Assassination Research And The Pathology Of Knowledge

Dennis Ford, Ph.D.
3247 Kennedy Blvd., Jersey City, NJ 07306

The Third Decade, Volume 8, #5, July 1992, pages 1–6

      At the end of the 48 Hours documentary on the JFK assassination, Oliver Stone took Dan Rather to task for the malign neglect of the Kennedy assassination exhibited by CBS News and by other national news organizations. Rather’s mealy-mouthed reply confirmed that many of the advances in the study of the assassination have come through the hard work of individual citizens. Almost no thanks is owed to the national media, whether television or print, in advancing our understanding of the events in Dallas; to the contrary, some of the national media have labored to confuse our understanding of events.[1]
That said, it is the contention of this paper that the quality of critical research in our field is in desperate need of improvement—now more than ever, since media giants have directed their efforts not to solve the murder of President Kennedy but to attack assassination researchers.[2]
This paper treats assassination research as if it were scientific research and hypotheses about the motivation and modus operandi of the assassination as if they were scientific hypotheses. (The terms “theory” and “hypothesis” are not pejorative; they simply mean statements that can be tested.) Several criteria in scientific theory-building will be offered and several flaws in assassination research relevant to these criteria will be noted. General prescriptions will be offered about avoiding such flaws. A final section will suggest that the current state of inquiry in the research community is perilously close to exhibiting pathological traits. Throughout the paper, psychological aspects of research will be emphasized.
The approach made in this paper is advanced because the author is an experimental psychologist and a teacher of psychology and because he believes that the logic of proof demanded in science is the strictest available. It is incumbent on the community of assassination researchers to maintain the most rigorous standards. Popularizations of assassination research, such as High Treason[3] and Crossfire,[4] ought to maintain the highest standards since they serve for many as the introduction to our field. We cannot expect people unfamiliar with the particulars of our research to be generous in their judgments, or to pursue research on their own, if their first contact involves confused and flawed presentations.
An important gain in the treatment of assassination research as scientific research is the analysis of evidence as the sole criterion of “truth.” Reliance on a more logically based and precise methodology may free the research community from the proliferation of theories and, what is better, provide a means of eliminating theories. Finally, a scientific orientation may reduce the insidious effects of sloppy arguments and flawed data; unwitting commission of such flaws in hypothesis-testing retards progress in our field, as it does in all fields.
Testability. The first criterion of scientific research is that hypotheses about the assassination must be, or become, testable. (Given our present state of knowledge, many of our hypotheses are not currently testable; some will probably never be.) Researchers ought to specify in advance the kind of evidence which, if it existed, would verify their hypotheses. More importantly, researchers ought to specify in advance the kind of evidence which, if it existed, would falsify their hypotheses. It may be a considerable burden to ask researchers to refute their hypotheses, but it is refutation and not verification which determines the empirical value of a hypothesis.[5] The operative logical principle is:

      —if a theory implies certain empirical observations,
—and if these observations do not occur,
—then the theory is false.

      For example, the documentary The Men Who Killed Kennedy[6] presents the hypothesis that Corsican mobsters shot the President. In support of this hypothesis, the filmmakers rely on the experiences of Gordon Arnold, who stood near the picket fence on the knoll and was approached by men posing as police. The hypothesis is intriguing: Arnold was there; he appears plausible, even compelling, as he is presented with photographic evidence of Badgeman’s existence. However, no direct standards of proof are offered by the filmmakers connecting Corsican mobsters with the men who approached Arnold.
The hypothesis is testable, and quite simply. Since these men were Corsican, they had to speak with accents. If these “policemen” spoke with foreign accents, the hypothesis is verified. If they did not speak with accents, the hypothesis is refuted and Corsican mobsters are eliminated as suspects. (Thanks to Richard Buckley for this example.)
The reliance on the principle of falsification, which has proven valuable in other fields,[7] may serve: first, as an objective means of separating among hypotheses (which can be tested, which not?); and secondly, as an equally objective means of eliminating hypotheses. Refutation is especially important in a field where fabricated evidence may exist, and it may free the research community from the enormous dangers of “plausibility.” Merely because certain events (or certain scenarios) are plausible does not make them true. In the same way that Lou Costello was able to prove that 2 + 2 = 5 on the walls of Sidney Field’s rooming house, so a person who is good with words and has a command of the evidence can “prove” any hypothesis.
Parsimony. An issue related to falsification is parsimony, an important criterion in scientific research which suggests simply that “less is more” and “small is better.” Precise hypotheses are to be preferred to complicated and vague hypotheses, even at the cost of losing explanatory power. (Since the empirical content of a hypothesis increases as a function of the degree to which it can be refuted, precise hypotheses are preferred since they are more easily falsified.) “Large” hypotheses about the assassination, such as those offered by Groden and Livingstone, Marrs, and by Oliver Stone in his movie, are often irrefutable and thereby unprovable. If a hypothesis cannot be falsified, or withstand fair attempts at falsification, it cannot be verified.[8] Researchers must acknowledge that not every piece of evidence can be incorporated in a theory about the assassination; if a theory accounts for all the evidence, it cannot be refuted, nor can it be proven. Researchers must also be careful not to salvage refuted hypotheses by engaging in retrospective analyses. Once a hypothesis is refuted it is, or ought to be, permanently removed from consideration.
To return to the Gordon Arnold example: since Arnold, who comes across in the film as a “good old boy,” makes no mention of the “policemen” speaking with accents, and if we do not presume that Corsican mobsters spoke unaccented English, we may suggest that the hypothesis underlying this portion of the documentary is refuted. A researcher has two options at this point. He or she may opt for the former possibility—Corsican mobsters did not shoot President Kennedy—or he or she may choose to salvage the hypothesis in an ad hoc and unparsimonious manner by entertaining the possibility that the policemen who approached Gordon Arnold were American lookouts for Corsican assassins. If the latter possibility is entertained, the researcher now has two hypotheses to prove rather than one, and one (it might be added) refuted in an expeditious manner.
Correlation. A pervasive flaw demonstrated by many researchers is an excessive reliance on correlational data. (Such data is often referred to as “associations” or “links.”) This reliance, which has often produced many worthy leads, can become a grave error when it interferes with a causal analysis of events leading up to and occurring in Dallas. Put simply, correlations cannot establish causation.[9] If we consider motivation and modus operandi to be independent variables and the murder of President Kennedy to be the dependent variable, we are faced with a situation where we have a plethora of independent variables, many of which are multiply correlated, which is another serious problem, and most of which have not been shown to have a causal connection to the shooting of the President.
Correlational analyses present us with other problems. They are often expressed as character analysis (or assassination)—which can take the researchers far afield from Dallas, given that the events on Nov. 22 are the beginning and end of our researches. Correlational analyses often involve slanderous accusations made against men many citizens find worthy of respect.[10] Often, there is a backward reading of the evidence which implies guilt by association. Such post hoc reasoning falsely assumes a causal connection merely because one event precedes another.
An example from one of my interests may demonstrate this flaw. Consider the following chain of connections in the names of assassination-related persons: the elderly “missionary” who accompanied Oswald to Mexico City was John Bowen, aka Albert Osborne; Oswald worked for a John Bowen (real name John DiGrassi) at Jaggers; the elderly tramp was Albert Alexander Osborne, aka Howard Bowen (unless his name was Gus Abrams or Fred Chrisman or Chauncey Holt); there was an “Officer Alexander” who visited Oswald at the rooming house; there was Assistant DA William Alexander; there was a Laurence Howard who volunteered himself as one of the visitors to the Odio house; there was a Mack Osborne who served with Oswald in Japan; there was a Dr. David Osborne present during the autopsy; Oswald used the alias “Osborne” to order FPCC leaflets; finally, and by no means last, there was Howard Osborne, chief of the CIA Office of Security.[11] However many Osbornes there are, these correlations do not add up to proof; none of them, perhaps not even Oswald, has been linked in a causal way with the murder of President Kennedy. Suspicious, yes. Proof, no. Coincidence? Probably not, but more needs to be done linking each Osborne with Dallas—and then attempting to refute such links.
Eyewitnesses. Researchers would not presume to build theories in defiance of modern physics or chemistry, but they seem to disregard the findings of experimental psychology, which has moved beyond the somewhat antiquated psychology concerning eyewitnesses presented in books on the assassination. Failure to consider these issues will result in critical abuse of assassination critics by defenders of the Warren Commission myth.
The first factor often disregarded by researchers is that of selection of eyewitnesses. It is a fact that the population of eyewitnesses to the assassination is finite and probably not random. Researchers rarely consider the possibility of bias among the witnesses who’ve come forward. It is unknown, and perhaps permanently unknowable, how similar the testimony of identified eyewitnesses would be with the testimony of witnesses who’ve not come forward. This issue is not moot, given the often irreconcilable differences in the testimonies of the identified witnesses.
Another factor rarely considered is the psychological assessment of the witnesses who’ve come forward. What were their powers of observation? What were their motives in coming forward? Why have some witnesses remained in the public spotlight for years? These are important questions, considering the psychological finding that volunteer subjects differ in motivation and personality from subjects who are reluctant to volunteer.[12] Consideration that the unidentified witnesses may offer different versions of the assassination than what is now known presents the chilling possibility that there is a third “myth” waiting to be discovered.
Another factor rarely considered is the effect of interviewer bias. It has been demonstrated in psychological research[13] that interviewers exert subtle influence on their subjects. The stories provided by eyewitnesses derive partly from the recall of events and partly from interpersonal effects occurring during the interview in which they report their observations. Interviewer bias has been shown to occur in police work,[14] in everyday interactions,[15] and in clinical fields.[16] Assassination research is not exempt.
An example of interviewer bias may have been exhibited during The Third Decade conference in June, 1991, when Harrison Livingstone described how he obtained different information after reinterviewing the autopsy witnesses previously interviewed by David Lifton (and others). It is probably unknowable which version of these interviews is correct. What is more certain is that the interviewees were responding to different questions, different interviewer styles and different nonverbal expressions, in addition to recalling autopsy events. (The autopsy witnesses favored by researchers also reflect serious selection bias.)
Another factor rarely considered by researchers is the possibility of fraud committed by an “eyewitness” who, for whatever reason, comes forward and passes fictions off as truth. There is enough known about events in Dallas for a person so inclined to dupe a psychologically naïve researcher. Paul Ekman has shown that the identification of liars is exceedingly difficult, even by trained law enforcement agents.[17] Since the motivation is high to “crack the case,” researchers may be at great risk of deception perpetrated by unscrupulous or abnormal individuals feigning special knowledge of the assassination.
Memory. Researchers do not give enough consideration to memory factors. Often, there is a naïve belief that witnesses saw what they saw, pure and simple. If skepticism is applied to eyewitness accounts, it is only to dissenting witnesses. Yet, memory research has shown that memory is not a copy of an event but a reconstruction.[18] Eyewitness reports are unreliable; contrary to common sense, stress constricts the focus of attention and reduces memory.[19] People remember what they want. People remember what is plausible. People remember a blend of observation and conversation about the observation. People remember what interviewers put in their heads.
Events happen and begin the fragile course of the human memory system from short-term memory to long-term memory to later retrieval. The encoding of events and their retrieval may be interfered with at any stage. Scenes of Dealey Plaza after the assassination show witnesses crying, conversing, and running about. Such human responses to tragedy may have prevented the accurate encoding of the assassination; such responses may also have distorted original memories, a process technically called “retroactive interference.”[20]
Consider the case of Jeanne Hill, an often quoted eyewitness. No doubt, she witnessed important events; she is also subject, on paper at least, to considerable memory distortion. She saw the President murdered, the event to be encoded. She then chased after suspicious characters and had scary encounters with police. She was taken to an upstairs room and interrogated; there, lawmen suggested she change her story. (It’s almost as if the lawmen were trying to induce retroactive interference.) Over the years she has told her story probably to hundreds of interviewers. An industrious researcher might track her story and chart how it changed over the years; this researcher might also conclude that her original report was the most accurate, a conclusion which is not automatically correct (above and beyond what the lawmen succeeded in distorting). Whether it’s the first recounting or the hundredth, memory is subject to similar kinds of distortion.
Perhaps greater care should be taken with regard to the testimonies of witnesses whose stories have not changed over the years, an unlikely situation given the experimental study of memory. That certain witness accounts have not changed demonstrates to a memory psychologist that the person wrote down and memorized what he or she believed to have happened—which may or may not be the case. Researchers should become aware that eyewitnesses may combine accurate, original observations with what they’ve read in books or been told, more or less directly, by interviewers. The latter possibility is not unusual, but can be experimentally induced in laboratory studies of memory.[21]
Similar care must be exercised with the reports of sightings of Ruby and Oswald before the assassination. (Some sightings predate the assassination by years.) It is unlikely that a person would have, or could have, retained specific memories, such as pieces of conversation, about Ruby or Oswald unless there was an extenuating circumstance to make such encounters memorable. People simply do not store, and usually do not need, such specific memories. Furthermore, memory is not an automatic process but requires effortful processing. It is unlikely that witnesses would have had reason to memorize specific deeds or statements about men who were not then famous; it is also unlikely that witnesses would have written down their memories of such encounters.
The Pathology of Knowledge. Researchers have lately revealed a well-earned tendency toward self-congratulation; at long last one’s life work is receiving public appreciation. Researchers have also revealed a less desirable tendency originally described, embarrassingly enough, in a psychological context by the philosopher and psychologist Sigmund Koch.[22]
This tendency, which Koch calls “the pathology of knowledge,” arises when groups of researchers become isolated from the mainstream of scientific investigation. This isolation, which frequently arises as a defense against criticism, causes the group to make epistemological errors. The combination of the intellectual isolation and flawed research leads to serious distortions in the pursuit of whatever knowledge the group is seeking. Perhaps most seriously, such a combination also leads to the group’s failure to appreciate anything is wrong. In fact, the reverse happens: the group becomes ever more impressed with its correctness and ever more defiant in its conclusions.[23]
Some of the characteristics of the “pathology of knowledge” include: a preoccupation with certain unchallenged leading ideas; a reluctance to consider alternative explanations; an unwillingness to submit hypotheses to fair tests; a reliance on ad hoc explanations; a reliance on authority figures who decide what is and is not the case; and a reluctance to accept criticism.
One of the themes of this paper has been that progress in assassination research will best occur within a scientific context. The scientific endeavor requires the free and open exchange of information; it demands the endless testing of ideas and a thorough allegiance to empirical investigation. Whatever its origin, criticism must be accepted; criticism must lead not to entrenchment, which fosters the pathology of knowledge, but to a deeper analysis and to the search for better evidence. Researchers much recognize that not every critic of the assassination community is a dupe of the Warren Commission. (It shows how near the research community is to the pathology of knowledge when critics within the community must make an act of faith up front that they do not believe in the Warren Commission.) There are kooks among researchers, as there are in every field. There are researchers who hold unconcealed political agendas; often, these agendas extend backward in time from the current administration to Dallas, 1963, and beyond. There are also researchers who are less interested in truth than in the almighty dollar. Not every critic of assassination research is deviously motivated; nor is the reach community universally blameless. Such ad hominem attacks to and fro indicate that a line of inquiry has reached a dead end. Criticism must not be met with verbal abuse but with harder and better research.
However pessimistic this assessment of assassination research has been, there is cause for hope. There are incredibly talented and dedicated people working to solve this case. Adjustments need only be of degree, corrections only of focus. There is so much to work with, and new methodologies become available at an increased rate.[24]
Failure to solve this case cannot be held against the research community; government agencies failed, if indeed they tried. What can be held against the community is entrenchment and isolation, and confused and flawed investigations. At all costs, the community must prevent sliding into the pathology of knowledge. Failure to prevent this slide will cause assassination research to resemble astrology or psychoanalysis—an endless Talmudic study barren of practical results.


[1] Robert Hennelly and Jerry Policoff, “JFK: How the Media Assassinated the Real Story,” Village Voice, March 31, 1992, pp. 33–39.
Arnold Harris, “The Mass Media, JFK and the 28th Anniversary of the Assassination,” The Third Decade 8 #2,3, January–March, 1992.
Robert Groden and Harrison Livingstone, High Treason (Baltimore: Conservatory Press, 1989).
Jim Marrs, Crossfire (New York: Carroll and Graf, 1989).
Karl Popper, Conjectures and Refutations (New York: Basic Books, 1959)
The Men Who Killed Kennedy (London: BCTV, 1988), television documentary which aired in America on A&E Network.
Paul E. Meehl, “Theoretical Risks and Tabular Asterisks: Sir Karl, Sir Ronald, and the Slow Progress of Soft Psychology,” Journal of Consulting and Clinical Psychology, 46 #4, 1978.
Popper, Conjectures and Refutations.
David G. Myers, Psychology (New York: Worth, 1992), pp. 622–636.
Craig I. Zirbel, The Texas Connection (Scottsdale AZ: Wright & Co., 1991).
Groden and Livingstone, High Treason, p. 155.
Robert Rosenthal and Ralph L. Rosnow, The Volunteer Subject (New York: John Wiley & Sons, 1975).
T.X. Barber, Pitfalls in Human Research: Ten Pivotal Points (Elmsford NY: Pergamon Press, 1976)
E.F. Loftus and G.R. Loftus, “On the Permanence of Stored Information in the Human Brain,” American Psychologist, pp. 546–574.
Myers, Psychology, pp. 546–574.
Harry Stack Sullivan, The Psychiatric Interview (New York: Norton, 1954).
Paul Ekman and Maureen O’Sullivan, “Who Can Catch a Liar?” American Psychologist, 46 #9, 1991.
John P. Houston, Fundamentals of Learning and Memory (New York: Harcourt Brace Jovanovich, 1991).
Kenneth A. Deffenbacher, “Eyewitness Research: The Next Ten Years,” in M.M. Gruneberg, P.E. Morris and R.N. Sykes, Practical Aspects of Memory: Current Research and Issues (New York: John Wiley & Sons, 1988), pp. 20–27.
Houston, Fundamentals of Learning and Memory.
Myers, Psychology, pp. 253–285.
Sigmund Koch, “The Nature and Limits of Psychological Knowledge,” American Psychologist, 36 #3, 1981.
Koch further suggests that the isolated research community gradually assumes the nature of a cult. It gains its holy books, its clergy, its novices, and its heretics; I hesitate to add its martyred President.
Martin Shackelford, “Report From Dallas: The AS Symposium, November 14–16, 1991,” The Third Decade 8 #2,3 January–March, 1992, pp. 2–3.