Conspiracy thinking reveals the utility of internal simulation, but also helps to explain the chronic ad hoc reasoning of the paranoid mind. Even Herbert A. Simon is willing to concede the point that, "A simulation is no better than the assumptions built into it" (Sciences 18). In effect, in every instance of the ad hoc addition of an assumption, the end result of the simulation will change; in principle there is no end to the process of ad hoc additions, hence no end to the utility of the paranoid's continual simulations (the domination of utilities over probabilities).

Unfortunately, utilities, unlike probabilities, are objectively ego-relative even when they are objectively assessed. A sudden freeze in Southern California is good if one is a ski lodge owner, bad if one is a strawberry farmer. Paranoia only exacerbates the inherently conflictual nature of human interests. Reflection upon the radical divergence of opinions between whites and blacks on O. J. Simpson's guilt illustrates the problem. The objective probability that detective Mark Fuhrman could have done what Simpson's defense lawyers accused him of doing is vanishingly low regardless of his racial animosity. Given the evidence of the case, it would have required of him frankly supernatural powers of prescience, organization, memory and dissimulation to perform what was asserted. However, a person belonging to the racial group of which Fuhrman had expressed genocidal desires might very well edit these improbabilities out of his consciousness, leaving nothing remaining but the overvalent idea: the thought that, "this man is on record as out to destroy me and my kind." Under such circumstances, the demand for survival dictates that the utility of overestimation exceed that of underestimation.

During the course of the trial in 1994­95, Simpson's defense team contended that a possible Los Angeles Police Department conspiracy existed, consisting of at least Detective Mark Fuhrman, whose intention it was to frame their client. The fact that Simpson's blood was apparently found on the infamous bloody glove entails that some person or persons in the Los Angeles police crime lab, having access to blood collected from Simpson at the jail, be involved in the plot as well.

Suppose that a defense team (or prosecutor, for that matter) should subpoena and examine N different hostile witnesses in a trial which it avers a conspiracy has occurred. For this set of witnesses there exist 2N-1 distinct subsets of them, to each of which there corresponds a set of testimonies that must remain internally consistent under interrogation, or else produce the appearance of a conspiracy. The proof is as follows: Designate the witnesses as W1, W2 . . . Wn. Each of them either does or does not belong to any given subset of them. Therefore there are 2x2x2...x2}N ways of forming such subsets. Substracting from this product that subset which contains none of the witnesses (that is, is empty in all places), you obtain 2N-1 distinct subsets of witnesses.

Each subset of size M In fact the real key is the size of the witness list, not the size of the purported conspiracy. Thus, for a witness list of size N and a hypothesized group of conspiring witnesses of size M A number of observations need to be made at this point, some cutting in one direction, some in the other. An intelligent search must surely cut down on the time that would otherwise be required to conduct a brute-force search for inconsistent sets of testimonies, probably by several orders of magnitude. This is simply the value of experience and knowledge to an attorney and his investigators. In a trial some witness testimonies are modular with respect to others, that is, no matter their content, they will neither reinforce nor contradict each other. Whether this degree of modularity­which can be represented as the average degree of interconnection between the nodes of a planar graph­is a function of the size of the witness list, that is, of the number of nodes in the graph­is unknown to me, and is probably a fit topic for a thesis by some mathematically inclined legal scholar. (Perhaps it has already been written).

The most important additional factor is the degree of accuracy of the testimonies given by non-conspirators, that is, of honest testimonies. The problem is that, given large witness lists and extensive investigatory resources, a good attorney can make a non-conspiracy simulate a conspiracy, simply by virtue of the fact that fallible human beings are subject to slips of the tongue, faulty memories and so on, so that honest witnesses can get confused and contradict themselves or each other. Once again, as the number of witnesses increases arithmetically, the likelihood of this disaster happening increases exponentially. In such instances the reason for the contradictions is innocuous enough, though the consequence is perfidious: this is exactly the result which a defense attorney representing a client he knows to be guilty would want to see. In principle it should be possible to distinguish between contradictions generated by honest inaccuracies and those generated by lies, since the average rate of occurrence of false testimonies by honest witnesses should be less than that of dishonest ones. For one thing, a liar is as likely as a truth-teller to also be inaccurate on those items he is not lying about, increasing his relative rate of inaccuracy. By how much? One suspects a great disparity here, though I have researched whether such statistics have ever been compiled. Of course, it is also likely that large qualitative differences exist between the kinds of contradictions generated by dissimulations and accidental factors, with the former being more consequential (in general) than the latter.

The conclusion seems to me certain: large witness lists and big, protracted trials favor guilty defendants for purely mathematical reasons, tending to evoke delusions of conspiracy where none, in fact, exist. In this case there would be no practical benefit to actually engaging in a conspiracy even if it were not uncovered, since it could not accomplish its aim anyway. The defendant will always wind up looking like the victim, and the cops will only put their own careers (if not their freedom) at risk by pursuing an absurdly unlikely gain. Large conspiracies simply cannot work when they are subject to extensive interrogation after the fact, nor is anyone likely to participate in one who realizes it's definitely going to be so subject after the fact, as in the case of a police conspiracy (or the assassination of a president).

In his letter to the New York Times the Unabomber continually referred to "we," which can be read as his attempt to create the notion of an organized conspiracy, that is, a vast network of conspirators working in unison even though the FBI consistently believed the Unabomber to be a lone, white male. For reasons detailed above, the FBI believed all along that the Unabomber was a lone individual, not a member of a large network of conspirators. There are simply too many possibilities for failure as more conspirators are brought into the fold, and the Unabomber simply never made one in seventeen years. However, because the prime suspect in the Unabomber case left writings which were later found by his brother, it only confirms how difficult conspiracies are to enact: all it took was one other person (two if you count the mother) to expose him. On the other hand, "underground," terrorist groups and organizations, following the lead provided by Gillo Pontecorvo's film The Battle of Algiers (1965), fragment into "cells" of three or four and operate without any contact with other cells. In such a way the whole organization remains active even if one particular cell is exposed or arrested; an analogy would be to the scientists who built the Saturn V moon rocket. No one scientist knew all the systems, though a large number of scientists were able to build the rocket because all were directed toward a common goal. Again I can turn to Herbert A. Simon, who in Sciences of the Artificial observes that there are levels of rational behavior "above the individual level" (48). In the field of economics, "markets" are one such mechanism to achieve such a form of rationality.

The fact that the Unabomber claimed to be part of a vast conspiracy is an example of his use of "white noise" strategy. Large conspiracies simply cannot work when they're subject to extensive interrogation after the fact, nor is anyone likely to participate in one who realizes it is definitely going to be so subject after the fact, as in the case of a police conspiracy (e.g., alleged in the O. J. Simpson trial) or a conspiracy to assassinate a president (JFK). Large conspiracies sometimes achieve their ends, of course, not when the goal is to convince an adversary of the truth of a false story, but rather when it is to confuse him about which one of several false stories is the true one: this "white noise" strategy was the one used by General Eisenhower on D-Day, in which he offered German intelligence a whole host of alternative possible invasion plans to react against, thus forcing them to dilute their defensive resources: a shell game. This is yet one more example of the Unabomber's intelligence (a prima facie case exists for his high intelligence in that he avoided being caught for almost 18 years). His letters attesting to the existence of a vast and ubiquitous conspiracy to undermine industrial society strike me as an attempt to simulate the existence of a conspiracy, and is an example of the "white noise," a game playing strategy used in situations of unmanageable conflictual interests. The Unabomber's methods are a form of game-playing in that he was trying to simulate or mimic random processes. His motive is intimately bound to the strategy of encryption, for the existence of dangerous enemies always provokes the development of encryption techniques, which explains their prominence in military and industrial espionage. As an example, think of Galileo's motives for encrypting messages in order to avoid the Inquisition.

WebGlimpse	Search Options
Search: The neighborhood of this page The full archive