Thinking About Science: Essays on the Nature of Science
As published in Skeptical Inquirer magazine
2003-2008
Massimo Pigliucci
Copyright 2009 by Massimo Pigliucci
Smashwords Edition
cover drawings, clockwise: Bacon, Descartes, Newton, and Galileo
some of the architects of modern science
Introduction
What follows is a collection of essays on the nature of science and its sometimes fuzzy distinction from pseudoscience. These essays were originally published as a regular column in the magazine Skeptical Inquirer, one of the best sources of information available on controversies surrounding pseudoscience. The column, entitled “Thinking About Science” (just like this collection) is still going at the time of this writing (early 2012), and I refer the interested reader to its future installments to follow the evolution of my own thoughts about how science works.
Science is a human activity, and as such it is hampered by all the typical human frailties. Scientists are no less interested than anyone else in glory, money, and sex, not necessarily in that order. Yet, as philosophers of science have argued for some time now, science as a social activity manages to be remarkably objective and truth-augmenting. Scientists may blunder, as in the infamous case of “Piltdown Man” recalled in one of these columns, but in the long run they seem to get it mostly right (after all, it was scientists, not, say, creationists, who uncovered the Piltdown forgery). This is very different from the situation with pseudoscience, where astrologers and paranormalists seem to be perennially stuck in the same place, always making the same arguments, and chronically short of empirical evidence to back them up.
These essays look at science from both the point of view of a scientist and that of a philosopher. This reflects my own dual background, with original training in evolutionary biology and the later addition of philosophy of science. The two disciplines have always had a difficult relationship, ever since science originated as natural philosophy and became independent in the 17th and 18th centuries. Scientists of the time, like Galileo and Newton, thought of themselves at least in part as philosophers, and figures that we count today as philosophers, like Descartes and Bacon, thought of themselves as scientists. But today’s academy all too often relishes the division, with scientists like physicist Steven Weinberg brazenly writing essays entitled “Against Philosophy,” and philosophers like Paul Feyerabend calling for “a formal separation between science and state” to guard society from the evils of science. My columns are written instead in the spirit that science and philosophy have much to gain from each other, with philosophy providing a broad view of how science works, and even criticism of specific scientific enterprises, and science returning the favor by informing philosophical debates with the best understanding of the facts of the universe that we can achieve at any particular moment.
I hope the reader will enjoy the quest as much as I do, and that readers will come to value honest human intellectual endeavor both for its own sake and for the good it can do to the human condition. As David Hume aptly put it, “What a peculiar privilege has this little agitation of the brain which we call 'thought.'”
—Massimo Pigliucci
Manhattan, January 2012
Table of Contents
(January/February 2003)
One of the most common fallacies committed by believers in the paranormal is what in philosophy is known by the Latin name of post hoc, ergo propter hoc, which loosely translates to ‘after this, therefore because of this.’ Surely you have heard some version of it: “I dreamed of my brother the other night, and the following morning he called me, though he rarely does.” The implication here is that there is some causal connection between the dream and the phone call, that one happened because of the other. We all know what is wrong with this argument: a correlation between two events does not constitute good enough evidence of a causal connection between them. In the case of the dream’s precognition, we probably dream of our relatives often enough, and most often the dream is not followed by their call; yet, because of an innate tendency of the human brain to remember hits and forget misses, we pay attention to the exceptions and charge them with special meaning.
But the good skeptic could go further and ask herself what exactly do we mean by causation to begin with. If a correlation is not the hallmark of a causal relationship, what is? The modern study of causation started with the Italian physicist Galileo Galilei, who viewed causes as a set of necessary and sufficient conditions for a given effect. According to Galilei, the dream can be considered a cause of the call only if every time the subject dreams of his brother, the following morning the brother actually does call. The problem with this idea is that it is too restrictive: many phenomena have multiple causes, a subset of which may be sufficient to generate the effect. The brother could call for other reasons than the dream, notwithstanding a true causal connection between dreaming and calling. Or, the dream may be causing the brother to have the impulse to call, but he can’t do it because he is at a vacation spot where there are no phones in sight (as hard as this may seem to believe).
Scottish skeptic philosopher David Hume made the next important contribution to our understanding of causality, one that many philosophers (and a few scientists) are still grappling with. Hume argued that we never actually have any evidence that causal connections are real, we only have perceptions of the likely association between what we call a cause and an effect. Here Hume was being a good empiricist, something that a skeptic ought to appreciate. For him, talk of “causes” sounded as strange as talking of action at a distance, which in pre-Newtonian times was an exercise for mystics, not scientists. So Hume decided to settle on a very pragmatic concept of causality. He suggested that we are justified in talking about causes and effects if three conditions hold: 1) the first event (say, the dream) precedes the second one (say, your brother’s call); 2) the two events are contiguous in time, i.e., your brother called the morning after the dream, not a month or a year later; 3) there is a constant conjunction between the two events, i.e., every time you dream of your brother, he will call. As the reader will have noticed, however, the latter clause is very similar to Galilei’s idea of necessary and sufficient condition, and will not actually help the scientist in real situations.
John Stuart Mill, well known as a utilitarian, proposed a concept that is at the basis of much modern experimental science and, hence, of skeptical investigations. Mill argued that causality simply cannot be demonstrated without experimentation. Essentially, Mill said that in order to establish a causal connection between two phenomena we have to be able to do experiments that allow us to manipulate the conditions so that only one factor at a time is allowed to change. A series of these experiments will eventually pinpoint the cause(s) of certain effects.
While Mill’s idea has been of fundamental importance for modern science, the problem with it is that it imposes on the investigators logistic requirements that are often too restrictive. What if it is not possible to control all variables but one during an inquiry? Carefully controlled manipulative experiments are possible only in certain fields and under very taxing conditions. Should we then give up the concept of causality for the much larger number of instances in which such manipulations are not possible, unethical, or simply too expensive? That would be problematic because, for example, we could not conclude that smoking causes cancer. It is simply not possible to do the right experiment, especially with human beings: there are too many variables, not to mention deep ethical issues.
What then? One of the most modern conceptions of causality is the so-called probabilistic one. According to probabilistic causality we can reasonably infer that, say, cancer is caused by smoking if the probability of getting cancer is measurably higher when the subjects smoke than when they don’t. Other factors here are taken into consideration statistically, not necessarily by experimental manipulation. That is, one carries out the investigation taking care of sampling individuals with different socio-economic backgrounds, diets, exercise habits, and genetic constitution. If, when these other variables are kept in check statistically, we still detect an increase in the likelihood of getting cancer in the smokers compared to the non smokers, we are justified in tentatively accepting a causal connection.
Notice, however, that while the probabilistic account of causality is indeed very powerful in practice, conceptually it brings us back towards Hume: the only reason we are talking about causality is because we perceive a series of regularities, not because we know that actual causes are at play. So, in science as in skeptical investigations we might have to admit that the most we can get is a certain probability of being right. Definitive truth is a chimera that does not belong to science after all.
Further reading: David Hume (1739-40/2007) A Treatise of Human Nature. NuVision.
The strange case of cathode rays and what counts for evidence
(March/April 2003)
In 1859, the year Darwin published The Origin of Species, German physicist Julius Plücker discovered what for some time were referred to as “cathode rays.” Plücker used a glass tube filled with air and containing a positive and a negative electrode. When he lowered the air pressure inside the tube to 0.001 mm of mercury and connected a source of electric potential to the positive electrode (the anode), the region of the glass near the negative electrode (the cathode) started glowing with green phosphorescence. Plücker’s conclusion was that something was being emitted by the cathode, and one of his students, Johan Wilhelm Hittorf, demonstrated in 1869 that a solid placed between the cathode and the walls casts a shadow: the mysterious cathode rays were traveling in straight lines. This is the beginning of a little known story that philosopher Peter Achinstein recounts to gain some surprising insight into what, exactly, counts for “evidence” in science. Let us follow Achinstein’s reconstruction of the events, as well as his interpretation of what this tells us about how science works.
Physicists soon split themselves into two camps providing different theories concerning the nature of cathode rays: on the one hand they were thought to be atoms or gas molecules inside the tube, that had become negatively charged. An alternative hypothesis was that they were not particles at all, but rather a type of wave moving through ether. Heinrich Hertz was among the physicists proposing the second scenario, and in 1883 he conducted a crucial experiment to demonstrate that cathode rays were not in fact charged particles. Hertz generated cathode rays inside an apparatus that included an electrometer to measure the electrical charge of the rays. He found that the electrometer did not register any charge, which led him to state that: “As far as the accuracy of the experiment allows, we can conclude with certainty that no electrostatic effect due to the cathode rays can be perceived.”
Notice that Hertz’s conclusion is very carefully stated: he did not say that cathode rays are not electrically charged, only that given the characteristics of his apparatus, there was no reason to believe so. In other words, his experiment was evidence for the hypothesis that cathode rays are not electrically charged.
In 1897, the British physicist J.J. Thomson repeated Hertz’s experiment, obtaining the same results. However, he then went further and lowered the pressure of the air inside the apparatus more than it was technically possible in Hertz’s days. Sure enough, Thomson did detect a clear deviation of cathode rays toward a positively charged plate, which he took as conclusive evidence not only that the rays are electrically charged, but that they are negatively so (because negative charges are attracted to positive ones). Indeed, today we refer to cathode rays as electrons, which are sub-atomic particles that are negatively charged.
We could end the story here, with another triumph of progressive science. But an important question lingers: when exactly do we take something to be evidence for a particular conclusion? This is obviously of crucial importance to skeptical investigations as well and it is worth pondering a bit. First off, notice that nobody here was disputing the facts. Thomson never believed Hertz’s conclusions (which is why he carried out his experiments), but he did not question his results. Indeed, Thomson used the same sort of conditions employed by Hertz to confirm his results before showing why they were misleading.
There are at least three ways of thinking about the Hertz-Thomson story that reflect three concepts of evidence of which we need to be aware. First, one could say that Hertz’s 1883 results were in fact strong evidence that cathode rays are not charged, but that the latter conclusion has turned out to be false in the end nonetheless. Second, we could claim that Hertz’s results were strong evidence for the neutrality of cathode rays between 1883 and 1897, but no longer so after Thomson published his experiments. Finally, one might advance the proposition that Hertz’s results were in fact never strong evidence that cathode rays are electrically neutral.
If we accept the first answer, Achinstein points out, we are saying that evidence is a matter of epistemic status, i.e., relative to the knowledge available in 1883, the first set of results were indeed good evidence to conclude for the neutrality of the rays. The fact that Hertz turned out to be wrong is not a problem, since one can be perfectly justified in believing a hypothesis even though it may not be true. Anybody in the same epistemic situation (i.e., with the same knowledge and understanding) of Hertz in 1883 would have been justified in accepting Hertz’s conclusions.
In the second case, we are accepting a subjective view of evidence, according to Achinstein. What we are saying is that evidence is relative to a particular person or group of persons: it was sensible to accept Hertz’s conclusions in 1883, but no longer in 1897. There is a subtle difference with the first case: subjective evidence requires that somebody actually does believe that the evidence in question supports a certain theory. The epistemic concept of evidence discussed above, on the other hand, is valid even without anybody actually believing that the evidence points to a certain conclusion: in the case of the epistemic situation if anybody had access to the knowledge that Hertz had at the time, she would have been justified in agreeing with his conclusions.
The last possibility considered by Achinstein is that one could say that evidence has to be veridical (i.e., has to carry truth-value), so that Hertz’s experiments were not good evidence for his conclusions simply because later research showed that they were flawed in an important respect.
Most scientists (and I suspect, many skeptics) are much more interested in the third (veridical) sense of evidence, while historians and philosophers are fascinated by the first two. The problem is that by the very nature of the scientific enterprise we can never be sure that some evidence is veridical in respect to a given theory: it could always happen that the next day somebody else publishes an experiment showing a flaw in the procedures followed thus far in order to demonstrate a certain conclusion. Just because we live in the 21st century, that doesn’t mean we have an a-historical, God’s eye view of things, so that we know that Thomson was definitely correct and that’s the end of the matter. All we can say is that, given our epistemic situation, it seems highly likely that Thomson did in fact reach the final conclusion on the nature of electrons. But we could be wrong.
Further reading: Peter Achinstein (2001) The Book of Evidence, Oxford University Press.
(May/June 2003)
It turns out that Sherlock Holmes never says “Elementary, dear Watson!” in any of Sir Arthur Conan Doyle’s original stories. Apparently, the phrase was introduced with the first theatrical representations of the fictional detective, with the consent of Doyle in reply to a letter by the actor William Gillette, who originally played Holmes on the stage. More importantly, as far as skeptics are concerned, the type of reasoning employed by Sherlock Holmes was not “deduction” at all, though Doyle referred to it that way. In fact, in all his adventures Holmes uses what in philosophy of science is referred to as “induction to the best inference.” It is important for anyone interested in critical thinking and science to understand the difference between deduction and induction, and the strengths and pitfalls of the two approaches.
Deduction is a form of reasoning with which one can go from general premises to specific predictions. It was first formalized in Western philosophy by Aristotle, who described a simple variety of it known as syllogism. From the two premises “1-All men are mortal” and “2-Socrates is a man” one can deduce the conclusion that “3-Socrates is mortal” (a conclusion that, unfortunately, was given final proof when Socrates was condemned to die by the Athenian state for teaching critical thinking to the local youth). Deduction is particularly useful in mathematics and to formulate predictions (hypotheses) based on general scientific theories. Notice, however, that while deduction is truth-conservative, it is not truth-ampliative. What this means is that, if the premises are true (and the deductive sequence is formally correct), then the conclusion is guaranteed to be true (i.e., the truth of the premises is conserved). However, a deductive reasoning does not augment our knowledge of the world (it is not ampliative), it simply makes explicit what is already contained in the premises. While the latter is often a valuable enterprise, it clearly is not the sort of thing that would help Holmes solve a crime, or a scientist to make generalizations about the world.
For the latter, we need a different kind of reasoning, often referred to as induction. Induction is, in some sense, the opposite of deduction: it seeks to go from particular facts to general statements, i.e. it attempts to be ampliative of truth. However, the price to pay for that is that induction is not necessarily truth-preserving: we can generalize from specifics, but the generalization may turn out to be wrong. Bertrand Russell, with his characteristically dry humor, nicely explained this: he imagined an “inductivist turkey” which is brought to a farm and fed regularly every morning at the same time. The turkey wishes to make predictions about his future, but-being a good inductivist-realizes that he needs a large sample of data before being able to do so confidently. So he collects data on when he is fed, how often, and with what. After 364 days, the turkey feels confident that he has enough particular examples to draw a general conclusion: he will be fed every morning at the same time, with the same amount and type of food. Sadly, the following day was Thanksgiving, and the turkey was instead slaughtered and brought to the farmer’s table. Such are the perils of the non truth-preserving character of induction.
And yet, induction is a quintessential component of scientific investigations. Indeed, the argument can be made that even deductive reasoning does, ultimately, rely on induction: after all, one has to get the premises from somewhere, and this is usually done by induction. For example, while we can establish that “Socrates is a man” by direct observation, the premise that “All men are mortal” is neither the result of simple observation not a necessary piece of logical inference. The only reason we think that all men are mortal is because every man we have seen so far eventually died. But-for all we know-we may be in the same position as Russell’s turkey and drawing unwarranted conclusions from insufficient evidence (notice, incidentally, that even “simple observations” are made possible by the inductive inference that our senses-usually-are reliable).
Skeptic philosopher David Hume was the first one to recognize this “problem of induction,” i.e. that all our knowledge, ultimately, is based on inductive generalizations, even though we have no independent validation of induction itself. One might think to get around the problem of induction by taking a pragmatic approach and stating that we use induction because it has worked well so far. Sure, but that in itself is a piece of inductive reasoning. You see the problem.
Famously, Karl Popper thought he had solved the problem of induction by stating that science does not seek to prove its theories, only to disprove them. This is the well-known principle of falsification. I will eventually devote a whole column to it, given its importance in skepticism and science, but for now let us say that most philosophers of science think that simple falsificationism cannot work for the reason that we can never test hypotheses in isolation. Every time we allegedly attempt to falsify a hypothesis, in reality we are also testing several corollaries and assumptions, and at least some of them will have been derived via induction.
Having said this, I wouldn’t throw away science or skeptical investigation just because induction cannot be justified from within science. There are solid philosophical reasons to think that science is on firm grounds as a method of inquiry, and there are plenty of results that assure us that it works. But the problem of induction is a good reminder that even skeptics and scientists do have to accept certain philosophical assumptions to do their work. If such assumptions make you a bit uncomfortable, this just adds to the fun.
Further reading: David Hume (1748 / 1956) Enquiry Concerning Human Understanding. Gateway.
(July/August 2003)
The idea of a thought experiment may seem like a perfect example of philosophical oxymoron: we usually think of experiments as things that are done manually, in practice, with the use of some measuring tools. So, how can one carry out a thought experiment, i.e. one that requires only sitting down and thinking really hard about the possible outcomes of a certain (hypothetical) situation?
And yet, thought experiments are the bread and butter not only of philosophy, but of science as well. The trick is to understand how they work and learn to distinguish good from bad thought experiments (just as there are good and bad empirical experiments). Let’s start by dispelling the potential skepticism of the reader while considering a clear example of a good thought experiment: Galileo’s refutation of the Aristotelian theory of gravity.
Aristotle held (in agreement with common, but fallacious, intuition) that heavier bodies fall faster than lighter ones. If we label a (imaginary) heavy body as H and its light counterpart as L, we then have that V(H) > V(L) (where V stands for velocity). Galileo invited us to consider a situation in which the two bodies are now connected to each other, for example with a rope. Since we now have a combination H+L, the new body should fall faster than either of its two components (because its weight is higher): V(H+L) > V(H). But, Galileo observed, the new body also has to fall at a slower pace, because of the dragging effect of L, so V(H+L) < V(H). Combining the two results one gets a contradiction, since the compound object is expected to both be faster and slower than the heavy object alone. Since the Aristotelian theory has led us into a contradiction it must be rejected. Furthermore, a moment’s reflection shows us what the solution is: V(H) = V(L), as physicists have indeed accepted (and then experimentally demonstrated, for example during the Apollo missions on the Moon) to be the case.
Mind blowing, isn’t it? Galileo, though he is popularly known as a real experimenter, actually made some of his most valuable contributions to science by simply thinking about certain problems! And he was certainly not the only one (or even the first). Other examples include Lucretius’ (99/94-55/51 BCE) argument attempting to show that space is infinite, Maxwell’s demon illustrating the second principle of thermodynamics, Einstein’s example of the elevator introducing the restricted theory of relativity, and of course Schrödinger’s famous half-alive and half-dead cat in the Copenhagen interpretation of quantum mechanics.
Naturally, there are also examples of bad, or at least uninformative, thought experiments. One of my favorites occurs in the field of philosophy of mind, where we are often asked to think about consciousness by considering the idea of a zombie (i.e., a dead person who again acquires motion and some sort of will, and yet is not conscious of what he is doing). What does our intuition tell us about the zombified condition, the philosopher is then apt to continue? Well, nothing, really, because we don’t have either any experience of zombies, or any plausible a priori expectations of what it is like to be one. So, whatever your intuition tells you about zombies vis-à-vis consciousness, it’s at best fit for the plot of a B movie, not for advancing our understanding of neurobiology.
Why is Galileo’s case a good example of a thought experiment, while the zombification of philosophy of mind doesn’t seem to lead us anywhere? It seems intuitive that a thought experiment has to be based on reasonable and informative premises in order to be fruitful. The textbook joke about thought experiments concerns the problem that starts with “Consider a spherical cow…” and goes on to derive all sorts of (irrelevant to real cows) properties of these imaginary animals.
A more satisfactory answer to what makes a thought experiment good or bad must come from an understanding of what, in fact, a thought experiment is. This is no easy task, judging from the rapidly increasing literature on the topic in philosophy of science. Ernst Mach, the physicist who first coined the word “thought experiment” (gedankenexperiment, in German), believed that they are possible because of a vast repertoire of empirical knowledge that we acquire instinctively. What a thought experiment does, then, is to bring such knowledge into sharp focus.
Another view of thought experiments has been advanced by J. Norton, who suggested that they are (disguised) formal arguments: they start with a premise (which is often grounded in experience) and proceed by a combination of deduction and induction (see last issue’s “Thinking about Science”) to achieve a certain conclusion. Not every philosopher agrees, however, and J.R. Brown has upheld thought experiments such as the Galileian one as examples of true new knowledge acquired without referring to experience at all, a rather Platonic view of the process.
The two schools represented by Norton and Brown are the extremes of a continuum of positions, which includes the idea that thought experiments are in some sense a limiting case of standard experiments, and the suggestion that thought experiments are a sort of mental model of the world. Ultimately, thought experiments by themselves are not considered satisfactory in science, and we are much happier when we can carry out a real check of a particular prediction. However, it seems that even at the stage of designing a real experiment one tries to simulate its set up and possible outcome in one’s own mind, which means that thought experiments are indeed a crucial component of the scientific method.
Further reading: Tamara Horowitz (1991) Thought Experiments in Science and Philosophy. Rowman and Littlefield.
When bias is good, when bias is bad
(September/October 2003)
How many times have you heard it said that skeptics are biased against the possible truths of claims of the paranormal? How many times have you accused creationists of being biased against the facts supporting the theory of evolution? This often makes for endless disputes, but as we shall see, not all biases are created equal.
Bias is a natural, and useful, feature of the way humans reason about the world. A bias in this case can be thought of as the seriousness with which one regards claims in favor or against the view that one holds at the moment. In general, we tend to be biased in favor of our own views and against conflicting ones, and this asymmetry helps us navigate the world without going insane. Imagine what would happen if you changed your mind every time you heard somebody defend a different opinion from the one you hold at that moment. After awhile you would find yourself unable to make any decision.
That is probably why nature built us as biased rational animals. However, it is equally obvious that too strong a bias will also get you into trouble. If you keep refusing to believe that casino games inevitably favor the house in the long run, you will lose everything you own.
So what distinguishes a good from a bad bias, and how do we decide whether to stick with or change our preconceived notions? How do we know when we are being reasonably skeptical rather than close-minded? One thing that-contrary to popular belief among skeptics-does not distinguish us from creationists is that we carefully consider the evidence while they simply dismiss it out of hand. Psychological research has shown that most people like to think of themselves as rational beings that give the evidence a fair shake. This is true for both skeptics and true believers. A classic example will illustrate the point.
A study discussed in Thomas Gilovich’s How We Know What Isn’t So (pp. 53-54) reports the results of a test of bias in which two groups of people where given literature containing data supporting and criticizing the effectiveness of the death penalty. When one compares data on the same state before and after capital punishment was introduced (within a certain historical period), it turns out that the frequency of homicides does significantly go down. This can reasonably be interpreted as evidence in favor of the deterrence effect. However, if one compares the same statistic (the rate of homicides) between states that have and do not have the death penalty, there is no significant difference, a result that just as clearly contradicts the deterrence hypothesis.
Predictably, the evidence did not sway opinions: proponents of the death penalty thought their conviction was bolstered by the data, and so did those who had declared opposition to capital punishment before examining the evidence. What was interesting was the reason why these predictable results obtained. It was not that subjects discarded the contrary evidence. Quite the opposite: they examined it more carefully than the evidence that favored their own views. As a result of this closer scrutiny, they were able to uncover (very reasonable) flaws in the study that did not fit with their bias, leaving the exercise convinced that they had thoroughly dismantled the opposite view. This sounds a lot like the sort of arguments I get from creationists: they read quite a bit of evolutionary literature, and they are convinced that it is reasonable to reject what scientists propose on the basis of the evidence itself. Of course, they do not pour over their Bibles with anything like the same sort of critical spirit they employ when reading articles written by evolutionary biologists.
Scientists, incidentally, show the same sort of bias. Gilovich’s example is the one of French anatomist Paul Broca (1824-1880), who refused to believe that the German brain cases he was studying were significantly larger than the French ones in his sample. He thus appropriately corrected for body size, which resulted in the difference evaporating. However, Broca somehow neglected to apply the same correction in the much more obvious instance of the difference in brain size between men and women, concluding that women were less intelligent!
The crucial point, of course, is that scientists no longer believe Broca’s ideas about gender differences in intelligence, while creationists still refuse to accept the overwhelming evidence in favor of evolution. Why? Peter Medawar explained it this way: science is “a rapid reciprocation of guesswork and checkwork, proposal and disposal, conjecture and refutation.” It is not that individual scientists are any less biased than any other person, or even that bias is inherently bad; rather, it is that the social activity of science as a community of people committed to the free market of ideas leads to the long-term improving of our understanding of the world as it really is. Creationists or pseudoscientists share no such commitment, and that is why they may end up rationally convinced that their conclusions are well founded, despite all evidence to the contrary.
Further reading: Thomas Gilovich (1991) How We Know What Isn’t So. Free Press.
(November/December 2003)
The United States is characterized by a peculiar mixture of science-worshipping and anti-intellectualism. On the one hand, the US is the world clear leader in science and technology, boasting achievements such as landing a human being on the moon (or, more questionably, inventing and using nuclear weapons). On the other hand, almost half of the American people don’t “believe” in evolution, and many espouse all sorts of doubtful or downright silly beliefs in paranormal phenomena. How is this possible?
Many explanations have been proposed, and undoubtedly several are needed. As it is often the case in complex sociological phenomena, many factors are at play simultaneously, and there is no simple answer to the problem. I’d like to focus here on what I think certainly is one of these factors, which when mentioned finds scientists and skeptics immediately on the defensive: the intellectual hubris of scientism.
Scientism is not a philosophical position that people espouse of their own choice. There is no National Association for the Advancement of Scientism, and in fact there is not even a word to label one that engages in scientism (engaging in scientistic behavior doesn’t make you a scientist). Indeed, the word is often hurled at people as an insult, often by philosophers at other philosophers, or by creationists at evolutionary biologists and other scientists.
Scientism is essentially an ideological position implying that science is the only key to solve any problem worth addressing, and that - given enough time and resources - science in fact will solve those problems. Let us consider some examples. In philosophy of mind, Patricia and Paul Churchland have proposed the rather radical idea that emotions do not exist. Their notion of “eliminativism” (see: Armstrong, D. M. 1999. The eliminativist theory, pp. 91-99 in The Mind-Body Problem: an Opinionated Introduction. Westview Press, Boulder, CO) aims at reducing all psychological talk in terms of neurobiology, and when one thinks of neurons and electrical potentials one does not need to bring up cumbersome and vague concepts such as emotion. Yet, one could object that if there is a problem when attempting to translate the complexities of human mental phenomena into current neurobiological parlance, perhaps it is the latter that is at fault for being too simplistic. The Churchlands, on the other hand, have faith in the fact that, eventually, psychology will be absorbed into biology, just as chemistry is now considered largely a branch of physics. Perhaps, but the jury is obviously still out there.
Another example of scientism can be found in the ambitious program that E.O. Wilson set up for himself when writing his Consilience: the Unity of Knowledge. In it, the famous biologist (already controversial enough for wishing to straightforwardly extend the sociobiology of ants to that of human beings) attempted to present the broad picture of a “consilience,” i.e., a unification, of all branches of human knowledge, from science to history, from religion to art. The problem was, rather than a unification, Wilson’s project increasingly took the shape of a program of academic imperialism in which science would eventually reduce and explain everything else.
Skeptics have their share of scientistic tendencies, real or perceived, as on those occasions in which they dismiss out of hand (i.e., without serious consideration, or based only on armchair investigations) new unusual phenomena. We should always remember that plenty of now accepted scientific discoveries were once thought to be “impossible” or to contradict established scientific principles (heliocentrism, the theory of evolution, and continental drift immediately come to mind).