Why "Who did it?" is the wrong question to ask first and maybe ever
26 February 1999

    Everybody who becomes interested in the JFK assassination starts by automatically asking "Who did it?" For example, how many books can each of us name that have titles like "Who killed JFK?" Although it is very natural to ask this question first, it is the wrong one to begin with because it presumes that we can actually determine the answer now or in the future. There are several reasons why we might never know who killed the president. First, there are matters of principle: (1) If the assassination was truly the work of one person, we cannot ever know that with certainty because it is impossible to prove that others did not participate. You can't just take the word of the presumed lone assassin because she might be lying. (A little bit of PC for our friends who get hung up on such things.) Rather, you have to check out all other possible coconspirators to verify that they were not involved. Proving individually that each of them was not involved is the classical problem of proving a negative—you can prove that they were involved, but not that they were not involved. Thus we can never really prove nonconspiracy. We should never forget that the inability to prove nonconspiracy is a high principle—it has nothing to do with the nature of the available evidence. (2) If, on the other hand, the assassination was the result of a conspiracy, it can be proven so, provided only that the available evidence is strong. Thus, conspiracy can in theory be proven but nonconspiracy cannot. But until we study the nature of the evidence for conspiracy, we cannot decide whether it is strong enough. Thus, before we begin to delve into the assassination, we need to understand that in principle, nonconspiracy can never be proven and conspiracy can or cannot, depending on the strength of the evidence for it.
    Now for two matters of practice. (3) In practice, proving conspiracy was dealt a major blow with the early death of Oswald. With him no longer available to finger others, proving conspiracy instantly became much harder. (4) The same holds for the later death of Jack Ruby. With both him and Oswald out of the picture, proving conspiracy became double difficult. If there had been a conspiracy, it is no wonder that no solid evidence for it has been produced in 35 years.
    The net effect of these four solid results, two of principle and two of practice, is that "Who did it?" is the wrong question to ask first. The right first question is a step before that: "Can we know who killed JFK?" But even this question can be resolved into finer components with subtle shades of meaning. For example, we might restrict the question to our present capability: "Can we determine from current evidence who killed JFK?" Then we might project into the future (i.e., take a wild guess at developments): "Is it reasonable to expect that we will ever be able to know who killed JFK?" We might temper this last version with hindsight: "Given that the last 35 years of intense research have not brought the JFK community any closer to agreeing who killed JFK, how likely is it that a solid answer is still out there waiting to be found?" (Shades of Essay #1) Or in slightly different words: "Given no progress in 35 years of trying, do we have any reason to think that the next 35 will be any different?"
    Another shading to the first question is to address how close to the answer we can come. This can also be broken down into present and future states of the evidence: "How close can we come now to determining who killed JFK?" and "How close might we ever come to determining who killed JFK?"
    These modified questions fit into a logical sequence that looks something like this:

            1. "Can we know now who killed JFK?"

a. If yes, then "Who killed him?"
b. If no, then "How close can we come to knowing now?"

            2. "Will we ever be able to know who killed JFK?"

a. If yes, then someday, "Who killed him?"
b. If no, then "How close will we ever be able to come to knowing?"

Note that only two of the four scenarios allow us to ask who did it, and only one of those two scenarios includes the present time. Thus if we give all scenarios equal weight, there is a 50% probability that we will never be able to ask who did it, and a 75% probability that we should not be asking it near the beginning of our research.
    Moral: some starting points that seem obvious may be seriously wrong!
    In closing, I wish to emphasize that this entire essay uses the word know as equivalent to prove, that is, implying 100% probability, or certainty. It must be understood that we may not know a particular thing but still have extremely high probabilities for it. For example, we may not be able to prove that Lee Harvey Oswald actually fired the shots that killed President Kennedy, but we can come extremely close to it: we know that he ordered the rifle, that he practiced with it, that he used it to try to kill General Walker six months before the JFK assassination, that he brought it to the Depository the day of the assassination, that someone looking very much like him was seen by at least two witnesses firing from the sixth floor, that he had been on that floor just before the assassination, that his rifle was fired that day, and that all the fragments of bullets that were recovered and properly tested were traced to his rifle. About the only thing we cannot do, and probably will never be able to do, is to put his finger on the trigger. Thus the web of circumstantial evidence against Oswald is extremely deep, to the point that he would have almost certainly been convicted in court. Most importantly, we have zero evidence of this type against anyone else. Do we know that Lee Harvey did it? No, we don't, and this cannot be stressed too much. But do we have evidence against him? Absolutely. We have massive circumstantial evidence, perhaps to the 95%–98% level. But we cannot prove he did it. I believe that each of us who study this case learn to think simultaneously along these two tracks of proof and probability.
    This is why when anyone asks me if I know who killed Kennedy, I always answer first , "No." If I have a chance to supplement that answer, I add two statements: (1) If not forced to choose between conspiracy and nonconspiracy, I answer truthfully that we do not know which it was; (2) If forced to choose, I state truthfully that since all the strong evidence in the case is equally consistent with conspiracy and nonconspiracy, we must choose the simpler option of nonconspiracy, with Oswald as the only reasonable choice for the assassin. To put it another way: we can be certain both that we cannot prove who killed JFK and that have extremely strong evidence against Lee Harvey Oswald and no one else.