Why Believe That for Which there Is No Good Evidence?

Robyn M. Dawes^*

I would like to begin this paper with a confession: I believe in the reality of global warming.  But why do I believe that?

First, I cannot assess the phenomenon directly myself.  In fact, the only subjectively compelling evidence I've personally encountered for the phenomenon consists of the unusually warm winters in Pittsburgh from 1987 on — which is probably about the lousiest bit of evidence available.  My knowledge of better evidence is weak; I've not taken a chemistry class beyond high school, and while I've taken many graduate courses in mathematics and statistics (originally to compensate for linguistic incompetence and a language requirement), I have never checked out the details of the models themselves — and even if I could, I would have no way of assessing the validity of the data on which they are based.

What I have done is to trust authority.  I read articles summarizing the models in such sources as Science and Scientific American and American Scientist.  When the authors of these articles assure me that the models are coherent — and that recent evidence is confirming the models — I believe them, even though I have no direct knowledge of their data, their qualifications, or their veracity. I trust.

In addition, all my close personal friends, including the woman with whom I live, believe in the reality of global warming.  None, for example, has anything but anger and contempt for President Bush's half-hearted denial that it is real.  In my everyday interactions with these people when the topic of global warming comes up, none has ever suggested that I retain an "open mind, questioning all authority" and check out the evidence for myself.  In short, two critical factors in my acceptance that the evidence is "good" are authority and social consensus.

Now I would like to consider a belief that I don't hold: The reality of an after-life.  I regard the religious authorities who attest to its existence as, frankly, "silly," and very few of the people with whom I interact, including the woman with whom I live, believe in its reality.  None has ever suggested that I adopt an "open mind, questioning all authority" and examine the evidence for an after-life on my own.

Moreover, just as in global warming (Pittsburgh is getting hotter in the winter), I find ersatz evidence of my own experience compelling.  People to whom I have been close who suffer severe neurological impairment from such conditions as Alzheimer's appear to me to be partially non-existent, as people; the inference that when their brain totally shuts down their personal qualities will be totally non-existent is compelling.  Further, my own "near death" experience (or was it just an ether dream?) was nothing like the ones reported by Kübler-Ross.  Every feeling of "losing" something valuable I have ever experienced before or after in my life was "rolled into one," although I didn't know what I was losing.  The universe itself was simultaneously disintegrating and contracting to a single point.  I kept screaming "There's such a thing as life," although I didn't know what the words meant.  A committed theologian would dismiss both my interpretations of what has happened to others and my own experience as totally irrelevant to the reality on being resurrected from the dead and judged by God.

But 80% of the people in the United States believe in an after-life (Gallup & Castelli, 1989), perhaps even some members of this audience.  Why?  Two important factors are the same as those that I cited for belief in global warming; trust in authority, and social consensus.  People who believe in the after-life are reassured of its existence periodically by learned ministers, and interact (at least in churches) with others sharing their belief.  Moreover, they may readily interpret ersatz "evidence" as supportive.

My first point is to suggest that belief in a high rate of "repressed" child sexual abuse and the widespread existence of satanic cults practicing sexual abuse on children is influenced by the same factors.  Authorities have attested to their existence.  Moreover, especially when people attend self-help groups or go into therapy with authority figures who believe this nonsense, they encounter a great deal of social consensus concerning such reality.  Finally, they can interpret non-diagnostic experience as compatible "evidence."

And why shouldn't we be trusting of authority and influenced by consensus? (at least so far as deciding what it is we don't need to investigate on our own).  As Stich and Nisbett (1980) point out, a total rejection of authority figures in our life would lead to an inability to function, totally.  We could not buy either a prescription or a nonprescription pill at a pharmacy if we had no trust in the authorities of the FDA, or for that matter eat food that was not grown in our own gardens.  (Even raising our own animals involves some trust in the people who are monitoring what the animals themselves are fed — as the recent popular book and movie Bitter Harvest pointed out, a trust that is not always warranted.)

Or how could we be here today without trusting a remarkable number of authorities? — beginning with the people who program the computers used by our reservation agents, continuing through those who design the airplanes or monitor their safety-and including the architect of this hotel.  (Again, this trust is occasionally misplaced — as when one discovers that a confirmed reservation has been "erased," or discovers that the left engine is on fire and hopes that the pilot will have enough expertise to realize that it is the left one.)  The completely open mind that questions all authorities would reside in a body that is a blithering mess.

(We might do well to take quite seriously the finding that people answering authoritarianism scales are inconsistently authoritarian but consistently anti-authoritarian, rather than to ascribe the result to a "response bias."  [For a critique see Rorer, 1965].  Perhaps authoritarianism comes naturally to people, while the "syndrome" indicated by consistency of aberrant responding — is anti-authoritarianism; "eternal vigilance" is, after all, akin to paranoia.)

Moreover, consider what life would be like if we did not accept consensus belief as a quite valid cue to reality.  An enormous amount of what we "learn" about the world is provided by the people around us (Simon, 1990), and to be oblivious to their beliefs would trap us in the position of the hypothetical child raised by an altruistic wolf.

Now trust in authority and consensual validation are not "rational" bases for belief.  What I am suggesting, however, is that they are the bases for most of what we believe most of the time — in fact, almost everything we believe almost all of the time.  Is it any wonder then that a contagion of beliefs can occur even though these beliefs do not withstand critical scrutiny (i.e., ours)?  We have certainly observed contagion of belief in the past concerning belief that we and others have come to debunk.  And we will in the future.

But why these particular beliefs in widespread repressed child sexual abuse and satanic cults?  I wish to hypothesize a major factor later.  Now, however, I would like to emphasize the difficulty of understanding — more specifically, believing — the evidence that these beliefs are erroneous.

1. The first evidence concerns the fallibility of retrospective memory.  That has been discussed elsewhere (Loftus, Korf & Schooler, 1989; Pearson, Ross & Dawes, 1992), and I don't wish to reiterate all the evidence concerning this fallibility.  The point I wish to make is how difficult it is to believe that what one remembers is not true.  The present is fleeting, the future does not yet exist; consequently, virtually all our conscious knowledge is based on our memory.  Moreover, we have good reason to believe that our memory is generally correct (if it weren't, we'd be dead), even memories involved in "motor programs" such as driving your car or playing a piano.  To ask people to question their own memory is tantamount to asking them to question their own interpretation of reality, which at the extreme is close to asking them to consider the possibility that they are schizophrenic.  Just as people couldn't function with a "perfectly open mind," they couldn't function if they were constantly to doubt that what they recalled as true was in fact true.

Moreover, it is perfectly reasonable (Dawes, 1989, 1990) to "project" on others our beliefs in the validity of our own memories.  Even though we may have been "influenced" to believe one thing or another, that does not make the memory less real to us — hence less of a basis for making decisions in our life.  So why should we believe that such factors account for the memories of others?

2. Another important bit of evidence is that confidence is not very highly correlated with accuracy. But how can I change my confidence — e.g., in a particular judgment or memory — on the basis of knowing that confidence is not a good cue to validity, when in fact confidence is confidence about how correct I am?

An anecdote: A close colleague has written a brilliant paper about "outside" versus "inside" judgment (Kahneman & Lovallo, in press).  The exemplar involved asking academics how long it will take for a particular committee to achieve its goal.  These same academics are then asked to think of similar committees they have been on with similar goals and how long it took these committees to achieve their goals.  The second time estimates are an order of magnitude greater than the first.  And they are more accurate.

When I was talking to this same colleague about the upcoming presidential elections, he stated that he was "absolutely sure" that Clinton could not win, and he gave me a "subjective probability" close to .98.  I couldn't resist asking him how often in general his political judgments were correct when he was that certain, and he snapped back "63% of the time."  We both laughed, but we also both agreed that it is extraordinarily difficult to be uncertain about things when one is certain.  Moreover, a knowledge of the "calibration" literature concerning confidence (see Fischhoff, 1988), doesn't help much.  Think of individuals "recalibrating" themselves on the basis of this knowledge: "I believe that on judgments about which I am this sure, I am correct 95% of the time; therefore, I believe that on judgments about which I am this confident, I will be correct 79% of the time."

And once again, when we ourselves treat confidence as an excellent cue to validity, it is perfectly reasonable to "project" that it is an excellent cue for others as well.  Why should we doubt someone who claims they are absolutely confident concerning an event (judgment perhaps being a slightly different matter).  If we were to adopt as a general principle that high confidence about recalled events is not a very good reason to believe that these events actually occurred, we'd be faced with a real dilemma about how to run our own lives.  "I am very confident that I know exactly what happened, but it probably didn't happen that way."

3. Even though we believe that "you can't teach old dogs new tricks" and that "the generals are always fighting the last war," we ourselves know that we get wiser as we grow older (although we may admit senescence in some particular areas).  Why?  I propose that in addition to all the self-serving reasons, there is a compelling cognitive basis for this belief.  When we are older we have all the information we had when we were younger plus more.  It is quite natural to believe that the more information one has, the better one's grasp of reality.  It is particularly compelling to believe that if information set A subsumes information set B, then a belief or judgment based on A must be superior to one based on B.  Ellsberg¹ pointed this assumption out with respect to the debates on the Vietnam War between people who did versus those who did not have access to classified or secret information.  Those who had access knew that they had knowledge others didn't — while also having access to the knowledge that others did have.  Hence, their judgment was naturally superior, only it wasn't.

In fact, additional knowledge can simply cloud judgment, although not confidence (Oskamp, 1965).  But that's hard to believe.  Now, we critics are claiming that when people believed earlier in their lives that they were not sexually abused they were correct, while they are incorrect in their later beliefs.  How can that be, given they have access to the same information they had earlier plus more, including the judgment of the authorities?  Certainly, we do not go around proclaiming that "now that I know more I know less" or believing that our knowledge is enhanced by literally throwing away information, and once again we would have little reason not to "project" that principle on others in general.

In short, asking people to doubt the conclusions concerning widespread childhood sexual abuse and satanic cults is asking them not only to reject the usual bases of authority and consensus for establishing reality, but in addition to accept principles that violate foundations of everyday functioning.

Now in point of fact we do ask people to accept such principles, and they do.  Few people, for example, believe that the world is flat, even though it appears to be, or believe that cigarettes and alcohol are good for them, even though both may have very pleasant effects.  We return once more to the efficacy of authority.  People who have no direct experience of the curvature of the earth believe that it is not flat, and even the greatest devotees of tobacco and alcohol believe that these drugs are harming them.  We accept what we have been told by "reputable authorities."  (We even accept what has been communicated by very minor authority figures, such as the person who draws a map that shows the Suez Canal to be longer than the Panama Canal.)

Here, we have a very serious influence on the epidemic of the current belief.  Specifically, the last 20 years have provided a veritable explosion of credentialed psychotherapists, with a simultaneous diminution in the scientific training that would lead them to know what they are talking about.  I don't want to go into all the statistical details here, but two statistics are relevant.  First, in the paper presented by Wakefield and Underwager (1992) the (admittedly non-random) sample yields the conclusion that 33% of the therapists supporting the validity of the repressed memory syndrome were psychologists and 8% psychiatrists.  While the field of clinical psychology has been doubling at the rate of once every 10 years and that of psychiatry has been doubling at the rate of once only 20 years (for a comparison, the rate of lawyers has been doubling once every 12 years), there are still only about 75% or so more practicing clinical psychologists than psychiatrists.  Why, then, the 4 to 1 ratio?

The answer I propose is that while the training of psychiatrists still involves required undergraduate courses in science and the first two years of medical school in the science of medicine, the scientific training of clinical psychologists has — I'm afraid — gone to hell.  Fewer and fewer (now somewhere between 13% and 18%) are being trained at the top 200 graduate institutions, while more and more (almost 40%) are being graduated from professional schools.  (The former figure was close to 40% in 1970 and the latter was 0.)

It is, of course, not true that a person obtaining an advanced degree from an institution of lesser status necessarily has had poorer scientific training that a person graduating from a place of higher status, or understands the scientific basis of psychology less well.  There is a great deal of overlap in actual training and understanding.  In general, however, as Lee Sechrest claims, "we are graduating people with only a peripheral knowledge of psychology" (Hayes, 1989).  (All the statistics supporting this conclusion are presented in the first chapter of my forthcoming book: Psychology and Psychotherapy: The Myth of Professional Expertise [in press] ().)

Without training in scientific psychology, it is little wonder that so many practitioners do not accept — or even know about — principles that violate our naive assumptions concerning everyday functioning.  Given they don't know, the "authoritarian" basis for holding or diminishing this epidemic is lost.  When they themselves have not been "inoculated" with scientific skepticism, there is little reason to expect them to inoculate their clients.

In summary, this epidemic of belief is consistent with our ways of functioning in the everyday world, and there is little to stop the epidemic.  We may understand the "irrationality" of much of our everyday functioning, but often only when it leads to a conclusion that we believe to be untrue.  We will have great difficulty teaching people in general about such irrationality, because our colleagues them selves do not understand it.

Appendix

I came across the following article in the most recent issue of Science (May 22, 1992).  It was entitled: "Open Season on Depression" (Vol.256, pg. 1137).

Mental illness may be taking another step out of the closet with a project recently launched by a Harvard psychiatrist: The first nationwide free screening program for depression.  Douglas Jacobs, who practices at McLean Hospital in Belmont, Massachusetts, says depression is a major public health problem, afflicting an estimated 10 million Americans in a given 6-month period, but only one-third of sufferers get diagnosed or treated for it.

So in 1990, Jacobs started a pilot program at McLean, which was expanded last year to 90 facilities in 44 states.  About 5,000 people attended lectures and discussions on depression last year at health facilities; 3,000 of them filled out self-report forms and had short meetings with mental health professionals who told them about treatment options.  Jacobs says half of the people who attended the screening had never had any treatment for depression, but the self-report forms indicated that half were probably clinically depressed.  Plans are to expand greatly the next screening, scheduled for October.  With the support of the National Depressive and Manic Depressive Association, a patient advocacy group, the organizers hope to reach up to 400 locations-including non-health facilities like shopping malls and libraries-and screen up to 20,000 people.

"It's time that psychiatry not be behind closed doors," says Jacobs.  He's already thinking about the next advance: A "national substance-abuse screening day."

That is allegedly science, but how is it really structurally different from the pseudoscience we have been discussing today?  I invite the reader to think of similarities and differences.

I would also like the reader to consider the following hypothetical experiment.  I am going to screen 5,000 people to determine whether they suffer from the newly discovered mental illness syndromes of aslantophelia and aslantophobia.  I will ask such questions as the following:

Do you ever think about the possibility that either you or your spouse partner/significant other/best friend) might die before the other does?

Has it ever occurred to you that even though you are doing your job well, the organization in which you work may be in trouble?

Have you ever wondered whether your personal life might be different if Lee Harvey Oswald — or whoever it was who shot President Kennedy — had missed?

When people answer "yes" to such questions, phrased positively as above, they will be diagnosed as suffering from aslantophelia, a feeling of great uncertainty and lack of stability in life.  When they answer "no," they will be diagnosed as suffering from aslantophobia, a denial of the degree to which life is uncertain and unstable.  (In the actual questionnaires, of course, positive and negative wordings will be balanced.)  Both syndromes will follow from their parents' — particularly their mothers' — inability to provide them as young infants with the necessary sense of security.  The primary recommendation for people suffering from either of these syndromes will be prolonged therapy to help them "get in touch" with those feelings they experienced as an infant so that they can "work them through."  How many takers would I have?  Could I set up support groups of people suffering from these syndromes?  If so, could there be an epidemic?

References

Dawes, R. M. (1989). Statistical criteria for establishing a truly false consensus effect. Journal of Experimental Social Psychology, 25, 1-17.

Dawes, R. M. (1990). The potential non-falsity of the false consensus effect. In R. M. Hogarth. (Ed.), Insights in Decision Making: A Tribute to Hillel T. Einhorn ()() (pp.97-110). Chicago: University of Chicago Press.

Dawes, R. M. (in press). Psychology and Psychotherapy: The Myth of Professional Expertise (). New York: The Free Press.

Fischhoff, B. (1988). Judgment and decision making. In R. J. Steinberg & E. E. Smith (Eds.), The Psychology of Human Judgment () (pp. 153-187). New York: Cambridge University Press.

Gallup, G., Jr., & Castelli, J. (1989). The People's Religion (). New York: Macmillan.

Hayes, S.C. (1989). An interview with Lee Sechrest: The courage to say "We do not know how." APS Observer, 2(4). 8-10. p.8.

Kahneman, D., & Lovallo, D. (in press). Timid decisions and bold forecasts: A cognitive perspective on risk taking. Management Science.

Loftus, E. F., Korf, N. L., & Schooler, J. W. (1989). Misguided memories: Sincere distortions of reality. In J. C. Yuille (Ed.), Credibility Assessment () (pp.261-174). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Oskamp, S. A. (1965). Overconfidence in case-study judgments. Journal of Consulting Psychology, 29, 261-265.

Pearson, R. W., Ross, M., & Dawes, R. M. (1991). Personal recall and the limits of retrospective questions in surveys. In J. M. Tanur (Ed.), Questions About Survey Questions: Meaning, Memory, Expression, and Social Interactions in Surveys (). New York: Russell Sage.

Rorer, L. G. (1965). The great response-style myth. Psychological Bulletin, 63, 129-156.

Simon, H. A. (1990). A mechanism for social selection and successful altruism. Science, 250, 1665-1668.

Stich, S. P. & Nisbett, R. E. (1980). Justification and the psychology of human reasoning. Philosophy of Science, 47, 188-202.

Wakefield, H.. & Underwager, R. (1992, June 20). Recovered memories of alleged sexual abuse: Lawsuits against parents. Paper presented at the Fourth Annual Convention of the American Psychological Society, San Diego, California. (A revised version is in press in Behavioral Sciences & the Law.)