Coincidence Theory

The below digression is essential reading for anyone who believes that it can easily be established that “coincidence” can explain the connections that I say point to a debasement theme in Measure for Measure.

co·in·ci·dence n. 1: an event that might have been arranged although it was really accidental (from www.dictionary.com)

The Shaksper threads demonstrate that at least some Shakespeare enthusiasts have trouble understanding how to evaluate an argument that is based on circumstantial evidence. One contributor in particular has been persuaded – probably correctly – that the similarities between the Earl of Oxford’s life and Shakespeare’s plays are coincidences. Unfortunately, he appears to have gone on to persuade himself that all events that can plausibly be attributed to “coincidence” must be coincidences.

Coincidence sometimes is the best explanation. But the determination that something is a “coincidence” cannot be divorced from the initial inquiry of how probable or improbable the “no coincidence” or “design” explanation is. The reason that we can be fairly sure that the parallels between Shakespeare’s work and the Earl of Oxford’s life are coincidences is that we have very good reasons – based on contemporary references – to believe that William Shakespeare of Stratford wrote the plays, plus at least one very good reason to believe that the Earl of Oxford didn’t (he was dead before several of the better ones were written). On the other hand, if it were independently established by incontrovertible documentary evidence that the plays had indeed somehow been written by the Earl of Oxford, then the parallels would no longer be coincidences – we could be fairly sure that the Earl was indeed writing about his own life in All’s Well That Ends Well, that Hamlet’s Horatio and pirates were based on the Earl’s friend and life-story as well.

In other words, before attributing a set of parallels to “coincidence,” it is critical that one look to extrinsic considerations that help one make that determination.

When conducting a murder investigation and encountering circumstantial evidence pointing to a given suspect, murder investigators don’t simply say – “I had a case like this last year, and the killer turned out to be someone else, so all this circumstantial evidence must be the result of coincidence.” No – instead, one considers whether there are other reasons to believe the suspect committed the crime – motive, opportunity, DNA evidence, proximity to the scene of the crime, etc.

An even more pernicious problem with assuming that because coincidences are common, “coincidence” is going to be the answer every time we confront a choice between “design” or “coincidence,” is that sometimes, the “coincidence” explanation, even apart from extrinsic evidence, is simply implausible. For example, if Joyce hadn’t confirmed with its name that “Ulysses” is based on Homer’s Odyssey, one might read the book without noticing the parallels, and even when some of the parallels are pointed out, ascribe them to coincidence. But at a certain point, after all the parallels have been pointed out, the “coincidence” theory becomes far less plausible than the design theory.

There is no mathematical model for determining when a given set of parallels between a literary work and something else (be it another literary work, a political issue, or putative author’s biography) should be ascribed to “coincidence” or “design.” We may assume, however, that for any two datasets, there is a “concentration” of parallels for which the “coincidence” explanation becomes so unlikely that the most logical explanation is “design.” In the case of Ulysses, the concentration of parallels exceeds this threshold.

Accordingly, it can be useful to ascertain whether the concentration of parallels that one has noted is uniquely high. This is in fact how we can be fairly sure that the parallels between the Earl of Oxford’s life and Shakespeare’s plays, as well as the parallels between Henry Neville’s life and Shakespeare’s plays, are likely to be coincidences. The fact that we can find a roughly equivalent “concentrations” of parallels between the lives of either candidate and the plays shows that neither is the sort of “uniquely high” concentration that should cause us to reject the coincidence theory in favor of a design theory.

How would one use this methodology to test the “debasement” theory of Measure for Measure? The best way would be to demonstrate, for some other Shakespeare play – preferably for any other Shakespeare play – that an equivalent concentration of parallels to a debasement theme could be found. Take a look. If you find one, get back to me.

Absent such a showing, we are left with the (for some) uncomfortable fact that the “design” explanation is more likely than the “coincidence” explanation. That doesn’t necessarily disprove the coincidence explanation – even implausible coincidences occur quite frequently – but it means that we must examine and evaluate the extrinsic evidence prior to rendering judgment.

In the context of the argument about Measure for Measure, therefore, it is not enough to simply say “coincidences are common, so this is coincidence.” Since the coincidence theory proponent cannot show that the concentration of parallels is still within the “expected” range (and thus, easily attributable to coincidence), the burden should be on the coincidence theory proponent to come up with some convincing reason that Shakespeare would not have put debasement into the play. This should involve evaluating the likelihood that Shakespeare intentionally placed this particular theme in this particular play, and presumably balancing this against the question of whether Shakespeare might have learned of Juan de Mariana’s views a year before their first confirmed publication, 400 years ago (which in turn requires an honest evaluation of the intrinsic evidence relating to Juan de Mariana). To make an informed judgment, take a look at Shakespeare’s Economics; Birdseye View – Measure for Measure; Predictive Value; Juan de Mariana.