Jump to content

- - - - -

John Dupré: The Disunity of Science

John Dupré is a professor of philosophy of science in the Department of Sociology and Philosophy at Exeter University in the UK, and also the director of Egenis, the ESRC Centre for Genomics in Society. I was able to ask him about several keys areas of his work and relate it to contemporary issues in both science and the philosophy of science.

- Interviewed by Paul Newall (2006)

PN: You have written that "mathematicism", or the extension of mathematics to areas of science where it can play no role, is a form of scientism and "a sociologically significant contributor to scientific prestige". Can you explain why prestige is important given that science is often thought of as an objective endeavour?

JD: Prestige is important in science for the very simple reason that resources for science are limited. As Philip Kitcher eloquently explains in his recent book, Science, Truth, and Democracy, there is an inexhaustible reservoir of truth, and what science should aim to deliver is significant truth. Particular sets of scientific concepts, methods, instruments, and so on, give us a particular path into the endless thicket of potential knowledge, and which of these we invest in determines which more or less significant truths we may discover.

For reasons that a number of philosophers of science including myself have emphasised in recent years the highest status mathematical work in science is not addressed to discovering truth at all, but at best to developing tools and methods. Even the "fundamental" mathematical laws in physics are not true generalisations about the world but abstract formulae that may or may not prove useful in specific applications carefully tailored to address real problems. (Nancy Cartwright's work over many years has been very important in making this point.)

The problem with the field in which mathematicism has been most dominant, economics, is that the fetishisation of this tool construction activity has often led to a lack of interest in actually attempting to apply these tools to real problems. This has begun to be partly addressed by the project of experimental economics, which has investigated the basic psychological assumptions underlying economists' mathematical theories, and has found that they are often largely false. This raises serious doubts as to whether the mathematical endeavours that have been at the centre of high prestige economics have any real value at all.

The basic position underlying my hostility to mathematicism is empiricism. The success of science has always depended on its connection to empirical reality. Of course, one must be a sophisticated empiricist nowadays. The world does not simply speak to the scientist, the scientist must develop concepts and tools to interpret the world. But if these are not constantly evaluated and refined in interaction with empirical testing, they have increasingly little value. My objection to mathematicism, then, is that it constitutes a divorce of scientific work from empirical reality. It is not, of course, a claim that mathematics does not have an essential role in science; it is a warning against mathematics becoming an end in itself.

Sophisticated empiricism, for me, is ultimately the recognition that pure reason has never told us anything about reality. The challenge is to find ways of applying reason to developing ways of interacting with the world that can give us insight into reality. But experience of the world is always the final arbiter.

PN: What other sociological factors play a role in science?

JD: Science is a social process, and societies are extraordinarily complex things. So I certainly won't offer an exhaustive list of social factors that might affect the development of science. I'll just mention the factor that has been central to my work from the beginning, the development of scientific conceptual schemes. The sciences that I have been concerned with, the biological sciences and the social sciences, aim to tell us about things in which we have had a profound interest that long predates modern science. We talk about ourselves and the living world in a language in which fact and value are inextricably intertwined. I tried to develop this point in my earliest work in which I traced the various interest-relative ways in which people classify biological organisms. Even strictly scientific interests in biology are diverse, for example evolutionary versus ecological, and can lead to different classificatory practices.

The problem that is not, in my opinion, sufficiently appreciated is that scientific findings cannot be purged of this interest relativity. Many scientific claims are presented directly in terms of familiar everyday language, imagining that the applications of these concepts can be investigated in a purely objective, interest-free manner. This is one of my main objections to evolutionary psychology, a "scientific" project that I've spent a good deal of time criticising. The idea that one can illuminate a human problem such as rape by observing the behaviour of ducks or flies is an extreme example of this insensitivity to the subtlety of human language. But of course the reason that these investigations interest people, in fact are a great source of best-selling books, is that people want to learn about rape (or violence, altruism, mate choice, and so on), so this simplistic kind of argument is largely unavoidable. We can formulate scientific findings in esoteric scientific languages that may reflect only scientific, not social, interests. But then if anyone is to care we need to translate the findings somehow into the language we understand.

It is sometimes presented as an objection to this line of thought that physics and chemistry, at least, are sciences that use entirely objective languages generated solely by scientific experience. So shouldn't the biological and social sciences aspire to the same objectivity? The point that this objection misses, I think, is that the objectivity of the physical sciences is based on the fact that the matters they discuss are not ones that anyone else has any interest in, apart from purely epistemic interest. Electrons, say, were developed as a scientific concept in an entirely esoteric scientific context. Of course we care about some of the applications of the physical sciences. But the idea that physics might be free of social values hardly suggests that there are value free atomic weapons or nuclear power stations. And as explained in my answer to the first question, what we find out about physics or chemistry may certainly reflect the interests to which we hope to apply our findings, as is often claimed by those who point with dismay to the extent to which the physical sciences have been funded by the military.

PN: How can science benefit from sociological study?

JD: This is a very large question of which I can only scratch the surface. As I just mentioned, whatever else science may be, it is a social process. Understanding the way this social process works is surely one fundamentally important approach to understanding science. Though I am not a sociologist of science, my current job is as director of a Research Centre mainly concerned with social aspects of contemporary biology, specifically genomics. There are a host of important questions about how the production of scientific knowledge interacts with the rest of society that require sociological investigation. As should be clear from my previous remarks, this is not just a matter of the reception of science by particular interested groups and publics, but also a matter of how these groups affect the science that is produced. This is surely something scientists as much as anyone else would benefit by understanding better.

Related to my remarks about the importance of analysing concepts, I am very interested in the transmission of ideas between different groups of people. A term such as "gene" is understood in very different ways by different groups of scientists, and as scientific findings disseminate to doctors, lawyers, genetic counsellors, and the public at large, different understandings again emerge. This semantic drift has been a major focus of study at Egenis.

PN: In your work The Disorder of Things you claimed that there are problems with falsificationism. What are they and what do you make of the reliance on it as a demarcation criterion by many in the debates over creationism and ID?

JD: One problem with falsificationism is that it is a relic of a philosophy of science that held science to be fully intelligible as a set of propositions (laws, theories, etc). As I have said already, I take a particular scientific project to involve a number of elements: methods, concepts, background assumptions, instruments, and so on. (This is the most important thing that philosophy of science has learned from Thomas Kuhn, though no doubt he is not the only person who has tried to make the point.) It is these diverse elements of a scientific field that make the falsification of a prediction impossible to interpret in any conclusive way - our background assumptions may have been partially mistaken, our instruments may not have worked in quite the way we expected, and so on. No one is going to abandon molecular genetics, say, because it makes a few false predictions, simply because it is currently an enormously productive and successful scientific endeavour.

Of course this leads me to conclude that falsification is not the right concept with which to confront creationists. But the general picture of a successful scientific project will serve very nicely. Creationism has, as far as I can tell, not a single scientific achievement to its credit. Its "instrument" is a book written by a large number of different individuals in the distant pre-scientific past. Its main arguments are generally highly confused claims about scientific problems that a very successful scientific programme has yet to solve - and of course, any science will have plenty of those. And, to be blunt, its central explanatory concept, God, is one without a shred of empirical evidence. As I argue in Darwin's Legacy, prior to the development of compelling evolutionary theories, our ignorance about the possible origins of life were such as to make an intelligent creator a hypothesis worth considering, though as David Hume brilliantly showed in his Dialogues Concerning Natural Religion, such a speculation was extremely tenuous and told us little we wanted to hear about the nature of God. As an empiricist I cannot see any conceivable role for the concept of God in science, at least pending His or Her decision to present Himself or Herself for empirical examination.

PN: Why do you think so many people, particularly scientists, have a "commitment to a universe amenable to one systematic and orderly description", and why do you not believe that such a universe exists?

JD: I might begin by saying that, at least among philosophers of science, I think that this commitment is in quite rapid decline. My main answer must, once again, be an appeal to empiricism: there is no evidence for such a systematic and orderly description. What we should certainly agree is that our description of the world should be coherent, non-contradictory. I suspect that underlying a lot of the insistence on a unitary account of the world is a vague sense that only such an account could meet the condition of being coherent. But I can see no credible grounds for this assumption.

There are a number of reasons why I do not believe a unified account of the universe is possible. I'll mention two. First, once again, is empiricism. While there are of course many points at which different parts of science interact and combine, I see no tendency towards unification. Unifying projects proposed by philosophers, such as the reduction of the mental to the neurophysiological, are not only exposed to powerful philosophical objections, but also are very diffusely reflected in scientific practice.

Second, there is a widespread assumption that causality flows only upwards from simple things to more complex things. This is of course a very metaphorical expression, and exactly what it means depends on an account of causality and what it is for it to "flow". Without trying here to sort out these notoriously difficult questions, it seems to me that this assumption expresses an ungrounded dogma. I cannot see why emergent properties of complex things should not cause the movements of their constituents, and it seems to me that this happens all the time. It is properties of me, including very complex relations to my social environment, that explain and, I would say, cause the movements of physical particles in my fingers as I type these words. Of course the arrangement of particles and molecules in my body are also necessary for the emergence of the powers displayed in my typing. This two way casual flow between levels of organisation is one of the things that makes biological processes so difficult to understand.

PN: What is reductionism and why do you object to it?

JD: Reductionism is the project of explaining the sciences of complex things, without remainder, by appeal to the laws that govern the behaviour of their constituents. I have already mentioned many of the reasons why I reject this position. Clearly it is incompatible with the downward causation I have just described.

Reductionism tends to come down in the end to the conviction that the laws of fundamental physics are universal in scope. If this were so, any laws claimed to apply to complex entities would have to be deducible in principle from these fundamental laws, or false. I can't address the details of the argument about fundamental physical laws. But one reason for being very sceptical about the conviction just mentioned is that it seems to be a hangover of the general view that science is ultimately directed at discovering universal laws, an idea that has proved to have little relevance to, for example, the biological sciences. A concept that has largely superseded this for most philosophers of science is that of a scientific model. Models provide partial representations of aspects of reality that provide more or less valuable, but always partial, insight into the complexities of the real world. Nancy Cartwright in particular has offered compelling arguments that so-called fundamental laws are higher levels of abstraction still that are useful in devising more concrete models. If this is right, the basic premise underlying reductionism is a non-starter.

PN: What do you consider to be the "limits of science"? Why is scientific imperialism (or "the tendency to push a good scientific idea far beyond the domain in which it was originally introduced, and often far beyond the domain in which it can provide much illumination") a problem?

JD: I certainly don't want to specify the limits of science a priori! My point, which follows from my remarks on reductionism, is only that scientific models are useful for a very specific set of questions. Evolutionary models, to take one example, have been enormously successful in helping us to understand the processes by which new forms of biological organisation emerged from earlier forms.

A saying of which I am rather fond is "if one has a hammer, everything looks like a nail". This gets at what I mean by "scientific imperialism". It is no doubt a natural tendency, when one has a successful scientific model, to attempt to apply it to as many problems as possible. But it is also in the nature of models that these extended applications are dangerous. The abstractions that work well in one context may eliminate what is essential in another. So human behaviour, for instance, though it is certainly an aspect of a biological entity that emerged from features of earlier biological entities, is very little illuminated by standard evolutionary models. Human development and human social organisation are two crucial aspects of the aetiology of human behaviour that standard evolutionary models abstract away from, and the attempt to make this particular "imperialist" expansion is to provide a parody of human behaviour.

I do think it is a reasonable hypothesis that the project of finding a general theory of human behaviour lies beyond the "limits of science". This is simply because the models that we have must abstract so drastically to provide useful and intelligible insights into very specific aspects of human behaviour that it is difficult to imagine how they could develop into one supermodel. I certainly won t rule this out a priori! But I think the onus of proof rests very much with the advocate of such an imagined achievement.

PN: What is your view of the manner in which concepts from the philosophy of science are employed in contemporary debates? Does it help for philosophers of science to explain the difficulties associated with demarcation, for example, or can questions of what is or is not science be addressed without them?

JD: As with scientists, philosophers will only provide useful interventions in public debates if they are able to express their findings in the terms in which those debates are framed. I don't think a concept such as demarcation is especially problematic, provided it is made clear that it is simply a word for distinguishing science from non-science. Having said that, I think that the concept of demarcation tends to suggest that there is a sharp line and some definite criterion that sorts the scientific sheep from the scientific goats. I think this is a highly misleading picture. History and philosophy are not (in the English speaking world anyhow) considered to be sciences, but they can be entirely credible grounds of knowledge. And sciences differ greatly in their epistemic credentials. As I argued in The Disorder of Things, I think we are much better to think in terms of epistemic virtues, features of an investigative practice that confer credibility. No doubt the cardinal empirical virtue is a proper connection with empirical evidence, which is the large grain of truth in the criterion of falsificationism. I don t know whether "epistemic virtue" is a good concept to apply to public debates, but the idea behind it is surely not especially esoteric.

PN: In your new book Darwin's Legacy, you are critical of the appropriation of evolutionary biology by psychologists. Why do you think evolutionary biology can tell us little about human nature? Why do others believe it can do more?

JD: As described above, evolutionary psychology is the paradigm of scientific imperialism. There is vastly more to human behaviour than an understanding of the evolutionary process that provided us with the capacities for behaviour so variously exercised by contemporary humans. Of course that is an interesting and difficult question, and a modest evolutionary psychology that attempts to address it - and there are modest evolutionary psychologists - is entirely respectable. My objection is to the project that supposes that specific behavioural dispositions can be inferred from evolutionary reflection, and that has usurped the name "Evolutionary Psychology".

My first objection is that this project is distressingly a priori. General reflections on evolutionary theory are taken to provide strong grounds for believing that humans have specific behavioural dispositions. Second, the project assumes a very crude view of genetics. The argument that our behaviour essentially evolved in the Stone Age presupposes the idea that evolution involves the gradual accumulation of genetic variations that generate behavioural dispositions. A quite simplistic view of the relation between genes and phenotypes is implicitly assumed. Thirdly, the vision of evolution - gradual accumulation of genes with definite phenotypic effects - comes from the 1950s. We now know that evolution can involve a great range of reorganisations of the developmental process, as well as of the genome, and can draw on a vastly greater range of developmental resources than this model assumes. These resources can range from the epigenetic - aspects of cellular chemistry that are transmitted in the maternal cytoplasm - to the construction of a developmental niche by former generations. In the human case the latter would include such things as schools and hospitals. Developmental cycles have had access to a great variety of potentially decisive changes in recent human history, and the assumption that our basic psychology is stuck in the Stone Age is entirely groundless.

It's difficult to speculate on why this programme is nevertheless so popular. No doubt a part of the problem is intrinsic to interdisciplinary work. It is difficult for a psychologist, say, to keep up to date with developmental and evolutionary genomics, so the danger of basing theories on antiquated science is ever present.

I do have more cynical suspicions, I must confess. Evolutionary Psychology has provided a wonderful recipe for best-selling science writing. Perhaps many people do like simplistic stories, and Evolutionary Psychology offers simple compelling Just So stories that mesh nicely with many stereotypes about human behaviour. The fact that these stories can have the effect of reinforcing such stereotypes and making them seem inescapable makes this activity not just bad science but, unfortunately, potentially socially harmful. That they can have these effects is not merely speculation, but is something that is being empirically explored and confirmed by some of my colleagues in social psychology at Exeter.

PN: What are some of the philosophical issues surrounding genetics and genomics and how do you think the philosophy of science can help address them?

JD: A fundamental issue is making sense of the concepts of "gene" and "genome". The former has been of increasing interest to both philosophers and historians of science. Historians have helped to understand the major disjunction between the classical "Mendelian" concept of the gene and the various concepts that have emerged from molecular genetics and genomics in the last half century. Philosophers have tried to make sense of this diversity of meanings and the relations between them, as well as the errors that can arise from conflating them. One noteworthy contribution has been the book by my colleague, Lenny Moss, What Genes Can't Do, in which he expounds a fundamental division between concepts of the gene directly related to phenotypic outcomes, and concepts grounded in molecular biology. He argues that the conflation of concepts of these two kinds has been central to public and professional misunderstanding of the concept. Another has been the project in "empirical philosophy" by Paul Griffiths and Karola Stotz, who have explored what scientists mean by molecular genes and have found very little consensus. As I mentioned above, another very important project is tracing the relations between these diverse meanings of "gene" in professional scientific contexts and the understandings that are imported into various contexts outside science itself.

The concept of a genome has received less philosophical attention, apart from a general understanding that widely disseminated metaphors such as blueprint, programme, and recipe are seriously misleading. These, in fact, reflect exactly the conflation or confusion between concepts of gene connected to phenotypes and concepts derived from molecular biology. Another fundamental issue is the relation between understandings of the genome as a purely informational concept, as in the idea of the human genome as a particular sequence of nucleotides, and its treatment as a physical object interacting in diverse and complex ways with the cellular environment.

One might wonder why these were questions for philosophers rather than scientists. Scientists in these fields typically work in very well-defined areas for which very specific understandings of the "genes" with which they are concerned are well entrenched. Reviewing the diversity of such entrenched meanings both contemporarily and historically is not something they are normally called on to do, but is very important for developing a coherent overview of this scientific activity and its epistemic and social significance.

PN: What do you think is the relevance of the philosophy of science today? What are the main issues you are interested in other than those in the philosophy of biology?

JD: Science is (rightly in my view) the major authority on how we understand the world today, but what it is, and why it (sometimes!) deserves that status are very poorly understood. For this reason I think philosophy of science is an extremely important area of intellectual activity. A rather different activity, though certainly not unrelated, is science criticism. The authority of science presents a danger, only at its most obvious when actors in white lab coats endorse dubious health aids on television. It is inevitably difficult for most people to assess what is a well-grounded scientific fact and what is pseudo-scientific speculation, and philosophers of science have a potentially vital role in trying to explain the basis for making such distinctions, and in criticising ungrounded scientific claims.

It has to be said that a lot of contemporary philosophy of science is rather distant from these vital public concerns, and while some of their more esoteric concerns may be valuable, in a world in which no one much thinks intellectual activity is worth supporting for its own sake, I think it is very important for philosophy of science to engage more visibly in public discussions of the value and nature of science. One of my general interests that goes beyond the philosophy of biology relates directly to this belief, and is the extent to which scientific theory inevitably reflects social values.

PN: What is Egenis and what is the extent of your involvement?

JD: Egenis is part of a very large investment by the Economic and Social Research Council in investigating the social impact of developments in genomics and related areas of biology. It includes two other research centres and the Genomics Research and Policy Forum in Edinburgh. While the elements of this investment form a closely interacting network, the research centres have particular strengths and expertise, and Egenis is distinctive for its focus on detailed analysis of the production and content of genomic knowledge. This is dependent on close collaboration between the core work in social science and biologists and philosophers of biology. In parallel with the development of Egenis, the University of Exeter has developed an outstanding group of philosophers and historians of molecular biology (with major support from the Arts and Humanities Research Council and the Wellcome Trust).

My current full-time job is Director of Egenis, including the work in philosophy of biology. This is an exciting interdisciplinary project, and though there is inevitably a fair amount of straight management involved, the presence of such a substantial and diverse group of experts in numerous aspects of genomics provides an extraordinary intellectual opportunity. And, fortunately, I do find some time to write philosophy, both independently and in collaboration with colleagues!

PN: What other projects are you currently working on?

JD: My work at the moment is very much concentrated on philosophy of genetics and genomics. I'm near completion of a collaborative book on the philosophy and sociology of genomics with my colleague (and Co-Director of Egenis) the well-known sociologist of science Barry Barnes. In the longer term I'm looking towards a book that will integrate this recent work with the ideas initially developed in The Disorder of Things, and will provide a wide-ranging ontological account of biology. At the time of writing I am on a short leave from Egenis as holder of the visiting Spinoza Professorship of philosophy at the University of Amsterdam. This requires me to deliver two public lectures, for which I have chosen the title "The Constituents of Life". I am hoping that these will provide a skeleton for this developing project, which may become a collaborative venture with some of my fellow philosophers of biology at Exeter.

PN: Who have been the main influences on your thought?

JD: For some reason I have never been comfortable with this question. One undoubtedly important influence has been Wittgenstein. I was an undergraduate at St. John's College, where the resident philosophers were Peter Hacker and the late Gordon Baker, renowned experts on Wittgenstein's thought, and serious early exposure to Wittgenstein has had important influences throughout my career.

I have always been impressed by two sometimes underrated philosophical virtues, clarity and humour. These partly explain my selection of two other influences, Hume and J. L. Austin. Though much of Hume's philosophical system is fairly clearly indefensible today, his commitment to empirical knowledge is still inspiring. And his Dialogues Concerning Natural Religion should be compulsory reading for all first year philosophy undergraduates (especially in the U.S.!). Austin is not a very fashionable philosopher today, but his combination of clarity, stylistic brilliance, wit, and common sense remains something to which I shall always aspire.

I should also mention that I am very proud to be associated with what has increasingly been referred to as the "Stanford School" in Philosophy of Science. I have mentioned Nancy Cartwright, who was my colleague at Stanford for many years, as was also Peter Galison, who has made decisive contributions to understanding the nature of experiment in physics. I only overlapped at Stanford very briefly with Ian Hacking, but his work in philosophy of science has been an inspiring model for me. All of these people, not to mention a good number of younger philosophers who did their PhD work at Stanford, share commitments to the diversity of scientific practices and the social embeddedness of science, while at the same time having a real interest in the detailed content of scientific knowledge. The opportunity to work in this philosophical environment has undoubtedly been a major influence on my philosophical development.


Recent Comments

Recent Discussions

Icon Recent Blogs