Rhetorical Probability


I always find it amusing when I am sitting in a doctor’s office during the middle of the day how the only thing on the TV is a show involving some manner of litigation. I don’t know if Judge Judy is still doing her thing, but there always seems to be some ridiculous sort of court case being televised.

Lest one think this to be something novel to our society, we have nothing on the fifth-century Athenians.

Athenians loved to litigate and argue and dispute with each other. In many ways, it was a form of entertainment, for the juries were large. (between 201 and 501 jurors) The Athenian democracy assumed that you were going to participate, and any man could speak if he wanted.

In fact, if you were involved in the litigation somehow, whether of a civil or criminal manner, you were expected to speak on your own behalf. During this period the idea of a having lawyer with a formal understanding of legal procedures and precedents was unknown. As such, one’s rhetorical ability became a rather important aspect of winning the case.

Litigation was largely characterized by a series of running speeches- the plaintiff would begin by addressing the jury, followed by the defendant’s rebuttal. In each case the primary form of evidence was oral rather than physical. Of course, in many cases the plaintiff would clearly have an advantage in that the defendant would be far more likely to have less experience in delivering any kind of oratory to such a large group of people, especially in criminal cases.

This eventually led to the production of handbooks on rhetoric that could be studied by unskilled speakers to learn the basic principles of oratory to help give them the edge they needed in the midst of their litigation. These handbooks have a fairly specific structure:

1. The proomion. This is the introduction in which the speaker attempts to bring the audience into a favorable impression.
2. The diegesis. This is the bulk of the speech, mostly narrative in function into which evidence, witnesses and probabilities are woven.
3. The pistis and epipistis. This section is where proofs and supplementary proofs are presented.
4. The elenchos and epexelenchos. This is where the speaker presents counter-arguments to his opponent.
5. The epilogos. A recapitulation of the entire argument.[1. George A. Kennedy, Classical Rhetoric & Its Christian and Secular Tradition From Ancient to Modern Times, p. 21-22]

Of all of these elements, the diegesis was the most important, and within this sub-section the notion of probability was seen by the rhetoricians as the crux of the entire argument. That is, one was far more likely to find success arguing one’s case if one could convince the jury of the probability of one’s argument. The handbooks of rhetoric suggest that the actual veracity is far less important than the ability to convince one of the probability of it. Socrates was not terribly impressed with this sort of thing:

…Tisias and Gorgias showed how much more probabilities were than facts and could make small things great, great things small, new things old and old things new and discuss anything for any length.[2. ibid. p. 21-22]

Yet the Athenians, even acknowledging that such a thing was occurring, seem to have preferred this manner of disputation and litigation. To our modern sensibilities, that can seem kind of strange. After all, forensic evidence has come a long way in which we are able to determine a number of things from physical evidence alone. Thus, to imagine a system of law in which juries decide by the weight of an argument from probability that may or may not have any verity with facts can seem like a road to nowhere.

The Athenians were no strangers to physical evidence, and it didn’t hurt to weave that into one’s argument from probability. However, experience is also a hard task-master. It taught them that evidence could be planted or faked, witnesses could be bribed, slaves could be tortured for false information. All of these mitigating factors left enough doubt in their minds about ‘facts’ that being able to show the probability that something did or didn’t occur was at least as reasonable a course. In the end, it wasn’t about if it actually happened in exactly this way, but rather if one could make it seem like it did to the jury. A short example of this sort of eikos:

If a weak and brave man, having beaten up a strong and cowardly man, is brought into court, neither must tell the truth. The coward must claim that he was not beaten by a single brave man: that is, he must claim to have been attacked by two or more; whereas the other must refute tho, insisting that the two of them were alone, in order to use the argument “How could a little one like me have attacked a big on alike him?”[3. ibid. p. 23]

With this kind of back and forth subtlety, it is no wonder that the Athenians found litigation so exhilarating! To attempt to tease out the truth from two litigants whom you know are trying to present a probabilistic account that anyone might deem reasonable might have been a rather sublime experience.

In the fifth century with the advent of the Sophists came the first attempt at creating schools of a sort for rhetoric. While ‘sophistry’ has a negative connotation in the modern world, it originally meant ‘wise’ or ‘expert.’ The schools of the sophists consisted mainly of listening to the sophist speak and attempting to memorize and mimic his style. In some cases example speeches were written to give the student something to study. Different situations and scenarios were developed so that a wide berth of potential judicial contingencies could be met and probabilistic arguments developed and adapted to the situation. A rather famous example comes from the Tetralogies involving a homicide held before the Areopagus:

Whatever actions result from plots by ordinary citizens are not difficult to prove, but if people of considerable ability are the perpetrators, experienced in the business, and at a time of life when they are the peak of their mental powers, they are difficult to discover and convict. Through the greatness of risk, they pay much attention to the safety of their schemes, and they take no action until they have provided against every suspicion. Knowing this, it is necessary for you [the jury] to put great trust in the probability you perceive… It is improbable that muggers killed the man, for no one who rant the risk of his life would have abandoned the object of his robbery when he had it in his hands. Yet the victims were found with all their property intact.

Nor did someone kill them in a drunken frenzy, for we would have information from his fellow drinkers. Nor did the murder result from argument, for they wouldn’t be arguing in the middle of the night in a deserted spot. Nor did the murderer kill the victim in mistake for somebody else, for he would not have killed both him and his slave.

Since these possibilities are dismissed, the fact of the death points to the man having died as a result of premeditation. And who is more likely to have set upon him than one who has suffered wrongs at his hands and was expecting to suffer still more? That man is the defendant. [4. ibid. p. 32]

It is easy to see that the actual physical evidence of the case (the dead slave who actually pointed out the defendant before dying, the lack of robbery, lack of inebriation, etc.) is secondary to the idea that the probability of these things occurring is so remote as to be non-existent, and in fact the probabilities themselves point to the defendant as the murderer, rather than the physical evidence.

The defendant then gets a chance to respond:

It is not improbable, as they say, but probable that a person wandering around in the middle of the night should be killed for his property. That he was not robbed is a sign of nothing. But if the assailants had not yet stripped him, but left in fear of someone who was coming, they were sensible and not at all mad to prefer safety to profit… Further, whoever allows probability the force of fact when it testifies to my guilt must on the same principle bear the following in mind as evidence of my innocence; it was more likely that, with an eye to carrying out my plot in safety, I should take the precaution of not being present at the scene of the crime than that the slave should recognize me distinctly just as his throat was being cut. [5. ibid. p. 32]

In an interesting twist, the defendant turns the tables, insisting that it is actually quite probable that people get mugged and killed at night when they are wandering around, since this a common experience with which the jury would be familiar. Additionally, by impugning the honor of thieves (whom he suggests will run at the slightest hint of danger) he not only shows a probability that they might actually leave without stealing the murdered man’s property, but by extension (since he is not behaving at this moment in a cowardly manner by addressing the jury) distances himself in degrees of probability from being that sort of person.


Unfortunately, we are not told how this trial ends. So imagine you are an Athenian juror in the fifth century.

Is he guilty or not?

Which is more probable?


  • My prowess in this area generally consists of “Colonel Mustard in the kitchen with the rope,” but here I go…

    I would say that both arguments are equal or very close in probability, so we must look at the consequences of each in the case that we’re wrong. If we convict the man and he’s innocent, then we’ve killed two men. If we let the man go and he’s guilty, he may kill again. In this case, the deciding factor becomes the accused’s character–has past behavior indicated that he would be capable of this?

    I’m probably breaking the rules a little here, but that’s how I would settle things.

  • I agree about the equality of probability, which is probably (oh!) why Antiphon used this particular example. Very tricky, that Antiphon.

    Interestingly, your approach is actually very similar to the way the jury would have settled things, from the very little that I know about this subject.

    I wasn’t able to present the entirety of the arguments in the interest of some measure of brevity, but the prosecution does argue at some point that if this man goes free, he will probably kill again and spread his pollution throughout the greater body politic.

    The defendant also makes the counter-argument about killing two men unnecessarily if he is convicted, the injustice involved, etc.

    So your method would not be breaking the rules, but rather you would be acting as a good Athenian jurist.

By deviantmonk

Be Social



Secret Archives