What_is_this_thing_called

Epilogue to the third edition

In this concluding section I offer some reflections on what has been achieved in the foregoing chapters. I raise three interrelated questions or problems that have concerned me during the writing of this book, and continue to do so.

1. Have I answered the question that forms the title of this book? What is this thing called science?

2. What is the relation between the historical examples given in the book and the philosophical thesis defended? Do the examples constitute evidence for my case, or are they simply illustrations?

3. How do the general claims made about science by the Bayesians and new experimentalists, discussed in chapters 12 and 13, relate to the case against method made in chapter 11? Isn’t it the case that if there is no general account of science then all further discussion of the issue is redundant?

My response is as follows: I reaffirm that there is no general account of science and scientific method to be had that applies to all sciences at all historical stages in their development. Certainly philosophy does not have the resources to provide such an account. There is a sense in which the question that forms the title of this book is misguided. Nevertheless, a characterisation of the various sciences at various stages is a meaningful and important task. In this book I have attempted to accomplish that task for the physical sciences from the time of the scientific revolution in the seventeenth century until the present time (although I have refrained from tackling the question of the extent to which modern innovations such as quantum mechanics and quantum field theory involve characteristics that are qualitatively new). This task involves displaying the nature of the physical sciences mainly by means of historical examples of the appropriate kind. The historical examples therefore constitute an important part of the case rather than being mere illustrations of it.

Although the account of the physical sciences presented falls far short of providing a universal definition of science, it is far from useless when it comes to debates about what is or is not to count as science, exemplified, for example, in disputes about the status of ‘creation science’. I presume that the main aim of those who defend creation science under that name is to imply that it has a character similar to that of acknowledged sciences such as physics. The position defended in this book enables that claim to be appraised. Having displayed what kind of knowledge claims are sought in physics, what kinds of methods are available for establishing them and what kind of success has been achieved, we have what we need as the basis for a comparison with creation science. Once the similarities and differences between the disciplines have been displayed, we have all that we need for a judicious appraisal of them, and will be in a position to appreciate whether anything can be legitimately read into the naming of creation science as a science. No universal account of science is necessary.

In the paragraph before last I indicated that my portrayal of the physical sciences is to be defended by reference to ‘historical examples of the appropriate kind’. Some elaboration is called for here. Examples of the appropriate kind are concerned with the way in which the physical sciences function as knowledge. They are concerned with the kinds of claims made about the world in the physical sciences and the kinds of ways those claims are brought to bear on the world and tested against it. They are concerned with what philosophers call the epistemology of science. Philosophy of science is conducted by way of historical examples that display and clarify the epistemological function of science. The kind of history of science involved is a selective kind of history, and certainly not the only kind of history of science that is possible or important. The production of scientific knowledge always takes place in a social context in which that aim is interrelated with other practices with different aims, such as those involving the personal or professional aims of scientists, the economic aims of funding agencies, the ideological interests of religious or political groups of various kinds and so on. A history that explores these connections is both legitimate and important, but, I claim, beside the point as far as the project of this book is concerned. There is a range of kinds of ‘social studies of science’ currently in vogue that imply that an epistemological study of the kind I have conducted in this book cannot be achieved without due attention to the full range of senses in which science is social. In this book I have not faced the challenge posed by these schools of thought head-on. I have been content to show that what they say cannot be done can indeed be done simply by doing it. My attempt to square accounts with contemporary social studies of science appears in my Science and Its Fabrication (1990), a book in which I hope I make it clear that I regard a study of the social and political aspects of science as of great importance. The point at issue is the epistemological relevance of such studies.

Let me now turn to the question of the status of Bayesianism and the new experimentalism in the light of my denial of universal method. Bayesianism appears as an attempt to give an account of scientific reasoning in general, as is clearly signalled by the title of Howson and Urbach’s 1993 text. However, this impression does not bear analysis. Even if we accept unquestioningly the Bayesian machinery, what that machinery gives us is a general way of adjusting the probability to be ascribed to beliefs in the light of new evidence. It does not single out scientific reasoning and distinguish it from other areas. Indeed, the most useful applications of Bayesianism are in gambling rather than science. Consequently, if Bayesianism is to tell us something distinctive about science in particular, then it will need to be augmented by some account of the kinds of beliefs and the evidence bearing upon them that occur in the sciences. I suggest that this can only be done by a careful look at the sciences themselves. What is more, I suggest that when that is done differences in the various sciences, and even qualitative changes within the methods of a single science, will emerge. That is, even if the Bayesian approach is the correct one, it does not stand as a threat to the denial of universal method, and is in need of the kind of epistemological history of science that I advocate.

The new experimentalists have certainly revealed some important features of experiment and its achievements within the physical and biological sciences. However, the account of science that this yields cannot be taken as providing the universal account of science. By way of examples, the new experimentalists have demonstrated the capabilities and achievements of experiment in the natural sciences during the last three hundred years and Deborah Mayo has provided a formal underpinning for much experimental reasoning by appeal to error theory and statistics. This does not amount to a universal account of science for two reasons. First, the emphasis on experimental manipulation involved in the new experimentalism renders that account largely irrelevant for an understanding of disciplines, especially in the social and historical sciences, where experimental manipulation is impossible or inappropriate. This conclusion could conceivably be avoided by identifying science with experimental science, but this would hardly serve to appease those who wish to call themselves political scientists or social scientists, for example. Second, as was argued in chapter 13, the new experimentalist account is incomplete insofar as it does not include an adequate account of the various crucial roles played by theory in science. The problem is very evident, I suggest, in Peter Galison’s 1997 text in which he gives a descriptively rich account of progress in twentieth-century microparticle physics by focusing on the particle detectors and counters, their capabilities and their evolution. What is left unclear in the book is the relation between the experimental detection of particles and the high-level theory, involving symmetry and conservation principles, by means of which the particles are understood and classified. At the time of writing this epilogue, I regard it as an outstanding and pressing problem in the philosophy of the natural sciences to augment the insights of the new experimentalists with a correspondingly updated account of the role or roles of theory in the experimental sciences, substantiated by detailed case studies.

The following historical reflection illustrates the difficulty of extracting some universal characterisation of or prescription for science from the work of the new experimentalists, and illustrates the kind of study I have in mind for clarifying the nature of the relationships between theory and experiment. The idea that one should attempt to understand the world by experimentally manipulating it was by no means novel at the time of the scientific revolution. Alchemy, understood broadly as the precursor to modern chemistry involving the purposeful transformation of matter rather than narrowly as the attempt to transmute metals into gold, dates back to antiquity and flourished in the medieval period. The practice was not particularly successful. That lack of success cannot be simply attributed to lack of guidance by theory. A range of atomistic and other matter theory informed the work of the alchemists. If one is inclined to ignore theory and look simply to experimental practice, then significant progress can be discerned in the craft traditions of the metallurgists and drug manufacturers of the sixteenth and seventeenth centuries. However, the knowledge involved can be seen as qualitatively different from the chemistry that was to emerge in the late seventeenth and eighteenth centuries. The latter did involve ‘theory’, but very low-level theory far removed from atomism. What was needed, and what was supplied early in the eighteenth century, was a notion of chemical combination and recombination of substances, involving the idea that substances, when combined, continue to exist in the resulting compound and are there to be extracted again by means of appropriate manipulations. The classification of substances into acids and alkalis, and the salts produced by the neutralisation of one by the other, offered a way of organising research in a way that made progress possible without the need for some atomistic or other matter theory. It was well into the nineteenth century before the time was ripe for such speculations to be linked with experiment. So the question of the role of experiment in science and its relation to theory is a complex and historically relative one even if we restrict the discussion to chemistry.

I conclude with some remarks about the relationship between the views on science explored in this book and the work of scientists. Since I have denied that there is a universal account of science available to philosophers and capable of providing standards for judging science, and since I have argued that an adequate account of various sciences is only to be had by way of a close look at the sciences themselves, it might be concluded that the views of philosophers of science are redundant and that only those of scientists themselves are of consequence. It might be thought, that is, that insofar as I have successfully made my case, I have done myself out of a job. This conclusion (fortunately for me) is unwarranted. Although it is true that scientists themselves are the practitioners best able to conduct science and are not in need of advice from philosophers, scientists are not particularly adept at taking a step back from their work and describing and characterising the nature of that work. Scientists are typically good at making scientific progress, but not particularly good at articulating what that progress consists of. This is the reason that scientists are not particularly well equipped to engage in debates about the nature and status of science, and do not typically do a good job when it comes to controversies about the nature and status of science such as are involved, for example, in the evaluation of creation science. This book is not intended to be a contribution to science, not even the physical sciences on which I have focused. Rather, largely by means of historical examples, I have tried to clarify what kind of things the physical sciences are or have been.