Eliasmith, C. (1999) Review of Craig Dilworth's 'The metaphysics of science: An account of modern science in terms of principles, laws and theories'. Dialogue, (37) 3: 656-8.

(Chris Eliasmith, Philosophy-Neuroscience-Psychology Program, Department of Philosophy, Washington University in St. Louis, Campus Box 1073, One Brookings Drive, St. Louis, MO 63130-4899, chris@twinearth.wustl.edu )


Review of Craig Dilworth's "The Metaphysics of Science''


Chris Eliasmith

The Metaphysics of science: an account of modern science in terms of principles, laws and theories
Boston Studies in the Philosophy of Science, vol 173
Dordrecht: Kluwer Academic Publishers, 1995.x + 235 pp. $98.00

 Most clouds have a silver lining. Unfortunately for this book, that consolatory treasure is found in chapter 1. The early appearance of some excellent work in the history of the philosophy of science only renders more stark the shortcomings of the bulk of the book. The reader is treated to an insightful discussion of the realist/empiricist debate, the result being an extremely helpful clarification of both what is being debated and the positions of those taking part. Despite a consistently clear writing style, however, Dilworth does not manage to sustain this high quality of content in the nine subsequent chapters.

 The stated purpose of The metaphysics of science is twofold: first, to determine how theories laws and principles are related; and second, to determine what the core principles of science are. By meeting these two goals, Dilworth hopes to provide a convincing picture of modern science. Throughout the course of the first five chapters, the principle-theory-law (PTL) model of scientific explanation is introduced. Under this model, scientific explanation ``consists in indicating how particular facts are but the manifestation of the principles of science'' (p. 127). Clearly, for such a model to be substantive, the principles of science must be determined. Modern science is seen by Dilworth to be centred on three principles constituting a conceptual paradigm. They are:


  1. The principle of the uniformity of nature (i.e. nature behaves predictably)
  2. The principle of substance (i.e. substance exists perpetually)
  3. The principle of causality (i.e. change is caused)

Laws and theories enter this picture of science as metaphysically secondary. Laws provide scientific knowledge (and are prior to theories), whereas theories, by linking laws to principles, provide scientific understanding.

 Perhaps the greatest theoretical success of the book is its convincing arguments for the importance of metaphysical considerations to science (pg. 71). On the PTL model, Dilworth shows why metaphysics plays, or at least should play, a central role in understanding modern science. If one accepts the notion that there are non-empirically testable (in the standard sense) principles on which all of science is based, one is also forced to admit metaphysics. Since these core principles are presuppositions of doing science in the first place, they become the ``transcendental heart of science'' (p. 2). The principles on which science is based are not conceived of as necessary truths, nor as axioms for the derivation of more principles; thus, they are not truly a priori. Rather, principles are understood as being assumed to be true for the purposes of doing science though they are subject to revision. The question clearly arises, is this sort of characterization of principles an attempt by Dilworth to have his metaphysical cake and eat it too? Clearly, this is the case, but the reason this aspect of the book is a success is that Dilworth does a convincing job of walking the fine line between the strong foundationalist and strong coherentist positions, carving out, in the process, an interesting place for metaphysics in science.

 Despite this success, and the excellent discussion of the empiricist/realist debates, The metaphysics of science falls disappointingly short of its main goal of providing a convincing picture of modern science. Though admittedly descriptive of the nature of science during the opening chapters of the book (p. 61), the promised culmination of this description in both an explanation of the nature of science and an evalution of science is, in short, disappointing. It is unclear why we should accept the provided description of the mechanisms and commitments of modern science over any others (see, for example, the discussion of technology and science on p. 84 or the status of empirical laws on p. 90). The `structure of science' is simply asserted throughout the course of chapters 3 and 4 with a surprising dearth of supporting examples; in fact, there are none until the last page of chapter 4 and that is only six lines long (p. 106). This lack of supporting instances, coupled with ample redefinitions of standard terms (e.g. fact (p. 79) and understanding (p. 105)) makes for a confusing and unconvincing examination of modern science.

 Dilworth's discussion of laws, theories, models, and analogies provides just one example of the sorts of difficulties which mar the book. On one occasion, theory and theoretical model are equated (p. 115). Subsequently, theoretical model and model are claimed to be interchangeable (p. 112). So, one can only conclude that theories and models are one and the same for Dilworth. There are, however, no reasons given as to why such a position is to be desirable. Furthermore, this is a position that many (perhaps most) concerned with the relation between laws, theories, models, and analogies would find unpalatable and thus deserves a defense (Hesse, 1972). This flaw, particularly when coupled with a number of other inconsistencies and inaccuracies 1, diminishes the confidence of the reader in the overall quality of the work.

 However, it is not, in the end, such minor flaws which make the book somewhat unrewarding. After the excellent, impartial introduction to the realist/empiricist debate in chapter 1, Dilworth assumes a realist position in the remainder of his discussion. Without having taken sides in the debate, he seems to have declared a winner: ``The aspect of modern science we shall consider now is precisely that which investigates the nature of real causes'' (p. 95). Though this lives up to the claim that the debate will be resolved (p. 48), it is strange that the resolution comes by fiat. For example, principle 3 provides a strongly realist conception of causality. However, there need be no commitment to such a realist conception of causes to do science (e.g. one may assume causes just are the regularities in nature, they don't `act' to change nature). Why then, does Dilworth consider this a core principle of the discipline? It is unclear. The strongly realist flavour of this principle seems to have stacked the deck against empiricists from the beginning of the analysis and is later employed to refute aspects of empiricism (p. 101). It is hardly fair to call such question begging a true resolution to the long-standing debate. Having built a realist position into these core principles of science, it is hardly surprising that Dilworth determines that science is a realist project. A realist ought to find this particularly distressing because he or she would be inclined to agree with the conclusion but is unable to add further argument to the realist arsenal.

 Dilworth has done a good job of presenting the debate which he wishes to resolve, and manages an interesting analysis of the role of metaphysics in science. However, the majority of the book is subject to flaws which make reading it somewhat less rewarding than one expects given the success of earlier chapters. It is not clear that the bit of silver one will find is worth the toleration of the cloud.


 Hesse, M. (1972). ``Models and analogies in science.'' The encyclopedia of philosophy. New York, MacMillan Publishing Co. & The Free Press. pp. 354-359.


 1. For example, in a seven page span, the conceptual priority of laws and theories seems to change (p. 96-103). As well, there is a misleading discussion of chaos theory in which he guards principle 1 against the supposed non-determinism of the theory (p. 189-190). Finally, there is a blurring of the distinction between theories, laws and principles (p. 114).