Book s review forthcoming in International Studies in the Philosophy of Science String Theory and the Scientific Method RICHARD DAWID Cambridge, Cambridge University Press, 2013 x + 202 pp., ISBN 9781107029712, 60.00, US$95.00 (hardback) Richard Dawid s new book, String Theory and the Scientific Method, is an interesting and sophisticated analysis of the conceptual novelties introduced in the traditional methodology of science by string theory. This book makes a major contribution to the contemporary philosophical debate about confirmation and assessment of scientific theories. Scientists are presently in total disagreement about the significance of string theory s achievements. Dawid puts forward a philosophical interpretation of this apparently insolvable conceptual tension. The dispute is between defenders and revisionists of the traditional empirical paradigm of theory assessment. The former are critics of string theory, whereas the latter are supporters. As Dawid says, the dispute fails to be productive because of the meta-paradigmatic rift between the two disputants. What he means is that this insolvable tension does not simply derive from conceptual differences between particular scientific theories; rather, it arises from the failure to recognize that a solution can be found only by focusing on the meta-level issue about the choice of viable criteria of scientific theory assessment. Dawid s book unravels this issue in a very accurate and interesting way. Dawid s well-structured and original notion of non-empirical theory assessment offers a robust conceptual framework within which string theory is rescued from the charge of not being a viable scientific theory. Overall, the author s mastery of intertwining philosophical and scientific perspectives on these foundational issues is impressive. The book starts with an intriguing introduction to string theory given in simple and clear terms, but still capturing all the theory s physical and conceptual features. Moreover, philosophyoriented readers will find extremely helpful the author s accurate presentation of the types of scientific underdetermination that figure most prominently in philosophy. To begin with, I shall focus on how Dawid relates scientific underdetermination to the issue of the non-empirical assessment of the theory. When dealing with a scientific theory so much in advance of experiments as string theory is, applying a conventional criterion of theory assessment based on direct empirical confirmation becomes problematic. More precisely, without challenging the role of empirical data as the ultimate judge of a theory s viability, we need to enrich the existing criteria with 1
conceptual tools defining what status a scientific theory can acquire in the absence of empirical confirmation. This is not an easy task and Dawid deals with it quite interestingly. What does scientific underdetermination mean? As Dawid says, the process of building a scientific theory is inevitably underdetermined by the available empirical data. If we have a successfully tested theory that not only fits available empirical data but also predicts new phenomena, what keeps us from completely trusting the idea that those not yet observed phenomena really exist? What reduces our trust here is the idea that there might be different, still unknown, scientific theories fitting the same available empirical data, but also predicting completely different new phenomena. According to Dawid, this is the sense in which scientific theories are underdetermined by the available empirical data. But what does scientific underdetermination have to do with string theory in the first place? The theory still remains unconfirmed by available empirical data, after all. In other words, given the lack of direct experimental access to the theory s physical content, how can we apply the notion of underdetermination by experiments? Dawid s answer to this question refers to an important theoretical feature of string theory. This feature also plays a crucial role in two approaches to the problem of justification of theories of quantum gravity developed respectively by Nick Huggett and Christian Wüthrich (2012) and by me (Vistarini 2013). Some features of these two approaches seem to resemble Dawid s view of non-empirical theory assessment. String theory can reproduce at low energy general relativity and gauge field theory, namely two successfully tested theories. According to the two approaches, although the empirical data confirming the latter do not directly confirm string theory, still they strongly contribute to string theory s justification claims. If string theory predicts, preserving self-consistency, low-energy theories that fit those empirical data, then we have good reasons to believe that the theory is a plausible model of explanation of why those observed low-energy physical dynamics are as they are. As long as the theory gives plausible answers to why-questions, hence offering a robust model of explanation, it fulfils justification criteria. But coming back to Dawid s answer to whether scientific underdetermination can be correctly applied to string theory, he says the theory cannot predict those empirical data directly, but its formal articulation relies on them in virtue of the theory s ability to reproduce the low-energy theories confirmed by those available data. This conceptual relation between string theory and the available empirical data makes scientific underdetermination applicable to the former. So, Dawid very interestingly defines a criterion of non-empirical assessment of string theory in terms of an assessment of scientific underdetermination. In other words, 2
our trust in string theory s explanatory power depends on how many theoretical alternatives to string theory meeting the mentioned criteria of underdetermination are out there. This assessment estimates string theory s chance of being viable. I can give here just a brief presentation of the author s line of reasoning. By considering theoretical alternatives to string theory, the author does not consider all logical possibilities. Some specific rules of scientific method are already built into the assessment procedure from the start. Quoting Larry Laudan (1996), the author calls them ampliative rules : Ockham s razor, the exclusion of ad hoc explanations, endorsement of some basic form of the principle of induction, and so on. Then, Dawid develops the assessment of scientific underdetermination in terms of the two main schematizations of scientific reasoning in present-day philosophy of science, namely inference to the best explanation and the Bayesian approach. The former approach underlies all the main arguments made by string theorists in favour of the theory s viability. Such arguments are also supported by the authors and presented as having the good property of significantly restricting the number of viable alternatives to string theory. First, the no-alternatives argument: there aren t many good alternatives out there, so not many viable unifying theories can compete with the self-consistency and explanatory power of string theory. Second, the unexpected explanatory coherence argument, namely the fact that a theory built to fix a problem, so far left unsolved by other theories, not only offers a solution to that problem but also unexpectedly reveals itself to be a plausible model of explanation of additional problems. Finally, the meta-inductive argument: the idea that if two research programmes are similar and one of them was successful before, then we have strong reasons to believe that the other will also work well later. Dawid seems to read this notion of similarity in historical terms. In fact, he claims that string theory s research programme develops from the standard model in particle physics. So it is plausible to think that if the precursor programme worked before, then its successor will do the same later on. Like Dawid, I accept the first two arguments. I take them as offering a strong reason for studying string theory, despite the lack of direct empirical support. This also relates to what I said earlier about the theory s justification. Originally unexpected predictions, like those of gravity and of crucial features of gauge field physics, rescue the theory from the charge of being unviable. In fact, as Dawid argues, this contributes to restricting significantly the number of plausible alternatives to string theory, hence reinforcing its viability. And I would also mention that the derivation from string theory of those successfully tested theories compensates to a great extent for our lack of direct empirical access to its physical content. 3
However, if I am reading Dawid correctly, I do not agree with his historical use of the meta-inductive argument. The history of string theory does not seem to intertwine significantly with the history of the standard model. Nevertheless, I think there is a way of supporting the meta-inductive argument without introducing historical considerations. The standard model programme is successful, and so will be the programme of string theory. But that is not in virtue of the theories histories rather in virtue of a relation of formal deduction of the standard model from string theory. Then, the similarity between the two theories may be seen in terms of synchronic formal derivation. Particular configurations of strings define string models of particles, which in a low-energy regime reproduce the chiral fermions and gauge bosons of the standard model. In this sense the standard model reveals itself to be an effective, phenomenological theory, deduced within a certain approximation from string theory. So, if the research programme of the deduced theory works well, we have strong reasons to believe that the research programme of the reducing higher-energy theory will work well too (for the notions of deduction and reduction mentioned above, see also Butterfield 2011). In conclusion, I support Dawid s main idea that string theory brings a new phase of progress in fundamental physics; a phase in which, instead of having a collection of separate theories, each fully describing physics at its own length scale, we see the gradual appearance of an all-inclusive theoretical scheme whose general features seem to qualify it as a candidate for a final theory. References Butterfield, J. 2011. Less Is Different: Emergence and Reduction Reconciled. Foundations of Physics 41: 1065 1135. Huggett, N., and C. Wüthrich. 2012. Emergent Spacetime and Empirical (In)coherence. Studies in History and Philosophy of Modern Physics 44: 276 285. Laudan, L. 1996. Beyond Positivism and Relativism: Theory, Method, and Evidence. Boulder, CO: Westview. Vistarini, T. 2013. Some Remarks on the Emergent Character of General Relativistic Spacetime in String Theory. To be submitted. TIZIANA VISTARINI Department of Philosophy Rutgers, The State University of New Jersey 4
106 Somerset Street, 5th floor New Brunswick, N.J. 08901-4800 U.S.A. E-mail: tv85@rci.rutgers.edu 5