as mentioned above, a philosopher of science with years of archaeological experience, suggests that the ways that archaeologists collect and interpret evidence places some real constraints on archaeological explanations: “the key to understanding how archaeological evidence can (sometimes) function as a semiautonomous constraint on claims about the cultural past is to recognize that archaeologists exploit an enormous diversity of evidence—not just different kinds of archaeological evidence, but evidence that depends on background knowledge derived from a number of different sources, that enters interpretation at different points, and that can be mutually constraining when it converges, or fails to converge, on a coherent account of a particular past context” (Wylie 2002, 192). Along a similar vein, archaeologist Bruce Trigger navigates between what he calls the twin myths of archaeological objectivity and archaeological multivo- Archaeological Responses to Philosophical Problems 109 cality. These sorts of trends can be used to provide some stability to archaeological explanations. Other archaeologists should find comfort in a position similar to that of Ian Hacking. Hacking considered realism to concern those entities that can be directly or indirectly observed. Although not all invisible entities fall into this category, some (such as electrons) are able to be “seen” indirectly. Electrons are said to be known by the vapor trails they leave in cloud chambers, and they can even be sprayed on a niobium ball in order to test for the presence of quarks. For Hacking, the fact that electrons can be used in experiments means that they are more than fictions, and that we can know something about them. If spraying something with electrons has an effect that can be tested, then something must be real and knowable about the electrons. To Hacking: “If you can spray them, then they are real” (Hacking 1988, 210). Many contemporary archaeologists focus their activity on those archaeological entities that they can relate to or be tested by other visible traces. Jonathan Reed an archaeologist working in the Galilee, provides a realist example of the ways that archaeological data can be used to understand the past. His realism is moderate due to his beliefs about the proper ends of archaeological fieldwork and analysis: While Reed believes that archaeology is uncovering real facts about the past, he stops short of claiming that archaeology can be used to reconstruct or explain that past in its full detail. In his book on the archaeological world of Jesus, Reed, along with coauthor Dominic Crossan provides a response to biblical archaeologists’ claims that archaeology can be used to prove the veracity of biblical stories or settle questions about the lives of particular people. Reed responds that archaeology can not (and should not) set as its goal the reconstruction of a particular figure’s life. To Crossen and Reed: ...combining archaeology and exegesis—doing parallel layering—is not just about cutting through the layers of the texts to find the authentic sayings of Jesus and cutting through the layers of the ruins to find the real places where Jesus stayed. Instead, it extends to examining the archaeological record of first century layers in all excavated cities, towns, villages, and hamlets, including those never mentioned in the gospels, like Sepphoris and Caesarea Maritima, so that the broader social context in which Jesus was building his kingdom can be understood. And the context of that kingdom in the earliest text stratum of the Jesus tradition has to fit interactively and dynamically with the archaeological context. (Crossan and Reed 2001, 94) Reed’s realism does not go as far as some Biblical archaeologists, a group who see the point of archaeology as the proof of the Biblical text. He does not believe that the archaeological record will give him information about for the guy who sold Jesus a falafel. Reed’s realism is in this way limited. However, Reed can talk about how changes in urban architecture affected urban life during the period when Jesus would have lived and died. This focus, on the period instead of the person, is a much more defensible position to archaeologists as scientists. 110 Chapter 5 The Future of the Realist/Antirealist Debate Although the goals and needs of archaeologists are different from those of physicists, the problems that practitioners of these fields encounter are similar. Philosophers of science would be valuable to archaeologists, providing insights about the philosophical and scientific implications of differing realist and antirealist stances. This would provide some structure, if not guidance, to archaeologists trying to move forward theoretically. While the realist/antirealist debate has been important to philosophers of physics, these issues have not yet been addressed in archaeology, a field that has, until now, struggled with realist issues without any help from the larger scientific and philosophical community. Clearly, the realist/antirealist debate is bigger than any single discipline. If philosophers of science hope to be able to address the larger issue, a general realist/antirealist discussion, it would be a good idea to explore these issues as broadly as possible, and new case studies (such as this one) would seem to be a good place to start such an analysis. Science, Values, and Archaeology On June 1st of 2003, 2 senior archaeologists working in Israel, Larry Stager of Harvard University and Israel Finkelstein of Tel Aviv University were brought to Los Angeles to, in essence, defend their reputations at an event billed as “The Great Debate.”49 At issue to both scholars is whether archaeological and historical evidence supports the notion that Kings David and Solomon ruled most of what is now called modern-day Israel in the tenth century BCE, an account supported by textual material as presented in the Hebrew Bible. While the debaters focused on their respective positions, I believe that this debate highlights two issues. First, the debaters’ positions presuppose a disagreement over a larger issue surrounding the role of certain sorts of evidence to scientific explanation. Second, key elements in this case study will be used to document the integral relationship between science and values.50 The science and values debate centers on the appropriate place of so-called subjective elements in science. My goal is to provide evidence to both philosophers of science and working scientists that their separate attempts to address these sorts of issues are unviable. Only by reengaging with each other can science and the philosophy of science hope to answer these questions and move forward. The Great Debate51 Archaeological finds are silent, and therefore, usually do not provide absolute dates. Archaeologists date different stratigraphic layers and/or assemblages based on the premise that the oldest material is on the bottom and the more recent on the top of an archaeological site. Archaeological Responses to Philosophical Problems 111 Fig. 6. A Hypothetical Tel52 Knowing that one assemblage predates another provides a general framework for relative dating of the phases of occupation at the site. Materials from different strata from various sites are then compared, and inter-site similarities in style can be used to link those sites chronologically, providing a relative chronology of the region. Because the Syro-Palestine region has been extensively excavated, there is a good general understanding of the different archaeological phases. The difficulty is tying archaeological material from a particular phase of occupation to a particular historical event, thereby providing a more specific, or absolute date, for that phase. In order to do this, we must either find something in the archaeological record with a date on it, or link archaeological materials to 112 Chapter 5 an event from some extra archaeological source, such as textual material, generally from Biblical or extra-Biblical sources. To do this, we must find compelling evidence from at least one site to make the connection between the archaeological data and the event. If this is found for even one site, because of the relationship of other sites’ material culture to that site, the same historical horizon should apply to them as well. No two archaeological sites went through the same set of events. Thus the archaeological remains at all sites (even contemporaneous ones) will differ to some extent. In sum, the controversy at hand is packaged to the archaeological community as whether there is sufficient evidence from the archaeological record to connect certain assemblages with the reigns of Kings David and Solomon.53 Evidence for the Traditional Chronology Archaeological evidence: Architecture The archaeology of Jerusalem is complicated because, among other things, a modern city with deep religious significance still exists there. The best candidate for Davidic architecture is a single building called the Stepped Structure, which may have been a part of a terrace, a foundation or extensions, or a retaining wall for David’s city. No remains dating to the reign of Solomon have been uncovered in Jerusalem. Gezer, Megiddo, and Hazor contain material that has been considered evidence of Solomonic architecture and these phases have thereby been dated to the tenth century BCE. All three sites have remarkably similar six chambered city gates, as well as other archaeological remains that the traditional school has attributed to the years of Solomon’s rule: Fig. 7. (Solomonic) 6 Chamber City Gates (not to scale) Archaeological Responses to Philosophical Problems 113 At issue is whether these architectural elements were dated to the 10th century because of internal archaeological evidence or because of the biblical text. Archaeological evidence: Ceramics The big archaeological question initially surrounded a conflict with intra site dating based on ceramic evidence. Assigning dates to different stratigraphic levels usually is based on pottery. Pottery is the most ubiquitous material found at any archaeological site in the region. Because pottery styles can be identified, and since these styles change frequently over time and place, archaeologists use ceramic the record as their primary method of relative dating. Ceramics found at each of the aforementioned sites (and connected to the gates) are believed to be contemporaneous with one another, making the case that if any of the gates are Solomonic, there is reason to believe that all of them are.54 Textual evidence: the Bible David’s escapades are pretty well known (first mentioned in I Samuel 16:13, and the focus of the bibilical text for the remainder of I and II Samuel). David is secretly anointed at a young age as the next king of Israel due to a power struggle between Saul, the current King, and the prophet Samuel. David is famous for a little incident with Goliath, the giant in the employ of the Philistines. After slaying Goliath, David is allowed to marry Saul’s daughter Michal for a reasonable sum of 100–200 Philistine foreskins. Saul and David’s relationship becomes acrimonious, distracting Saul for the remainder of his life. Saul dies, and King David finishes conquering Israel, extending its territory significantly. He moves the secular and religious center of power to Jerusalem, and apparently spends his time watching women bathing. Solomon (the subject of I Kings). is credited in the Biblical account with a number of spectacular (and spectacularly expensive) building projects, including the Temple in Jerusalem. He was also famous for his wisdom in court, and for his large number and variety of wives. After Solomon’s death, and in part due to his fiscal policies, (reportedly the first example of trickle down economics) the Kingdom is split in two, with Israel in the North and Judah in the South. Textual evidence: Extra-Biblical sources Extra-biblical textual materials are written contemporary testimonies to the events purported in the bible. As most of these references are by Israel’s enemies it can be argued that they have no ideological stake in mentioning biblical characters as individuals if they did not exist. Of course, these documents too are subject to numerous interpretations. For example, an Aramaic stele from Tel Dan was discovered in the early 1990s. The inscription includes a reference to “Bit David” or the “House of David.” The stele is terribly broken and the context of this reference is missing but it is the earliest non-biblical reference to the house of David from the ancient world. A second text, the Harris Papyrus, written during the reign of Ramses III (of Egypt’s twentieth dynasty ca. 1196–1166) describes an attack on Egypt by a 114 Chapter 5 group of “Sea Peoples,” among whom are the Philistines, a group also featured in stone reliefs at a temple to Ramses at Medinet Habu. Ramses claims to have settled these Sea Peoples in Canaan, and indeed, sites attributed to the Philistines appear on the southern coast of Israel. As the Philistines were supposed to have arrived in Canaan at roughly the same time as the Israelites, their arrival date is a major pillar in dating remains in Iron Age Israel. Another example lies in a number of monuments commissioned by Sheshonq I (Shishak). Sheshonq I, the founder of Egypt’s twenty second dynasty, claims in his monument at Karnak to have destroyed 154 towns in 925/926 BCE. Steles with Shishak’s name have been found at Megiddo and Byblos. Because Shishak’s attack is corroborated in the Bible and he claims to have wrought so much havoc, archaeological destruction layers at many sites are dated to his rampage through the region. Scientific evidence: Carbon-14 An article in Science55 seems to accept a new argument, that c14 dates can now replace archaeologists’ reliance on pottery. Traditionally, archaeologists shied away from dependence on C14 dates because the method lacked the required precision and archaeologists could date more closely using pottery. Some claim that advances in the procedure have been made, despite the continuing problems (mentioned in chapters 1, 2, and 4) that persist in the calibration and utility of radiocarbon dating. Recently, archaeologist Ami Mazar has provided a series of C14 dates from Tel Rehov, which he claims proves the traditional interpretation that there is ample evidence for a United Monarchy. Science quotes this study, where a colleague of Mazar says that this evidence “puts the nail in the coffin of Finkelstein’s theory” (Ibid.), discussed below. Evidence Against the Traditional Chronology Archaeological evidence: Architecture Finkelstein questions the motivations for calling the aforementioned architecture “Davidic” or “Solomonic.” The stones making up the aforementioned pieces of architecture can rarely be absolutely dated (unless a coin or piece of otherwise datable material falls into the stones, bricks, mortar, etc.), so the evidence must be based on other sources of data. Although Stager insists that the other source is based on archaeological evidence (such as ceramics or Carbon14), Finkelstein claims that the traditionalists are merely forcing the archaeological record fit the Biblical account. Archaeological evidence: Ceramics Finkelstein cites as evidence for his position the analysis of a locus of ceramics from an enclosure at another site, Jezreel. These pieces are similar in type to pottery found at Megiddo in a Temple previously thought to be Solomonic. These finds have been dated to the ninth century BCE.. If the Megiddo Solomonic palace shifts from the tenth century to the ninth century, then should Archaeological Responses to Philosophical Problems 115 it take the rest of the Solomonic architecture with it? Finkelstein believes this must be so. Scientific evidence: Carbon-14 Not to be outdone as a scientist, Finkelstein also brought to the debate a series of C14 samples that he claims dates his material conclusively to the ninth century, and has similarly proclaimed victory. Finkelstein’s reply to Stager is also presented in the article in Science. He concludes that: “The diehards of the conventional dating are in a desperate situation (Ibid. 231).” In The End… The question presented to the audience was: which chronology should we believe? However, this question is unanswerable, given the current level of debate. In chapter 3, one of the problems of this debate, the role of textual evidence in archaeological (and generally scientific) explanation was introduced, and here, beliefs about the differences between particular texts are highlighted. One way to reframe that debate is by focusing on the role (or value) of textual data for the investigator. This focus, on the proper relationship between value judgments and scientific explanation, will be the subject of the rest of this section. Science and Scientific Evidence Both Stager and Finkelstein can (and do) claim to be working as scientists. They gain their data in much the same ways, through the same techniques, etc. Theoretically, the expectation is that two scientists, working on the same materials, should come to the same conclusions on their materials. So, how can science be considered authoritative in any way if its practitioners can come to such widely divergent conclusions? If scientific explanations are rational, cognitive, and “objective,” then this divergence must be due to other (possibly outside) factors that traditional philosophers of science have considered “subjective” or non-cognitive. Traditionalist accounts of science (e.g. Laudan 1984) seek to eliminate or constrain subjective, non-cognitive values, believing that these elements, once removed, will result in a “pure” science. While others (e.g. Lacey 1999) have made a point to include non-cognitive values in science, all traditional theoreticians are committed to the idea that there are “black boxes” or safe regions for cognitive values to lead to objective science.56 116 Chapter 5 Fig. 8. The Logical Positivistic Explanatory Model of Science Black boxes, such as those demonstrated in figure 8, provide a way to save scientific objectivity. For the logical positivists and their heirs, the logical, objective phase of science, called the justification phase, is separated from the initial, subjective, or discovery phase. People could use whatever apparatus they wanted to come up with hypotheses. Hypotheses could come from well reasoned questions or from drug induced hallucinations. Regardless of their source, the hypothesis would then be tested. The testing (or justification) phase, unlike the discovery phase, is thoroughly logical, based on an approved explanatory apparatus, such as Hempel’s covering law models. the goal of this model of science is to retain objectivity while recognizing the role of subjective values in theory formation. However, the strong separation between discovery and justification is, in my opinion, untenable, as real science can not be neatly separated into subjective and objective stages. Neither Stager nor Finkelstein’s work can be separated into objective and subjective components. Each of them come to conclusions based on a much more complicated model of scientific explanation. This model, a variant of the model being developed by Alison Wylie, explains how values can be found throughout science, on what she calls both the “receiving” ends and “shipping” ends of scientific exploration. Throughout her Thinking From Things, Wylie develops a model of explanation that focuses on the effect value judgments have, both in the so called discovery phase (the shipping end) and in the resulting explanation (the receiving end). On the receiving end, a number of factors (or values) direct each scientist’s research: many of which are interrelated, many of which will set the conditions for good science, and many of which will determine what the proper answers are to archaeological questions. So, the shipping end of science cannot be value free, if the questions asked are value laden. Archaeological Responses to Philosophical Problems 117 My version of this model, represented by figure 9, adds a valuable feedback mechanism to Wylie’s description of source and subject side factors, (see the bottom of the figure). The feedback loop added to the explanatory model is designed to help scientists working from competing perspectives to come to consensus, or at least to provide a model for relative confirmation.57 Fig. 9. A Modified Version of Wylie’s Explanatory Model of Science Currently, exchanges like those between Finkelstein and Stager have resulted in both parties devoting their energies to trying to prove each other wrong and to entrenching their positions. This behavior, completely reasonable given the stakes for the winner, provides no means for settling their conflict. Rather than focusing on the competing theory, creating critical tests or Popperian style calls for falsify-ability of the enemy’s position, which merely result in escalating attacks on the competing theory, and result in combativeness and entrenchment rather than in resolving disputes, my interest is in self-regulation. Conclusions should be plugged back into the source-side portion of a tentative explanation in an effort to self-regulate, improve, or disprove it. This additional step should help to shift archaeologists from thinking about their models as being either entirely correct or entirely wrong, in a sort of Popperian framework, and move them toward a Lakatosian model of science. For Lakatos (1999), scientific models are always in competition with one another. Unlike Kuhn, who sees science as having periods of unification (or normalcy), Lakatos’ world is one of continual conflict. Models feature a hardened (protected) core and a mass of auxiliary hypotheses, designed to protect the core. These hypotheses are modified, exchanged, and pitted against competing models, and if the desired model gains more acceptance against its competitors, the research program is considered to be moving in a good direction, and is said to posses a positive heuristic. Alternatively, a failing program is said to posses a negative heuristic. If scientists are forced to expend effort in critiquing and analyzing their own positions (espe- 118 Chapter 5 cially in the face of competing theories), instead of attempting all or nothing attacks on each other, there is at least some reason to believe that better theories will be the result. In the case study at hand, an examination and rigorous critique of the auxiliary hypotheses at the foundation of the two positions is a necessary first step if those archaeologists have any hope for rapproachmont. A Call for Re-evaluation Instead of focusing on ways to apply particular subjective texts to an objective science, we should instead take the time for a systematic reevaluation of the uses of textual and archaeological data in Syro-Palestinian Archaeology. Just as differences in scientific practice result in differing explanations of the past, information gathered from text (and from other receiving end values) needs to be explicitly understood and utilized if it is going to have a positive impact on our quest to understand the past. In this example and in the previous study of archaeological realism, archaeologists encounter problems that could as easily be understood in philosophical terms. One conclusion from this fact is that it would benefit archaeologists to take advantage of the gains that have been made in philosophical explanation in the half century since the rise and fall of processual archaeology. However, I would counter that archaeological questions, qua philosophical questions, should arouse the attention of philosophers of science as well. Archaeological science has its own focus. Archaeological realism is not entirely like realism for physicists, and archaeological values might differ from biological values. As such, archaeology is a fertile and relatively untapped resource for philosophers of science. Past attempts to fit archaeology into “established” philosophical models, or to fit philosophical theories to archaeological practice, have ended in failure. Only a truly interactive approach, with archaeologists looking at philosophical problems and philosophers looking at archaeological problems, may succeed where past attempts at dialogue have failed. This can only occur, however, if the two groups talk to each other, instead of each field trying to recast the other in its own image. The goal of the final chapter of this book is to show how and where this interaction might start. Notes 46. This is especially relevant, since that ancestor must have been real for the author to exist. Archaeological Responses to Philosophical Problems 119 47. These explanatory models run the gamut from postmodern rejections of archaeological objectivity to neo-processual attempts to “fix” the science of archaeology. 48. Whether Hempel did or did not actually consider himself a realist (or possibly a pragmatist) in regard to theoretical entities is contentious. This question (however interesting) goes beyond the scope of this book. 49. The Great Debate, held at the UCLA campus, was sponsored by The Cotsen Institute of UCLA, the American Schools for Oriental Research, and the Tel el-Farah, South archaeological excavation. 50. Portions of this section were developed at the 2003 NEH summer institute on Science and Values. 51. Tammi Schneider of the Claremont Graduate University provided the background information for this case study (in the form of her opening remarks for this event). 52. From Ogden and Chadwick, p. 131 53. Whether this is the real question, or even a good question will be the subject of the remainder of this chapter. 54. As I will note below, the original dating of these ceramics to the 10th century has recently come under fire. 55. Science. Volume 300. 11 April, 2003, p 229-231 56. Even the otherwise very progressive scientific accounts of Kitcher (2001), Longino (2002), and Solomon (2001) try to divide science into objective and subjective parts (although each fights to draw that line in a slightly different place), and so fail to be able to account for the complexity and interconnectedness between science and values. 57. These modifications to Wylie’s Model were first presented and discussed during a session devoted to her work at the 2004 PSA Conference. 120 Chapter 5 Chapter 6 Future Studies: Archaeological Explanation and the Philosophy of Science Where Do We Go From Here? One goal of chapters 1–5 has been to assert that both philosophers of science and archaeologists have benefited from a their partnership, a partnership that is seeking new members and fresh perspectives. The objective of this final chapter is to move beyond the more traditional issues of chapter 5, and provide an example of a new issue for future study by both groups. Clearly, it would be impossible to provide an exhaustive treatment (or even an exhaustive introduction) to the potential of this interaction. The following case study (as well as the study of archaeological realism and antirealism from chapter 5) is only meant to be a sampling of what the author hopes are things to come. Dirt archaeologists, archaeological theorists, and archaeologically minded philosophers of science routinely question the connections between their disciplines in a manner that parallels more general discussions about the relationship between philosophers of science and scientists. For archaeologists specifically, one question revolves around the supposed unity of archaeological practice that was promised by Hempelian philosophy of science. For example, in the 1960s and 1970s, archaeologists around the world to adopted the methodological practices of American new archaeologists, believing that archaeological practice needed to be unified to meet the goals of new archaeology. The unification thesis (at least in terms or archaeological methodology) will be rejected after examining differences between the archaeologies of different regions. A project newly underway will be cited as evidence of the potential for archaeological growth in a non-American context. Theory, Geography, and the (Dis)unity of Science An issue touched upon throughout this book has been the unity of science, or more specifically, the unity of archaeological science. Although it is true that archaeology was to have been unified (or mostly unified) by the new archaeological platform, the question remains whether such unity is possible today, or whether it should even be a goal of contemporary archaeology. One facet of the unity of science debate centers around scientific practice. If archaeology is to become an objective study of the past, some might posit that unifying archaeological practice would result in the repeatability and predictability desired by the scientific community. 122 Chapter 6 It could be argued that Binford’s (1962, 1965) idea of archaeological science implied a standardization of archaeological practice. While the main goal of Binford’s new archaeology was to find a theoretical foundation for scientific archaeology, the resultant theoretical changes led archaeologists to push for a standardization of archaeological practice. The sciences are seen as valid due to their known, documented, and agreed upon methods that result in predictability and repeatability, the hallmarks of objective experimental science. As such, archaeologists believed that, in order to become a science, archaeology would need to leave behind its haphazard, idiosyncratic practices in favor of new, objective, and ideally, universal methods. The push for methodological unification did not start with Binford and Hempel. These thoughts come from the wellspring of logical positivism, the Vienna Circle. In its manifesto, the Circle’s members are clear in their desire to unify (and reduce) scientific practice: The scientific world conception is characterized not so much by theses of its own, but rather by its basic attitude, its points of view and direction of research. The goal ahead is unified science. The endeavour is to link and harmonise the achievements of individual investigators in their various fields of science. From this aim follows the emphasis on collective efforts, and also the emphasis on what can be grasped intersubjectively. (Der Wiener Kreis, 327–8) Without focusing on issues surrounding reduction,58 unification was seen as essential to the creation of universal laws for any science. Even the immature sciences, such as History (the subject of Hempel’s 1942 paper), would benefit from unification, and would eventually attain the sort of precision currently found in the most complete science: “The methodological problem of the application of axiom systems to reality may in principle arise for any branch of science. That these investigations have thus far been fruitful almost solely for physics, however, can be understood from the present stage of historical development of science: in regard to precision and refinement of concepts, physics is far ahead of the other branches of science” (Ibid. 334). The new archaeologists, understanding the goals of the Vienna Circle through the lens of Hempel and Binford’s work (the focus of chapters 1-3 of this book), saw their task of creating an archaeological science as being tied to scientific (and therefore unified) methods of enquiry. Binford’s vision of a new archaeology coupled a change in theoretical focus (from the collection of things to the examination of extinct cultural systems) with a methodological change in focus (from excavating sites and setting chronologies to surveying regions). Binford’s theoretical vision came from an improved understanding of the complexity of human behavior. By setting rigid chronologies at sites, based on areas of comparative (and artificial) stability, new archaeologists believed that their predecessors forfeited the ability to understand normal fluctuations of communal behavior within a larger context. These fluctuations are better understood from a regional survey approach. To Stephen Plog, “[the] wholesale adherence to methods of dating sites to stage or phase systems…has obscured patterns of short-term change that may provide critical clues for understanding that [area’s] Future Studies 123 history” (1986, 3).59 Since American archeologists (rightly) saw that their previous focus on the (static) site distorted the real (fluid) picture of the region, the new, philosophically informed, archaeologists recognized that their goal of creating the deductive/universal laws envisioned by Hempel would be best achieved by maximizing their data set. Culture history (the previous archaeological paradigm) focused primarily on site excavation. This focus made perfect archaeological and philosophical sense, as the goal of this archaeology was to catalog and understand local changes in order to set that site’s chronology. The new archaeology’s change in archaeological focus also reflected a philosophical emphasis. As Hempelian philosophy of science would require large amounts of data (seen as the only way to create general laws), changes would need to be made on the ground in order to provide those data. Further, in order to make the data more representative, what constituted good archaeological practice changed to reflect the new philosophical paradigm. As individual sites represent idiosyncratic activities, the excavation of sites could be seen as counterproductive to the construction of good science. If universal laws are the objects of archaeology, archaeological science should ignore the local site and focus on the region. As an emphasis on and excavation of sites (the small picture) would distort the new (more realistic) regional picture, sites were de-emphasized, becoming a part of the landscape and being and excavated in that context. In this case, a theoretical shift would seem to necessitate practical changes that would have serious effects on the way that localized archaeologies are practiced. The geography of the American Southwest lent itself well to the philosophical changes that characterized theoretical archaeology in the mid twentieth century. In some cases, the new emphasis led scientific archaeologists in the field to make sweeping changes to reflect the new focus on regions. For example, some have advocated the increased use of mechanized equipment in excavation. Traditionally this equipment is to be used to clear off topsoil in order to identify archaeological remains. However some investigators claim that mechanized trenching of sites has some advantages over excavating, both for both theoretical and practical reasons. Since excavating by hand can take years, and still results in very little (in many cases, less than 10 percent of the site) exposure, archaeologists such as Herr, Clark, Swartz, and Hall believe that the backhoe can produce superior results for the new archaeologist’s purposes. “By revealing archaeological remains at a scale more substantial than allowed by hand excavation, the backhoe facilitates the reconstruction of settlement, community, and regional population dynamics.”60 Others have used backhoes to trench through regions, including any sites found in the region.61 In addition, changes in the field led to changes in the literature. While texts used to focus on ways to maximize and control data acquisition at a site, newer texts (such as F. Plog 1974, M. Schiffer 1982, and S. Plog 1986) contained sections focusing on better survey strategies. This is not to say that the new archaeologists completely abandoned excavation. However, if the goal of the new archaeology is to develop general (universal) laws, then the de-emphasis (or for practical purposes, abandonment) of site 124 Chapter 6 analysis seems like the next logical step. In fact, some new archaeologists, both in the Americas, and more importantly for the purposes of this paper, in the Middle East, believed that the move to a survey approach was the next logical step in the quest to make archaeology scientific, and it is this trend that I wish to use as my case study against methodological unification. In fact, although Binford may have been correct in his desire to transform archaeology into a science, I would argue that specific geographical and anthropological elements of ancient Canaan give good reason to fight against the contention that archaeological practice should be unified (in any manner that goes beyond making grand generalizations about archaeology’s focus on soil, ceramics, and the like.62 In the Americas, much of the archaeology done is on comparatively lightlystratified remains.63 These remains are either found at sites that have a few levels of occupation, such as abandoned buildings or agricultural installations, or in a larger region. In American excavations, the region to be surveyed or excavated is chosen for a variety of reasons, including: the remains of villages or buildings, such as those found at Carter Ranch, Broken K Pueblo, the Mesa Verde cliff cities, or sites such as Caral (Peru), or distinguishable features spanning a given area, such as the Hopewell mounds in Ohio, or changes in tombstone iconography in the Boston area, documented by the Plymouth Colony Archive Project, or a significant amount of artifactual evidence throughout an area, such as the widely distributed petroglyphs found throughout the Puako site in Hawaii. Fig. 10. Puako Petroglyph Archaeological District, Hawai’i Future Studies 125 Although geography such as mountain ranges or significant bodies of water can be used to set the boundaries of an excavation, determining the boundaries of a region can be quite difficult. In large areas where artifact densities are very low (such as one or two pieces in an acre of land), it can be impossible to set any real boundary between inhabited and uninhabited spaces. In many cases, a site’s boundaries are set for the purposes of the project, or need to be argued for as a part of the dig’s rationale. For Syro-Palestinian Archaeologists, these types of questions are (for the most part) far less important. As discussed in previous chapters, in ancient Canaan, the dominant archaeological feature is the tel: Fig. 11. Tell el-Far’ah (South), viewed from wadi Besor 126 Chapter 6 As a reminder, tels are typically well-demarcated flat-topped mounds that reflect a very different way of life from that of the Native American groups, mentioned above. Tels are human constructions. They are built up from bedrock in layers. People built upon older sites (for reasons including defense, the proximity of food and water sources, and the like), and tended to stay there until something forced them to leave. When they left, another group settled in the same area, and either added on to the previous tenants’ architecture, or knocked it down, leveled it off, and built on top of it. As a result, tells can easily contain thousands of years of more or less continuous occupation, broken up by almost constant conflict, shifting populations, and the occasional fire, flood, or earthquake. During times of prosperity, when the tel became overpopulated, cities expanded to the surrounding landscape. Just as we see today, we can (depending on the level of trust we give to ethnographic analogy64) imagine contact between these cities, including cultural and material exchange; we can further postulate the existence of larger groups (or peoples) who inhabited numerous sites in a region (or even the whole land) and who considered themselves to be citizens of that larger whole. In addition, several cultures of semi-nomads moved about in a fashion similar to that of the aforementioned groups in the American Southwest. The recognition that Canaanite archaeology is different than American archaeology, and that different places need different excavation plans reaches back to the late 1800s. By the time Sir William Flinders Petrie published Methods and Aims in Archaeology (1904) he had established that the tel-centric geography of regions such as Israel is best suited to a specific archaeological method, involving stratigraphic analysis, ceramic typology, etc. Although all of the other methodological tools of new archaeology have substantially changed field practices in the intervening years, the geography of the land has not changed. Tels are still the centers of most activity, so their excavation remains the focus of Syro-Palestinian archaeologists. Although theoretical archaeology has moved forward (and much to archaeology’s benefit, in my opinion) since the introduction of the new archaeology, I would argue that archaeological methodology has continued to focus on those features of new archaeology, mentioned above, that are in tune with scientific unification. In fact, a number of Syro-Palestinian archaeologists came to believe that the future of Syro-Palestinian archaeology was to be found in a move to what they believed was the American-style standardization of archaeology.65 These theoretical Syro-Palestinian archaeologists cast themselves and their projects in the terminology and the goals of American archaeology, calling themselves new archaeologists and positivists.66 Believing that theory should dictate methodology, some argued that Syro-Palestinian archaeology’s future would include a similar shift in focus toward an emphasis on regional survey and a deemphasis on (and in a couple of rare cases, elimination of) excavations. These theoretical Syro-Palestinian archaeologists cast themselves and their projects in the terminology and the goals of American archaeology, calling themselves new archaeologists and positivists,67 and they claimed that their data will be useful in producing general laws of archaeology. In addition to this theoretical shift, it is becoming more and more difficult to reconcile the expenses associated with Future Studies 127 excavating a tel. Coupling the claim that they are more scientifically fruitful (and certainly more economical) than the large scale, excavation of single sites, fewer and fewer site licenses (and more and more surveys) are being granted by the Israel Antiquities Authority. However, I believe that this trend, the well meaning, seemingly theoretically informed attempt to move the Archaeology of Canaan toward a unified notion of science, ignores specific features of Canaan that are crucial to its scientific analysis.68 The problem here is a result of the mistaken belief of Syro-Palestinian archaeologists that a theory developed elsewhere (however effective in its native homeland) should be applied uncritically in a new context. This belief is not uncommon, and results from another feature of Hempelian science (and science in general), the belief that the proper relationship between theory and application is top down. As science is the universal quest for truth, the important (and in many cases, the only documented step) is the development of the scientific model. Once the theory is perfected, it is simply applied. In fact, even in the case of the new archaeology in the Americas, the relationship between theory and methodology was never one sided. Landscape features and settlement patterns have traditionally governed the methodological approach of the American archaeologists as much as it does for the archaeologists of ancient Canaan. The fact that Binford’s adaptation of covering law explanations was tied to a shift in emphasis toward regional surveys was more than coincidence. American archaeology was not simply grafted onto Hempel’s models. There was an interplay between theoretical desires and a geographical reality, which resulted in a new archeology that was well suited to the American landscape. It makes perfect sense that a theoretical archaeology of ancient Canaan could only emerge after engaging in a similar struggle, and it will probably be very different than its American counterpart. This should not be surprising. After all, if there are different theoretical models of physics for large and small entities, why should anyone assume that the entire field of archaeology should be unified in its approach. It is worth stressing that although advances in archaeological theory continue to be made, (by both archaeologists69 and philosophers of science70) the relationship between theory and methodology has not been a major focus for archaeologists. In other fields, intuitions about the complex relationship between theoretical and practical science are finding a voice. For instance, biologist Robert Kohler (2002) provides a history of the two components of his field (lab biology and field biology), and examines the changing relationships between the two. His conclusions about the need for an interactive approach, exemplified in what he calls “border practices,” in which laboratory and field techniques are combined, should give archaeologists and philosophers of science a lot to think about. Instead of centering on the idea that fieldwork should blindly follow from explicitly stated theoretical goals and methods, archaeologists should be aware that issues surrounding geography and methodology are relevant to archaeological theory development. An analysis of archaeological field/lab border practices would hopefully cause archaeologists to conclude that archaeology as a field need not be monolithic, and that there may be several sciences of archaeology, 128 Chapter 6 or several models that, working together, comprise a looser, interconnected, and more comprehensive theoretical science of archaeology.71 An example of the emerging Canaanite approach is utilized by the Philistine Countryside Project and Tell el-Far’ah, South excavations, directed by Lehmann and Schneider. 72 Lehmann and Schneider’s project combines an intensive excavation of major sites in the region (focusing for the present on Tell el-Far’ah, South) and a separate survey of the Northern Negev countryside: The Philistine Countryside project will apply satellite/airborne Remote Sensing and Geographic Information Systems (GIS) as well as other advanced nondestructive technologies for regional archaeological investigations such as an archaeological survey, which will contribute to an understanding of the longterm interaction between the Bronze and Iron Age cultures and their physical environment. The project will also conduct an excavation at Tell el-Far'ah (South)…The excavations will provide stratified scientific samples, ceramic finds, and other relevant artifacts, which will also help dating the finds of the survey. Together with the artifacts, environmental data from the excavations, including faunal and floral remains, will provide the necessary depth of information usually not available through surveys. (Lehmann 2000) Although Schneider and Lehmann (rightly) avail themselves of regional surveys (which have been done extensively all over Israel for well over a century), they recognize that site excavations provide important data that are unavailable from surveys. They reason that retaining an emphasis on tel excavation will result in a better understanding of the peoples of ancient Canaan than they could get by shifting their emphasis to survey, relegating tels to a part of the landscape, as some believe the new archaeology’s focus on universal laws and data maximization would suggest. For these archaeologists, following the advice of philosophers like Merrilee Salmon (2001), local interpretive issues are more important than blind adherence to a formal structure. Here, local archaeological features, such as the geography and settlement patterns of this region, and not borrowed theoretical approaches, should guide local archaeological practice. Schneider and Lehman have chosen an emphasis on regional geography over a loyalty to a borrowed, ill fitting, set of universal guiding principles of archaeological science. It remains to be seen whether Schneider and Lehmann’s methods will provide better explanations than those who prefer a shift in emphasis toward survey. Although it would seem that a comparison would easily answer this question, there is a significant problem standing in the way. Many Syro-Palestinian archaeologists do not seem to put a high priority on publication, so there are not many examples of contemporary survey-centric projects to test against Schneider and Lehmann’s methods. It is one of this team’s goals to make their data available, and to analyze the Philistine Countryside project and Tell el-Far’ah, South excavations’ results in light of other projects as their results are published. In closing, this case study shows the importance of understanding features relevant to a science before attempting to set theoretical or methodological constraints on it. In this case, both philosophers of science and archaeologists were Future Studies 129 so intent to make archaeology a science that they have hobbled their ability to explain or understand their subject matter. Here, methodological unification proved to be detrimental to archaeologists working in Syro-Palestine, causing them to devalue data that are central to their discipline. This is not to say that archaeology should return to the historically based enterprise that it once was, but it does mean that questions surrounding the relationship between theory and methodology have more complicated answers than Hempelians and new archaeologists once thought. If this means that instead of one philosophy of archaeology, that there are multiple theoretical archaeologies, so be it. More generally, if unification serves to help move a discipline forward, then unification is a worthy goal. However, unification for unification’s sake, or making unification a central goal of science serves to save only an oversimplified notion of science, to the detriment of real scientific progress. Moving Forward Throughout this project it has been the clear intent of the author to propose that both archaeologists and philosophers of science have a lot to learn about scientific explanation. Archaeologists have come to realize that science is not as unified as its image would lead us to believe. Regardless of its application, Hempel’s vision of explanation was an idealization of the scienctific enterprise. Hempel never claimed that scientists, working in the lab or the field, actually did science by using his methods, or that theories were explained or confirmed only by referencing Hempelian models. Hempel’s models were instead ways to recreate the scientific process, and make science look more scientific. When archaeologists attempted to use Hempelian explanation to make an archaeological science, they found this idealized version lacking. In the place of the covering law models given by Hempel and Binford, archaeologists, working with archaeologically informed philosophers of science, have begun to embrace a large variety of ways of doing science. The previous chapters have shown how archaeological science has evolved, and how issues surrounding the nature of archaeological laws, on the epistemological status of theoretical entities, and on the relationship between theory and practice, have come to the foreground. Hopefully, this project has served to demonstrate that many archaeological problems have philosophical analogues. If this is true, it is truly exciting, because despite the fact that archaeologists and philosophers of science have approached scientific issues from very different perspectives, the amount of common ground should lead both groups to believe that they are doing something right. Of course, this is not to say that the issues of today will result in a complete theory of scientific explanation. The hope is that the common threads addressed in this work will encourage more interaction, and lead more archaeologists and philosophers of science to work with one another. Finally, it is time to ask the question in the title of this book. Can there be a philosophy of archaeology? This was the question my dig director Sy asked me one afternoon in the hot sun of southern Israel. Certainly, in the new archaeolo- 130 Chapter 6 gy there was (at one time) a philosophy of archaeology. But, could there be one now, and what would that mean? This is a tough question, one which remains open. On a formal level, even if the future brings another model with the archaeological charisma of the new archeology, it is difficult to imagine that the complex philosophical and archaeological issues that have been the subject of this work will be easily amenable to unification. However, if archeologists and philosophers can develop a more hermeneutic approach to deal with theoretical and methodological issues, there are some real opportunities for future development. Could this lead to (or toward) a philosophy of archeology? That is an interesting question. Notes 58. The reduction of all sciences into the one true science, physics, was the ultimate programme of many of the Vienna Circle’s members. 59. For more examples, see the work of Fred Plog (1974) and Schiffer, Lyman, and O’Brian (2005). 60. Herr, Susan, et. al. (2002). See also the Excavations of Fort Clatsop by Paul J. F. Schumacher (reference in bibliography) 61. This practice is most commonly used by BLM/CRM (Bureau of Land Management/Cultural Resource Management) groups in order to find sites. However, in some cases, both known and unknown sites are trenched as a means to gather data about the larger landscape. 62. And such statements are so general as to be theoretically and methodologically worthless. 63. This is not to say that American archaeology is un-stratified, and that there are no examples of complex stratigraphy in American sites. However, I think there is little question that the amount of stratigraphy in an American site is far less than is generally seen in sites in the Ancient Near East. 64. Discussed in chapters 3 and 4. 65. Standardization of Locus sheets and other data collecting forms was explicitly called for to this end. The stated goal was to make our data compatible with other sites throughout the region. Hempel was often cited when we were told that the goal was that these data would be used to construct regional (and eventually) universal laws or archaeology. 66. For example, William Dever, a well-known figure in the field, routinely refers to himself as a new archaeologist. 67. Standardization of Locus sheets and other data collecting forms was explicitly called for to this end. The stated goal was to make our data compatible with other sites throughout the region. Hempel was often cited when we were told that the goal was that these data would be used to construct regional (and eventually) universal laws or archaeology. 68. Many of these “new archaeologists” have since withdrawn their explicit endorsement of universal laws of archaeology, but they have done so without giving up the Future Studies 131 title of new archaeologist or acknowledging the problems with methodological unification. 69. Examples include Bruce Trigger, Lewis Binford, Ian Hodder, and Steven Mithen. 70. Examples include Merrilee Salmon, Alison Wylie, and Peter Kosso. 71. This new, multi-vocal approach to archaeological theory can be found in a number of newer works from very different perspectives, including Wylie 2002 and Hodder, 2004. 72. Also see: The Landscape Archaeology Project at Dor (Gibson, Kingsley, and Clarke), The Wadi Faynan Landscape Survey (Barker, Adams, Creighton, Daly, Gilbertson, Grattan, Hunt, Mattingly, McLaren, Newson, Palmer, Pyatt, Reynolds, Smith, Tomber, and Truscott), and The Sataf Project of Landscape Archaeology in the Judean Hills (Gibson, Ibbs and Kloner). 132 Chapter 6 Bibliography Ayer, A.J. Language Truth and Logic. New York: Dover Publications, Inc. ———. “Verification and Experience.” In Logical Positivism. Edited by A. J. Ayer. New York: The Free Press, 1959, 228-243.First published in Proceedings of the Aristotelian Society 37 (1936-1937) (page citations are to the reprint edition). Binford, Lewis. “Archaeology as Anthropology.” American Antiquity 27:4 (1962): 217-225. ———. “A Consideration of Archaeological Research Design.” An Archaeological Perspective. New York: Seminar Press (1972): 135-163. First published in American Antiquity 29 (1964): 425-441 (page citations are to the reprint edition). ———. “Smudge Pits and Hide Smoking: The Use of Analogy in Archaeological Reasoning.” American Antiquity. Salt Lake City: The Society for American Archaeology 32:1 (1967): 1-12. ———. “Archaeological Perspectives.” An Archaeological Perspective. Binford. New York and London: Seminar Press (1972): 78-104. First published in New Perspectives in Archaeology. Edited by S. and L. Binford. (Chicago: Aldine Publishing Co., 1968) (page citations are to the reprint edition). ———. “Post-Pleistocene Adaptations.” In New Perspectives in Archaeology. Chicago: Aldine Publishing Company, 1968a, 313-342. ———. “Archaeological Reasoning and Smudge Pits—Revisited.” In An Archaeological Perspective. New York: Seminar Press, (1972), 52-58. Caldwell, J.B. “The New American Archaeology.” Science 129 (1959): 303307. Carnap, R. “Testability and Meaning.” In Theories and Observation in Science. Edited by Richard Grandy. Englewood Cliffs: Prentice-Hall, Inc (1973): 2746. First published in Philosophy of Science 3 (1936): 419-471 and 4 (1937): 1-40. (page citations are to the reprint edition). ———. “The Concept of Confirming Evidence.” In The Concept of Evidence. Edited by Peter Achinstein, 1983, 79-94. First published in Logical Foundations of Probability (University of Chicago Press, 1950), 462-464, 468-481 (page citations are to the reprint edition). ———. “Empiricism, Semantics, and Ontology.” In The Philosophy of Science. Edited by Boyd, Gasper, and Trout, 1991, 85-97. First published in Meaning and Necessity, enlarged edition (Chicago: University of Chicago Press, 1956), 205-221 (page citations are to the reprint edition). ———. “The Elimination of Metaphysics Through Logical Analysis of Language.” In Logical Positivism. Edited by A. J. Ayer. New York: The Free Press, 1959, 60-81. First published as “Überwindung der Metaphysik durch Logische Analyse der Sprache,” in Erkenntnis, volume 2 (1932) (page citations are to the reprint edition). Cartwright, Nancy. “The Reality of Causes in a World of Instrumental Laws.” In The Philosophy of Science. Edited by Boyd, Gasper, and Trout, 1991, 379- 134 Bibliography 386. First published in PSA (1980, volume 2): 38-48 (page citations are to the reprint edition). ———. How The Laws of Physics Lie. Oxford: Clarendon Press, 1983. Charlton, Thomas. “Archaeology, Ethnohistory, and Ethnology: Interpretive Interfaces.” Advances in Archaeological Method and Theory 4, 1981, 129176. Churchland, Paul. “The Anti-Realist Epistemology of van Fraassen’s The Scientific Image.” In Readings in the Philosophy of Science: From Positivism to Postmodernism. Edited by Theodore Schick, Jr. Mountain View: Mayfield Publishing Company, 2000, 273-279. First published in Pacific Philosophical Quarterly 63 (July 1982) (page citations are to the reprint edition). Coffa, J. The Semantic Tradition from Kant to Carnap To the Vienna Station. Edited by L. Wessels. Cambridge: Cambridge University Press, 1991. Collingwood, R.G. “Human Nature and Human History.” In The Philosophy of History. Edited by Gardiner, Patrick. London: Oxford University Press, 1974, 17-40. Cowgill, George. “Archaeological Applications of Mathematical and Formal Methods.” In American Archaeology Past and Future: A Celebration of the Society for American Archaeology 1935-1985. Edited by David Meltzer, Don Fowler, and Jeremy Sabloff. Washington: Smithsonian Institution Press, 1986, 369-394. Deetz, James and Edwin Dethlefsen. “Death’s Head, Cherub, Urn and Willow.” Online http://etext.lib.virginia.edu/users/deetz/Plymouth/deathshead.html . First published in Natural History 76:3 (1967): 29-37 (page citations are to the reprint edition and are therefore approximate). Dever, William. “The Impact of “New Archaeology” on Syro-Palestinian Archaeology.” BASOR, number 242 (Spring 1981): 15-30. ———. “Syro-Palestinian and Biblical Archaeology.” In The Hebrew Bible and its Modern Interpreters. Edited by Knight and Tucker. Philadelphia: Fortress Press, 1985. ——— What Did the Biblical Writers Know & When Did They Know It? Grand Rapids: William B. Eerdmans Publishing Company, 2001. Dothan, Trude and Moshe Dothan. People of the Sea: The Search for the Philistines. New York: Macmillan Publishing Company, 1992. Dray, William. Laws and Explanation in History. London: Oxford University Press, 1957. Drinkard, Jr., Joel, Gerald Mattingly, and J Miller, ed. Benchmarks in Time and Culture: An Introduction to Palestinian Archaeology. Atlanta: Scholars Press, 1988. Edwards, A.W.F. Likelihood. Cambridge: Cambridge University Press, 1972. Feyerabend, Paul. “How to Defend Society Against Science.” In Introductory Readings in the Philosophy of Science. Edited by Klemke, Hollinger, and Rudge. Amherst: Prometheus Books, 1998, 54-65. First published in Radical Philosophy 11 (1975): 3-8 (page citations are to the reprint edition). Flannery, Kent. “Archaeological Systems Theory and Early Mesoamerica.” In Contemporary Archaeology: A Guide to Theory and Contributions. Edited Bibliography 135 by Mark Leone. Carbondale and Edwardsville: Southern Illinois University Press, 1972, 222-234. First published in Anthropological Archaeology. Edited by Betty Meggers (Washington DC: Anthropological Society of Washington, 1968), 67-87 (page citations are to the reprint edition). ———. “The Cultural Evolution of Civilizations.” Annual Review of Ecology and Systematics 3 (1972): 399-426. ———. “Archaeology With a Capital ‘S’.” In Research and Theory in Current Archaeology. Edited by Charles Redman. New York: John Wiley and Sons, Inc., 1973, 47-53. Flannery, Kent and James Schoenwetter. “Climate and Man in Formative Oaxaca.” Archaeology 25:2 (1970): 144-152. Fodor, Jerry. “Special Sciences.” In The Philosophy of Science. Edited by Boyd, Gasper, and Trout. Cambridge, Massachusetts: The MIT Press, 1991, 429441. First published under the title “Special Sciences, or The Disunity of Science as a Working Hypothesis.” Synthese 28 (1974): 77-115 (page citations are to the reprint edition). Fritz, John, and Fred Plog. “The Nature of Archaeological Explanation.” American Antiquity 35 (1970): 405-412. Gardiner, Patrick. The Nature of Historical Explanation. London: Oxford University Press, 1952. Garfinkel, Alan. “Reductionism.” In The Philosophy of Science. Edited by Boyd, Gasper, and Trout. Cambridge, Massachusetts: The MIT Press, 1991, 443-459. First published in Forms of Explanation. New Haven: Yale University Press,1981 (page citations are to the reprint edition). Gero, Joan. “Producing prehistory, controlling the past: the case of New England beehives.” In Critical Traditions in Contemporary Archaeology. Edited by Pinsky and Wylie. Albuquerque: University of New Mexico Press, 1995, 96-103. First published in Forms of Explanation (New Haven: Yale University Press,1989), 49-74 (page citations are to the reprint edition). Giere, Ronald. Science Without Laws. Chicago: The University of Chicago Press, 1999. Hacking, Ian. Logic of Statistical Inference. Cambridge: The Syndics of the Cambridge University Press, 1965. ———. “Experimentation and Scientific Realism.” Philosophical Topics 13 (1982): 71-87. Hahn, Hahns. “Logic, Mathematics and Knowledge of Nature.” In Logical Positivism. Edited by A. J. Ayer. New York: The Free Press, 1959, 147-161. First published as the first four sections of the pamphlet “Logik, Mathematik und Naturerkennen,” (Vienna 1933) (page citations are to the reprint edition). Hayes-Roth, Frederick, Donald A. Waterman, and Douglas B. Lenat, ed. Building expert systems. Reading: Addison-Wesley Publishing Co., 1983. Hempel, Carl. “The Empiricist Criterion of Meaning.” In Logical Positivism. Edited by A. J. Ayer. New York: The Free Press, 1959, 108-129. First published in Revue Internationale de Philosophie (1950) (page citations are to the reprint edition). 136 Bibliography ———. “Aspects of Scientific Explanation.” In Aspects of Scientific Explanation. New York: The Free Press, 1965, 331-496. ———. “The Function of General Laws in History” (1942). Reprinted in Aspects of Scientific Explanation. New York: The Free Press, 1965a, 231-243. A modified version of the original text, first published in The Journal of Philosophy 39 (1942): 35-48 (page citations are to the reprint edition). ———. “Inductive Inconsistencies.” In Aspects of Scientific Explanation. New York: The Free Press, 1965b, 53-79. First published in Synthese 12 (1960): 439-469. ———. Philosophy of Natural Science. Upper Saddle River, NJ: Prentice Hall, 1966. ———. “The Logic of Functional Analysis.” In Readings in the Philosophy of Social Science. Edited by Michael Martin and Lee McIntyre. Cambridge, Massachusetts and London: the MIT Press, 1994, 349-375. First published in Symposium on Sociological Theory. Edited by Llewellyn Gross. (Harper & Row Publishers, 1959) (page citations are to the reprint edition). Hempel, Carl and Oppenheim, Paul. “Studies in the Logic of Explanation.” In Aspects of Scientific Explanation, 1965, 245-295. First published in Philosophy of Science 15 (1948): 135-175 (page citations are to the reprint edition). Hill, James. “Broken K Pueblo: Patterns of Form and Function” in New Perspectives in Archaeology. Edited by Sally and Lewis Binford. Chicago: Aldine Publishing Company, 1968, 103-142. ———. Broken K Pueblo: Prehistoric Social Organization In the American Southwest. Arizona: University of Arizona Press, 1970. Hodder, Ian. Reading the Past. Cambridge: Cambridge University Press, 1991. ———. Theory and Practice in Archaeology. Cambridge: Routledge Press, 1992. Kelley, Jane and Marsha Hanen. Archaeology and the methodology of science. Albuquerque: University of New Mexico Press, 1988. Kitcher, Philip. Science, Truth, and Democracy. New York: Oxford University Press. 2001. Kohler, Robert. Landscapes and Labscapes: Exploring the Lab-Field Border in Biology. Chicago and London: The University of Chicago Press, 2002. Kosso, Peter. Knowing the Past: Philosophical Issues of History and Archaeology. New York: Humanity Books. 2001. Kuhn, Thomas. The Structure of Scientific Explanations. Chicago: The University of Chicago Press, 1970. ———. The Road Since Structure. Edited by James Conant and John Haugeland. Chicago: University of Chicago Press, 2000. Kyburg and Smokler. Studies in Subjective Probability. Huntington, New York: Robert E. Krieger Publishing Company, (1980). Lacey, Hugh. Is Science Value Free? Values and Scientific Understanding. New York: Routledge. 1999. Laudan, Larry. Science and Values: The Aims of Science and Their Role in Scientific Debate. Berkeley: University of California Press. 1984. Bibliography 137 Lakatos, Imre, and Paul Feyerabed. For and Against Method. Chicago: University of Chicago Press, 1999. Leblanc, Charles (ed). “Two Points of Logic Concerning Data, Hypotheses, General Laws, and Systems.” In Research and Theory in Current Archaeology. New York: John Wiley & Sons, 1973, 199-214. Lehmann, Gunnar. E-mail conversation with author, 8 December, 2000. Leone, Mark ed. Contemporary Archaeology: A Guide to Theory and Contributions. Carbondale and Edwardsville: Southern Illinois University Press, (1972). Lightfoot, Kent G. “Culture Contact Studies: Redefining the Relationship between Prehistoric and Historical Archaeology.” American Antiquity 60:2 (1995): 199-217. Lischka, Joseph. “Broken K Revisited: A Short Discussion of Factor Analysis.” American Antiquity 40:2 (1975): 220-227. Longacre, William. Archaeology as Anthropology. Tucson: The University of Arizona Press, 1970. Longino, Helen. The Fate of Knowledge. Princeton: Princeton University Press. 2002. Lyman, R., M. O’Brien, and R. Dunnell. The Rise and Fall of Culture History. New York: Plenum Press, 1997. MacDonald, Starkey, and Harding. Beth Pelet II: Prehistoric Fara and BethPelet Cemetery. London: British School of Archaeology in Egypt, 1932. Maher, Edward. Field lecture on the uses and limits of faunal data for the purposes of archaeological reconstruction. Tell el-Far’ah, South, Israel. July 1999. Mandelbaum. “Societal Facts.” In Theories of History. Edited by Patrick Gardiner. New York: The Free Press, 1959, 476-487. Martin, Paul. “The Revolution in Archaeology.” In Contemporary Archaeology: A Guide to Theory and Contributions. Edited by Leone, Mark. London: Southern Illinois University Press, 1972, 5-13. First published in American Antiquity 36:1 (1971): 1-8 (page citations are to the reprint edition). Mayo, Deborah. “Novel Evidence and Severe Tests.” Philosophy of Science 58 (1991): 523-552. ———. Error and the Growth of Experimental Knowledge. Chicago: University of Chicago Press, 1996. Maxwell, Grover. “The Ontological Status of Theoretical Entities.” In Introductory Readings in the Philosophy of Science. Edited by E.D. Klemke, Robert Hollinger, and David Rudge with David Kline. Amherst, New York: Prometheus Books, 1998, 363-373. First published in Minnesota Studies in the Philosophy of Science, volume 3 (Minneapolis: University of Minnesota Press, 1962), 3-14 (page citations are to the reprint edition). Meehan. Explanation in Social Science-A Systems Paradigm. Homewood: Dorsey Press, 1968. Mithen, Steven. The Prehistory of the Mind: A Search for the Origins of Art, Religion, and Science. London: Thames and Hudson, Ltd., 1996. 138 Bibliography ———. “Evolution of the Imaginative Mind.” Paper delivered at Imagining Minds: A Bradshaw Conference. Claremont Graduate University. Claremont, California, February 8, 2003. Morriss, Charles. “The Unity of Science Movement in the United States.” In Science and Philosophy in the Twentieth Century-Volume 2: Logical Empiricism at its Peak. Edited by Sahotra Sarkar. New York: Garland Publishing, Inc (1996): 303-308. First published in Synthese (1938) (page citations are to the reprint edition). Munson, Patrick. “Comments on Binford’s ‘Smudge Pits and Hide Smoking: The Use of Analogy in Archaeological Reasoning.’” American Antiquity 34:1 (1969): 83-85. Neurath, Otto. “Protocol Sentences.” In Logical Positivism. Edited by A. J. Ayer. New York: The Free Press, 1959, 199-208. First published in Erkenntnis volume 3 (1932-1933) (page citations are to the reprint edition). Oakes, Michael. Statistical Inference: A Commentary for the Social and Behavioral Sciences. Chichester: John Wiley & Sons, 1986. O’Brien, Michael, R. Lee Lyman, and Michael Schiffer. Archaeology as a Process: Processualism and Its Progeny. University of Utah Press. 2005. Ogden, D. Kelly and Jeffrey Chadwick. The Holy Land: A Geographical, Historical, and Archaeological Guide to the Land of the Bible. Jerusalem: HaMakor. 1990. Petrie, Sir William Flinders. Methods & Aims in Archaeology. New York: Benjamin Blom, Inc., 1904 (latest reprint 1972). ———. Beth Pelet I (Tell Fara). London: British School of Archaeology in Egypt, 1930. Pinsky, Valerie, and Alison Wylie, ed. Critical Traditions in Contemporary Archaeology: Essays in the Philosophy, History and Socio-politics of Archaeology. Albuquerque: University of New Mexico Press, 1995. Plog , Fred. “Diachronic Anthropology.” In Research and Theory in Current Archaeology. Edited by Charles Redman. New York: John Wiley & Sons, 1973, 181-198. ———. The Study of Prehistoric Change. New York: Academic Press, 1974. ———. “The Keresan Bridge: an Ecological and Archaeological Account.” In Social Archaeology: Beyond Subsistence and Dating. New York: Academic Press, 1978, 349-372. Plog, Stephen ed. Spatial Organization and Exchange: Archaeological Survey on Northern Black Mesa. Carbondale and Edwardsville: Southern Illinois University Press,1986. Popper, Karl. “Science: Conjectures and Refutations.” In Philosophy of Science: Contemporary Readings. Edited by Yuri Balashov and Alex Rosenberg. London and New York: Routledge, 2002, 294-301. A lecture given at Peterhouse, Cambridge (1953). First published as “Philosophy of Science, a Personal Report.” In British Philosophy in Mid-Century. Edited by C. A. Mace (1957) (page citations are to the reprint edition). Quine, W. “Two Dogmas of Empiricism.” Philosophical Review 60 (1951): 2043. Bibliography 139 ———. “Epistemology Naturalized.” In Naturalizing Epistemology. Cambridge: The MIT Press, 1997. Rathje. “Praise the Gods and Pass the Metates: A Hypothesis of the Development of Lowland Rainforest Civilizations in Mesoamerica.” In Contemporary Archaeology: A Guide to Theory and Contributions. Edited by Mark Leone. London: Southern Illinois University Press, 1972, 365-392. Renfrew, Colin and Bahn. Archaeology: Theories, Methods, and Practice. London: Thames and Hudson, Ltd, 1991. Reisch, George. “Planning Science: Otto Neurath and the International Encyclopedia of Unified Science.” In Science and Philosophy in the Twentieth Century-Volume 6: The Legacy of the Vienna Circle. Edited by Sahotra Sarkar. New York: Garland Publishing, Inc (1996): 131-154. First published in BJHS 27 (1994): 153-175 (page citations are to the reprint edition). Rouse, Irving. “Analytic, Synthetic, and Comparative Archaeology.” In Research and Theory in Current Archaeology. Edited by Charles Redman. New York: John Wiley and Sons, Inc., 1973, 21-31. Salmon, Merrilee. Explanation and Archaeology. New York: Academic Press, 1982. ———. “On the Possibility of Lawful Explanation in Archaeology.” In Readings in the Philosophy of Social Science. Edited by Martin and McIntyre. Cambridge: MIT Press, 1994, 733-746. First published in Critica. 22 (1990): 87-112 (page citations are to the reprint edition). ———. “Explanation in Archaeology.” In Explanation: Theoretical Approaches and Applications. Edited by G. Hon and S.S. Rakover. The Netherlands: Kluwer Academic Publishers, 2001, 231-248. Salmon, Wesley. “Confirmation and Relevance.” In The Concept of Evidence. Edited By Peter Achinstein. Oxford: Oxford University Press (1983): 95123. First published in Induction, Probability and Confirmation, Minnesota Studies in the Philosophy of Science, volume 6 (Minneapolis: University of Minnesota Press, 1975): 5-24, 27-36 (page citations are to the reprint edition). ———. “Scientific Explanation: Causation and Unification.” In Philosophy of Science: Contemporary Readings. Edited by Yuri Balashov and Alex Rosenberg. London and New York: Routledge, 2002, 92-105. First published in Critica: Revisita Hispanoamericana de Filosofia 22: 66 (1990): 3-21 (page citations are to the reprint edition). ———. Causality and Explanation. New York: Oxford University Press, 1998. ———. “The Spirit of Logical Empiricism: Carl G. Hempel’s role in TwentiethCentury Philosophy of Science.” Philosophy of Science 66: 3 (1999): 333350. Salmon, Wesley, R. Jeffrey, and J. Greeno. Statistical Explanation & Statistical Relevance. University of Pittsburgh Press, 1971. Schiffer, Michael ed. Advances in Archaeological Method and Theory: Selections for Students from Volumes 1 through 4. New York: Academic Press, 1982. 140 Bibliography Schlick, Moritz. “The Foundation of Knowledge.” In Logical Positivism. Edited by A. J. Ayer. New York: The Free Press, 1959, 209-227. First published as “Über das Fundament der Erkenntnis” in Erkenntnis volume 4 (1934). ———. “The Turning Point in Philosophy.” In Logical Positivism. Edited by A. J. Ayer. New York: The Free Press, 1959, 53-59. First published as “Der Wende Der Philosophie” in Erkenntnis volume 1 (1930-1931) (page citations are to the reprint edition). Schmidt, Peter and Thomas Patterson, ed. Making Alternative Histories: The Practice of Archaeology and History in Non-Western Settings. Santa Fe: School of American Research Press, 1995. Schnapp, Alain. The Discovery of the Past. Translated by Ian Kinnes and Gillian Varndell. New York: Harry N. Abrams, Inc., 1997. Schneider, Tammi and James Bentley. “Statistics and Archaeology in Israel.” In Computational Statistics and Data Analysis 32 (2000): 465-483 Scriven, Michael. “Explanations Predictions and Laws. In Introduction to the Philosophy of Science. Edited by Arthur Zucker. Upper Saddle River: Prentice Hall (1996): 95-103. First published in Minnesota Studies in the Philosophy of Science, volume 3 (Minneapolis: University of Minnesota Press, 1962), 170-230 (page citations are to the reprint edition). Semmelweis, Ignaz. The Etiology, Concept and Prophylaxis of Childbed Fever, trans. K Codell Carter. Reprinted in Medicine: A Treasure of Art and Literature. Wisonsin: the University of Wisconsin Press, 1983. Shanks, Michael and Christopher Tilley. Re-Constructing Archaeology. Great Britain: Cambridge University Press, 1987. ———. Social Theory and Archaeology. Albuquerque: University of New Mexico Press, 1987b. Solomon, Miriam. Social Empiricism. Cambridge: MIT Press. 2001. Snead, James. Ruins and Rivals: The Making of Southwest Archaeology. Tucson: The University of Arizona Press, 2001. Steel, Daniel. “Bayesian Statistics in Radiocarbon Dating.” Philosophy of Science: Supplement to Volume 68, Number 3, (2001): S153-S164. Steward, Julina. “The Direct Historical Approach to Archaeology.” American Antiquity vii:4 (1942): 337-343. Stich, Stephen. “A Pragmatic Account of Cognitive Evolution.” In Naturalizing Epistemology. Cambridge: The MIT Press, 1997. Strevens, Michael. “The Bayesian Treatment of Auxiliary Hypotheses.” The British Journal for the Philosophy of Science 52 (2001): 515-537. Strong, Wm Duncan. “Historical Approach in Anthropology.” Anthropology Today. Chicago: The University of Chicago Press (1953): 386-397. Taylor, Walter. “A Study of Archeology.” American Anthropologist: American Anthropological Association 50:3, part 2 (1948). Taylor, Walter. “Old Wine and New Skins: a Contemporary Parable.” In Contemporary Archaeology: A Guide to Theory and Contributions. Edited by Mark Leone. Carbondale: Southern Illinois University Press, (1972) 28-33. Trigger, Bruce. Artifacts & Ideas: Essays in Archaeology. New Brunswick: Transaction Publishers, 2003. Bibliography 141 Tuggle, D. A. Townsend, and T. Riley. “Laws, Systems, and Research Designs.” American Antiquity 37 (1972): 3-12. Upham, Steadman. “The Tyranny of Ethnographic Analogy in Southwestern Archaeology.” In Coasts, Plains and Deserts: Essays in Honor of Reynold J. Ruppé. Arizona State University Anthropological Research Papers No 38 (1987). van Fraassen, Bas. “Constructive Empiricism.” In Readings in the Philosophy of Science: From Positivism to Postmodernism. Edited by Theodore Schick, Jr. Mountain View: Mayfield Publishing Company, 2000, 267-272. First published in The Scientific Image. (Oxford University Press, 1980) (page citations are to the reprint edition). ———. Laws and Symmetry. Oxford: Oxford University Press, 1989. Watson, Patty Jo, Steven LeBlanc, and Charles Redman. Explanation in Archaeology: An Explicitly Scientific Approach. New York: Columbia University Press, 1971. Willey, Gordon and Jeremy Sabloff. A History of American Archaeology. Great Britain: Thames and Hudson, 1974. Willey, Gordon and Philip Phillips. Method and Theory in American Archaeology. Chicago: Phoenix Books, 1958. “Wissenschaftliche Weltauffassung: Der Wiener Kreis” [The Scientific Conception of the World: The Vienna Circle]. In Science and Philosophy in the Twentieth Century 1: The Emergence of Logical Positivism. Edited by Sahotra Sarkar. Garland Publishing, Inc.: New York & London (1996): 321341. First published as a pamphlet (1929) (page citations are to the reprint edition). Wissler, Clark. “The New Archaeology.” Natural History 17:2 (1917): 100-101. Wylie, Alison. “A proliferation of new archaeologies.” In Archaeological theory: who sets the agenda?. Edited by Yoffee and Sherratt. Great Britain: Cambridge University Press, 1993, 20-26. ———. “Reasoning About Ourselves: Feminist Methodology in the Social Sciences.” In Readings in the Philosophy of Social Science. Edited by Martin and McIntyre. Cambridge: MIT Press, 1994, 611-624. First published in Women and Reason. Edited by E. Harvey and K. Okruhik. (The University of Michigan Press, 1992). ———. “The Reaction against Analogy.” In Thinking From Things: Essays in the Philosophy of Archaeology. Berkeley: University of California Press, 2002, 136-153. First published in Advances of Archaeological Method and Theory 8 (1985): 63-111 (page citations are to the reprint edition). ———. “The interpretive dilemma.” In Thinking From Things: Essays in the Philosophy of Archaeology. Berkeley: University of California Press, 2002, 117-126. First published in Pinsky, Valerie, and Alison Wylie, ed. (1995). Critical Traditions in Contemporary Archaeology: Essays in the Philosophy, History and Socio-politics of Archaeology. (Albuquerque: University of New Mexico Press, 1989) (page citations are to the reprint edition). 142 Bibliography ———. “Philosophy in Practice: Evidence Stabilizing Technologies in Archaeology.” Paper presented at the annual meeting of the Philosophy of Science Association, Milwaukee, WI, November 2002. Yoffee, Norman and Andrew Sherratt, ed. Archaeological theory: who sets the agenda? Great Britain: Cambridge University Press, 1993. Index analogical reasoning, 52, 68, 87, 88 analogy, ethnographic, 50, 51, 68, 133, 138, 141 antirealism, entity, 14, 103, 105, 107 epistemic, 105, 106, 107 ontological, 105, 106, 107 ontological perspective, 105 archaeology, culture history, 31, 32, 33, 34, 35, 36, 37, 38, 47, 56, 123 archaeology, landscape, 44, 45 archaeology, new, ix, x, 1, 2, 3, 4, 29, 31, 32, 33, 35, 39, 43, 45, 46, 47, 49, 55, 56, 62, 63, 64, 65, 66, 67, 68, 69, 74, 76, 77, 81, 82, 99, 101, 102, 121, 122, 123, 126, 127, 128, 130 new archaeologists, the, 1, 2, 32, 39, 40, 41, 43, 44, 45, 49, 62, 64, 65, 67, 69, 73, 81, 103, 104, 121, 122, 123, 124, 126, 129, 130 archaeology, post-processual, 22, 69, 97, 108 archaeology, processual, ix, x, 2, 31, 32, 37, 41, 42, 69, 102, 105, 118, 119 Binford, Lewis, viii, 1, 2, 4, 22, 31, 32, 39, 40, 41, 42, 43, 47, 49, 50, 51, 52, 53, 54, 55, 56, 57, 60, 61, 62, 63, 65, 70, 72, 73, 86, 88, 91, 97, 102, 106, 122, 124, 127, 129, 131, 133, 136, 138 Broken K Pueblo, viii, 56, 61, 88, 124, 136 Canaan, 45, 89, 114, 124, 125, 127, 128 Cartwright, Nancy, viii, 76, 84, 85, 87, 133 cognitive fluidity, 87 confirmation, 1, 7, 8, 16, 19, 54, 76, 80, 117 constructive empiricism, 95, 107 unobservable entities, 101, 104 unobserved entities, 101 dating, absolute, 111. See also radiocarbon dating. dating, relative, 35, 46, 111, 113 death’s head, cherub, willow and urn, 70 Deetz, James, 68, 69, 70, 72, 73, 74, 77, 134 dendrochronology, 2, 35 Dethlefsen, Edwin, 68, 69, 70, 72, 73, 74, 77, 134 discovery, 35, 46, 47, 51, 57, 58, 65, 73, 76, 80, 81, 86, 91, 116 equifinality, ix, 2, 28, 29, 54, 55, 66, 84 ethnocentricity in science, 86 ethnography, 46, 47, 48, 49, 50, 51, 52, 58, 60, 61, 64, 65, 73 explanation sketch, 23, 24, 25, 26, 28, 61, 62, 63, 64, 66, 68 explanation, causal, 77, 78, 79, 80, 83, 84, 85, 86 explanation, functional, 83 explanation, statistical models, 1, 16, 17, 18, 19, 42, 45, 49, 67, 78, 79, 81, 82, 83 feminist archaeologies, 97, 106, 141 Feyerabend, Paul, 29, 76, 93, 94, 97, 134 functionalism and functional explanation, 83, 84 Giere, Ronald, 76, 94, 95, 135 models as maps, 94, 95 great debate, the, 110, 119 144 Index Hacking, Ian, 96, 105, 109, 135 Hempel, Carl, vii, viii, 1, 2, 3, 4, 5, 10, 12, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 31, 38, 39, 40, 44, 47, 50, 51, 57, 63, 64, 65, 67, 68, 69, 74, 75, 76, 77, 78, 81, 83, 84, 85, 91, 93, 99, 102, 103, 104, 105, 116, 119, 122, 123, 127, 129, 130, 135, 136, 139 hermeneutics, 91 Hill, James, viii, 47, 49, 50, 53, 56, 57, 58, 59, 60, 61, 65, 68, 73, 88, 136 Hodder, Ian, 69, 91, 97, 108, 131, 136 logical positivism, 7, 28, 40, 73, 75, 122 justification, 47, 50, 51, 53, 54, 57, 58, 65, 83, 91, 116 black boxes, 116 testing phase, viii, 3, 7, 8, 37, 39, 46, 47, 50, 51, 53, 54, 55, 56, 57, 58, 60, 61, 63, 65, 73, 77, 81, 82, 86, 95, 102, 116 radiocarbon dating, viii, 2, 17, 29, 35, 36, 46, 83, 96, 114 realism, entity, x, 14, 95, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 119, 129 Reed, Jonathan, 109 regional survey, viii, 19, 27, 35, 37, 44, 45, 48, 49, 56, 73, 111, 113, 114, 123, 124, 125, 126, 128, 130, 146 Kosso, Peter, 76, 97, 104, 131, 136 Kuhn, Thomas, viii, 36, 43, 76, 91, 92, 93, 97, 117, 136 paradigmatic science, 16, 92, 93, 123 pre-paradigmatic science, 92 Lacey, Hugh, 115, 136 Lakatos, Imre, 117, 137 laws, ceteris paribus, 44, 84, 85, 95 laws, covering, viii, 5, 14, 15, 16, 17, 18, 20, 21, 22, 25, 26, 27, 28, 29, 32, 39, 43, 44, 45, 50, 62, 63, 64, 67, 69, 73, 74, 75, 76, 77, 78, 84, 85, 86, 91, 92, 99, 102, 103, 116, 127, 129 laws, general, 1, 2, 14, 15, 16, 17, 20, 21, 22, 23, 26, 43, 44, 45, 62, 63, 66, 69, 73, 76, 77, 85, 91, 102, 123, 126 Lehmann, Gunnar, 128, 137 marxist archaeologies, 97, 106 middle range theory, 50, 91 Mithen, Steven, 86, 87, 88, 97, 131, 137 Petrie, Sir W.F., viii, 35, 45, 126, 138 Philistines, vii, 89, 90, 113, 128 philistine pottery, 89 philosophy of archaeology, 99 philosophy of archaeology, the, xi, 1, 2, 129, 130 pottery seriation, viii, 35, 37 Salmon, Merrilee, 74, 76, 77, 78, 81, 85, 128, 131 Salmon, Wesley, 13, 76, 78, 79, 85, 86, 96, 135, 139 Schneider, Tammi, xi, 82, 119, 128, 140 science and values, x, 99, 110, 119 non-cognitive values, 115 science, reduction of, 9, 10, 11, 12, 13, 26, 40, 122, 130 science, unity of, vi, x, 3, 6, 9, 11, 13, 26, 29, 40, 92, 117, 121, 122, 124, 126, 129, 130, 131 Scriven, Michael, viii, 75, 76, 91, 140 Semmelweis, Ignaz, 102, 103, 140 site excavation, 123 smudge pits, 44, 49, 52, 53, 54, 72 Index statistical-relevance model, the, 78, 79, 80, 85, 86 stratigraphic method, the, viii, 35, 36, 37, 48, 110, 113, 126, 130 Syro-Palestine. See Canaan Taylor, Walter, viii, 32, 92, 140 texts, 64, 65, 66, 68, 73, 77, 109, 115, 118, 123 as archaeological data, 35, 64, 65, 110, 112, 113, 115, 118 theoretical entities. See realism, entity Trigger, Bruce, 106, 108, 131, 140 145 under-determinism. See equifinality van Fraassen, Bas, viii, 76, 94, 95, 107, 108, 134, 141 verification, 6, 7, 8 Vienna Circle, the, 5, 11, 141 Wheeler-Kenyon method. See stratigraphic method, the Wylie, Alison, vii, xi, 22, 46, 68, 76, 88, 97, 106, 108, 116, 117, 119, 131, 135, 138, 141 values on the receiving end, 116, 118 values on the shipping end, 116 About the Author William H. Krieger serves as a field director for Tell el-Far’ah South excavations in Southern Israel, and is an assistant professor of philosophy at the University of Rhode Island. He has been working in both archaeology and the philosophy of science (separately and together) since 1993. Dr. Krieger has conducted research and has publications and conference papers that span the fields of philosophy, religion, and archaeology. He currently serves as the regional secretary for the Western Commission for the Study of Religion (WECSOR), as well as the vice president for the pacific southwestern region of the American Schools of Oriental Research (ASOR). William holds a Ph.D. in philosophy from the Claremont Graduate University, a B.S. in biology from Columbia University, and a B.A. in Jewish Philosophy from the Jewish Theological Seminary of America. When he is not in the classroom or doing research (during the year) or on a digsite (during the summer), Dr. Krieger spends his time skiing as a member of the Mt Baldy Ski Patrol, scuba diving wherever the water is warm and clear, riding his motorcycle to points unknown, and taking photographs of anything that will stay in frame. Can There Be a Philosophy of Archaeology? provides a historical and philosophical analysis of the rise and fall of the philosophical movement know as logical positivism, focusing on the effect of that movement on the budding science of archaeology. Significant problems resulted from the grafting of logical positivism onto processual, or new archaeology, and as a result of this failure, archaeologists distanced themselves from philosophers of science, believing that archaeology would be best served by a return to the dirt. By means of a thorough analysis of the real reasons for failures of logical empiricism and the new archaeology, as well as a series of archaeological case studies, Krieger shows the need for the resumption of dialogue and collaboration between the two groups. In an age where philosophers of science are just beginning to look beyond the standard examples of scientific practice, this book demonstrates that archaeological science can hold its own with other sciences and will be of interest to archaeologists and philosophers of science alike. "Krieger provides a highly accessible account of the active, sometimes fractious interchange between archaeology and philosophy, detailing the fortunes of positivist ideals, the New Archaeology that embraced them, and the post-positivist models of research practice that have taken shape in archaeology and in philosophy of science in the last 30 years.... Archaeologists and philosophers both stand to benefit from the intellectual partnership that Krieger details, particularly when the scope of the archaeological discussion is expanded to include the rich and diverse research traditions he considers." —Alison Wylie, University of Washington William H. Krieger serves as a field director for Tell el Far'ah South excavation in Southern Israel, and is an assistant professor of philosophy at the University of Rhode Island.