Metascience (2012) 21:65–69 DOI 10.1007/s11016-011-9527-3 ESSAY REVIEW
The theory of everything? Brian Ellis: The metaphysics of scientific realism. Durham: Acumen, 2009, x+179pp, £16.99 PB Emma Tobin
Published online: 6 April 2011 Ó Springer Science+Business Media B.V. 2011
The aim of this monograph is to develop a metaphysics for scientific realism, which has the characteristics of a first philosophy. In the first instance, this monograph is an excellent chronological and philosophically honest account of the avenues, which led Ellis to his scientific essentialism. This book takes his essentialism further than ever before. It presents a systematic version of metaphysical realism, which accommodates not just all of science, including the quantum world, but any discipline, which makes assumptions about what there is in the real world, including mathematics, ethics, freedom of the will and many other thorny metaphysical issues. This is a very ambitious project and even after putting the parts of the puzzle together into an integrated first philosophy, the reader is still left with many questions about the overall picture. Nevertheless, this is a book not to be missed by metaphysicians. Chapter 1 provides a discussion of the claim that metaphysics is required for scientific realism. This is justified on the basis of a particular construal of the relationship between science and metaphysics, namely that scientific knowledge has limitations. The question ‘‘what sort of world would be required to accommodate quantum field theory’’ is a relevant question, which quantum field theory cannot answer (20). Ellis rejects semantic and epistemic accounts of scientific realism based on theories of truth and reference, because there is no relevant theory of truth to take on the burdensome task of explaining how truths supervene on reality. The correspondence theory of truth cannot do the job and the semantic theory of truth entails a commitment to realism about possible worlds. Ellis’s project then is an enquiry into the truthmakers of science, and he argues that a metaphysical theory of reality is required to accommodate those truthmakers. There is a logically inescapable circle in metaphysical reasoning: namely, any E. Tobin (&) Science and Technology Studies, University College London, Gower Street, London WC1E 6BT, UK e-mail:
[email protected]
123
66
Metascience (2012) 21:65–69
postulated existent is ontologically plausible only if it fits into an adequate metaphysical theory, but any metaphysical theory is only adequate if it accommodates all the things that exist. Nevertheless, Ellis makes a simple inference to the best explanation: a metaphysics for scientific realism is the best explanation for realism in science. In chapter 2, Ellis assesses Smart’s original argument for scientific realism. This argument was also an inference to the best explanation: if the world behaves as if entities of the kind postulated by science exist, then the best explanation of this is that they really do exist (30). Ellis favours this argument over more recent arguments for scientific realism, all of which are committed to descriptivism, the view that the aim of science is to give a true, generalised description of reality (25). This kind of descriptivism is rejected on the grounds that scientific realism should aim to account for why the theoretical entities of science explain the world, not just describe it. Ellis offers a further causal argument for the existence of theoretical entities: causal connectivity characterises real entities because they have effects other than those effects that were used to define them. This causal argument is compelling, because we can infer two further arguments from it: the first is an argument for the existence of properties, the properties of these entities must be capable of causally interacting in the way that they do, while the second is an argument for the existence of forces, fields, spatio-temporal relationships and so on. But this sophisticated realist ontology for science is inadequate for a more general theory of what there is. The aim of ontology, according to Ellis, is to say what kinds of elements must be thought to exist fundamentally and to explain the ‘‘kinds of things we really need for an ontological reduction of everything there is’’ (34). Chapter 3 concerns scientific essentialism. Ellis formulates a more general theory of what there is, viz. an essentialist account of natural kinds. The existence of a natural kind hierarchy is not an empirical discovery, but more of a metaphysical hypothesis being proposed in order to explain the exact and structured nature of the entities that are found in science and the precise hierarchical arrangements between them. Natural kinds are intrinsically disposed to behave in the way that they do, because of their intrinsic causal powers. This ontology has the added benefit of providing truthmakers for the laws of nature. The conception of natural kinds is more liberal than in traditional substantive accounts of natural kinds. Three categories of natural kind are distinguished: substantive, dynamic and tropic. Where a natural kind is a generic kind, composed of different species (e.g. chlorine and its isotopes), then any species must be a member of the generic kind. In other words, it is impossible for there to be a species of chlorine, which could be a species of some other element. In contrast, generic kinds in biology do not satisfy this constraint, because the species of any genus overlap if we go far enough back in evolutionary history. Ellis discusses some concerns with this earlier ontology. (1) He accepts that his earlier system was too lavish. (2) He was incorrect to include relations as natural kinds, because many relations (e.g. numerical and spatio-temporal) relations lack a definitive causal role. (3) Natural kinds of substance must now be construed as ongoing physical processes, so that the category of natural kinds of objects is a
123
Metascience (2012) 21:65–69
67
subcategory of natural kinds of processes. At the fundamental level, the category of natural kinds of processes must be split into two distinct kinds: the first, temporally extended continuing processes (e.g. energy transmission processes) and the second, spatially extended instantaneous ones (e.g. entangled electrons will remain quantum-linked until some interaction breaks the link). This ontology while more parsimonious retains the explanatory power of its predecessor. Chapter 4 examines the claim often made that quantum mechanics (henceforth QM) is a problem for scientific realism. The problem associated with realism and the two-slit experiment in QM is that we have to accept both that Schro¨dinger waves are, on the one hand, real and that, on the other hand, they do not behave as waves (they act locally and behave like classical particles). Ellis argues that the best option for a scientific realist is to say that the transmitted wave has the potential to reconstitute itself into a particle instantaneously and act as a classical particle would. However, this has extreme consequences since the known laws of mechanics are timesymmetrical, but the scientific realist description of this quantum mechanical process is temporarily irreversible, and thus, it contradicts the time-symmetry thesis. Ellis argues that a version of quantum mechanical realism is nevertheless possible, on the basis of a distinction between time-reversibility and time-symmetry. The time-symmetry thesis states that unless there are basic laws of a radically different kind yet to be found, there is nothing to stop the world from evolving in the temporally reversed direction. Ellis claims that we must distinguish between two fundamental kinds of processes: continuous states of affairs and instantaneous states of affairs. The time-symmetry thesis is only true for continuous states of affairs. In contrast, because the ontology required for quantum mechanics includes instantaneous states of affairs (because there is an instantaneous change, a quantum leap from the state of indeterminacy to that of determinacy in the collapse of the wavefunction) and because the process appears to have no temporal inverse, Schro¨dinger waves are not time-reversible. This is quite a move, and of course, as Ellis acknowledges, an acceptance of QM realism involves the rejection of one of the core tenets of Einstein, viz. that the world is essentially symmetrical in time. Ellis gives very little time to the alternatives to his version of Schro¨dinger wave realism discussing Bohr, de Broglie, Everett and Frisch in less that five pages of text. Moreover, there is no discussion of recent collapse theories in the philosophy of physics. At the end of this underdeveloped section, Ellis urges that the time has come to accept QM realism, but the reader is left rather unsatisfied, given that Ellis has given his opponents insufficient consideration. Equally, problematic is the fact that the subsequent account of causation is based on an acceptance of QM realism, namely if Schro¨dinger waves are taken to be real and directed (that is, if Ellis is correct in rejecting Huw Price’s time-reversibility thesis for Schro¨dinger waves), then elementary causal processes are real and directed too and have nothing to do with the direction of entropy increase, nor pace Humeanism, with human psychology. Ellis’s account of causation is similar to the proposal of Dowe and Salmon where the concept of a causal process is envisaged as a physical event, involving a change in energy distribution. Nevertheless, Ellis claims that his account of physical causation can be applied to all causal processes.
123
68
Metascience (2012) 21:65–69
Chapter 5 addresses properties. Ellis appears to move from property dualism to property trialism. He accepts (1) causal powers (e.g. the power of one negatively charged electron to repel another); (2) categorical dimensions (e.g. quantities, shapes and molecular structure); and (3) those that are neither causal powers nor categorical dimensions (e.g. the changes due to collapsing wavefronts and radioactive decay processes) (99). Though Ellis makes this distinction and claims that (3) is an interesting metaphysical category he does not discuss it at any length. This is unsatisfactory, given the fact that QM realism is motivated on the grounds of an interpretation of the collapse of the wavefunction and also given that the subsequent metaphysical account of causation is based on the latter. Surely, it is fundamental to the arguments that have gone before that Ellis provides a realist account of the third category. Equally, given that at the end of chapter 3, Ellis promised to economise his lavish 6-category ontology but is now expanding the processes and properties subcategories, the reader is left wondering whether the new metaphysics for scientific realism is just as lavish as his original scientific essentialism. Ellis argues for the inclusion of (1) and (2) above on the basis that the problems that both face can be overcome, viz. the Meinongian argument and quidditism, respectively. Taking the former first, the chief problem associated with dispositional monism is the Meinongian argument; that is, anything that is the bearer of an unmanifested disposition must point towards something that is a non-existent but possible object: a mere possibilium. Ellis argues that this argument is flawed in that it is committed to a possible world analysis of the truth conditions for causal conditionals. On the converse, the main argument associated with categorical monism is quidditism, i.e. that if the identity of a property depends not on what its bearer is disposed to do, but merely on what it is, then we are necessarily ignorant of it. Ellis claims that quidditism can be accepted, because a quiddity can be something that even though it has no effects to be observed, is nevertheless observed accidentally. For example, ‘the identity of a molecule depends on its structure and the identity of the structure of a molecule depends essentially on what it is, not on what it does. In [Alexander] Bird’s terminology, it is a quiddity’. (106). At first sight, this might appear removed from scientific essentialism because the essences of molecular kinds are their atomic structures and these atomic structures are categorical. Thus, it looks as though Ellis is committed to the claim that the laws of chemistry are contingent. Moreover, the categorical monist will claim that there is nothing to stop this also being true at a more fundamental level: the identities of atoms might also depend on subatomic constitutions. But Ellis claims that ‘‘this process must have a natural limit. At some stage one will have to admit that the basic constituents of the world must behave as they do, because it is of their natures to do so, that is the basic tenet of dispositional essentialism must be upheld’’ (114). But on what grounds should we adhere to this basic tenet? There are several assumptions being made here; that would not convince any of the essentialist’s opponents: (1) this process may not have a natural limit; there may be no fundamental level to be relied upon; (2) it is equally possible, as a structural realist for example might hold, that the subatomic constitutions are all entirely structural.
123
Metascience (2012) 21:65–69
69
Why are categorical dimensions not reducible to lower-level categorical dimensions or to structures (to satisfy the categorical monist and the ontic structural realist on the one hand) or to more fundamental dispositional powers (to satisfy the dispositional monist on the other)? If indeed it is to the third category mentioned above, viz. those properties which are neither causal powers nor categorical dimensions, then more needs to be said about it. Perhaps, Ellis has structures themselves in mind, since he argued in chapter 4 that Schro¨dinger waves of particle realisation potentials are entities that are defined by their structures (84). But categorical dimensions are understood to be structures and relations too (93). Thus, the discussion of properties is inconclusive and the reader is left extrapolating the details from a comparison of the two chapters. Moreover, this issue about the relationship between levels has implications for the issue of the reduction of chemistry to physics. It is interesting that Ellis uses molecular shape as the least controversial example of a categorical dimension. Yet, recent discussions of reductionism in philosophy of chemistry have highlighted molecular shape as a problem for reductionism. Quantum chemists have argued that molecular shape cannot be described straightforwardly in terms of Coulombic Hamiltonians. Thus, even if we grant QM realism, Ellis must provide an explanation about the relationship between QM realism and molecular realism. In chapters 6 and 7, ‘‘Realism as first Philosophy’’ and ‘‘Realism in Ethics’’, Ellis takes the project even further showing how an ontology developed for scientific realism could act as a first philosophy for any kind of discipline which makes assumptions about what there is in the real world. What does this include? Just about everything: realism about mathematics, time, freedom, personal, collective and social responsibility to mention but a few. Though the space given to these debates does not do them justice, one cannot help, but enjoy the grand theory of everything presented. The philosophical honesty in this book, from one of philosophy’s major authors in metaphysics, is refreshing and it makes it a book well worth reading. Acknowledgments I am grateful to Anjan Chakravartty for helpful comments on this review.
123