Tuesday, 14 August 2018

Acts 12 or 15?

No doubt, the single most pivotal question for establishing a Christian chronology of the 30s and 40s is whether Paul's second journey to Jerusalem reported in Gal. 2:1-10 refers to the journey narrated in Acts 11/12 (usually abbreviated simply to "Acts 12" in chronological discussions) or Acts 15. The majority of scholars hold that it refers to the journey narrated in Acts 15, whereas a minority of scholars hold to that of Acts 12. Other solutions--that it refers to the journey narrated Acts 18, for instance, or that there is no correspondence between Acts and Paul on this matter--are generally non-starters, for reasons that need not distract us in this post. When we take into account such matters as validity, scope, and parsimony, the hypothesis that Galatians 2:1-10 refers to either the journey narrated in Acts 12 or that in Acts 15 remains that which can best explain the data that we find in both the Lukan and Pauline material. But beyond that, should we prefer the Acts 12 hypothesis, or the Acts 15?

Let us begin with the Acts 15 hypothesis, as it is the majority report. The Acts 15 hypothesis depends upon supposed parallels between this text and Gal. 2:1-10. In Gal. 2:1-10 Paul describes a meeting between himself and "the pillars" in Jerusalem--James, John, and Peter--in which the latter affirm that the former had been entrusted with the gospel for the Gentiles. The Acts 15 hypothesis sees in this an impressive parallel with the Jerusalem Council of Acts 15, which narrates a discussion about whether or not Gentile converts should be circumcised. Now, prima facie, the parallels are indeed strong. There are however at least two flies in this ointment. One, the Acts 15 hypothesis has a hard time accounting for Gal. 2:11-14. Gal. 2:11-14 describes a conflict in Antioch over the issue of whether Jewish and Gentile persons can eat together between Paul on the one side and "men from James" on the other, with Peter and Barnabas in the middle, while Acts 15:1-2 states that the conflict in Jerusalem began with a dispute over circumcision in Antioch. Probably most iterations of the Acts 15 hypothesis suppose that the events of Gal. 2:11-14 precede those of Gal. 2:1-10, with the former paralleling Acts 15:1-2. This is deeply problematic, as there is frankly no hint whatsoever in Galatians that Paul intends the reader to think that the events the he reports in 2:11-14 precede those of 2:1-10. Quite the opposite: the "But when" (ὃτε δὲ) with which Paul initiates 2:11 most naturally suggests a temporal progression. Ultimately, although probably the most popular solution, the hypothesis wherein Gal. 2:1-10 parallels Acts 15:3ff. and 2:11-14 parallels 15:1-2 seems to flounder on the data. That said, a variant iteration of the Acts 15 hypothesis is probably more viable on this matter. In this iteration, Acts 15:1-2 has no parallel in Galatians, Acts 15:3ff. parallels Gal. 2:1-10, and Gal. 2:11-14 narrates a conflict that emerged subsequent to the council. This second iteration does not require us to gratuitously suppose that Paul is narrating the course of events out of temporal order, but it also does not escape the second challenge faced by the Acts 15 hypothesis, namely the number of journeys to Jerusalem undertaken by Paul.

In Gal. 1-2, Paul is much concerned to show that he had little contact with the leadership in Jerusalem. He lists the times that he went, specifying with whom he met and how little they contributed to his understanding of the gospel. In 1:18, he mentions that he went to Jerusalem after three years and met with no apostles but Peter and James. This journey can be quite unobjectionably be identified with that narrated in Acts 9:26-30. In 2:1-10, he states that he went to Jerusalem after fourteen years, and met with Peter, James, and John. On the Acts 15 hypothesis, this second journey mentioned in Galatians is actually the third narrated by Acts. One either has to conclude that Paul has failed to mention that second journey, or that Luke has introduced a journey that never happened. The former hypothesis is often advanced, on the basis that Paul is only concerned to narrate instances in which he interacted with the leadership in Jerusalem. Unfortunately for this hypothesis, Acts 11:30 of course specifies that Barnabas and Paul were sent to meet with the elders on the second journey. Now, yes, elsewhere Luke will distinguish between apostles and elders, and Paul specifically in Gal. 1-2 refers to meetings with apostles. But if anything that weakens the Acts 15 argument, for if Paul is concerned to show how little interaction he had with the apostles in Jerusalem then surely it would serve his rhetorical purpose to mention a journey to the holy city in which he didn't even meet with them (if it is objected that this construal of Pauline intent is speculative, I would observe that it is no more so than the argument that Paul is only interested in narrating journeys to Jerusalem in which he met specifically with the apostles). And in any case, there is enough slippage between the elders and the apostles that we probably do not have adequate warrant to conclude that the elders of Acts 11:30 must exclude the apostles.

By contrast, the Acts 12 hypothesis not only reads Gal. 2:1-14 in sequential order, has a very straightforward explanation for why Paul only mentions two visits to Jerusalem: as of the time that we was writing Galatians, he'd only taken two visit. In this understanding, Gal. 1:18 refers to the visit of Acts 9, 2:1-10 to the visit of Acts 12, and 2:11-14 to the events of Acts 15:1 (although not of Acts 15:2; rather, we need to assume that Paul is writing as he prepares to travel to Jerusalem for the council). In short, it combines the best of both worlds. Indeed, there is virtually only one substantial challenge that can be raised against the Acts 12 hypothesis, and that is that Luke doesn't mention any sort of conflict or discussion that takes place during that visit. Such an argument however is limited in its strength, for at least two reasons: one, Luke is always highly selective in what he presents; two, as recognized since at least F.C. Baur, Luke tends to emphasize the irenic side of inter-apostolic relationships, and as such is not necessarily inclined to report every single conflict that took place. Indeed, given the first of these reasons, lacunae in Luke's accounts should generally not be taken as evidence of absence (this, incidentally, is quite the opposite situation of the problem with the number of journeys, wherein on the Acts 15 hypothesis it is Paul who has a lacuna precisely where we'd expect there not to be one).

For comparative purposes, I have devised a handy chart.


Acts 12
Acts 15/Gal. 2 sequential
Acts 15/Gal. 2 non-sequential
Gal. 2:1-14 in sequential order?
Yes
Yes
No
Both are Paul’s second journey?
Yes
No
No
Acts 15:1 has Galatian parallel?
Yes
No
Yes
Acts 15:2 has Galatian parallel?
No
No
Yes
Both mention a conflict?
No
Yes
Yes

Given that it is unlikely that Paul is narrating events out of sequence in Galatians 2:1-14 and the Lukan tendencies to elide details and present an irenic front, the sequential order and the number of journeys should probably be taken as the most definitive factors. And when this is recognized, Acts 12 presents as the strongest hypothesis.

And once again, that is how historians do.

Thursday, 9 August 2018

The Date of the Crucifixion

As long-time readers of this blog will know, I am mildly obsessed with the question of chronology, that being a mildly huge understatement. For more than a year now, the process of interviewing for, being hired at, and moving to the Lonergan Research Institute in Toronto--while welcome and wonderful and a genuine honour--has kept me from really indulging that obsession, but now that I'm settled in the compulsion to precisely date events that occurred in early Christian history has returned with a vengeance. So, pulling together some recent conversations and work, here is my take on that particularly significant event of Christian history, Jesus' crucifixion.

My general procedure when seeking to date an event is to begin with the data that permits an absolute lower date and an absolute higher date, then work with the data that leads me to raise and lower those dates respectively. An absolute date is one with a numbered year, and contrasts with a relative date. For instance, if I say that Germany invaded Poland in 1939, that is an absolute date. If I say that Germany invaded Poland twenty-five years after the First World War broke out, that is a relative data. In the case of the crucifixion, the relevant absolute dates are 27 and 35. We know Jesus did not die before 27 for a cluster of reasons. First, he died in April, at Passover; the Passover at which he died had to be after Pilate arrived in Judea in 26; it had to be after the beginning of the fifteenth year of Tiberius (Luke 3:1), which began no earlier than October of 26; thus 27 is the earliest April in which Jesus could have died. We know that he did not die after 35 because that is almost certainly the last Passover that Pilate spent in Judea. 26 and 36 are just barely permissible, if we ignore the data from Luke 3:1 and play loose with Pilate's chronology, but in reality are improbable to the point that we can treat them as non-starters.

For the better part of a century, the primary way that scholars sought to narrow this down was to employ astronomical data. Jesus not only died at Passover, he died at a Passover that fell on a Friday. This yielded 30 and 33 as the most likely years, and scholarship probably for the most part opted for the former number, with the latter constituting a very respectable minority report. I no longer think that we can rely so heavily on the astronomical data. The difficulties of lining up modern astronomical observations with the practical realities of an ancient lunar calendar obviate the confidence we can place in such data. That said, I would not dispense with it entirely, and we will come back to it. We can however reasonably raise the lower date up to 29, as John reports that Jesus' ministry spanned at least three Passovers, including that at which he died, and Passover of 29 would be the third after October of 26.

So, how do we narrow it down, then? This requires that we turn to Pauline chronology. We know that Paul converted to Christianity subsequent to Jesus' death, so that event can serve to establish the latest possible date for the crucifixion. In order to establish the date of the conversion, we must first determine when the visit to Jerusalem described in Gal. 2:1-10 occurred. This depends in large part upon whether we associate that visit with the one mentioned in Acts 15, as do the majority of scholars, or with the one mentioned in Acts 12, as do a minority of scholars. (Other options, such as identifying it with the visit of Acts 18, are basically non-starters). If the former, then the visit of Gal. 2:1-10 probably occurred around 48; the latter, then sometime between 40-44. From this we, look at Gal. 1:18, which tells us that Paul first went to Jerusalem three years after his conversion, and Gal. 2:1, which tells us that he again went after fourteen years; we recognize that three years probably here means "more than two but no greater than three," and fourteen "more than thirteen but no greater than fourteen"; that the three years could be included in the fourteen years, or in addition to; and thus the conversion occurred thirteen to seventeen years before the crucifixion. This means that the conversion happened no earlier than 23--which can be moved up to 29, as it must postdate the crucifixion--and no later than 35.

So, how does this help us? In terms of establishing an undeniable, "it definitely couldn't have happened any year other than this," not at all. But in terms of allowing us to make informed judgments--which is where historical reasoning matters the most (anyone can establish possibilities; the hallmark of an actual thinker is the ability to take the risk of being wrong and say that this possibility here seems more likely)--quite a lot, for now everything depends upon whether one thinks that the Jerusalem trip of Gal. 2:1-10 is that of Acts 12 or of Acts 15. If one thinks that it is the trip of Acts 15, then one is still left with 29 to 35 as possible years for the conversion and thus the crucifixion. But if one thinks, as do I (for reasons I'll not get into here, lest the post gets even more bloated), that the trip of Gal. 2:1-10 is that of Acts 12, then one must judge that the conversion happened between the latter part of 29 through 31 (44 less thirteen yielding 31, and 40 less seventeen yielding 23, but again, moved up to 29). Now, although it is possible that Paul converted later in the same year as Jesus died, that seems a bit tight chronologically. The Christian movement proper doesn't get started, according to Acts, until May or June of that year, at Pentecost, leaving very little time for the events and developments that we can infer from Acts 3-8 if the conversion occurred in the same year. Thus, I'm disinclined to opt for a crucifixion in 31. As such, I'm inclined towards either 29 or 30 as the probable years of the crucifixion.

Here is where I would tentatively turn to the astronomical data. Helen Bond has recently challenged whether or not Jesus actually died at Passover or on a Friday, and while she makes some interesting observations ultimately I think that Brant Pitre's treatment of the matter in Jesus and The Last Supper shows that this likely remains the case. And on the astronomical data, I think it more likely that ancient persons would have ended up celebrating Passover on a Friday in 30 than in 29. Thus, I opt for Passover in 30 as still the most probable date for the crucifixion, but with Passover of 29 rather than Passover of 33 as a close second.

And that, my friends, is how historians do.

Nota bene: Douglas Campbell has argued that on the basis of 2 Cor. 11:32, correlated with data from Josephus, the conversion should be dated with certainty to 33 or 34. It is a well-argued position, but ultimately I do not think that the data permits the confidence which Campbell has in his chronology. I have a article on the matter, which is forthcoming in Journal of Biblical Literature, so I won't steal my own thunder here. Alexis Bunine published a solid rejoinder in Revue Biblique, which has not had the exposure that it deserves, probably because it was written in French.

Thursday, 2 August 2018

It's Inference, People

A few weeks ago, a friend of mine who was staying in Toronto for a bit came to visit me at the Lonergan Research Institute. Around 11 am, he texted me to say that he was at Bay and Bloor. I immediately apprehended three things. One, he had taken the subway to get to the LRI. Two, he had taken the Bloor-Danforth line. Three, he had just exited the Bay St. subway station, which is located at Bay and Bloor. How did I know all this? Well, because I know the city, and I knew that he knows the city. I know that Bay and Bloor is the closest station on the Bloor-Danforth line to the LRI. I know that the only reason one would specifically text from that location to say one is in the area is if one was just getting out of the station. I further know that he had exited the station because I know that wi-fi is virtually nonexistent in Toronto subway stations. This is all data, and when presented with the new datum that was this text message, led me to correctly apprehend my friend's past actions, even though I neither witnessed them nor was told what they were.

And that is history. That's how we make judgments. We take our relevant knowledge about the world, and use that to infer what happened in the past. The study of the ancient past does not differ from this methodologically. It tends to be more difficult, not for methodological reasons but rather empirical. Quite simply, our data on the past is spottier than that on the present. That increases the work that must be undertaken to arrive at historical judgments, and requires us to recognize more fully that historical judgments are always implicitly if not explicitly probabilistic. Some can be described as certain, or if we're exercising an abundance of caution virtually certain. For instance, using a modern example, we can be virtually certain that on 31 August, 1939, Hitler ordered German forces to invade Poland on the following day, and that this resulted not only in Polish defeat but also in the start of the Second World War in the European theatre. In the ancient world, we can be virtually certain that the Battle of Issus occurred on 5 November, 333 B.C.E., resulted in Macedonian victory, and was a crucial moment in the conquest of the Achaemenid empire. Other matters will allow comparable certainty, but less specificity. We can be virtually certain that Jesus died on a cross at Passover, sometime between 27 and 35 (after the earliest that we can date the fifteenth year of Tiberius, but before Pilate left Judea); that Paul was converted no earlier than the summer of 27 (as his conversion postdated Jesus' death) and no later than 35 (thirteen years before the Jerusalem conference), and that he died sometime between early 62 and mid-68 (after the end of his two years at Rome, but before Nero's death as we know that he died under said emperor). We can marshal evidence to present arguments for particular dates within those specified possibilities, but barring new evidence no one argument is likely to exceed the level of possibility (although that said, in each case we can suggest that both the very early and the very late dates are far less likely than those in the middle). But the crucial point is that while our capacity to speak with certainty to the specific details of the lives of a Jesus or Paul might be less than our capacity to speak to the life of a Hitler or an Alexander, that is not due to methodological but rather empirical limitations. The fundamental procedure of historical judgment remains the same, whether inferring from someone's message that they took the subway to meet up with you, or inferring from data scattered across a number of relevant ancient texts that certain events had to have happened within a specified range. It's inference, people.