Thursday, February 11, 2016

Science, social science, and archaeology: Where do we stand?

This is the first of a planned three posts on my view of a “scientific archaeology,” within the domain of the social sciences. The second will be, “Why is it important to strive for a more scientific archaeology,” and the third, “Why is a scientific archaeology so hard to achieve?” I have long been distressed that (1) some archaeologists dismiss science as a model for our academic discipline; and when they do that, (2) they often show a faulty or partial understanding of what kind of science might be appropriate for archaeology. My short answer is that “social science” is an appropriate model. My views will target the archaeology of complex societies, both because that is my domain and I know it well, and because most of the anti-science rhetoric and practice seem to be in this domain. My critiques of archaeological practice probably have less relevance to the study of hunter-gatherers; in that field, archaeologists do science and they don’t go around wringing their hands about whether that is an appropriate model or not.

What is science?

My reading of the literature of scientific methods and the philosophy of science, coupled with my experience in archaeology and transdisciplinary projects, lead me to the following definition of science:

  1. Science is a method to gather accurate knowledge about the natural and social world:
    1. It gives primacy to reason and observation.

  1. Science has a critical spirit:
    1. Constant testing of claims through observation and experiment;
    2. Findings are always tentative, incomplete, and open to challenge.

  1. Science is complex:
    1. It consists of an interconnected network of diverse evidence and theory;
    2. Its content and findings are judged by communities of scientists.

I developed this definition after close consultation with these and other sources: (Bunge 1999; Gerring 2012; Kosso 2009; Little 1998; Sokal 2006; Wylie 2000). For a similar succinct way of framing this kind definition, John Gerring (2012:11) defines science as follows:

“Inquiry of a scientific mature, I stipulate, aims to be cumulative, evidence-based (empirical), falsifiable, generalizing, nonsubjective, replicable, rigorous, skeptical, systematic, transparent, and grounded in rational argument. There are differences of opinion over whether, or to what extent, science lives up to these high ideals. Even so, these are the ideals to which natural and social scientists generally aspire, and they help to define the enterprise in a general way and to demarcate it from other realms.”

If this is at all strange to you, or if you want something to assign to your undergraduates, check out the chapter on epistemology (called “How do you know what you know?”) in Ken Feder’s excellent text, Frauds, Myths, and Mysteries: Science and Psuedoscience in Archaeology (Feder 2002).

Please note that there is nothing in these definitions about experiments or laws. Hempel is nowhere to be seen. This definition does not coincide with the way the New Archaeologists viewed science (Watson et al. 1971), something I will take up in my third post in this series. I think their faulty views of science and explanation caused great harm to archaeology. Nor does my approach coincide well with the various traits that Matthew Johnson (2010:38-41) includes in his discussions of “Definitions of Science” and “Positivism.” His discussion, quite similar to that of the New Archaeologists, resonates with an older literature in the philosophy of science, but it doesn’t have much relevance to the way I think about the question of a scientific approach to archaeology.

Please note also that this definition is about epistemology, not about methods narrowly defined. That is, science is a way of doing research that may or may not include the use of “scientific” or archaeometric techniques from chemistry, physics, or other disciplines. Conversely, “scientific” techniques can be employed in the pursuit of science as well as in the pursuit of non-scientific (and even anti-scientific) ends. I discuss this in a number of previous posts, where I draw a contrast between Science-1 (a scientific epistemology) and Science-2 (the use of “scientific” techniques): Try  here, and  here.

What is social science?

For most of my career, before moving to ASU in 2005, I had no idea that the social sciences had any relevance to archeology. Apart from the contentious issue of the relevance of cultural anthropology to archaeology (my views on this issue are here: Smith 2011b), the notion that sociology or political science or economics might be useful to me was foreign, not even on my radar. The notion that there might be a body of methodological and theoretical work that is extremely relevant to archaeology (far more so than just about anything in cultural anthropology) did not even cross my mind.

After moving to ASU, I realized that ancient and modern cities could be compared and analyzed in common frames of reference; I discovered transdisciplinary research; and I discovered an epistemological literature in the social sciences that fit rather precisely with my own views of how to pursue scholarship. I have blogged about these issues on and off for a number of years, in this blog ( here) and in Wide Urban World (here,  and here).

One way to highlight what is distinctive about the social sciences is to contrast them with the natural sciences on one hand, and the humanities on the other. The former are often said to focus on instrumental knowledge, and the latter on reflexive knowledge. As described by sociologist Michael Burawoy, this leaves the social sciences in the middle:

“The social sciences are at the crossroads of the humanities and the natural sciences since in their very definition they partake in both instrumental and reflexive knowledge. The balance between these two types of knowledge, however, varies among the social sciences” (Burawoy 2005:22).

Here is a chart, abstracted from a longer table from Jerone Kagan (2009:4-5):

 While this is obviously a greatly simplified scheme, it does indicate nicely the position of the social sciences between the natural sciences and the humanities. Half a century ago, C.P. Snow (1959) could describe scholarship as a choice between two cultures: the natural sciences and the humanities; today there are clearly three cultures of relevance (Kagan 2009). But archaeologists have been slow to get the news.

What are the major goals of social science research?

Here is a handy list of the major goals of the social sciences, from the very nice methods textbook by Charles Ragin and Lisa Amoroso  (Ragin and Amoroso 2011:35-56). As archaeologists, do we do these things?

  1. Identifying general patterns and relationships
  2. Testing and refining theories
  3. Making predictions.
  4. Interpreting culturally or historically significant phenomena
  5. Exploring diversity
  6. Giving voice
  7. Advancing new theories

My work mostly concerns points 1, 2, 4, and 5, but there is an element of all of these goals in what I do. I would guess that this scheme could be used to organize the nature of social research by archaeologists. Ragin and Amoroso use this scheme to organize their textbook. Check it out.

Social science ontology

Its about time that I quote from my favorite social scientist, my intellectual hero, Charles Tilly (NO, this is NOT the phenomenologist Christopher Tilley!). Tilly (2008:6-7) lists the following as the four major social science ontologies. I’m sure you can find your niche in this list:

  1. “Methodological individualism insists on decision-making human individuals as the basic or unique social reality.”  (Focus on persons, one at a time.)
  2. “Phenomenological individualism refers to the doctrine that individual consciousness is the primary or exclusive site of social life.” (No assumptions of rationality. Speaking of Christopher Tilley......)
  3. “Holism is the doctrine that social structures have their own self-sustaining logics. In its extreme form—once quite common in social science but not unfashionable—a whole civilization, society, or culture undergoes a life of its own.” (World-systems analysis, and studies of large-scale social institutions fit here.)
  4. “Relational realism, the doctrine that transactions, interactions, social ties, and conversations constitute the central stuff of social life, once predominated in social science.” (Marx, Weber, networks. This is Tilly’s preferred ontology, and I find it very attractive).

I think this list covers most of the terrain of archaeology. I can fit my views into this scheme, and it helps me make sense of why I find the work of some writers attractive and others less so.

Social science epistemology

Then Tilly (2008:8) gets down to the major social science epistemologies, or what he calls “logics of explanation”:

  1. Covering laws
  2. Specification of necessary and sufficient conditions
  3. Statistical regression accounts (one variable “accounts for” another)
  4. Locations of structures and processes within larger systems (functionalist)
  5. Stage models. Invariant growth sequences.
  6. Identification of individual or group dispositions just before a point of action.
  7. “Reduction of complex episodes, or certain features of those episodes, to their component mechanisms and processes”

The situation for archaeology is somewhat different, now. I can find my personal logics of explanation here. I’ve used #4 and #5 in the past, but now I favor #7 (although I haven’t really published a major mechanisms-based analysis yet). But much of archaeology today cannot be incorporated into this scheme. Where would materiality, actor-network theory, or structuration be accommodated?

These latter abstract, philosophical theoretical frameworks cannot be accommodated into the standard social science epistemologies because they pertain more to the humanities than to the social sciences. As evidence for this, consider the question of “how would you know when you are wrong?” This is a fundamental issue in the social sciences. I use this as the organizing principle for my paper on arguments in archaeology (Smith 2015). This question, which is basic to many domains of scholarship, derives from the second point of my definition of science at the top.

Perspectives like materiality or practice theory or the social production of space cannot be disproven. They are so abstract that they cannot be tested and confirmed or rejected (Smith 2011a, 2015). They are more appropriately considered as part of the humanities than as part of the social sciences. This does not make them useless or bad; it just means that they have little role to play in developing causal models of past societies, or in understanding the hows or whys of specific social trajectories of past societies, or in relating archaeological findings to work in other disciplines on the major social problems of today. If this does not sound right to you, I suggest reading some social-science epistemology. This is pretty basic stuff (although I must admit that I was ignorant of these ideas and sources until about eight years ago). I recommend a number of relevant sources in my two papers just cited. Or here are a few suggestions: (Abend 2008; Bunge 2004; Gerring 2012; Little 2011; Mjøset 2001, 2009).

So, why is it important to strive for a more scientific archaeology? See my next post, coming soon.


2008  The Meaning of "Theory". Sociological Theory 26: 173-199.

1999  Social Science under Debate: A Philosophical Perspective. University of Toronto Press, Toronto.

2005  For Public Sociology. American Sociological Review 70 (1): 4-28.

2002  Frauds, Myths, and Mysteries: Science and Psuedoscience in Archaeology. 4th ed. Mayfield, Mountain View, CA.

2012  Social Science Methodology: A Unified Framework. 2nd ed. Cambridge University Press, New York.

2010  Archaeological Theory: An Introduction. 2nd ed. Blackwell, Oxford.

2009  The Three Cultures: Natural Sciences, Social Sciences, and the Humanities in the 21st Century. Cambridge University Press, New York.

2009  The Large-Scale Structure of Scientific Method. Science and Education 18 (1): 33-42.

1998  Microfoundations, Method, and Causation: On the Philosophy of the Social Sciences. Transaction, New Brunswick.

2001  Theory: Conceptions in the Social Sciences. In International Encyclopedia of the Social and Behavioral Sciences, edited by Neil J. Smelser and Paul B. Baltes, pp. 15641-15647. Elsevier, New York.

2011  Constructing Social Research: The Unity and Diversity of Method. 2nd ed. Sage, Thousand Oaks, CA.

Smith, Michael E.2011    Why Anthropology is too Narrow an Intellectual Context for Archaeology.      Anthropologies 3: (online).

1959  The Two Cultures and the Scientific Revolution. Cambridge University Press, New York.

2006  Pseudoscience and Postmodernism: Antagonists or Fellow Travelers? In Archaeological Fantasies: How Pseudoarchaeology Mispresents the Past and Misleads the Public, edited by Garrett G. Fagan, pp. 286-361. Routledge, New York.

2008  Explaining Social Processes. Paradigm Publishers, Boulder, CO.

1971  Explanation in Archaeology: An Explicitly Scientific Approach. Colombia University Press, New York.

2000  Questions of Evidence, Legitimacy, and the (Dis)unity of Science. American Antiquity 65: 227-237.


Richard M. said...

Long time reader here, first time commenter. I appreciate the effort you put into defending scientific archaeology, especially in such an accessible medium as this blog. I also wanted to share a few of my own thoughts. My perspective may be considered derisive and ignorant by some, but I sincerely hope no one takes it personally.

I came into the field of archaeology/anthropology with a science background and naively assumed that all its practitioners had a social science epistemology like yours, Dr. Smith. In undergrad, when I first encountered various postmodernist ideas, I found them to be incomprehensible and therefore easily dismissed them. I hadn’t yet formulated my own view nearly as explicitly as you have here, however, and my fellow students and I would have been well served by an emphatic and rational defense of scientific archaeology such as this. I’m a few years out with my master’s working in CRM now, but when I attend conferences or look at recent publications I’m still baffled by these debates in academia. I find especially interesting your perspective that it’s anthropology/archaeology specifically, uniquely among the social sciences, that is still struggling with science and these debates. I think I can partly see why.

When reflecting upon my own education, I feel as though had I not started out with a strong scientific worldview, I might have fallen down a “slippery slope” to less rigorous thinking due to many factors, some of which are unique to the field. I think anthropological pedagogy should be examined for its role in creating this divergence. Here’s my own, highly simplistic multi-part hypothesis for why some archaeologists might come to find postmodernist ontology and epistemology so alluring:

1) When I was introduced to the anthropological axiom of cultural relativism, I understood it as part and parcel of a scientific approach in order to remain objective and non-ethnocentric while studying cultures outside my own, but also recognized it as an ideal that is almost impossible to completely achieve that serves to make us continually aware of our own bias. I believe some of my fellow students saw it differently; they internalized it as a literal challenge, as an ethical imperative, and as an anthropological endorsement of relativism more generally. Perhaps because more is better, some then make a profound ontological leap and abandon all notion of objective truth, believing reality is merely composed of subjective experiences. When absolute relativism is coupled with a strong sense of social justice and a concern with dismantling hegemony, scholarship must be about creating “safe spaces” for others to share their “own truth.” To this peculiar postmodern world of solely ideas, which denies an objective reality by which to empirically evaluate them, scientific criticism and debate are antithetical. Perhaps this is appealing because it allows one to deal in solely emotional, factless rhetoric, which I argue comes easily, especially when you feel as though you are defending some noble ethic of fairness implicit in certain versions of moral relativism. This brings me to my second point, and I'm hitting the comment length limit so I'm posting it as a separate comment below:

Richard M. said...

2)The alternative, science, is hard. You must present ideas that are open to scrutiny and may later be proven wrong. Furthermore, archaeology is especially difficult, because what you’re trying to study are past societies and your data are pits and potsherds, a fragmentary archaeological record over which you have very little control. I remember being dismayed at simultaneously studying big ideas like changing human societies in one course, and detailed methods like ceramic typologies and seriation in another, and being at a loss for connecting the two. I recall being briefly inspired when learning about processualism in archaeological history and Binford’s definition of middle range theory (which I know you object to, Dr. Smith), thinking, aha, now, with a better understanding of formation processes, surely someone will help explain to me how to work backwards from pits and pots and settlement patterns to institutions and social complexity and ritual. I was still thinking that someone could lay it all out for me programmatically, e.g. “Here: this is what moieties,” for instance, “definitively look like in the archaeological record; when you observe X then you can infer Y as established by middle range theory Z.” Turns out, making defensible inferences is harder than that, something that I continually struggle with and hope to get better at. But I think the key is to continually look to the data for patterns, rather than counting on the next fad theory to suddenly satisfactorily explain all results.

3)Academia is high pressure. Publish or die. Sometimes, it’s simply easier to take some high-level abstract theory, “apply it” to some data, a viola! You’re “creating meaning.” And it can't be wrong, because all of our data are subjective and based on the theoretical lens through which we view it, right? Also, by tying yourself to some specific scholar’s high-minded but subjective and untestable theory, you have automatically joined a community of supporters who may reference your work. Win-win, right?

Michael E. Smith said...

@Richard - I also came into archaeology from a math/science background, and it took me a long time to recognize that science and a scientific approach might need defending in archaeology. But many of those who defended science vocally and publicly (Binford and the new archaeologists, for example), had a simplistic and limited view of science which was then easy for the postmodernists to dismiss or make fun of. Biology would have been a much better model for archaeology than was physics and chemistry. I agree with your analyses. This is a complicated issue, but one that needs more discussion. When a prominent archaeologist can publish the idea, in 2013, that the theoretical alternatives for archaeologists are limited to processualism and post-processualism, or when a prominent archaeologist can say that archaeology must be either a natural science or the humanities, this signals to me that our theoretical and epistemological situation is pretty bad.

Simon said...

I'm struggling to make sense of this post. The starting-point seems to be a rejection of an old logical-positivist vision of science - so general laws are out, and experiment is no longer crucial. But the result is a definition of science that is so broad it is hard to distinguish from any reasonable inquiry: science gives primacy to reason and observation; has a critical spirit, and involves a diverse interconnection of theory and evidence. For sure you might find some wild postmodernists who would define what they do differently, but I don't really see that this new and improved definition excludes any inquiry worth its salt, including anything that should pass muster in the humanities.

The table from Kagan 2009 suggests that what makes the humanities distinct is a focus on meaning. Maybe, but you seem to be encroaching on semantics with major goal of social science number 4: 'Interpreting culturally or historically significant phenomena'.

Possibly such interpretation is to be non-semantic, which would mean (I assume) carried out purely on the level of quantifiable correlations. But I question if an archaeology so restricted can really produce the goods.

The most established social science is economics, which proceeds by way of a method of abstraction. Specifically (and as first formulated by J.S. Mill and Alfred Marshall) economics places a large ceteris paribus clause around all factors that cannot be reduced to a reductive system of measurable relationships. The idea is that this simplified model still tells us something about the world we live in. Maybe this makes sense, in economics (maybe).

But to apply such a reductive prison to the distant and long period past would seem to remove from the picture by assumption a very substantial part of what needs to be examined and explained. Put crudely, you are surely going to end up with economic explanations of such human revolutions as the origins of agriculture or the advent of urban life, and this because all the non-economic factors are identified as semantic and dismissed as the proper subject of the 'humanities'.

I am aware of my presumption in writing all this, and I suspect that most of my objections can be dismissed with some fairly standard arguments. But for what it is worth, the idea of a scientific archaeology set out in the post seems to me to simply bypass the problem of subjectivity (dropping it into a litter bin labelled 'humanities' and 'postmodernism') rather than grappling with it.

Michael E. Smith said...

@Simon - Let me think about this a bit. I am not really an expert thinker in this area. I do write about it, though, because it seems that much of archaeology is really screwed up about theory and epistemology, and someone has to talk about these things from a scientific viewpoint. I'm not sure I have it on the best definition of science. Also, I don't want to just dump on the humanities as having bad explanations or being irrelevant, and I will go into that further in my next posts. There is very rigorous scholarhip in the humanities, but it differs from science, and I will try to put my finger on this. I'll try to reply again later when I've thought this through more fully.