<<

Science versus Austrian (and later British) philosopher of Sir Karl Popper; we begin with his conceptualization. 1 2 Scott O. Lilienfeld, Steven Jay Lynn, Popper and Rachel J. Ammirati1 According to Popper (1959), scientific claims 1Emory University, U.S.A. and 2Binghamton University, U.S.A. are falsifiable: They are in principle capable of being proven false (according to Popper, scientific claims can never be proven true; The prefix “pseudo” means “false.” Accordingly, theycanonlybefalsified).Incontrast,Popper pseudoscience is false science. Yet equating maintained, are impossible pseudoscience with false science is not entirely to falsify, either because their proponents (a) accurate, because most false scientific claims advance claims that are so vague that they are are not pseudoscientific. Instead, we can best untestable or (b) continually invoke view pseudosciences as fake that hypotheses—escape hatches or loopholes—that display (often intentionally, but sometimes effectively immunize their claims from refu- unintentionally) the superficial appearance of tation. Eventually, the overuse of ad hoc science but largely lack its substance (Hines, hypotheses can place claims outside of the 2002). Nevertheless, distinguishing “fake” purview of scientific testing. For example, science from both genuine science and other Popper argued that much of Freudian forms of false science is more easily said than is pseudoscientific, because no conceivable done (Pigliucci & Boudry, 2013). The question evidence could disprove it. For example, Pop- of how to differentiate pseudoscience from per considered Freud’s concept of the Oedipus other disciplines bears significant implications complex (the idea that young children are for clinical and allied mental health romantically attracted to the opposite-sex fields, because numerous authors contend that parent) as unfalsifiable and therefore pseu- these fields’ scientific foundations are being doscientific. If a psychoanalytic researcher increasingly eroded by pseudoscientific prac- weretofindthatmostadultsrecallatime tices (e.g., Lilienfeld, Lynn, & Lohr, 2003; when they were romantically attracted to McFall, 1991, 2001). Moreover, mental health their opposite-sex parents, she could interpret consumers, practitioners, and researchers this finding as confirming Freud’s hypothesis would benefit from guidance in separating of the Oedipus complex. In contrast, if this scientific from pseudoscientific claims, largely researcher were to find that most adults do because the latter assertions can fool even not recall such a time, she could interpret this well-trained individuals (Herbert, 2002). finding as implying that adults find memories of romantic attraction toward their parents The Demarcation Problem to be so threatening that they repress them. The “demarcation problem” refers to the This “heads I win, tails you lose” reasoning, challenge of distinguishing science from non- according to Popper, renders Freudian theory science, including pseudoscience (Pigliucci & difficultorimpossibletofalsifyandtherefore Boudry, 2013). To this day, there is no uni- pseudoscientific. versally agreed upon solution to this problem. Nevertheless, few scientists believe that Pop- Perhaps the best-known attempt to resolve per’s criterion, important as it is, the demarcation problem was offered by the succeeds as a necessary or sufficient criterion

The Encyclopedia of , First Edition. Edited by Robin L. Cautin and Scott O. Lilienfeld. © 2015 John Wiley & Sons, Inc. Published 2015 by John Wiley & Sons, Inc. DOI: 10.1002/9781118625392.wbecp572 2 SCIENCE VERSUS PSEUDOSCIENCE for distinguishing science from pseudoscience. retreat” to the protective belt, thereby invoking For one thing, certain pseudoscientific claims ad hoc hypotheses that could explain away the do appear to be capable of refutation. For negative findings. In doing so, these advocates example, although astrology is a prototypical tinker with the theory’s auxiliary hypotheses pseudoscience, many of its claims, such as in an effort to salvage their theory from fal- the proposition that astrological signs are cor- sification. In certain cases, strategic retreats related with personality traits, are falsifiable, arejustifiableinscience,especiallywhenthey and have been essentially falsified (Carlson, enhance the theory’s content, predictive power, 1985). Moreover, several critics have argued or both. Yet, when strategic retreats to the that contra Popper, many scientific claims protective belt are merely “pasted on” to the are difficult or impossible to falsify. Specifi- theory in desperate attempts to rescue it from cally, because legitimate scientists can almost falsification, such retreats become problematic. always invoke ad hoc hypotheses to rescue According to Lakatos (1975), in progressive their claims from refutation, these claims may research programs, theory tends to precede never be strictly disproven, a view known data.Thehardcoreofthetheorygenerates as the Duhem–Quine thesis after the two novel predictions that are generally corrobo- philosophers who propounded it. rated by findings. In contrast, in degenerating research programs, data tend to precede the- Lakatos ory. Negative findings—those that challenge An alternative perspective on the demarcation thetheory’shardcore—repeatedlyemerge, problem derives from Hungarian philosopher and the theory’s advocates continually invoke ofscienceImreLakatos.AccordingtoLakatos auxiliary hypotheses within the protective belt (1975), scientific , or as he called them in an effort to keep the sinking theory afloat. “research programs,” possess two major com- The theory is not moving forward; it is merely ponents: (a) a “hard core” of fundamental trying to keep up with the accumulating influx presuppositionsthatisintegraltothetheory of negative findings. Yet, as Lakatos noted, and (b) a “protective belt” of auxiliary (sec- because some degenerating research programs ondary) hypotheses that are not central to the do eventually right themselves by amending theorybutthathelptoprotectthetheoryfrom their protective belts (cognitive dissonance refutation. For example, the existence of the theory is arguably a potential example in unconscious comprises part of the hard core of psychology; see Greenwald & Ronis, 1978), psychoanalytic (Freudian) theory. It is difficult it is not necessarily illegitimate to cling to a to imagine calling oneself psychoanalytic in degenerating research program, just so long theoretical orientation without endorsing the as one honestly admits to oneself and others core assumption of unconscious processing. that it is degenerating. What is illegitimate is to In contrast, the existence of the defense mech- pretend that a degenerating research program anism of reaction formation (the supposed is in good scientific shape. tendency to deal with an anxiety-provoking Lakatos’ framework provides a useful way emotion, such as hatred, by transforming it of conceptualizing the difference between into its opposite, such as love) almost certainly science and pseudoscience. Specifically, from a resides in the protective belt of psychoanalytic Lakatosian perspective, we can define pseudo- theory. One could legitimately call oneself psy- sciences as degenerating research programs that choanalytic, although perhaps not an orthodox its proponents claim are progressive.Inthecase Freudian, even if one rejected Freud’s concept of pseudoscientific theories or methods, claims of reaction formation. are not supported by evidence, yet adherents According to Lakatos, when data repeatedly of these theories or methods insist that they conflict with a theory’s hard core, proponents are. As a consequence, these adherents are of this theory typically perform a “strategic deceiving themselves, deceiving others, or SCIENCE VERSUS PSEUDOSCIENCE 3 both. As helpful as Lakatos’ conceptualization dividing line between day and night (hence the is, it does not offer a strict set of criteria for equally fuzzy concepts of dusk and dawn), we demarcating progressive from degenerating can still distinguish morning from night for research programs. For example, when a the- most pragmatic purposes (Leahey & Leahey, ory’s predictions are met with mixed results, 1983). Similarly, we can differentiate clear-cut how do we decide whether it is progressive or sciences from clear-cut pseudosciences, even degenerating? Nevertheless, this may not be a though the distinction inevitably becomes fundamental flaw within a Lakatosian frame- murky in certain cases. work, because strict criteria for distinguishing According to the family resemblance per- science and pseudoscience may not exist. spective, pseudosciences are characterized by a set of useful, albeit fallible, indicators (Bunge, Pseudoscience: A Family 1984). We can think of these indicators as Resemblance View “warning signs” of pseudoscience. Although no single criterion is dispositive of a theory’s Indeed, some eminent philosophers of science status as a pseudoscience, the more criteria are (e.g., Laudan, 1983) have declared the demar- present, the more skeptical of this discipline we cation problem dead, because there appears to should typically become. Here we present 10 be no single criterion—or set of criteria consid- warning signs of pseudoscience that are useful ered jointly—that can invariably be used to dis- for distinguishing pseudoscientific from scien- tinguish science from pseudoscience. Some of tific claims in clinical psychology, psychiatry, thesescholarshavegonefurther,recommend- social work, and allied fields (Lilienfeld et al., ing that the very concept of pseudoscience be 2003). abandoned. Still, there is an alternative. According to Overuse of Ad Hoc Hypotheses Designed a family resemblance view of pseudoscience to Immunize Claims from Falsification (Pigliucci & Boudry, 2013), there is no “bright Whentakentoanextreme,adhochypotheses line” distinguishing science from pseudo- can provide virtually impermeable barriers science. Yet just as first-degree relatives within against falsification. In other words, when one a biological family typically share a loosely overuses ad hoc hypotheses, one is attempting correlated set of facial features, allowing us to to explain away findings that are inconsistent distinguish them from members of other fam- with a claim. For example, proponents of a ilies, pseudosciences tend to exhibit a loosely newpsychotherapymightinsistthatnegative correlated set of characteristics. According to findings are due to inadequate training of the this view, pseudoscience is an “open concept” individuals delivering the treatment. Although (Pap, 1953). Open concepts are characterized this might be true, it cannot by (a) fuzzy boundaries, (b) an indefinitely stand on its own; it itself requires systematic extendable indicator list, and (c) an unclear testing. Otherwise, it becomes what Dawes inner nature. If pseudoscience is an open (1994) termed an “argument from a vacuum:” concept, the distinction between science anassertionadvancedintheabsenceofany and pseudoscience is one of degree rather supportive evidence. than kind. The fact that one cannot draw a clear line Absence of Self-Correction in the sand distinguishing science from pseu- Like the assertions within pseudoscientific doscience does not imply that most scientific disciplines, the assertions within most sci- claims cannot be distinguished from most entific disciplines are incorrect in certain pseudoscientific claims in clinical psychology respects. Nevertheless, in the long run, or related disciplines. As the psychophysicist S. most scientific research programs tend to S. Stephens observed, although there is no clear learn from and correct errors, whereas most 4 SCIENCE VERSUS PSEUDOSCIENCE pseudoscientific research programs do not. strategy is not an efficient means of rooting out Consequently, intellectual stagnation is a errorinone’swebofbeliefs.Infairness,how- hallmark of most pseudoscientific research ever, —the tendency to seek programs. For example, astrology has changed out evidence consistent with one’s hypotheses, remarkably little over the past 2,500 years and to deny, dismiss, or distort evidence that despite enormous advances in planetary and is not—is not unique to pseudosciences and is stellar science (Hines, 2002). a serious problem in some sciences, including medicine, as well (Ioannidis, 2008). EvasionofPeerReview Many proponents of pseudoscience avoid Reversal of the Burden of Proof subjecting their work to the crucial safeguard The burden of proof in science typically falls of peer review, which is an essential means squarely on individuals who put forth a claim, for identifying errors in authors’ reasoning, not on critics of this claim. Proponents of pseu- methodology, and analyses. They may do so doscience often insist that skeptics demonstrate on the grounds that the peer-review pro- beyondareasonabledoubtthataclaim(e.g.,an cess is biased against findings that contradict assertion regarding the efficacy of a novel psy- well-established paradigms or that their asser- chotherapy) is false. This error is similar to the tions cannot be evaluated adequately using logician’s ad ignorantium fallacy (the argument standard scientific methods. For example, from ignorance), the mistake of assuming that advocates of graphology (handwriting analy- a claim is likely to be correct because there is no sis intended to infer personality traits) have convincing evidence against it. rarely subjected their claims to peer review. Absence of Connectivity By avoiding or circumventing the peer-review In contrast to scientific research programs, process, proponents of pseudoscience forfeit pseudoscientific research programs tend to lack an invaluable opportunity to obtain corrective “connectivity” with other scientific disciplines feedback from informed colleagues. (Stanovich, 2012). In other words, pseudo- Emphasis on Confirmation rather than sciences often purport to create novel para- Refutation digms out of whole cloth rather than to build on extant, well-supported ideas. For example, -winning physicist Richard Feyn- many proponents of extrasensory perception man (1985) maintained that the essence of (ESP) argue that it is a genuine—although science is a bending over backwards to prove heretofore undetected—physical process of oneself wrong. Popper’s student Bartley (1962) perception, even though reported cases of ESP similarly argued that science entails the max- violatealmosteverymajorlawofphysical imization of constructive criticism. Ideally, signals (e.g., ESP purportedly operates just as scientists subject their cherished claims to strongly from thousands of miles away as it the risk of refutation (Meehl, 1978; see also does from a few feet away). Although scientists Ruscio, 2002). In contrast, pseudoscientists should remain open to the possibility that tend to seek only confirming evidence for their an entirely novel paradigm has successfully claims. For example, if the developers of a overturned all preexisting paradigms, they treatment for a psychological disorder request must insist on high standards of evidence feedback from clients to help them “prove” before drawing this conclusion (Wagenmak- that the treatment is effective, they will avoid ers, Wetzels, Boorsbom, & van der Maas, important information that could show that 2011). the treatment is ineffective, or even harmful. Because a determined advocate can find at Overreliance on Anecdotal Evidence leastsomesupportiveevidenceforvirtually Anecdotal evidence or “anecdata” can be any claim, a confirmatory hypothesis-testing quite useful in the early stages of scientific SCIENCE VERSUS PSEUDOSCIENCE 5 investigation. Nevertheless, such evidence is for the efficacy of eye movement desensiti- almost always more helpful in the context zation (EMDR), a controversial treatment of discovery (hypothesis generation) than for posttraumatic stress disorder and related in the context of justification (hypothesis conditions: testing; see Reichenbach, 1938). Proponents We propose that the repetitive redirecting of of pseudoscientific claims frequently invoke attention in EMDR induces a neurobiological reports from selected cases (e.g., “This treat- state similar to that in REM sleep, which is ment clearly worked for Person X, because optimally configured to support the cortical Person X improved following the treatment”) integration of traumatic memories into general as a means of furnishing dispositive evidence semantic networks. We suggest that this integra- tioncanthenleadtoareductioninthestrength for these claims. For example, proponents of hippocampally mediated episodic memories of unvalidated treatments (e.g., facilitated of the traumatic event as well as the memories’ communication, secretin, gluten-free diets, associated, amygdala-dependent negative affect. chelation therapy) for autistic spectrum dis- (Stickgold, 2002, p. 61) order have often pointed to uncontrolled case reports of improvement as supportive Absence of Boundary Conditions evidence. As Gilovich (1991) observed, case Most well-supported scientific theories pos- reports almost never provide sufficient evi- sess boundary conditions: well-articulated dence for a claim, although they often provide limits under which hypothesized phenomena necessary evidence for a claim. For example, shouldandshouldnotoccur.Incontrast,many if a new form of is efficacious, pseudoscientific phenomena are purported to one should certainly expect at least some operateacrossanexceedinglywiderangeof positive case reports of improvement. Nev- conditions and individuals. As Hines (2002) ertheless, such case reports do not provide noted, one common characteristic of fringe adequate evidence that the improvement was psychotherapiesisthattheyareostensibly attributable to the psychotherapy, because this efficacious for almost all disorders regardless of their causes. For example, some proponents improvement could have been generated by of thought field therapy, a widely used “energy a host of extraneous influences (e.g., placebo therapy,” have proposed that this treatment is effects, regression to the mean, spontaneous beneficial for virtually all mental disorders. The remission, maturation; Lilienfeld, Ritschel, developerofthistreatmenthasevenposited Lynn, Cautin, & Latzman, 2013). that it is effective not only for adults, but for Use of Obscurantist Language “horses, dogs, cats, infants, and very young Many proponents of pseudoscience resort to children” (Callahan, 2001, p. 1255). Although impressive-sounding or technical jargon in it is certainly possible that a therapy derived an effort to provide their disciplines with the from a well-supported could effectively treat a wide variety of disorders (e.g., superficial trappings of science. Such language cognitive behavioral therapy; Butler, Chapman, may seem persuasive to individuals unfamil- Forman, & Beck, 2006), therapies of this nature iar with the scientific underpinnings of the are rare. claims in question, and may therefore lend these claims an unwarranted imprimatur of The Mantra of Holism scientific legitimacy. The ascendancy of neuro- Proponents of pseudoscientific claims, espe- scientific explanations of behavior (Satel & cially in mental health, often resort to the Lilienfeld, 2013) over the past two decades has “mantra of holism” (Ruscio, 2002) to explain rendered questionable explanations that invoke away negative findings. When invoking this neurological concepts especially popular. For mantra, they typically maintain that scientific example, consider the following explanation claimscanbeevaluatedonlywithinthecontext 6 SCIENCE VERSUS PSEUDOSCIENCE of broader claims and cannot be judged in must be something to it”). Nevertheless, isolation. For example, some proponents of researchers have not examined whether these theRorschachinkblottesthaverespondedto or other putative indicators of pseudoscience criticisms of this technique by asserting that we have delineated are intercorrelated, as clinicians virtually never interpret results from would be anticipated of a family resemblance a Rorschach in isolation. Instead, in actual model. practice clinicians consider numerous pieces The distinction between science and pseudo- of information, only one of which may be science has been one of the most contemplated a Rorschach protocol. There are two major and debated issues in of science, difficulties with this line of reasoning. First, it and it bears potentially important implications implies that clinicians can mentally integrate for distinguishing well-supported from poorly a great deal of complex psychometric infor- supportedclaimsinclinicalpsychologyand mation from diverse sources, a claim that is allied disciplines (Lilienfeld et al., 2003). The dubious given the research literature on clinical question of how to differentiate science from judgment. Second, by invoking the mantra of pseudoscience is not fully resolved, and the holism, proponents can avoid subjecting their difference is probably one of degree rather claims to the risk of falsification. In other than kind. Nevertheless, psychologists and words, if findings corroborate the validity of a other mental health professionals need not given Rorschach index, proponents can point wait for philosophers of science to solve the to these findings as supportive evidence, but if demarcation problem to make use of the these findings are negative, their proponents pseudoscience indicators delineated here to can explain them away by maintaining that ascertain the validity of novel and untested “clinicians never interpret this index in isola- claims. tion anyway” (see Merlo & Barnett, 2001, for an example). SEE ALSO: Clinical Science Model; Duhem–Quine According to the “open concept” view Thesis; Hypothetico–Deductive Model; Meehl, described above, these 10 warning signs do Paul E. (1920–2003); Popper, Karl (1902–94); not exhaust the list of possible indicators of Science–Practice Gap; Scientific Metaphysics and pseudoscience. To the contrary, they are only Ontology apartiallistofpotentialfeaturesofpseudo- science (for a list of useful websites containing References information on additional warning signs, see Bartley, W. W. (1984). The retreat to commitment. Lilienfeld, Lohr, & Morier, 2001). Indeed, New York: Open Press. other writers have proposed additional or Bunge, M. (1984). What is pseudoscience? Skeptical alternative useful indicators of pseudoscien- Inquirer, 9, 36–46. tific disciplines, such as exaggerated claims Butler,A.C.,Chapman,J.E.,Forman,E.M.,& (e.g., “This weight-loss program is a miracle Beck, A. T. (2006). The empirical status of cure!”), frequent appeals to authority (e.g., cognitive-behavioral therapy: A review of “This new psychotherapy must work because meta-analyses. Clinical Psychology Review, 26, Professor X claims it works”), reliance on the 17–31. bandwagon fallacy (e.g., “This figure draw- Callahan, R. J. (2001). Raising and lowering of heart rate variability: Some clinical findings of Thought ing test must be valid because thousands Field Therapy. Journal of Clinical Psychology, 57, of clinicians use it”), frequent ad hominem 1175–1186. attacks against critics (e.g., “I’ll bet that Carlson, S. (1985). A double-blind test of astrology. most of the critics of recovered memory Nature, 318, 419–425. therapy are child abusers themselves”), and Dawes, R. M. (1994). House of cards: Psychology and arguments from antiquity (e.g., “Astrology psychotherapy built on myth.NewYork:Free has been around for millennia, so there Press. SCIENCE VERSUS PSEUDOSCIENCE 7

Feynman, R. P. (1985). "Surely you’re joking, Mr. McFall, R. M. (2001). Elaborate reflections on a Feynman!":Adventuresofacuriouscharacter. simple manifesto. Applied and Preventive New York: W. W. Norton & Company. Psychology, 9, 5–21. Gilovich, T. (1991). Howweknowwhatisn’tso:The Meehl, P. E. (1978). Theoretical risks and tabular fallibility of human reasoning in everyday life. asterisks: Sir Karl, Sir Ronald, and the slow New York: Free Press. progress of soft psychology. Journal of Consulting Greenwald, A. G., & Ronis, D. L. (1978). Twenty and Clinical Psychology, 46, 806–834. years of cognitive dissonance: Case study of the Merlo, L., & Barnett, D. (2001). All about inkblots. evolution of a theory. Psychological Review, 85, , 283, 13. 53–57. Pap, A. (1953). Reduction-sentences and open Herbert, J. D. (2002). The concept of pseudoscience concepts. Methodos, 5, 3–30. as a pedagogical heuristic. Scientific Review of Pigliucci, M., & Boudry, M. (2013). Philosophy of Mental Health Practice, 2(2), 102–105. pseudoscience: Reconsidering the demarcation Hines, T. (2002). Pseudoscience and the paranormal. problem. Chicago: University of Chicago Press. Amherst, NY: Prometheus Books. Popper, K. R. (1959). The logic of scientific discovery. Ioannidis, J. (2008). Perfect study, poor evidence: London: Hutchinson. Interpretation of biases preceding study design. Reichenbach, H. (1938). Experience and prediction: In Seminars in hematology (vol. 45, no. 3, pp. An analysis of the foundations and the structure of 160–166). Philadelphia, PA: W. B. Saunders. knowledge. Chicago: University of Chicago Press. Lakatos, I. (1975). Falsification and the Ruscio, J. (2002). Clear thinking with psychology: methodology of scientific research programmes. Separating sense from nonsense. Pacific Grove, In S. Harding (Ed.), Can theories be refuted? (pp. CA: Wadsworth Publishing Company. 205–259). Dordrecht: Springer. Satel, S., & Lilienfeld, S. O. (2013). Brainwashed: The Laudan, L. (1983). The demise of the demarcation seductive appeal of mindless neuroscience.New problem. In R. S. Cohen & L. Laudan (Eds.), York: B asic B o oks. Physics, philosophy and psychoanalysis: Essays in Stanovich, K. E. (2012). How to think straight about honor of Adolf Grünbaum (pp. 111–127). psychology.NewYork:HarperCollins. Dordrecht: D. Reidel Publishing. Stickgold, R. (2002). EMDR: A putative Leahey, T. H., & Leahey, G. E. (1983). Psychology’s neurobiological mechanism of action. Journal of occult doubles: Psychology and the problem of Clinical Psychology, 58, 61–75. pseudoscience. New York: Nelson-Hall. Wagenmakers, E. J., Wetzels, R., Borsboom, D., & Lilienfeld, S. O., Lohr, J. M., & Morier, D. (2001). van der Maas, H. L. (2011). Why psychologists The teaching of courses in the science and must change the way they analyze their data: The pseudoscience of psychology: Useful resources. case of psi: Comment on Bem (2011). Journal of Teaching of Psychology, 28, 182–191. Personality and Social Psychology, 100, 426–432. Lilienfeld, S. O., Lynn, S. J., & Lohr, J. M. (2003). Science and pseudoscience in clinical psychology. Further Reading New York: . Shermer, M. (2002). Why people believe weird Lilienfeld,S.O.,Ritschel,L.A.,Lynn,S.J.,Cautin, things: Pseudoscience, superstition, and other R. L., & Latzman, R. D. (2013). Why many confusions of our time (2nd ed.). New York: clinical psychologists are resistant to Henry Holt and Company. evidence-based practice: Root causes and constructive remedies. Clinical Psychology Review, 33, 883–900. McFall, R. M. (1991). Manifesto for a science of clinical psychology. The Clinical Psychologist, 44, 75–88.