I’ve encountered this argument enough in NT studies that I’m really shocked that scholars think that is has so much weight. Simply put, scholars think that just because a saying or pericope in the gospels (like Mark 5.41) has an Aramaic source that this means that it must go back to Jesus himself. Sure, if some saying or episode was originally penned in Greek (like John 3.1-30) then it is pretty much impossible for it to go back to an Aramaic speaking Jesus. But the converse is not necessarily true; an Aramaic saying possibly goes back to Jesus. But just because something is possible doesn’t mean it’s probable.
I don’t want to state the obvious, but I have to. The “Aramaic Therefore Jesus” fallacy is literally a logical fallacy. It’s a smaller version of Affirming the consequent:
P1: If Jesus said something, then it was in Aramaic
P2: Something is in Aramaic in the gospels
C: Therefore, Jesus said it
It’s really sad that textbook logical fallacies are being trumpeted as foolproof evidence for the historical Jesus.
The thing is, if NT scholars were thinking like Bayesians, then they wouldn’t fall into this trap. Even though it’s a logical fallacy, it’s not necessarily a fallacy of probability. If we take into account alternative hypotheses for the source of an Aramaic saying or episode, then we might see that the presence of an Aramaic source isn’t strong enough Bayesian evidence to rest a conclusion on. Aramaic could be evidence for a historical Jesus, but Aramaic could also be evidence for a preaching and exorcism performing Cephas (Mark 6.7-13). It’s not like Cephas didn’t speak Aramaic.
Because of all of the other possible sources for Aramaic sayings within the early (Aramaic speaking) Christian community, an Aramaic pericope/saying is very, very, very weak probabilistic evidence for a historical Jesus.
H1 = Jesus said it
H2 = some other early Christian said it
~H = every other explanation
E = Aramaic saying
P(E | H1) = probability of an Aramaic saying given Jesus said it
P(E | H2) = probability of an Aramaic saying given some other early Christian said it
P(E | ~H) = probability of an Aramaic saying given some other explanation
P(~E | H1) = probability of some other language given Jesus said it
P(~E | H2) = probability of some other language given some other early Christian said it
P(~E | ~H) = probability of some other language given some other explanation
Obviously, under the assumption that the historical Jesus only spoke Aramaic, P(~E | H1) has to be zero. This, in turn, means that P(E | H1) has to be 1.00. Next, under the assumption that the majority of some other early Christians only spoke Aramaic, P(~E | H2) would be a bit more than zero, since there probably were early Christians who spoke some other language. So while not as extreme as an Aramaic-only speaking Jesus, it’s still pretty extreme. We would have to know the percentage of how many of the first Christians only spoke Aramaic. Since I don’t know that number, I’ll say that most (c. 90%) spoke only Aramaic. Thus P(E | H2) would be .9. Under some other explanation, there really doesn’t seem to be any constraints on that one. Given a mythical Jesus, there’s no language that would falsify it; sayings could have originated in Aramaic, Greek, Latin, or whatever. Thus P(~E | ~H) seems to be no different than P(E | ~H) so P(E | ~H) I would generously place at .5
There’s another way of going about divvying up the conditional probabilities, and that would be to find the Total Probability. That would just be a frequentist approach of counting up all of the sayings/pericopae in the gospels and seeing how many of those are originally in Aramaic. Whatever percentage that is would be the denominator of Bayes’.
Now that I think about it, not only is the “Aramaic therefore Jesus” fallacy an Affirming the Consequent fallacy, it’s also a Prosecutor’s Fallacy. Two fallacies for the price of one! (Or maybe all Affirming the Consequent fallacies are Prosecutor’s Fallacies?).
So let’s say I was unsure of whether a saying went back to Jesus or went back to some other original apostle (i.e. both are .4 or 40%), with the alternative hypothesis taking up the rest. How much would the presence of Aramaic increase the probability of a historical Jesus?
P(H1 | E) = P(E | H1) * P(H1) / [P(E | H1) * P(H1)] + [P(E | H2) * P(H2)] + [P(E | ~H) * P(~H)]
= 1.00 * .4 / [1.00 * .4] + [.9 * .4] + [.5 * .2]
Going through all of the math, the probability of a historical Jesus given an Aramaic saying/pericope went up from 0.4 to 0.465. But, the probability of some other Aramaic Christian saying it given an Aramaic saying/pericope also went up from 0.4 to 0.417. And the probability of some other explanation (i.e. mythicism) went down from 0.2 to 0.17.
So again, it looks like this is another case of a logical fallacy being weak Bayesian evidence. Obviously looking at it this way, we can see that the problem is differentiating between a saying going back to Jesus and a saying going back to some other early Aramaic speaking Christian. Hell, I didn’t even attempt to factor in an Aramaic saying going back to some non-Christian altogether, I just lumped it under ~H. A difference of .46 and .41 isn’t very strong evidence; certainly not strong enough to put all of your weight behind it, like what certain recent scholars have done.