Some Implications of the Demise of the Demarcation Problem

Two court cases, McLean v. Arkansas (“Arkansas”) in 1980 and Kitzmiller et al. v. Dover Area School District et al. (“Dover”) in 2005, showed how decisive philosophy can be when wielding the demarcation argument, as both creation science and intelligent design were denied victories because they were judged to be unscientific based on demarcation arguments. However, since the Arkansas decision and before Dover, the demarcation problem has generally come to be viewed by philosophers of science as intractable (i.e., “unsolvable”). The corollary of the intractability of the demarcation problem is that anything and everything can claim to be science without fear of being proved otherwise and that, therefore, the term “science” has no meaning. This has some clear implications for the creation project, including the renewed prospect for success in the courts and powerful answers to anti-creationist rhetoric. * Tom Hogan, MS, Shawnee, KS, [email protected] Accepted for publication September 9, 2009 cial opinion. Demarcation arguments attempt to distinguish between science and nonscience; demarcation is used by creationists, who argue that the theory of common ancestry is not science because it has not been observed or documented in the fossil record, and by evolutionists, who argue that creation isn’t science because only arguments based on methodological or metaphysical naturalism are “science.” The intractability (i.e., “unsolvability”) of the demarcation problem was generally accepted by philosophers at the time of Dover (Meyer, 1994). The philosophical strategy used by the defense in Dover was sound, but the defense made several tactical errors that prevented the court from discerning the status of the demarcation problem. One of the foremost experts on the problem, Larry Laudan, published an article about demarcation (Laudan, 1983) that is widely regarded by philosophers as definitive (Fales, 2005; Koperski, 2008; Meyer, 1994; Monton, 2006). Robert Pennock, a philosopher who was an expert witness for the plaintiffs, testified about Laudan’s work on the demarcation problem. Pennock answered the question equivocally, thereby giving two false impressions: 168 Creation Research Society Quarterly (1) That demarcation is not intractable and was used by most philosophers to distinguish science from non-science, and (2) That Larry Laudan would have considered intelligent design to be nonscience because Laudan would have used demarcation to do so, which was contrary to fact, since Laudan had shown the error of that strategy (Laudan is an anti-creationist who thinks that creation science is bad science; cf., Laudan, 1983). This paper will review and assess Laudan’s (1983) paper, “The Demise of the Demarcation Problem,” and will discuss some implications that follow from it that may impact future creationist/intelligent design court cases and the creation/evolution debate. Revisiting Laudan’s “The Demise of the Demarcation Problem” Following the 1980 Arkansas case, Laudan reviewed the history of the demarcation problem in 1983. (All references to Laudan in this section refer to his 1983 paper unless otherwise indicated.) According to Laudan, science was initially defined as those disciplines in which there was apodictic (demonstrable) certainty. It was knowledge, not mere opinion. In contrast, “craft,” which seeks methods to accomplish a specific purpose, is also demonstrable, and philosophers of science needed a way to distinguish science from craft. The earliest philosophers (e.g., Aristotle) used first causes as their demarcation criteria. First causes were thought to be those events that were independent of other causes, such as natural laws. Laudan (1983) used astronomy as an example of something that could be both science and craft, depending on whether it searched for first causes. Seventeenth-century scientists discarded the first causes idea as essential for science when Newton was able to describe the behavior of gravity without finding its cause. Eighteenth-century philosophers of science maintained demonstrability as the essential criterion for defining “science.” Laudan notes that the view that all theories may be fundamentally flawed— a view called fallibilism—became prominent in the nineteenth century. Laudan found that foundationally organized progress of scientific theories was no longer a certainty, as many theories were overturned or seriously amended during that period. These theories had been developed gradually and systematically, with newer theories built on the earlier ones. With the rejection of some of the more foundational, earlier theories, the whole progressive view of science was in jeopardy. Science was no longer apodictically certain, but was a work in progress. Philosophers then began to rely upon methodology—such as methods of reasoning and researching observable entities—for their demarcation criteria. However, this is not what scientists were actually doing. Nevertheless, despite the philosophic failure in demarcation, rhetoric was used against what was perceived as superstition under the pretense that effective demarcation criteria existed. After examining historical attempts at defining demarcation criteria, Laudan lays out questions that he believes must be considered in demarcation: 1. How shall we determine if the demarcation criteria are adequate? 2. What are the necessary and sufficient conditions for demarcation? 3. What rationale for calling things “scientific” or “unscientific” is precise enough to be useful? 4. What is our motivation for applying demarcation? Laudan concludes that the question of adequacy must allow for current scientific practice, although I believe he is begging the question. The demarcation plan must identify the methodological or epistemic criteria sufficiently clearly to give the desired result. In dealing with the second and third questions, Laudan shows that sufficient conditions are used to make sure that things thought to be “scientific” are not excluded by the demarcation criteria and that a failure to meet necessary conditions is used to exclude things that are in fact “scientific.” Both types of conditions are essential for any demarcation plan and must be precisely specified in order to be useful. In answering the last question, Laudan uses historical examples of demarcation attempts. He evaluates the motivations of Aristotle, the logical positivists (e.g., Carnap), and Karl Popper in their efforts to determine demarcation criteria. Laudan states that Aristotle used demarcation to attack the followers of Hippocrates, the logical positivists did the same against the metaphysicians, and Popper aimed at discrediting Marx and Freud through demarcation. Laudan recognizes the power of labeling and categorizing—stressing that philosophers should be careful to do this correctly. Laudan notes that the motives of philosophers should not corrupt their philosophical reasoning when applying demarcation. He might have been thinking of Michael Ruse’s testimony in Arkansas as an example of motives corrupting correct reasoning. In the final section of his paper, Laudan turned to modern demarcation attempts, beginning with the logical positivists. Logical positivists insisted that scientific statements must be exhaustively verifiable. However, many scientific conclusions cannot be exhaustively tested (e.g., all universal physical laws), while many other ostensibly nonscientific laws are testable (e.g., the statement that the earth is flat). Popper’s sophisticated falsificationism lacks discrimination—many things that are ostensibly “unscientific” (e.g., the flatVolume 46, Winter 2010 169 earth hypothesis) are falsifiable, while some scientific propositions (e.g., that atoms exist) are not falsifiable. Laudan considers whether “scientificity” might have the property of degree and whether theories might be compared based on their testability. He notes that there are technical difficulties with this approach, and only Popper’s theory is articulated well enough to evaluate “testability.” Furthermore, it can only be used when one theory implies another competing theory and is, therefore, useless for demarcation. Another concern is that Popper sees testability as a semantic notion, while demarcation relies on epistemic principles. Again, if anything is falsifiable, according to Popper, it is scientific, even if it is falsified, including the flat-earth theory or the statement that the moon is made of Limburger cheese. Next, Laudan looks at some current (as of 1983) candidates for demarcation strategies. These include fruitfulness and progressive development, whether a theory has been well tested, accuracy of predictions, pragmatism, and coherence. He asks (and I paraphrase), “Can ostensibly ‘unscientific’ things be well tested, including literary theory, carpentry, and football strategy?” (Laudan, 1983, p. 325). Laudan observes that some disciplines that are speculative and investigate first principles may not be well tested when compared with disciplines that are not speculative and are well tested. He also shows that many ostensibly “unscientific” disciplines make cognitive progress and many ostensibly “scientific” disciplines do not make much progress by comparison. The conclusion of Laudan’s survey paper is that the current-as-of-1983 status of the demarcation problem in the philosophy of science is that it is intractable (i.e., unsolvable). Laudan’s conclusion is still generally uncontroversial in philosophy of science (Fales, 2005; Koperski, 2008; Meyer, 1994; Monton, 2006). His conclusion means that philosophy cannot specify what is not science. This startling development has a number of major implications for creationists and intelligent design (ID) advocates. One of the most important is that the basis for the defeat of ID and creation in Arkansas and Dover was philosophical malpractice. Philosophical Malpractice After Ruse testified against creation science in Arkansas in 1980, Laudan (1982) responded, essentially accusing Ruse of philosophical malpractice. Intelligent design philosopher Stephen Meyer (1994) also found fault with Ruse’s work. Ruse (1982) answered Laudan, arguing that the overwhelming need to defeat “creationism” can excuse any philosophical errors. Philip Quinn (1984) responded to Ruse’s defense by caustically inquiring whether it was proper behavior for a philosopher to perjure himself in a secular setting as long as he washed his hands in academia. Robert Pennock likewise committed philosophical malpractice in Dover and was taken to task by Fales (2005), Monton (2006), and Koperski (2008). Plantinga (2006) likewise attacked Pennock’s idea that methodological naturalism as a shaping principle in science can be used as a demarcation criterion. In both Arkansas and Dover, heavy reliance was placed on the categorization of creation and intelligent design as nonscience, based on demarcation criteria. The judicial decisions were based on the philosophical error that valid demarcation criteria have been found and that one of them—methodological naturalism—excludes the supernatural from science. This makes Laudan’s assessment that demarcation is intractable vitally important to both creationists and proponents of intelligent design. If the current inability to define clear demarcation criteria were properly communicated in court, increased prospects of success in obtaining advantageous court decisions for creationists and intelligent design advocates may result. If it can be shown that secularists cannot adequately and conclusively define “science,” then their options for prevailing in other cases will rest on naked judicial fiat. Let us be very clear—as far as philosophy of science is concerned, demarcation is dead and long buried these twenty-odd years since Laudan’s 1983 paper was published, notwithstanding the philosophical malpractice by Ruse and Pennock at the bar. No adequate reply to Laudan has been published. Heterogeneity of Science and Laudan’s Proposed Legal Review of Science As part of his overall argument, Laudan (1983) also made two other important, but ultimately problematic points. First, that science is wildly heterogeneous, and second, that some a posteriori demarcation is possible between good and bad science based on evidential arguments. In the second statement, Laudan implicitly suggested that the courts should judge between good and bad science. His first statement is absurd, and his second statement holds the potential for disastrous consequences. If the courts were given the power to judge between good and bad science, they would be taking on the responsibility of professional peer review—which is likely to stultify science. Science has thrived for centuries on the principle of peer review. Since law and science are different disciplines, applying different methods, review by lawyer rather than review by peers would eliminate the cooperative and piecemeal nature of scientific discovery that has proven fruitful over many centuries. There is no reason to think that judges would be more competent than scientists in reviewing scientific work. If legal review were to be applied to creation and ID, it could have the effect of stifling open 170 Creation Research Society Quarterly inquiry; scientists and professors might forever be under the threat of legal action for their communications. Political considerations would drive science, as was seen with the ideas of Lysenko in the Soviet Union (Bergman, 2009; Feyerabend, 1975). This situation would be undesirable for both science and law. Laudan’s acceptance of the heterogeneous status quo in the definition of “science” ignores the damage it does to disciplines that have built their reputations on technological achievements— e.g., physics, chemistry, and biology. When the technological contributions and epistemic credibility of some disciplines are diluted by association with disciplines that have not produced many technological achievements, such as economics, political science, sociology, psychology, meteorology, and paleontology, the reputation of the technologically-prolific disciplines suffers from the association. We see this in the complaints of physicists and chemists when climatologists and experts from the “social sciences” claim to speak for science; physics and chemistry have a cornucopia of technological benefits to their credit, but climatology and the “social sciences” not so much. Conversely, many disciplines have enhanced their reputations simply by claiming to be “science,” but without gaining their credibility by making significant technological contributions. Hence the historical increase in “science’s” heterogeneity and the accompanying dilution of its reputation. Asserting that a discipline’s reputation can be enhanced by technological fruit does not rely on a claim that demarcation criteria exist. Such claims ignore Laudan (1983). This paper makes no claim that physics, chemistry, and biology are “science” or that technological contributions by a discipline constitute demarcation criteria; this paper merely asserts that there are certain ethical reasons for these disciplines to be jealous of their reputations; those ethical reasons flow from their technological contributions. In order to show that Laudan’s statement, “science is wildly heterogeneous,” is absurd, we must revisit the demarcation (of science from nonscience) problem. Laudan (1983) concluded that philosophy of science has so far been unable to define a list of properties that will always separate science and nonscience. This means that philosophy of science cannot clearly distinguish science from nonscience. This vagueness means that philosophy of science will be unable to prevent anything from claiming to be “science” since philosophy of science has shown that demarcation criteria do not exist that can be used to prevent anything from claiming to be “science” (e.g., astrology, carpentry, and art could claim to be “science,” which is counterintuitive; we shall examine astrology and art later in this paper). If this is true, then “science” ceases to have a clear or even useful meaning. If the demarcation problem is intractable, then we must consider whether “science” therefore has any meaning. This is a very counterintuitive question, because we think of specific disciplines such as physics, chemistry, etc. as “science.” However, philosophy deals with knowledge; our intuition about what is “science” may merely be opinion. Without demarcation criteria, how can philosophy of science rationally support any definition of “science”? Laudan (1983) has shown that no demarcation criteria exist to prevent any craft, art, or philosophy from claiming to be “science.” Suppose that sewing and football coaching wanted to designate themselves as sciences alongside art, astrology, and carpentry. How should we stop them? We have seen that no demarcation criteria exist to do so. What meaning does “science” have if the list of things that are acknowledged as “science” includes diverse human interests that may have no common properties? Aren’t we forced to admit that if this were to happen, “science” would have no real meaning? Yet we know that due to our current knowledge about the status of the demarcation problem, we cannot prevent this scenario from occurring. Thus, even without the list of “sciences” being expanded absurdly, we see that the definition of “science” is so vague as to be meaningless. If “science” cannot be defined in a way to give it meaning, then Laudan’s (1983) statement about the “heterogeneity of science” is clearly meaningless, since it relies upon “science” having meaning. In fact, “philosophy of science” and “history of science” also would be meaningless phrases. Hence, Laudan’s (1983) statement about the “heterogeneity of science,” being meaningless, is absurd. Why Laudan’s Failure to Find Demarcation Criteria Was Predictable Laudan examined demarcation criteria against examples of science from history. Laudan essentially granted history the privilege of defining “science.” This created inevitable problems for philosophy, because the history of science had no governing criteria for demarcation; rather, the reputation of “science” likely led to many disciplines riding the coattails of “science” by claiming to be “sciences.” Thus, where originally there were only physics, chemistry, biology, and geology, we see in history that many other disciplines later claimed to be “sciences,” including: • the “social sciences” of psychology, sociology, and political science, • economics, • engineering, • medical science, and, lately, • history. This conflation, which occurred historically, was independent of any plan to maintain meaning for “science” through common criteria. Thus, it was Volume 46, Winter 2010 171 to be expected that Laudan’s historical tests of demarcation criteria would be unsuccessful. Laudan’s Implicit Acceptance of Scientism Since Laudan accepted the heterogeneity of “science,” he also necessarily accepted the claims of various disciplines to be “sciences.” The aim of such claims is to enhance the disciplines’ reputations. We see from the failure of demarcation criteria that there is no necessary reason for such a claim based on a set of common criteria from a definition of “science.” Rather, the more likely reason for these claims is due to the reputation enhancements that they gain from the assumption that science has epistemic power; by successfully claiming to be “science,” a discipline is able to claim this same epistemic power, and its reputation is enhanced. The original aim of scientism was to rely only on “science” for epistemic truth. A weaker version of scientism claims that “science” produces the strongest claims for epistemic truth, rather than exclusive claims. In practice, this weaker version is almost indistinguishable from the stronger version, since “science” is advancing into many areas of human life, including explanations about religion and sexual attraction. By accepting the historical status quo (heterogeneity) for his definition of “science,” Laudan therefore implicitly must accept scientism. Meaninglessness, Demarcation Criteria, and the Definition of “Science” Meaninglessness Theorem 1. For any definition of something, we should be able to find at least one thing that does not fit that definition. 2. If a definition cannot find anything that is excluded by the definition, then that definition is meaningless. Rather than try to prove this theorem at this point, let us assume its validity and use it to prove that the word “science” is meaningless. Suppose we take a field that intuitively is removed as far as possible from science, such as astrology, and attempt to find actual demarcation criteria to exclude it from science. Laudan (1983) showed that any potential demarcation criteria that we might use to exclude astrology from science will fail, because the same criteria also would eliminate other fields that we would intuitively call “science.” Bear in mind that with demarcation, we are applying criteria based on a set of properties in order to dismiss something as nonscience so that empirical tests are unnecessary. If we cannot exclude something like astrology as nonscience, how can we exclude anything? If we cannot exclude anything as nonscience, then the definition of science is meaningless. What are the potential demarcation criteria that we might apply to astrology? Does astrology make risky predictions? Certainly. Is it falsifiable? Sure. Does it have a specified causal mechanism? No, but neither does gravity. Does its knowledge progress? Astrologers think so. Does it investigate phenomena? Yes, it investigates the relationships between astronomical bodies and their effects on people’s lives. Is astrology quantifiable? In some ways. The degrees of angle of the planets and stars from the sun can be measured. Some properties in biology cannot be measured, such as properties like “cat-ness.” Are there anomalies in astrology? Sure, but there are also many anomalies in other things we think of as science, including the placebo effect, evidence for variation in the fine-structure constant, and cold fusion, which was recently reproduced by a number of researchers (Brooks, 2007). Thus, we see that no demarcation criteria exist that can be applied to astrology that do not also exclude things that we consider to be scientific. Let’s consider another example. Suppose an artist were to assert that art is science. He might argue that art uses geometry, as does astronomy, because it studies perspective and form. The artist asserts that he is very empirical—he studies his subjects carefully, examining form, shadow, color, and lighting. He says that he performs experiments. He experiments with pen studies before committing to paint and experiments with mixing different colors of paint before committing the paint to the canvas. It might be asked, “What knowledge is gained from art?” The artist would reply that he investigates human conceptions of beauty, form, nature, and the human experience. His studies sometimes overlap with biology, geology, psychology, and sociology, which intuitively are regarded as sciences. Thus, we see that even art cannot be excluded by demarcation criteria. We’ve seen that philosophy has problems trying to define “science.” Do nonphilosophers do any better? Problems with the Definitions of “Science” Outside Philosophy Let’s start our examination of attempts by nonphilosophers to define “science” by considering the idea that “science” might be defined as a list of disciplines: physics, chemistry, biology, etc. Whose list are we to use? The list of the National Academy of Sciences (NAS), which includes economics, sociology, psychology, and engineering? The list of the National Science Foundation? The list that physicists might create? The list that creationists might create? The list that evolutionists might create? What is the rationale for which items to include in the list? Will that list be universally acceptable? Will that list even be useful in the future, when new disciplines arise? What justification can then be given for refusing to accept applications for 172 Creation Research Society Quarterly entry from current disciplines that are not on the list, seeing as we’ve just seen philosophy’s failure to validate any demarcation strategy? Is the determination to be based simply on ad hoc rationale and brute force? A couple of textbooks define “science” in terms of the “scientific method.” (Dickson, 1995; Thompson and Turk, 1998). Let’s next consider the idea that we can look at definitions from various disciplines for a definition of “science.” Science is the concerted human effort to understand, or to understand better, the history of the natural world and how the natural world works, with observable physical evidence as the basis of that understanding (Railsback, 2009a). This definition relies on the claim that one can use “science” to understand “history.” This definition forces a conflation of disciplines. Furthermore, it limits the domain of “science” to the natural world, which conflicts with the list from NAS, which includes disciplines that don’t primarily study the natural world in whole or in part (e.g., engineering, economics, and much of sociology). Let’s continue looking at various definitions: 1. the systematic observation of natural events and conditions in order to discover facts about them and to formulate laws and principles based on these facts. 2. the organized body of knowledge that is derived from such observations and that can be verified or tested by further investigation. 3. any specific branch of this general body of knowledge, such as biology, physics, geology, or astronomy (Morris, 1996). Science is an intellectual activity carried on by humans that is designed to discover information about the natural world in which humans live and to discover the ways in which this information can be organized into meaningful patterns. A primary aim of science is to collect facts (data). An ultimate purpose of science is to discern the order that exists between and amongst the various facts (Gottlieb, 1997). The investigation of natural phenomena through observation, theoretical explanation, and experimentation, or the knowledge produced by such investigation (Pickett