Definition of Scientific Realism
Arguments pro and contra Scientific Realism
The “no miracle” argument
The “pessimistic meta-induction”
The Bandwidth of Scientific Realism
Standard Scientific Realism
Conjectural Scientific Realism (Karl Popper)
Structural Scientific Realism (John Worrall)
Experimental Scientific Realism (Ian Hacking)
Active Scientific Realism (Hasok Chang)
Summary: Flavors of Scientific Realism
Critique of Chang’s Active Scientific Realism
Methodological or Epistemological Concept?
Chang versus Hacking
Humility or Confidence?
In his recent book, Is Water H20?, Hasok Chang presents a detailed analysis of scientific realism and enunciates a new concept of it, which he names “active scientific realism”. It is a view of scientific realism that accentuates experimental activity for learning about reality rather than armchair philosophy in the search for utmost metaphysical truth. Chang puts it in a nutshell as follows: “If the buzzword for standard realism is truth, it is progress for active realism.” (Chang 2012, 223)
This term paper attempts to critically look at this new concept, put it in the perspective of other realist concepts and find answers to questions like the following:
- How does Chang’s concept fit into the existing landscape of scientific realism?
- What are the roots of the concept?
- What is new and attractive in it?
- What are the weaknesses of the concept?
First I’ll try to define scientific realism as a metaphysical and epistemological position as opposed to anti-realism. In the next chapter I’ll present the main arguments for and against scientific realism, the “no miracle” argument and the “pessimistic meta induction”, and also look at them from Hasok Chang’s angle of view.
Then a brief overview of common realist positions in philosophy of science will be given, including Hasok Chang’s new conception. In the following chapter I’ll try to look critically at some aspects of Chang’s “Active Scientific Realism” and balance the strengths and weaknesses of the concept.
Definition of Scientific Realism
Scientific realism is a philosophical position postulating that scientific theories are supposed to reveal the truth about the external world. A contraposition would be instrumentalism or anti-realism, which considers scientific theories only to be instruments for predicting phenomena (cf. Ladyman 2002,104).
Scientific Realism has several aspects, depending on the philosophical angle of view. Ian Hacking summarizes the field of debate as follows: “Most of today's debate about scientific realism is couched in terms of theory, representation, and truth.” (Hacking 1983, 31). However, he makes it clear that no final argument for or against realism can be found at the level of representation. It is rather the experiment, the intervention, which “provides the strongest evidence for scientific realism” (1983, 262).
James Ladyman defines scientific realism in terms of three commitments, which he considers to be involved in the concept:
"scientific realism involves three kinds of philosophical commitment: a metaphysical commitment to the existence of a mind-independent world of observable and unobservable objects; a semantic commitment to the literal interpretation of scientific theories and a correspondence theory of truth; and finally an epistemological commitment to the claim that we can know that our best current theories are approximately true, and that they successfully refer to (most of) the unobservable entities they postulate, which do indeed exist." (Ladyman 2002,159, my italics).
Failing to accept all three of those commitments results, according to Ladyman, in antirealism, which can have different motives, such as skepticism, reductive empiricism, social constructivism, or constructive empiricism (cf. ibid.). But anti-realism is not identical with skepticism. Skepticism is rather, as John Heil puts it, “a meditation on realism: skepticism is realism reflected in the mirror of epistemology” (Heil 1998, 69). The skeptic doesn’t deny the existence of an external world, but he can’t “provide an epistemically noncircular defense of our conviction that we have such knowledge or justified belief” (ibid., 68).
As far as empiricism is concerned, Hasok Chang deplores that empiricism “is sometimes taken as a rather passive or defensive doctrine, emphasizing that the only source of knowledge we can have is experience and that we should avoid treating other things as legitimate sources of knowledge” (2012, 217). By contrast, in his view, empiricism is as much an active concept as active scientific realism. And if realism sounds just like empiricism, that’s just like it should be, according to Chang (cf. ibid.).
Arguments pro and contra Scientific Realism
Two important lines of argument pro and contra scientific realism are the “no miracle” argument introduced by Hilary Putnam and the “pessimistic meta-induction”, ascribed to Larry Laudan1.
The “no miracle"argument
Scientists and engineers have developed and built nuclear power plants and Global Positioning Systems (GPS) according to the current state of the art in natural science and engineering. The scientific know-how for these projects included unobservable entities like atoms as well as abstract theories like Einstein’s theory of special relativity. Isn’t it almost compulsory to assume that the underlying entities and theories are correctly representing reality if the power plant delivers exactly the output as calculated in advance and the GPS is able to determine the exact location of a receiving unit on earth with an accuracy of a few Inches? That is exactly the point of the “no miracle” argument, which Ladyman defines as follows:
In particular, it is the success if science in predicting new and surprising phenomena, and in the application to technology that realists argue would be miraculous if the theories were not, in general, correctly identifying the unobservable entities and processes that underlie what we observe. (2002, 213)
The argument sounds convincing, but Hasok Chang, among many other philosophers, has a number of objections against the “argument from success”, as he terms this kind of reasoning. He asks how successful modern science really is and answers, that it is “only clearly successful in comparison to older science, and to various dubious enterprises such as fortune-telling, witchcraft, investment banking, and politics”2 (Chang 2012, 228). He claims that we have no guarantee for any lasting success of scientific endeavors. Most of what the “man in the street” finds impressive about science is not really science anyway, but more often just technology or engineering. But inference from the success of a practical device to the truth of any of the underlying theories is not sound. Which theories should we consider to be proven by the success of the GPS system, if its construction required a myriad of different theories, some of them even contradicting each other? 3
Another question Chang raises is about the nature of scientific explanation. What standard of explanation do we accept as a scientific explanation? A deductive-nomological (D-N) explanation will not be applicable in most cases, because “a D-N explanation of success by appeal to truth would require a law enabling the deduction of success from truth” (Chang 2012, 231). And such a law is hardly to be found in any science. On the other hand, for a causal explanation, the causal power of a theory to achieve something like success would have to be substantiated, which is not possible in most cases. All in all, Chang presents a number of arguments to demonstrate that inference to and explanation of truth from the success of science is hard to establish. But he has a better idea: why don’t we look for explanations other than truth? Obviously, the suggested alternative consists of looking for active realist values, whatever that means.
Why do we have to explain the success of science at all? What is the benefit of explaining success? “Why can’t we simply enjoy success and leave it at that?” asks Chang (2012, 232) and concludes that the truth-based explanations of success by standard realism do not deliver useful advice, “since ultimate truth is not an operative category” (ibid.).
Chang concludes his reasoning against the no-miracle argument with the statement “[t]he success-truth link is fundamentally suspect” (2012, 233). It is not that theories have success, but what scientists or engineers do with them, the epistemic activities they undertake. “And thinking in terms of activities and systems takes us naturally away from thinking in terms of truth” (ibid.). That draws a bow to Chang’s active scientific realism, which is “to let success be, and to do what we can to have more and better success; [...]” (ibid.).
There are many more lines of counterargument to “no-miracle”, which cannot be dealt with in all details in this paper; only the most prominent will be mentioned: the “pessimistic metainduction”.
History of science shows that many successful theories that were current for many years, even centuries, were overthrown later and replaced by new - sometimes directly contradicting - theories. Who can guarantee that our current theories will not face the same fate some day? James Ladyman formulates the consequence to be drawn as follows: “By induction, we have positive reason to expect that our best current theories will be replaced by new theories, according to which some of the central theoretical terms of our best current theories do not refer.” (Ladyman 2002, 237).
This argument goes back to Larry Laudan, who compiled a long list of overthrown theories, which, he contends, could be extended ‘ad nauseam’. (Laudan 1981, 33). And he concludes that “[t]he fact that a theory's central terms refer does not entail that it will be successful; and a theory's success is no warrant for the claim that all or most of its central terms refer.” (1981, 47)
Hasok Chang offers an “optimistic rendition of the pessimistic induction” (2012, 226): Let’s enjoy the fact that we can be successful without knowing the truth, because then “the pessimistic induction can make us happy, if we learn to turn it on its head” (ibid.). If history tells us about scientific theories that were successful over long periods, let’s appreciate that fact! The only thing that is wrong about the pessimistic induction in Chang’s opinion as a strong supporter of pluralism is “the notion that a successful theory should be rejected if another successful theory comes along” (2012, 227).
The two contradicting views at scientific realism, no miracle and pessimistic induction, gave rise to a variety of different flavors and the following chapter will provide a feeling for the bandwidth available in the spectrum of realist theories in philosophy of science.
1 Hasok Chang also credits Thomas Kuhn and Mary Hesse for the basic idea (Chang 2012, 225).
2 Chang could have as well spared the coquetry of comparing investment banking and politics with fortunetelling and witchcraft - or did he want to demonstrate that he is ignorant about the former topics?
3 Hasok Chang also used the GPS system as an example for pluralism in his inaugural speech at Cambridge on October 11th, 2012: "GPS uses satellites we keep in place by Newtonian physics, an atomic clock ruled by quantum mechanics and corrected by special and general relativity; it maps the surface of the round earth on a geostatic grid and gives advice to people on the ground from a flat earth point of view. It all works beautifully." (Internet: http://www.voutube.com/watch?v=zGUsIf9qYw8, accessed on December 30th, 2012).