The Carrier Problem: Why the Evidence will not Bed, Part 2
In the first part of this article, we examined why Richard Carrier’s Bayesian methodology failed. Specifically, we looked at the reference class he used to derive his prior probability and saw how it was not only manipulated but stacked in favor of Mythicism.
While the argument in the first article would be sufficient to undermine the central claims of On the Historicity of Jesus, it doesn’t fully settle the question, as a defender of Carrier might respond that even a defective prior could, in principle, still be supported by sufficiently strong evidence.
Carrier himself argues that the documentary record does point that way; that when the actual texts are read correctly, they still support a celestial-Jesus model rather than historicity.
In this second part, we will examine this claim in depth. We will work through Carrier’s treatment of the most important bodies of evidence in turn: the undisputed letters of Paul, the canonical Gospels, the Josephan Testimonies, and the Tacitean reference. In each case, the pattern is the same. Carrier produces a reading of the evidence that is technically possible, but is not probable, and requires a unique reading.
The mythicist case fails at the level of evidence for the same reason it fails at the level of methodology: it relies on selecting the interpretation that supports the desired conclusion.
Carrier on Paul: The Celestial-Jesus Hypothesis
Quite possibly the most important evidence for Jesus as a historical figure are the seven undisputed letters of Paul: Romans, 1 and 2 Corinthians, Galatians, Philippians, 1 Thessalonians, and Philemon.
These are the earliest records we have of Jesus, written by an individual who had contact with the original disciples of Jesus and his brothers.
Carrier understands that the Pauline evidence is probably the best support for a historical Jesus and, realizing this, devotes a substantial portion of On the Historicity of Jesus to neutralizing it. His strategy, inherited from Earl Doherty and developed in considerably more detail, is to argue that the Jesus Paul writes about is not a recently deceased Galilean preacher at all but a purely celestial figure. A divine being who underwent his crucifixion in some sublunar mythological realm at the hands of demonic powers, who was never thought by Paul to have walked the earth, and whose later “historicization” in the Gospels represents a fundamental transformation of the original Christian belief.
This is an extraordinary claim, and the evidence that would be required to support it would have to be just as extraordinary. As we will see, though, it’s not. The hypothesis instead depends on reading every Pauline passage that suggests a historical Jesus as either an interpolation, a mistranslation, or a veiled mythological allegory, while reading every passage that can be construed in celestial terms as being straightforward and decisive. Carrier’s selectivity is not so subtle.
Let’s consider the most decisive passage, Galatians 1:18-19. As a backdrop, we have to understand that Paul is writing to a community in Galatia, where rival missionaries have been moving in and claiming greater apostolic authority than Paul. In response, Paul is attempting to defend the independence and authenticity of his own commission.
In the course of this, he recounts his past in the movement and his acquaintance with the Jerusalem leadership. Paul recalls how three years after his conversion, he went up to Jerusalem and spent 15 days with Peter. The Greek here makes it clear that Paul wasn’t simply chit-chatting with Peter, but becoming more acquainted with him.
Besides Peter, Paul states that he saw no other apostles during his visit except James, the Lord’s brother. A few verses later, in Galatians 2, he describes a subsequent visit during which he again met with James, Peter, and John, whom he names as the pillars of the Jerusalem church.
This passage alone is enough to argue that Jesus was a historical figure. Here we have someone who is a contemporary not only of the disciples of Jesus, but also the very brother of Jesus. Paul isn’t arguing for Jesus’s humanity. He’s not engaged in any apologetic defense of Jesus at all. Instead, he’s making a completely separate argument, but during all of this, he drops a piece of what seems to be common knowledge that didn’t need any defense. Jesus had a living brother named James, who was now leading the Church.
Carrier’s response, following Doherty, is to argue that “brother of the Lord” is a spiritual or fictive-kinship title intended to refer to certain individuals within the early Christian movement, similar to how Christians today might address one another as brothers or sisters in Christ. Carrier’s argument is that James was not actually a biological brother of Jesus but rather a devoted member of this new movement that would become Christianity.
There are many issues with this interpretation, and, as a whole, examining them helps illustrate the broader pattern of Carrier’s arguments here.
Now first, we do have to be aware that Paul does, at times, use the term “brother” in a generic sense for fellow members of this early Christian movement, and he does this rather frequently throughout his letters. But the specific phrase, “the brother of the Lord,” or its plural form, “the brothers of the Lord,” as seen in 1 Corinthians 9:5, seems to be reserved for very specific individuals.
It may be worth briefly looking at the 1 Corinthians passage, as it helps break this all down. Here, we have Paul speaking about the right to have a wife, and in doing so, he lists three groups: the apostles, the brothers of the Lord, and Peter. This is somewhat similar to what we see in Galatians, where we find Peter, the other apostles, and James, the Lord’s brother.
If in either of these passages, brother(s) of the Lord were just a generic term, it would make what Paul says less coherent. It would amount to Paul basically repeating himself, as if the term simply meant “fellow Christians”; then that would also include people like the apostles and Peter himself.
Now, one could point out that in 1 Corinthians, Paul mentions the apostles and Peter separately, even though he would still be one of the apostles. The reason really comes down to authority. What Paul is doing isn’t just naming random individuals within the faith, but appealing to those seen as authoritative.
We can apply this context to both 1 Corinthians and Galatians here, as the framework is largely the same. In both cases, Paul is attempting to appeal to authorities. He wants to bolster his claims. So, in part, he appeals to the authority of the apostles. But we have to realize what an apostle is, as well as what it isn’t.
Often, when we think of an apostle, we conflate them with a disciple. For Paul, though, this isn’t the case. Apostles are a larger group, those sent out with the gospel. But within this group, others could be identified. For instance, we know Peter as one of the 12 disciples.
To be clear, Paul doesn’t call Peter a disciple, nor does he even use the term. But what he does do is signify that there was a group of 12, with Peter standing out among them.
So while Peter is an apostle, he is also more than just that. He stands out as having more authority, and part of the reason is that he is one of the 12.
Paul does the same thing with James and the brothers of the Lord. These individuals are apostles. In Galatians, it is clear that Paul is calling James an apostle. But he isn’t just an apostle. He is something more, and that is where we get the signifier, the brother of the Lord. It’s not a generic idea; it’s something quite specific, which is meant to elevate James to a higher place of authority.
It is also worth noting that only in these two places do we get the specific phrasing: the brother (s) of the Lord. While Paul calls other people brothers and sisters in Christ, it’s always in a different manner, with a different language. All of this together makes it clear that Paul isn’t applying a generic title to James in Galatians.
Continuing on, second, we can see in the Synoptic Gospels an independent attestation to the idea that Jesus had biological brothers, one of whom was named James. We have another independent attestation to this, that Jesus had a brother named James, in the work of Josephus.
In Book 20, Chapter 9 of the Antiquities of the Jews, Josephus, speaking of how the high priest had a group of people stoned, mentions James as one of those individuals and clarifies which James he was referring to by stating that he was the brother of Jesus, who was called Christ.
While Carrier argues against this passage, virtually all scholars within relevant fields have accepted it as being authentic. Now, in the future, we will touch on Carrier’s argument against this passage in Josephus; for the time being, it’s not worth getting into.
Third, we have Christian tradition, which had no issue distinguishing between the biological brothers of Jesus and the broader categories of followers. The Carrier-Doherty reading requires us to believe that Paul had a private code accessible to his original readers, but then lost to every subsequent interpreter for nearly two millennia. We have to imagine a situation in which the phrase “brother(s) of the Lord” did not appear to mean what everyone thought it meant until mythicists discovered the truth.
Now, Carrier’s reading is not impossible, but it is awkward; it requires additional hypotheses to keep it afloat, and it conflicts with the natural reading of these passages. In comparison, the historicist reading is simple, natural, and consistent. It’s the one that makes sense in context, and it doesn’t require additional assumptions to be made.
The Galatians passage specifically is the single most damaging piece of evidence against the mythicist case, but it isn’t the only material from Paul that resists Carrier’s reading. While Paul doesn’t tell us much about Jesus, he does mention a few things that presuppose that Jesus was a recent historical person.
He tells us that Jesus was born of a woman, under the law (Galatians 4:4); that he was descended from King David according to the flesh (Romans 1:3); that he instituted a meal on the night he was betrayed (1 Corinthians 11:23); and that he had a number of teachings from which Paul occasionally cites.
Most importantly, though, he tells us that Jesus was crucified, which was specifically a Roman form of execution, which was practiced in the first-century Levant. And placing this in the near past, he speaks of people who knew Jesus, and seemingly were part of his actual mission.
To accommodate all of this into the celestial-Jesus hypothesis, Carrier must treat each passage as either an interpolation (without actual manuscript evidence for such), as veiled mythological allegory (without actual contextual warrant), or as referring to events in some sublunar realm rather than on earth (with no support from the actual Greek text). These additional hypotheses require a staggering amount of additional evidence, which is lacking.
By the time that Carrier finishes reinterpreting the Pauline corpus, he really has created a new Paul who is saying something other than what the natural reading of his Greek suggests. The Pauline letters, as Carrier reads them, simply do not exist in any manuscript. They only exist in Carrier’s reconstruction.
It’s not really a defensible position. Instead, it’s similar to the mathematical theater we discussed in our first article. It is this sort of theater where one has to accumulate reinterpretations that are individually possible but, together, result in the natural reading of the text being completely dissolved, and only then declare victory in the resulting void.
Carrier on the Gospels: Allegory All the Way Down
Carrier’s treatment of the canonical Gospels follows a similar strategy. The Gospels, in his reading, are not a historical memory of a real person with legendary material added. Instead, they are allegorical fictions composed to imitate Hebrew Biblical narratives, with Jesus serving as a symbolic figure drawn from the Septuagint (the Greek translation of Hebrew scripture), Hellenistic biographies, and mythological writings common at the time.
In his reading, the Gospel of Mark, which is regarded as the earliest of the Gospels, does not even attempt to record historical events. Instead, it is constructing a fictional character for theological purposes. The later Gospels, Matthew, Luke, and John, then elaborate on this fiction in various directions.
What makes this reading more persuasive is that there is a kernel of truth buried in this argument. Mainstream scholars have long recognized that the Gospels are shaped by their authors’ theological agendas, that they make extensive use of tropes from, and sometimes even echo, various narratives from the Hebrew scripture, and, because of all of this, they should not be read as straightforward modern biographies.
But there is a large leap from the Gospels being theologically shaped narratives that at times echo Hebrew Scripture to the idea that the Gospels are allegorical fictions about a made-up character. What makes this leap even less defensible is its disregard for the historical context.
Virtually every historical text from antiquity that we possess is shaped by the theological, political, or rhetorical agendas of its author. Every ancient biography employs these same basic ideas. If we take Carrier’s logic here to an extreme, we could conclude that Suetonius’s Twelve Caesars is an allegorical fiction about non-existent emperors, or that Tacitus’s Annals are mythological narratives about a non-existent Tiberius.
The presence of literary shaping in a narrative is not, by itself, evidence against the historicity of an individual. It is simply that the narrative was written by an ancient author rather than by a modern journalist.
Carrier’s specific argument that the Gospel of Mark is an allegorical fiction depends heavily on the work of literary critics who have identified extensive intertextual relationships between Mark and Hebrew Scripture. One of the prominent scholars in this regard is Dennis MacDonald, who has argued for Markan dependence on Homer. But other scholars have also tried to trace parallels between Mark and the Elijah-Elisha cycle.
These observations can be rather interesting and, in many cases, rather persuasive, at least when it comes to compositional techniques. But they don’t bear the weight Carrier wants them to.
That Mark may have consciously patterned his narrative on earlier literary models does not mean the author invented a character out of whole cloth. At most, it tells us that he is shaping his account in literarily significant ways, which is what almost every ancient biographer did.
There is a larger problem, though. Carrier’s hypothesis of a celestial Jesus does not account for how multiple, independent narrative documents that all situate Jesus in a specific Galilean and Judean geography, in a specific Roman-administered political context, that have him interact with multiple named historical figures, would be produced within just a few short decades.
This historicization, on Carrier’s account, had to be both rapid and total. It also had to leave no trace in the documentary record of the original celestial belief. We simply have no surviving texts from any branch of early Christianity that present Jesus as the purely celestial figure that Carrier’s Paul is supposed to be writing about. Instead, the earliest narrative sources that we have are all set on earth, in the same geographical area, at the same time. Even our earliest non-narrative source, Paul, refers to Jesus in very earthly terms and relates him to a very earthly brother.
The celestial Jesus exists nowhere in the actual documentary record. He exists only in Carrier’s reconstruction of what the documentary record must originally have said before it was corrupted into its current form.
The mythicist reading requires a systematic reinterpretation of every surviving piece of evidence in ways that run counter to the natural reading of the texts. In addition, elaborate secondary hypotheses must be developed to buttress this view by describing what the original belief must have been and how it must have been transformed, even though there is no direct documentation for this.
On the other hand, the historicity reading requires no such reconstruction. Instead, it takes the surviving evidence and deals with it as it is. It treats these works as any other similar works.
The Cumulative Pattern
This same pattern plays out in Carrier’s reading of other sources, such as Josephus and Tacitus, two accounts we will return to later in much greater depth. What we end up seeing, though, is a clear and consistent pattern.
In every individual case, the mythicist reading is technically possible. The Galatians reference can, with enough effort, be reinterpreted as a generic title for members of the Christian movement. The Pauline allusions to Jesus’s earthly life can, with enough effort, be relocated to a sublunar mythological realm. The Gospels can, with enough effort, be read as allegorical fictions about a made-up character.
But let’s notice what the cumulative case requires. To maintain the mythicist hypothesis, every single piece of evidence that points towards historicity must be reinterpreted in the same direction: against the natural reading, the manuscript tradition, and the scholarly consensus.
The probability of being right in any one such reinterpretation might be reasonably high. The probability of being right in all of them at the same time is vanishingly small. This is, ironically, a point that Bayesian reasoning itself should make obvious. When a hypothesis requires you to be the contrarian on a dozen independent questions at once, the prior probability that you are right on all twelve collapses toward zero, even if our prior probability of being right on any one of them is moderate.
Why Carrier Has Not Persuaded the Field
If we combine the methodological critique in part 1 of this article with the evidential critique presented here, the explanation for Carrier’s failure to convert mainstream scholarship comes clearly into view. It is not, as Carrier himself sometimes suggests, that the field is too confessionally committed to engage his arguments fairly. The field contains a host of scholars who have no Christian commitments to defend.
It is not, as some of his more credulous followers suggest, that the field is too mathematically illiterate to follow his Bayesian setup. Quite a few specialists in relevant fields are perfectly capable of following the math, and several of them have explicitly addressed it. After all, Carrier isn’t the first to introduce Bayes’ Theorem into this discussion. Generally, before, it’s been more about attempts to prove Jesus was resurrected.
It is, rather, that the field has examined Carrier’s case and found it wanting at every level. The Bayesian methodology fails at the construction of the prior. The evidential readings fail at the level of the texts. The cumulative case requires a degree of systematic contrarianism that no serious historian could responsibly endorse.
Carrier has produced the most rigorous-looking mythicist book in the literature. He hasn’t produced a persuasive one. And his failure to persuade is not a measure of the field’s bias. It is a measure of the field’s competence.
