The Scientific Attitude

by Lee McIntyre

  Might a similar move now be made in defense of the notion of warrant? As we have seen, what gives a scientific theory warrant is not the certainty that it is true, but the fact that it has empirical evidence in its favor that makes it a highly justified choice in light of the evidence. Call this the pragmatic vindication of warranted belief: a scientific theory is warranted if and only if it is at least as well supported by the evidence as any of its empirically equivalent alternatives. If another theory is better, then believe that one. But if not, then it is reasonable to continue to believe in our current theory.36 Warrant comes in degrees; it is not all or nothing. It is rational to believe in a theory that falls short of certainty, as long as it is at least as good or better than its rivals. By Reichenbach’s logic, belief in a scientific theory may not be validated, but it is at least vindicated. Even if it is not provably true (or even more likely to be true), it is still warranted for belief in light of the evidence.37

  At this point, philosophical questions will inevitably arise over what it means to say that a potentially false belief could still be warranted, and whether it is allowable to rely on inductive inferences made from empirical observations when we cannot even be sure that our senses are reliable or that the patterns we have discovered are the proper ones to project into the future. A set of skeptical worries over the foundations of empirical knowledge have bedeviled philosophers from René Descartes to Nelson Goodman.38 But, for the work of most scientists, none of these concerns hold much water. Yes, it could be that an evil genius is deceiving me, and my apparatus is not showing what the data suggest, but without reason to believe that this is true, why should I take it seriously? Yes, it is possible that the empirical evidence I am gathering could support an infinite number of alternative made-up predicates (and theories), but why should I worry about those unless I have reason to think that they would offer a superior explanation of the phenomena I am studying? The philosophical questions here are far from settled, but it is useful to remember that philosophy starts in a different place from science. We do not need to solve every outstanding problem in epistemology for science to go forward. Indeed, it is important to observe that science begins after one moves past Cartesian worries about the reliability of sensory data and accepts the premise that we can learn about the world from sensory experience. While it may be true that in certain situations our sensory abilities will be compromised, it is just too much to expect that science cannot go forward because we might—in any given case—be deceived or insane. Indeed, isn’t this why science relies on the critical scrutiny of others, to provide just such a check on our individual credibility? Or perhaps all of us are compromised. Yet unless there is reason to believe that we really are all asleep, or that the theoretical terms we are using are arbitrary, might scientists not simply elect to move forward with the conviction that the burden of proof here lies with the skeptic?39

  If scientists could learn to live with this sort of pragmatic philosophical defense, I think they would be better for it. No longer would they have to be embarrassed by the open-ended nature of science. No more would they have to feign certainty that their theories were true, when they knew deep down that they were—as they must be—defeasible. This would require a delicate shift not so much in how scientists do their work, but in how they defend it. Positive instances could still count, even if they counted not toward truth but only warrant. Scientists would still have support for their theories, while admitting that they were capable of being overthrown in the face of future evidence. I maintain that this is a completely honest way of proceeding because for the most part scientists already embrace exactly these values of comfort with uncertainty, skepticism, tentativeness, and a desire not to overstate their conclusions.40 Where scientists sometimes go astray is when they are challenged to defend themselves against partisan know-nothings and get goaded into saying something rash that feeds into one of the misconceptions about science. But, if scientists were prepared to give up leaning on the myth of proof and certainty—and substitute instead the idea of warrant—how much easier would their lives be?

  I maintain that philosophers of science might usefully play a role here in articulating an account of science that celebrates its uncertainty, rather than being embarrassed by it, while championing the importance of empirical evidence for warranted belief. If we can do this, it would go a long way toward defanging the denialists, ideologues, conspiracy theorists, and other critics of science. Indeed, what poetic justice that they would suddenly be required to defend the warrant of their beliefs, rather than merely exploit the uncertainties of science! The fact that scientific theories are inherently uncertain does not make nonscientific theories more likely to be true; in fact, it only highlights the epistemic gap between them. Without evidence, there would seem to be no warrant at all for nonscientific beliefs.

  So what is special about science? More than any logical principle, I believe (and will defend in the next chapter) that the most important thing that a scientist must possess in order to navigate these rocky waters is the proper attitude toward empirical evidence. Remember Feynman’s earlier quotation? “If [your theory] disagrees with experiment, it’s wrong.” I maintain that there is something called the “scientific attitude” that is crucial both to doing science and to explaining what is most distinctive about it. It is something that scientists know instinctively and that philosophers should embrace as the best means for articulating why science is a legitimately privileged way of knowing. This is neither a method nor a recipe for doing science, but without it science could not go forward. As I will argue next, the scientific attitude frames and shapes the mindset of scientists as they both build and critique the theories that are at the heart of scientific explanation.

  Notes

  1. It is important to note that Popper took pains throughout his life to defend himself against the charge of naive falsificationism. For more, see chap. 2, note 10.

  2. Popper, “Replies to My Critics,” in The Philosophy of Karl Popper, vol. 14, ed. Paul Schilpp (La Salle, IL: Open Court, 1974), 984.

  3. Some will be tempted to recall here the famous story that, upon learning that his theory had been confirmed, Einstein was asked what he would have done if the experiment had shown otherwise. He replied, “then I would have been sorry for the dear Lord; the theory is correct.” Despite Einstein’s confidence, does anyone really believe that if his prediction had been disproven the theory would have been accepted? Perhaps further tests would have been necessary, or it might have been revealed that Eddington’s measurements were flawed, but if not the theory would at least had to have been modified.

  4. Samir Okasha, Philosophy of Science: A Very Short Introduction (Oxford: Oxford University Press, 2016), 15.

  5. This is the observation that in its revolution around the Sun, Mercury does not precisely retrace its orbit, but instead shifts by a minuscule amount each time. This could only be to the result of some gravitational force.

  6. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1962).

  7. See Hilary Putnam, “The ‘Corroboration’ of Theories,” in The Philosophy of Karl Popper, 223.

  8. Popper uses this phrase many times in The Logic of Scientific Discovery (New York: Basic Books, 1959), including in a section heading called “The Positive Theory of Corroboration: How a Theory May ‘Prove Its Mettle.’ ”

  9. As Tom Nickles writes in “The Problem of Demarcation: History and Future,” in Philosophy of Pseudoscience, ed. M. Pigliucci and M. Boudry (Chicago: University of Chicago Press, 2013): “In [Popper’s] view even a general theory that has passed many severe tests (and is thus highly corroborated) still has probability zero. According to Popper … ‘in an infinite universe … the probability of any (non-tautological) universal law will be zero.’ (Popper’s emphasis)” (107–108).

  10. “Falsifiability as a criterion of demarcation is a purely logical affair and does not depend on our (non-existent) empirical or practical
ability to falsify a statement conclusively” (emphasis in original). Personal communication from Karl Popper to Lee McIntyre, March 26, 1984.

  11. See Popper, Conjectures and Refutation (New York: Harper Torchbooks, 1965), 41, note 8.

  12. For instance, the notions of phlogiston, ether, and caloric.

  13. Richard Feynman, “The Essence of Science in 60 Seconds,” https://www.youtube.com/watch?v=5v8habYTfHU.

  14. This is to say that even if falsification fails as a criterion of demarcation, it may get something right about science, by focusing our attention on the idea that comparing a theory with evidence is crucial to what is special about science. Remember too that, despite those who would seem to deliberately misunderstand him, Kuhn also champions the role of evidence in science.

  15. Another example, which we will explore in chapter 3, is Semmelweis’s discovery of the cause of childbed fever, which was later supported by the germ theory of disease.

  16. It is an interesting question, though, whether what defeated Bode’s law was its lack of theory or its later failed predictions. See McIntyre, “Accommodation, Prediction, and Confirmation,” Perspectives on Science 9, no. 3 (2001): 308–323.

  17. See Alberto Guijosa, “What Is String Theory?” http://www.nuclecu.unam.mx/~alberto/physics/string.html.

  18. The most readable book on this subject is Brian Greene’s The Elegant Universe: Superstrings, Hidden Dimensions, and the Quest for the Ultimate Theory (New York: Norton, 2010).

  19. It is interesting to note here that “string theory” was originally called the “string hypothesis.” See Ethan Siegel, “Why String Theory Is Not a Scientific Theory,” Forbes.com, Dec. 23, 2015, http://www.forbes.com/sites/startswithabang/2015/12/23/why-string-theory-is-not-science/.

  20. Richard Dawid, String Theory and Scientific Method (Cambridge: Cambridge University Press, 2014). Some of this may sound like Kuhn, but it is important to note that Kuhn’s “extra-empirical” criteria were intended to supplement empirical evidence, not substitute for it.

  21. See in particular remarks at the conference by David Gross (a Nobel Prize–winning physicist) who “classified string theory as testable ‘in principle’ and thus perfectly scientific, because the strings are potentially detectable.” David Castelvecchi, “Is String Theory Science?” Nature, Dec. 23, 2015, https://www.scientificamerican.com/article/is-string-theory-science/. For more perspective on the conference see Natalie Wolchover, “Physicists and Philosophers Hold Peace Talks, If Only for Three Days,” Atlantic, Dec. 22, 2015, https://www.theatlantic.com/science/archive/2015/12/physics-philosophy-string-theory/421569/.

  22. See Lee Smolin, The Trouble with Physics: The Rise of String Theory, the Fall of a Science, and What Comes Next (New York: Mariner, 2007), and Peter Woit, Not Even Wrong: The Failure of String Theory and the Search for Unity in Physical Law (New York: Basic, 2007).

  23. Peter Woit, “Is String Theory Even Wrong?” American Scientist (March–April 2002), http://www.americanscientist.org/issues/pub/is-string-theory-even-wrong/.

  24. Theodosius Dobzhansky, “Nothing in Biology Makes Sense except in Light of Evolution,” American Biology Teacher 35, no. 3 (March 1973): 125–129.

  25. See also Larry Laudan, Progress and Its Problems: Towards a Theory of Scientific Growth (Berkeley: University of California Press, 1978).

  26. For an account of how to explain the eye, and other “organs of extreme perfection,” see Richard Dawkins, Climbing Mount Improbable (New York: Norton, 2016). For an account of the relatively recent discovery of a fossilized fish with shoulders, elbows, legs, a neck, and wrists—named Tiktaalik—see Joe Palca, “The Human Edge: Finding Our Inner Fish,” NPR, http://www.npr.org/2010/07/05/127937070/the-human-edge-finding-our-inner-fish.

  27. See the discussion of intelligent design in chapter 8.

  28. There is another virtue for this point of view as well, which is that it allows us to dispute the idea that scientific theories are weak unless they can be proven. While it is true that you cannot prove an empirical statement, this does not mean that we are justified in believing whatever we want because it “might” be true. The Flat Earth hypothesis “might” be true, but where is the evidence? Even if one guessed correctly (like the proverbial stopped clock that is right twice a day), this is not science. To say that we are not warranted in believing “anything” we want does not mean that we are warranted in believing “nothing.” Science is about having a theory that is tested against the evidence, which is what gives us justification for believing in it.

  29. This is usually called pessimistic induction. Note, though, that this should be carefully distinguished from something called counterinduction, which is the idea that because something has happened in the past, it is bound to fail in the future. Consider the inductivist dice player who has rolled three sevens in a row and says “I’m bound to do it again!” This is not valid, but neither is his counterinductivist friend who says “no you’ve used up all your luck; it’s nearly impossible you’d do it again.” If the dice are fair, the odds are the same for each throw, no matter what has happened in the past. By contrast, pessimistic induction is merely the observation that, given the way induction works, it is nearly inconceivable to think that our limited experience will lead us to discover a true theory; thus, virtually every scientific theory will someday be overthrown.

  30. One of the most intriguing essays on this subject was written by the eminent pragmatist philosopher Charles S. Peirce. In “The Scientific Attitude and Fallibilism,” Peirce links the sorts of “virtuous characteristics” we might associate with someone who has the scientific attitude with respect for the idea that empirical knowledge must always be incomplete. The person who has the “scientific attitude” must never block the way of inquiry by thinking that their knowledge is complete, because in a field as open-ended as science there is always more to learn. See https://www.textlog.de/4232.html.

  31. See D. H. Mellor, “The Warrant of Induction,” in Matters of Metaphysics (Cambridge: Cambridge University Press, 1991), https://www.repository.cam.ac.uk/bitstream/handle/1810/3475/InauguralText.html?sequence=5.

  32. One might wish, however, to allow that it is perfectly acceptable for scientists to continue to use the word “true”—without having always to insert the fallibilist asterisk that future evidence may show their belief to be false—as long as one remembers that in every discussion of empirical knowledge this is surely the case.

  33. David Hume, A Treatise of Human Nature (London, 1738), Book VII.

  34. Although Reichenbach’s own work is enlightening, the most approachable discussion of this idea can be found in Wesley Salmon, “Hans Reichenbach’s Vindication of Induction,” Erkenntnis 35, no. 1 (July 1991): 99–122.

  35. It is crucial to note that the word “vindicated” is used here, as opposed to saying that induction is “verified.”

  36. The principle of conservativism in changing one’s theory is a deeply held norm in science. As we have seen, Popper held that—other things being equal—we ought to give heavy deference to those theories that have survived longer or are better “entrenched.” Quine too has held (on practical grounds) that we may legitimately prefer those ideas that do least violence to our other existing beliefs.

  37. There are still several problems to be worked out here. For one thing, even if pragmatic vindication may provide a bulwark against certainty, what can it say about whether warranted beliefs are more probable in light of the evidence? Remember that probability is undermined by the inductive observation that we cannot be sure that the piece of the world we have sampled so far is representative of the rest of it. But couldn’t we say that, if it is not, then no belief is warranted, but if it is then our belief is at least as well warranted as any other? For further details, see my “A Pragmatic Vindication of Warranted Belief” (work in progress).

  38. Descartes worried that our senses were potentially unreliable owing to the possibility that we were perpetually dreaming or could be duped by
an evil genius. Goodman—through his famous “New Riddle of Induction”—worried that we could not even be sure that the predicates we were using (like “green” and “blue”) were superior to fabricated ones (like “grue” and “bleen”) that may be equally well-supported by empirical evidence. See Nelson Goodman, Fact, Fiction, and Forecast (Cambridge, MA: Harvard University Press, 1955).

  39. If the point of defending the warrant of induction is not just to show that scientists know X, but to show that they must know that they know X, isn’t this just to give in to the demand for certainty that is based on one of the misconceptions about science?

  40. See, e.g., the letter signed by 255 members of the US National Academy of Sciences that I mentioned on the first page of this book.