In dealing with claims alleging paranormal events, a basic question concerns the degree of evidence needed to establish such claims. As the paper by Laurent Beauregard (in this issue) points out, this is not a simple matter and is one which must take into account one's prior orientation to the world. In the simplest terms, I have expressed the matter in the statement: "An extraordinary claim requires extraordinary proof." Any serious consideration of the literature on paranormal phenomena soon makes it apparent, however,that claimants of the paranormal often do not fully appreciate this requirement. Thus, parapsychologists quite frequently have pointed out that the controls and care used in their studies commonly exceed the requirements placed upon most other work in experimental psychology. To this, the critic must simply reply: "Of course!" This is because the critic considers the claims of parapsychologists to be more extraordinary than most other claims in psychology. But from the parapsychologist's standpoint, the critic sometimes requires evidence of such extraordinary character as to make the proponent of the paranormal believe that nothing would ever convince so extreme a skeptic. This becomes particularly apparent when a critic suggests experimenter fraud as an explanation even though there may be no direct evidence of fraud, merely a possibility of fraud. Thus, on the one hand we may have the claimant offering evidence that is insubstantial for the critic, and on the other hand we may have a critic giving insubstantial indication of what it would take to force the critic to accept the evidence. I suggest that matters might be helped somewhat by considering more exactly what we mean by the use of the term "extraordinary" in this context.

Extraordinary Events Versus Extraordinary Explanations

To place matters somewhat in perspective, before going directly into the question of exact definition, we should note that one can generally separate events from their explanations: and we commonly speak of both events and explanations as ordinary and extraordinary. (I am here using "extraordinary" in the broadest sense of meaning "unexpected," particularly theoretically unexpected. More careful consideration to the term will be given in later discussion.) This results in the following simple matrix of orientations.

  Explanation Offered
ordinary extraordinary
Event Claimed ordinary A C
extraordinary B D

This matrix represents analytic, "pure" types which might only seldom be found empirically. Extraordinariness and ordinariness are often a result of an interweaving of both the event and the explanation. But I would suggest that analytic separation may prove conceptually useful.

Cell A represents ordinary events being given ordinary explanations. This is what routinely takes place in "normal" science. In dealing with claims of the paranormal, we are usually concerned with allegations of an extraordinary event (e.g., the sighting of a monster, a non-chance statistical frequency, etc.) which we then seek to explain in terms of either an ordinary explanation (e.g., error in reporting, fraud, etc.) as in Cell B, or sometimes through an extraordinary explanation (e.g., visitation from Mars, psi, astrological forces, etc.) as in Cell D. We frequently forget that there is also the orientation represented in Cell C wherein ordinary events may be given extraordinary explanations (e.g., seeing pure chance coincidence explained by the acausal principle of synchronicity). To a degree, these four cells may correspond to rather distinct psychological proclivities. Thus, there may be persons who want to structure their worlds in terms of one or another of the orientations represented by these cells.

The Cell A type of "personality" is insistent on the complete ordinariness of things. The anomaly itself is denied, quite aside from any issue of its explanation. It is simply argued here that nothing extraordinary happened at all. This sometimes results in a posture that may be perceived by the proponents of the anomaly as a dogmatic denial rather than a simple skepticism. For example, rather than accept the existence of an extraordinary correlation or non-chance statistical finding, this individual may deny the real existence of the finding by claiming it is a hoax or fraud. (In a sense, claiming fraud constitutes an "ordinary" explanation of an extraordinary claim-- as found in Cell B--but without some external basis for such an explanation, it amounts to denying the alleged extraordinariness of the event itself; so I see this as an example of Cell A when argued this baldly.)

Cell B represents the case where a perceiver accepts an event as extraordinary but seeks to explain it through ordinary principles. When a skeptic says "Nothing extraordinary happened," he may be speaking of matters in this sense rather than in terms of Cell A. A verified anomaly may be the result of normal processes. Thus, a monster may simply be a rare mutation; an astrobiological correlation may be the result of a common third variable which somehow jointly produces the two states mistakenly viewed as directly causally linked; a high ESP-test score may be produced by non-verbal and unconscous communication as in the case of the famous "mind-reading" horse Lady Wonder.

The Cell C perspective is commonly found among adherents of cults. Thus, an ordinary event, say a crop failure, may be interpreted as the result of the gods; a death from drowning may be attributed to a curse; or a simple coincidence resulting within the limits of chance may be seen as the result of mysterious "forces." It should be noted that logic does not preclude ordinary events from having extraordinary explanations, but the doctrine of parsimony within science (the view that the simplest adequate explanation is the one that must be accepted) makes such extraordinary explanation untenable if an equally adequate ordinary explanation can be provided.

Cell D represents the most extreme of the positions in the sense that both the event and its explanation are extraordinary. For example, the event of reports of the sighting of a little green man emerging from a saucer-shaped craft may be explained as a visitation from an alien anthropological mission from a distant galaxy. The orientation represented by Cell D may commonly be viewed as the most "far out," but it is actually quite scientifically proper if all ordinary explanations for an established extraordinary event have been found inadequate. Thus, if a truly replicable psi experiment were produced that would convince any reasonable person that significant non-chance guessing scores took place, and if such any reasonable person could be convinced that all ordinary explanations are inadequate, an extraordinary explanation (such as that a psi process like telepathy was at work) could be invoked and considered to explain the extraordinary guessing scores. Not only would ordinary adequate explanations need to be shown inadequate before the extraordinary explanation could be considered, but such consideration would not in any sense allow one to leap to the conclusion that the suggested extraordinary explanation was valid. Once the door has been opened to the consideration of extraordinary explanations, one must consider all extraordinary explanations that might be presented with any degree of plausibility. Thus, the explanation of "telepathy" would have to compete with other extraordinary explanations such as PK, demonic possession, etc. Falsifying certain explanations does not automatically validate another explanation. Despite these limitations, the approach represented by Cell D is scientifically acceptable under proper conditions whereas the orientation represented by Cell C which may appear more reasonable (since only one element, the explanation, is extraordinary) is methodologically eliminated from serious science.

I would suggest, then, that the proper scientific approach to an alleged anomaly is first to see if we can view it in terms of the orientation found in Cell A. If, and only if, the extraordinariness of the event is established, we should move to see if we can view matters in terms represented by Cell B. Because of the rule of parsimony, we should completely avoid Cell C. And if, and only if, we can eliminate proposed adequate alternatives of explanation in Cell B, should we move into considering matters in terms of the orientation found in Cell D. Finally, once properly looking at matters in terms of the perspective in Cell D, we should consider alternative extraordinary explanations and not simply accept the one most prominently offered.

The Extraordinary As Relative And Mesurable

Though a dictionary definition of extraordinary states that it means "going beyond what is usual, regular or customary" or that it simply refers to that which is "remarkable" or "exceptional to a marked extent," this term must have more specialized meaning for any serious scientific consideration of anomalies and the paranormal. Otherwise such terms could easily be confused with the merely rare or abnormal. In a most fundamental sense, something is extraordinary when it is unexpected. But such extraordinariness (which I here equate with anomalous) can be both of a general and a theoretic variety. Thus, if we are shown a picture of a 30-inch tall adult, that might strike us as quite amazing and unexpected. Many would call such a person extraordinary. But a remarkable midget of this size would not constitute a paranormal phenomenon, merely an abnormal one. The scientist would probably not be so surprised by the appearance of such a midget as would the general public, for such a small person may be within the experience and certainly within the theoretical possibilities known to the scientist. On the other hand, a scientist stumbling across a strange and unexpected species of animal might regard such a beast as extraordinary while the native population, which has commonly seen the beast around for years, may perceive it as quite ordinary. This is because the scientist has theoretical reasons for expecting not to find such a beast (e.g., science may define the beast as extinct). Thus, a general anomaly for (most of us) may not be a theoretical anomaly for the scientist and vice versa. The question of extraordinariness, then, is relative to one's frame of reference, and when we are concerned with extraordinariness in a scientific context-- as we are here--such extraordinariness must be measured against theoretical expectations provided by the general body of scientific knowledge at the time. In addition, things are rarely simply just ordinary or extraordinary, for some things are more extraordinary (and by the same token, sometimes more ordinary) than others. Thus, we are not dealing here with a simple dichotomy but a continuum expressing degrees of expectation and surprise.

All of this becomes quite important when we consider specific paranormal claims. We tend to confuse our psychological surprise in seeing some things (the general definition of the anomalous) with the expectation level that scientific theory would produce. Thus, for most of us, an "abominable snowman" or a person who drinks human blood (a vampire) seems initially more extraordinary (and thus unlikely to actually exist) than someone's prophesying a well-loved political figure's assassination. The Loch Ness Monster seems a more "strange" possibility to many of us than the notion that some people may be capable of telepathy. A unicorn may seem more improbable than someone's being cured of a physical malady through faith healing. Obviously, the degree to which each of us may be surprised by a strange event is rather relative to our own experience and background. But though we may be able to say relatively little about people's general expectancy levels (what will constitute general anomalies for most people), it is far easier to make reasonable assessments of extraordinariness in relation to existing bodies of scientific knowledge and theory. And I would suggest that scientists should carefully make such assessments before judging the likelihood of some phenomenon's actual occurence. I think that this will quickly reveal that some paranormal claims are far less unlikely than others, and this has very important implications for the amount and quality of proof a scientific skeptic should demand before accepting such claims.

In examining the relevant literature, I have been amazed to find that many (if not most) scientists would probably rank the likely truth of various paranormal claims in quite unreasonable ways (if one accepts my rationale above as reasonable). Thus, the claims of the parapsychologists are generally and incorrectly perceived as more "reasonable' by many scientists than are the claims of those proponents of the existence of a Sasquatch (Big Foot) or lake-inhabiting large creature. To most of us, the existence of "monsters" seems more bizarre than the possibility of telepathy. Yet, the implications of telepathy's actual existence are far more revolutionary for contemporary psychology than a new species of ape or sea creature (which may simply be an ancient species incorrectly thought extinct) would be for zoology. In similar fashion, the various claims of the parapsychologists have quite different sorts of implications for the rest of science. Simple telepathy would not necessarily radically change our view of physics even if it caused major reconceptualizations in psychology and physiology. But the existence of clairvoyance and/or precognition would have quite revolutionary effects upon fundamental ideas in physics and almost all of science in so far as it might force alteration of our ideas about space and time. Yet both the proponents and critics of such claims commonly fail to consider the degrees of extraordinariness involved in the different anomalies discussed, and, of course, the differential implications they may have for what would constitute acceptable proof for the scientist.

In corresponding with a major critic of claims of the paranormal I was amazed to find that he considered the likelihood of parapsychological claims as more reasonable than the claims of the astrobiologists (such as Michel Gauquelin). Even though the remarkable correlations claimed by the astrobiologists are presented merely as anomalous correlations--no causal explanations are suggested--the association of such astrobiological "findings" with the causal claims of the astrologers seems to have been enough to make this critic view the astrobiological correlation as less likely than the claims of the parapsychologists. Even though confirmation of claims such as those by Gauquelin would not directly threaten any important theories in astronomy or biology (since no claim is made of any direct causal link), the association of such ideas with classical astrology is apparently enough to stigmatize such ideas and make them seem almost completely implausible. The parapsychologists have generally been much aware of this kind of "guilt by association" and have gone out of their way to dissociate themselves from what they and others call "occultists." But I would suggest that our views of many paranormal claims should be re-examined to avoid such theoretically irrelevant associations. An empirical claim should be examined on its own merits, for its truth is frequently quite irrelevant to the other ideas of its supporters.

Recent critics of claims of the paranormal have suggested that the acceptance of some paranormal effects may open the door to the acceptance of all sorts of irrational thinking. It is this too simplistic kind of black/white, either/or thinking that may be creating the current impasse between some critics and the proponents of the paranormal. It is urgent that if progress is to be made in any dialogue between the proponents and their critics that both sides must carefully hammer out the kind of fundamental decision criteria needed to make responsible scientific evaluations of not only what sorts of evidence would be acceptable but also the quality and quantity of evidence that may be needed relative to the degree of extraordinariness of a paranormal claim.

Dimensions of the Reasonable

In examining the discussions of paranormal phenomena, it is important that we locate the locus of that which is purportedly extraordinary. I have found that proponents and critics commonly concentrate on different elements. Thus, we find some, like Michael Polanyi, speaking of the plausibility of revolutionary ideas as a central consideration while other writers speak of the credibility of the experimenters, probability of the events themselves. I would suggest that terms like "ordinary", "plausible," "likely," "probable," "reasonable," and "credible" frequently are treated synonymously when they may refer to quite different things; and this in turn confuses the debates between proponents and critics.

On the most general level, we can analytically separate three elements that are involved when we deal with alleged paranormal events. First, we can speak of the event itself, and I would suggest that this should be referred to on a scale of ordinary-to-extraordinary. As described earlier, this simply refers to the degree to which the event was theoretically expected or unexpected. It is important to note that whatever we may psychologically think about an event, whether we expected it or not, in the final analysis events simply exist or do not exist, independent of our desires. Second, we deal with a witness or narrator of the event. I would suggest that we refer to the character of such a witness/narrator (and of course the narrator may not be the original witness) as varying on a scale with credible at one end and non-credible at the other. Here, credible simply means believeable. Obviously, a number of different factors go into this designation including the witness/narrator's honesty, perceptual abilities, motivations, carefulness, training and knowledgeability as an expert, etc. But note that this terminology does not allow us to refer to the event itself as "credible" (or what we may be inclined to at first think of as its opposite, incredible). The term must only apply to the character of the witness/narrator. Third, we have the narrative or description-report of the event. I would suggest that this varies on a continuum of plausible-to-implausible. Again, note that we should, by this terminology, refer to events or witnesses/narrators as plausible or implausible; only narratives about events should be thus described. We thus have the three elements and the dimensions used to describe their reasonableness as follows:

Event: ordinary........ extraordinary
Witness/Narrator: credible.......non-credible
Narrative: plausible......implausible

These three elements can result in eight different combinations. At one extreme, we have an ordinary event, narrated plausibly, by a credible witness/narrator. This is the sort of case we hope to usually find in "normal" science. At the other extreme end of the spectrum of combinations, we would have an extraordinary event, narrated implausibly, by a non-credible witness/narrator. This last form is the most easily rejected kind of paranormal claim and would commonly be branded nonsense or quackery. But between these two extremes we have six other combinations and these are not so easily dealt with.

Four of the eight total combinations deal with ordinary events. But in dealing with ordinary events we usually have little reason to be suspicious about the plausibility of the narratives or the character of the witness/narrator. When we are concerned about the narrative, it is usually in a courteous methodological way that concentrates on what is in the report rather than upon what may have been left out of the report. And we tend to presume credibility of the witness/narrator if he is a member of the scientific community, certified by its merit system and socialized through its training process to produce standardized reports. This probably means that many seemingly plausible narratives in "normal" science are actually poorly done (would be evaluated as implausible if we knew the full truth about how the research was conducted), and we have a good bit of evidence to support this judgement. We also probably have a reasonable amount of error in "normal" science in our judgements about the credibility of witness/narrators. The history of science and recent polls would indicate that some fraud does go on in ordinary research. But since we are dealing with ordinary events, that is theoretically expected phenomena, the disturbance created by such errors is relatively minor in relation to the general scientific progress being made; and such errors will probably be corrected through replication and later work within the "normal" science community. But when we are dealing with extraordinary events, the type we would call paranormal, the implications of such events can be quite revolutionary for general science theories, so the social controls within science become far more extensive and sometimes somewhat exaggerated.

The four combinations in which the event is extraordinary are all situations we might find in dealing with the paranormal. The hardest case to dismiss is that found when the event is extraordinary but the witness/narrator is credible and the narrative is plausible. In fact, I would argue that in this case we should not dismiss the case but should be forced to at least tentatively accept it (at least until replications are conducted which confirm or falsify our belief in the credibility of the witness/narrator and the plausibility of the narrative-report). For example, if a reputable (therefore credible) scientist goes through easily checked procedures and comes up with an astounding correlation (a plausible narrative-report of an extraordinary event), such a case should have a higher probability of being a true picture of things than a case where we have doubts about the plausibility of the narrative and/or the credibility of the witness/narrator. I would go further and suggest that this case should be perceived by us as having a higher probability of being true than some ordinary event which is connected with an implausible narrative and/or non-credible witness/narrator (assuming no other credible and plausible other witnesses and narratives exist).

Since we can rarely be the witness/narrator for an extraordinary event ourselves, we must rely on plausible narratives by credible others. We should, therefore, concentrate less on whether the event is extraordinary or not and more upon the factors of plausibility and credibility. Unfortunately, some have confused these various dimensions with one another. For example, some critics of parapsychology have, in effect, argued that the extraordinary character of the event (in this case a significant, non-chance guessing score) has in itself created the presumption that fraud must be present. In other words, the extraordinary event has been used to measure the character of the narrator. As numerous defenders of parapsychology have argued, the presumption of fraud without any evidence for it is a non-falsifiable claim that has no place in science. Put this way, I would have to agree. A more common criticism, however, seeks to equate the possible flaws in the narrative with the character of the narrator. Thus, critics have frequently said that if a psi experiment is not completely controlled for alternative explanations, and if one such alternative explanation might be fraud by the investigator (the narrator), we must presume that fraud took place. In its extreme form, where the argument states that the only alternative (non-psi) explanation might be fraud and therefore we must presume fraud, is of course again non-falsifiable and therefore inappropriate to scientific discourse. Here I would contend that the critics of psi like C.E. M. Hansel and some others may go beyond the evidence and outside of proper scientific argument. At least such is the case if my above interpretation is accurate. On the other hand, I think it can be argued that extraordinary events can properly force the scientist into greater caution in his examination of the factors of plausibility and credibility than might be the normal case in examining claims of ordinary events. And where the controls on credibility (formal training, peer pressures, etc.) are absent, and there may be a history of past fraud in an area, this should certainly affect the evaluator's degree of caution in presuming the credibility of the narrator of a paranormal (extraordinary) event. On the other side of the matter, many parapsychologists seem so convinced of the credibility of their fellow narrators that they seem willing to lower their standards for plausibility of psi reports. I would suggest then that while skeptics may at times overdo their skepticism, claimants of the paranormal are very frequently not properly skeptical enough.

All things considered, I would urge that our main attention be given to the narratives themselves. We should not allow our perception of the event as extraordinary to too greatly color our analysis, for ultimately our theories must fit the facts rather than vice versa. The degree to which an event is seen as extraordinary should certainly affect how much evidence we should demand for its proof, but we should be careful to separate the event from its narration and narrator, for we otherwise prejudge matters in a way which denies the basic inductive attitude of science. We should also recognize that judgements about the credibility of the narrator must in most cases end up inferred only after our judgement about the plausibility of the narrative. Reputable scientists can make mistakes and fools can sometimes be right. A non-credible narrator (e.g., a witness whom we know has committed past fraud) may cast a shadow of doubt on reports from him, but it does not logically follow that these reports are false.

What then should we do with plausible reports by witnesses of extraordinary events? In most cases we simply must be patient, recognize that a single such report may simply not be enough to let us make a final judgement. Though we have a right to simply ignore such reports (giving them a low probability of later confirmation) and give them low scientific priority for our time, we do not have the right to dismiss such reports. Since the burden of proof is on the claimant in science, we must state that the evidence is inconclusive and remain skeptical (though certainly less so than we were before this plausible narrative entered the debate). But to say something is unproved is not the same as saying it is disproved. Until more and convincing evidence comes in (and this must be proportional to extraordinariness of the paranormal claim in the theoretical sense I have described), we need to remain skeptical and agnostic about matters paranormal. At least those which have not been explicitly disproved (as we have seen many things like the Bermuda Triangle and pyramid power falsified). It is this kind of scientific skepticism with an open mind that I think being a true zetetic is all about.