The Scientific Attitude

by Lee McIntyre

  3    The Importance of the Scientific Attitude

  Many thinkers have tried to identify what is special about science based on its allegedly unique methodology. This approach has been criticized because it has been shown that many scientists do not actually follow the steps that philosophers of science have used to justify their work.1 This does not mean that there is nothing important about what scientists do that might have great bearing on science’s privileged epistemic status. It is just that perhaps we should be looking less at the method by which science is justified and more at the attitude that its practitioners have in mind while they are doing it.

  As we saw in chapter 1, there is no recipe for doing science. Likewise, there may be no logical means of distinguishing between the type of reasoning that scientists use to think about the empirical world and that which is used elsewhere. Nonscientists can certainly be rigorous and careful in their consideration of evidence, and scientists can occasionally rely on subjective, social, and other criteria to decide between theories. There is, however, an important feature of scientific work that is seldom talked about in philosophical circles, which is the attitude that guides scientific inquiry. Even if scientists cannot always rely on a set of rules to follow, it is clear from the history of science that they must rely on something. An ethos. A spirit of inquiry. A belief system that tells them that the answer to empirical questions will be found not in deference to authority or ideological commitment—or sometimes even in reason—but in the evidence they gather about the subject matter under investigation. Such a creed, I maintain, is the best way of understanding what is special about science. I will call this the scientific attitude.

  The scientific attitude can be summed up in a commitment to two principles:

  (1)  We care about empirical evidence.

  (2)  We are willing to change our theories in light of new evidence.

  This, of course, does not rule out the idea that other factors may sometimes weigh in. As Thomas Kuhn’s work demonstrates, even when we care about evidence it can underdetermine theory choice, which opens the door to extra-empirical considerations. What must be ruled out, however, is wishful thinking and dishonesty. In the pithiest version of his ongoing attempt to capture what is distinctive about science, Richard Feynman tells us that “science is what we do to keep from lying to ourselves.”2 Perhaps there is no better rendering of the proper mindset behind the scientific attitude than this.

  Such talk about the attitudes and values of science may be dismissed by some as too vague and unrigorous to be helpful, so let me now be more specific about what this approach might entail. What does it mean to care about evidence? Maybe the best way to think about this is to examine what it means not to care about evidence. If one does not care about evidence, one is resistant to new ideas. One is dogmatic. Such a person might hold onto their beliefs no matter what the evidence shows. When the scientific attitude says that we must “care about evidence” the idea is that we must be earnestly willing to seek out and consider evidence that may have a bearing on the grounds for our beliefs. In some cases this will improve our justification, but in others it may undermine it. Scientists must be open to either.

  To care about evidence is to be willing to test our theory against a reality that might refute it. It is to commit to hold a belief not because it makes us feel good, seems right, or even coheres with other things that we believe, but because it fits with the data of our experience. While there is a vast literature in the philosophy of science that shows just how difficult it is sometimes to decide between theories on this basis—necessitating all sorts of other considerations such as simplicity, fruitfulness, and coherence—this does not change the underlying creed of science: where it is available, evidence should drive scientific theory choice.3

  Of course, with some topics, we might not care about evidence because it is irrelevant. If the subject is math or logic, then evidence will not make a difference because anything at issue can be resolved through reason. But when an empirical subject is under investigation, this kind of rejection of evidence is anathema to rigorous inquiry. In science, we seek knowledge from experience in order to see what the world is like. Caring about evidence is fundamental because it is the only way to shape our knowledge closer and closer to the reality that we seek to know.

  One might imagine here a list of traits. The person with a good scientific attitude is humble, earnest, open-minded, intellectually honest, curious, and self-critical. The danger here, however, is that we cannot simply make the scientific attitude a matter of individual psychology, nor can we leave it to the judgment of the individual whether he or she possesses these traits. For one thing, what to do about the denialist or pseudoscientist, who might say that they care about evidence—or even actually believe it—when it is obvious to the rest of us that they do not? Such a person may simply be lying to us, but they may also be lying to themselves. If the scientific attitude were just a matter of how one feels about whether one cares about evidence, it would not be possible to differentiate between the genuinely earnest person who is searching for a way to test their beliefs against experience versus the ideologues who are deluded into thinking that they care about evidence merely because they cherry pick facts that confirm their prior beliefs. Instead, the scientific attitude must be measured by our actions, and these actions are best judged not by the individual, but by the larger community of scientists, who share the scientific attitude as a guiding ethos.4 To care about evidence is thus to act in accordance with a well-vetted set of practices that have been sanctioned by the scientific community because they have historically led to well-justified beliefs.

  This is not to say that the process of science is perfect. Even when we fully embrace it, the scientific attitude probably cannot eliminate all of the denialists and pseudoscientists who say that they have it, even when they do not. Whether they are fooling themselves or others, it is sometimes hard to tell.5 Likewise, there may be scientific researchers who sometimes get too close to their own theories and refuse to believe what the data tell them.6 Where is the line between these two camps? Even if the scientific attitude cannot draw a firm logical or methodological division between science and its imposters, it can at least expose a basic lacuna in values that are illuminated by how we behave in the face of contravening evidence.

  What is evidence? It is probably impossible to define all of the different things that may count as scientific evidence. Statistical, qualitative, or even historical evidence may exist across different empirical endeavors. Evidence is the data we get from experience that affects our rational degree of belief in a theory. Sometimes this data is quantitative and can be measured directly. Other times it is diffuse and must be interpreted. Either way, scientists should agree that evidence is crucial in choosing or modifying a scientific theory.

  There are, however, many competing theories of what it means for scientists to use evidence in a rational way. As Peter Achinstein writes:

  Scientists frequently disagree about whether, or to what extent, some set of data or observational results constitute evidence for a scientific hypothesis. Disagreements may be over empirical matters, such as whether the data or observational results are correct, or whether other relevant empirical information is being ignored. But conflicts also arise because scientists are employing incompatible concepts of evidence.7

  Entire books have been written about these different concepts of evidence, and their various strengths and weaknesses in explaining how the facts of experience either do or do not lend support to a scientific theory.8 It might shock those outside the philosophy of probability and statistics to learn that there are competing accounts of what it is appropriate to infer from one and the same piece of evidence. There are furious debates, for instance, over the “subjectivist” approach to probability favored by Bayesians versus the “frequentist” approach offered by Deborah Mayo and others.9 The scientific attitude may be thought of as compatible with many different co
ncepts of evidence. No matter your theory of the proper way to use evidence, the scientific attitude toward evidence is one where you are committed to the idea that evidence is paramount in making up your mind about whether a theory is worthy of belief.

  Of course, the best way to appreciate the importance of the scientific attitude is to see it in action, and I’ll soon give a few examples. First, however, I would like to address two possible misconceptions. First, the scientific attitude is not meant to be a solution to the problem of demarcation.10 The goal of the demarcation project is to find a logical criterion by which one can sort all and only science into one camp, and everything else in the other. That is a tall order and, as we have seen, virtually every attempt to do this has failed. This leaves science open to misunderstanding and criticism by those who do not fully appreciate what it is about. The goal of identifying the scientific attitude as an essential feature of science is not to wall it off from other disciplines but to show that unless those who make empirical claims are willing to follow the rigorous standards that define scientific reasoning, they will fall short of the best way the human mind has ever devised of coming to know the empirical world.

  A second possible misconception may involve my intentions. I am not here attempting to give a descriptive account of what scientists actually do because, in any given lab on any given day, one’s commitment to the scientific attitude may be in flux. Scientists may occasionally violate the norms of science and then later, one hopes, come back into line.11 Instead, I offer the scientific attitude as a normative ideal, by which we may judge whether some individual scientist or entire field of inquiry is living up to the values of science. As we’ve seen, science is not based on some formula, as the scientific method promised. Neither is it strictly a matter of some all-or-nothing judgment about logic or methodology. Science is defined by a set of practices that are embedded in the values upheld by the people who perform it.

  This is not to say that science is solely justified (or not) by what it does. Practice is important but it is not the only thing that matters in judging science. I say this because one strain of argument against the methodological approach to the philosophy of science over the last few decades has been that, because science does not always follow the precepts that the logic of science would dictate, science must be no better or worse than any other form of inquiry. I believe this conclusion to be misguided. Although I myself eschew a methodological defense of science, my approach too is rooted in the idea of rational justification. I may not agree with the traditional idea that in order to defend science one must draw a rigid distinction between facts and values, but I do not believe that science is hopelessly subjective either. Although objectivity is important, values play a role by guiding our practice and keeping us on track. This is to say that even though the practice of science may sometimes fall short, it is still possible to justify science as a whole based on the goals of its aspiration.

  To recognize the role of practice in understanding science does not diminish the importance of its ideals. Though some may cheat or do sloppy work, this does not mean that science is unjustifiable. Just as it does not undermine the logic of science to say that individual scientists are sometimes irrational, it does not undermine the values of science to point out that some practitioners have occasionally betrayed the scientific attitude. Indeed, this is why it is important to champion the role of group scrutiny in judging scientific work. The standards of science are upheld not just by the individual, but by the community of scientists, who have developed a set of tools to keep it honest. This is why the scientific attitude is a normative rather than a descriptive thesis. Humans sometimes cheat on an ideal even when they believe in it. In such a case, it is up to others to offer correction. And this is exactly what the scientific attitude allows one to do. What makes science distinctive is not merely what it does, but what it aims to do. Despite any mistakes made by the individual, it is the ethos of science that affords it such great epistemic authority.12

  Two Examples of the Scientific Attitude

  I promised at the outset of this book that one could learn the most about science by looking not just at its successes but also at its failures. I also promised not to use examples exclusively from the history of physics and astronomy. Given that, I will now pursue one example that illustrates the virtues of the scientific attitude drawn from medicine, followed by a “failed” example from chemistry (cold fusion), which demonstrates what can happen when the scientific attitude is compromised.13

  For this strategy to be credible, it must square with the belief that it would have been easy to find other examples from the history of physics and astronomy that would also demonstrate the merits of the scientific attitude. I think this is not an unreasonable assumption. One could, for instance, turn to Newton’s theory of gravity or, better yet, to Einstein’s general theory of relativity. Given the head start we have from Popper’s reliance on this example, this might serve well to illustrate the power of having the proper mindset when one is testing a theory. But I will leave it to the reader to imagine what Popper might have said about the scientific attitude of Einstein. Instead I will here share one of my own favorite examples from the history of science: Semmelweis’s theory of childbed fever. This example was made famous within the philosophy of science by Carl Hempel, who used it in his 1966 book Philosophy of Natural Science to illustrate the virtues of scientific explanation.14 For my own purposes, I will take pains to highlight the way that Semmelweis’s theory demonstrates the scientific attitude, rather than the logical empiricist account of science within which Hempel frames it. This will also tie in quite well with what I have to say about how the scientific attitude transformed modern medicine in chapter 6.

  Given the unquestionably scientific status of modern medicine, it is hard to believe that more than two hundred years after the start of the scientific revolution in the seventeenth century, medical care was still in the dark ages. As late as 1840, medical care still had not enjoyed the discovery of anesthesia (1846), the germ theory of disease (1850s), or antiseptic surgery (1867). One problem is that even when discoveries were made, there were few agreed-upon routes for disseminating the information or overcoming the objections of skeptics.15 Experimental methods took a back seat to intuition and tradition. This makes it all the more remarkable that in 1846, at the Vienna General Hospital, we see one of the greatest examples of the scientific attitude in full flower.

  Ignaz Semmelweis was a lowly assistant physician in the world’s largest maternity clinic, which was divided into two wards. In Ward 1, childbed fever (also known as puerperal fever) was rampant and the mortality rate was as high as 29 percent; in the adjacent Ward 2, the rate was only 3 percent.16 Another piece of relevant information was that women who delivered their babies at home or even on the way to the hospital in a “street birth” had a much lower incidence of childbed fever. What was so different about Ward 1? Various hypotheses were offered. One was that Ward 1 was overcrowded. When Semmelweis counted up the patients, however, he noted that the overcrowding was in fact much worse in Ward 2 (perhaps because of all those women avoiding the notorious Ward 1). It was then noted that, because of the physical layout in Ward 1, the priest who was summoned to give last rites to women who were dying of childbed fever was required to pass by many other beds—all while ringing a bell—which might put great fear into the other women and perhaps increase their chances of contracting childbed fever. In Ward 2, the priest had direct access to the sickroom. Semmelweis decided to try an experiment in which he asked the priest to take a different, silent, route to the sickroom in Ward 1, but the mortality rate from childbed fever stayed the same.

  Other tests involving whether the women lay on their sides or their backs while giving birth were similarly fruitless. Finally it was noted that one of the main differences was that in Ward 1 the deliveries were handled by medical students, whereas in Ward 2 they were performed by midwives. Were the medical students giving rougher examinations? Af
ter the medical students and the midwives changed places, the mortality rates followed the medical students, but still no one knew why. After instructing the medical students to use gentler techniques, the mortality rate still did not improve.

  Eventually, enlightenment came in 1847 when one of Semmelweis’s colleagues received a puncture wound during an autopsy on a woman with childbed fever, and died of an illness that appeared to have the same symptoms.17 Could childbed fever be contracted by someone other than pregnant women? Semmelweis realized that there was a difference in where the medical students were before they came to the maternity ward; they came directly from performing autopsies, with unwashed hands and instruments (remember that this was before antisepsis and the germ theory of disease), straight to the maternity ward, leading to the hypothesis that childbed fever may have to do with the transfer of “cadaveric matter” to the pregnant women. As a test, Semmelweis ordered the medical students to wash their hands in chlorinated water before performing their deliveries. The mortality rate plummeted. He now had an explanation not only for why the incidence of childbed fever was so much greater in Ward 1, but also for why “street births” saw such a low incidence of childbed fever. Eventually, Semmelweis was forced to broaden his hypothesis to include the idea that childbed fever could also be transferred from putrified living tissue, after he and his colleagues examined a woman with cervical cancer, and then a dozen other women in succession, eleven of whom died of childbed fever.18

  Use of the scientific attitude in this example is obvious. Semmelweis did not assume that he already knew the answer to the question of what caused childbed fever; he examined the similarities and differences between the two wards, then learned what he could through observation and controlled experiment. He came up with various hypotheses, then began to test them one by one. When a hypothesis flamed out, he moved on to the next one, leaving himself open to learning new information along the way. Finally, when he found the answer—and later broadened it—he changed his ideas based on the new data.