With thanks to Cambridge University Press for an inspection copy.
Nina Livesey’s The Letters of Paul in their Roman Literary Context raises questions that go beyond the authenticity and date of Paul’s letters. If we no longer discern a wandering charismatic preacher, one who is competing with other preachers, and planting house churches in Asia Minor and Greece as he works his way, via a thriving Jerusalem, to Rome, then what do we have in his place?
The argument structure of The Letters is as follows:
— an explanation of the origin of the current consensus that the New Testament letters of Paul include some that are authentic, mid-first century, writings to real churches; Nina Livesey (NL) shows that the arguments giving rise to this view [that is, the historicity of Paul, the authenticity of the letters, their first century date, and the related “home churches”] are circular and grounded more in conservative piety than independent evidence;
— a comparison of the letters of Paul with letter-writing more generally at this time (the Roman world of the first and second centuries); NL explains how the training of authors prepared them to create characters, both of apparent authors and recipients, and situations that attracted readers because they seemed “so real” and “personal”; NL further compares Seneca’s use of fiction to teach applied Stoic philosophy through artificial letters with the Pauline correspondence, pinpointing many similar literary devices. A case is made that Paul’s letters were a collection intended for general publication from the point of their creation by “a school” of a highly educated elite.
— independent evidence that explains the contents of the letters does not appear until the wake of the Bar Kochba war that (132-135 CE), far more than the first Jewish war of 66-70 CE, saw a genocide of the inhabitants of Palestine and even a denial of their name for their homeland as an ultimate punishment. In the context of Judea and Jewish practices like circumcision becoming a byword for all that Rome found contemptible, “Christian” teachers migrated to Rome where they set up “schools” not unlike other philosophical schools. It was from here that one such teacher, Marcion (later relegated by the “proto-orthodox” teachers as a “heretic”), identified with “Paul” and purportedly produced the letters under his name around 144 CE.
Further, the letters point to intertextuality with Acts and the gospels, indicating that the authors of all these works knew one another. Indeed, in Acts one finds the name of Paul emerging in the context of a work with a cluster of other fictional names, double-names and cipher (or symbolic) names (e.g. Stephen, the first martyr, meaning “crown”).
I look forward to discussing some aspects of NL’s book in more depth. This post is only an introductory overview.
NL’s overall argument does not identify an indisputable, concrete piece of evidence that directly places the letters of Paul (PL) in the mid second century and no doubt many readers will prefer to fall back on their “gut feelings” about the epistles. What NL offers is an argument that has fewer unsupported assumptions than are required by those who trust in at least their partial authenticity. The NL view appeals more directly and simply to the context of the external evidence. This external evidence is used to offer more direct explanations of the contents, the style and the known first appearances of the PL. Most simply:
— there is no first century external evidence to explain the contents and traditional beliefs about PL
— there is second century external evidence that does explain the contents and style of the PL
— what is known of literary education of the time further explains the PL as consisting of literary devices to teach a philosophical or theological set of beliefs; many inconsistencies and other difficulties within the PL that have engaged scholars who read the PL at face value are resolved by NL’s hypothesis of a second century school producing them.
Not too long ago I posted a very lengthy series on three books by Thomas Witulski proposing a Bar Kochba War context for the Book of Revelation. Witulski understood not only that war but the rebellions and massacres of Jews in the eastern Mediterranean under Trajan (prior to Hadrian) had a major impact on “Christians” at that time that was expressed in the “four horsemen” chapters of Revelation preluding the Bar Kochba revolt. Revelation expresses a remarkably different kind of Christianity that we know from the gospels and PL (see Couchoud’s discussions), even pointing an accusing finger at Christians who appear to embrace customs that surface in the PL (e.g. eating meat sacrificed to idols). Joseph Turmel (=Henri Delafosse) considered the “Man of Sin” Antichrist figure of 2 Thessalonians (see 2 Thess at his commentary page) to have been Bar Kochba but I wonder if a better case could be made for it being Hadrian, especially given Revelation’s favourable view of Bar Kochba (Witulski). How that interpretation might fit with NL’s arguments is a question I’d like to think through. Certainly Hermann Detering’s scenario of the “Little Apocalypse” prophecy of Mark 13 (and Matthew 24 and Luke 21) being best explained in Hadrianic times comes to the fore, as does his evidence (much drawn from Rudolf Steck) for Paul’s opponents belonging to the second century. The surviving writings of Justin (post the Bar Kochba War) also strongly suggest — contrary to conventional attempts to read his knowledge of our canonical gospels into his works — a time when there was a free-for-all scope for interpreting Jewish Scriptures as prophecies of “Christianity”.
So you can see how NL’s book ties in with many ideas I have been toying with for some years now. I look forward to discussing some of its details.
Livesey, Nina E. The Letters of Paul in Their Roman Literary Context: Reassessing Apostolic Authorship. Cambridge: Cambridge University Press, 2024.
Nina E. Livesey is Professor of Religious Studies, Emerita at the University of Oklahoma.
Today I received a review copy from Cambridge University Press of Nina E. Livesey’s The Letters of Paul in their Roman Literary Context: Reassessing Apostolic Authorship. I have already read the Introduction and Chapter 4 and highlighted along the way a few dozen other works referenced by Livesey to follow up. But what I found most inviting is that quite a few other references I have already posted about — and in some cases translated — here on Vridar. It is encouraging to meet a friend who likes and has found value in my other friends.
The first reference that stands out is Patricia Rosenmeyer. I posted on one of Rosenmeyer’s works in 2006 and it is one Livesey refers to often (as I also have done in subsequent posts):
Livesey references many scholars I have discussed here (some more extensively than others) but some names stand out as being more “radical” than others — to name but a handful some long-term blog readers may recognize:
Bruno Bauer — whose relevant works discussed by Livesey I translated and made available here: see his Pauline Letters and Christ and the Caesars. Livesey writes of BB, “his arguments are more sophisticated than those typically found in current Pauline scholarship” (19). Nice.
Paul Louis Couchoud — a very nice surprise to see him make an appearance
Rudolf Steck — my translation of one of his works is online at https://vridar.info/
Joseph Tyson — I posted at length on his work on Marcion and Luke-Acts
Herman Detering — whose scholarship is vastly under-rated by at least one prominent name who notably failed to do a basic Bayesian analysis of his work (see Staged Forgeries — another work I have translated)
Markus Vinzent
Richard Pervo
Boyarin, Daniel — of whom Larry Hurtado expressed distinct discomfort for his forays into New Testament studies
Nina Livesey argues that the Pauline letters all date from the mid second century — after the Bar Kochba War.
I feel a little ashamed that till now I have only allowed myself to wonder if all of the New Testament writings should be dated to the post Bar Kochba war period. Livesey takes that step boldly.
Drawing on Rosenmeyer and numerous others I look forward to reading Livesey contextualizes the Pauline letters within the ancient custom of “schools” and teachers writing letters in the name of others in order to teach and persuade. Paul’s letters are not the product of a “wandering charismatic preacher” but of someone belonging to the wealthy elite.
Such a portrait, however, poorly suits an individual both trained and socially positioned to produce such letters. On the other hand, there is ample evidence of creative literary activity and production in schools (haereses). As I argue in what follows, a second-century social and political context and a school setting, such as that of Marcion, are suggestive of a viable location for the creation of doctrinal exhortative letters written in the name of the Apostle Paul. (xif)
And the thesis extends beyond the letters:
While certainly a contentious and debated issue, the dating of NT writings plays an important role in my thesis. Not only Acts, but also the canonical Gospels are more recently considered not first- but second century writings. If we consider – as did the Dutch Radicals – that the Pauline letters were produced alongside of and in a complex and dynamic relationship with the Gospels and Acts, the forward shift in the dating of the latter lends further support to a second-century provenance of the letters. (27f)
So it’s back to Marcion and the post Bar Kochba era for “everything”.
“Christian” teachers arrived in Rome in the wake of the Bar Kokhba revolt and established schools under Roman authority near one other. “Christian” literature, including gospel texts, flourished during this time, with compositions reflecting a post-Jewish temple and post-Judaea social and political reality. Marcion’s publication of what has been interpreted as the First New Testament, consisting of a gospel (Evangelion) and a collection of Pauline letters (Apostolikon), is likely one of the earliest among these compositions. (251)
I look forward very much to reading the work in full and posting about it as opportunity permits.
Livesey, Nina E. The Letters of Paul in Their Roman Literary Context: Reassessing Apostolic Authorship. Cambridge: Cambridge University Press, 2024.
For the ordinary lay person who does not have the background to draw upon to enable a confident “vetting” the arguments of a consensus, I thought the three-part conditions set out by Avazier Tucker were a good rule of thumb for when to justify appeal to a consensus. It certainly provides a good answer to anti-vaxxers. Similarly, it offers good guidance to conspiracy theorists of various types. (Not that many of them would be convinced, of course, but it is nonetheless good to have “an answer” out there for those who are ready to change their minds.)
Richard Carrier thought Tuckers’ three part program was no answer at all to the problems he raised. It’s my fault entirely. I sneakily hid Tucker’s antidotes to the very problems Carrier raised in between the title and the last line of the post so anyone can be excused for missing them.
These conditions cannot be met in captured fields (e.g. you will never ever see a “consensus” in biblical studies by this definition that Jesus did not rise from the dead and is not God or the literal Son of God), so it is not useful as a metric.
Oh dear — my fault entirely. I should not have hidden the fact that Tucker’s three point proposition is explaining exactly why what is regarded as a consensus in biblical studies is not a justifiable or trustworthy consensus. I really do have to stop hiding the main points of my posts beneath their titles.
But more to the point, and by way of demonstrating how biblical studies fails on Tucker’s point 2 — the issues of Thomas Thompson and Thomas Brodie certainly illustrated the failure of Tucker’s points 1 (coercion) and 3 (coercion but also alert to the public about the heterogeneous character of the opposition to the consensus) in the field of biblical studies — more to the point, as I said, I must point to a work by Michael Alter published in the SHERM journal, Dataset Analysis of English Texts Written on the Topic of Jesus’ Resurrection: A Statistical Critique of Minimal Facts Apologetics —
This article’s abstract:
This article collects and examines data relating to the authors of English-language texts written and published during the past 500 years on the subject of Jesus’ resurrection and then compares this data to Gary R. Habermas’ 2005 and 2012 publication on the subject. To date, there has been no such inquiry. This present article identifies 735 texts spanning five centuries (from approximately 1500 to 2020). The data reveals 680 Pro-Resurrection books by 601 authors (204 by ministers, 146 by priests, 249 by people associated with seminaries, 70 by laypersons, and 22 by women). This article also reveals that a remarkably high proportion of the English-language books written about Jesus’ resurrection were by members of the clergy or people linked to seminaries, which means any so-called scholarly consensus on the subject of Jesus’ resurrection is wildly inflated due to a biased sample of authors who have a professional and personal interest in the subject matter. Pro-Resurrection authors outnumber Contra-Resurrection authors by a factor of about twelve-to-one. In contrast, the 55 Contra-Resurrection books, representing 7.48% of the total 735 books, were by 42 authors (28 having no relevant degrees at the time of publication). The 42 contra authors represent only 6.99% of all authors writing on the subject.
The article is available at the link above. The book referred to with the complete study is A Thematic Access-Oriented Bibliography of Jesus’s Resurrection. I don’t know how Michael had the stamina to undertake such a study, but again, it’s good to have things like this done and available.
Consensus per se is not a reliable sign of rationality. It depends on how consensus is reached. All sorts of methods can yield consensus: brain-washing, the threat of the rack or burning at the stake, totalitarian control of the sources of information. Consensus reached by these means does not guarantee rationality. — Goldman
Recent discussions have sent me back to a second look at Aviezer Tucker’s Our Knowledge of the Past. (I posted about this book back in 2013.) A detail I had forgotten in the meantime is Tucker’s interesting explanations about when a consensus is more likely to be trustworthy than not. Here is part of what he writes:
In the real world, there is no universal consensus on beliefs because some people always dissent. There are still people who deny that the earth is round or that it revolves around the sun and others who deny that there was a Holocaust. This led philosophers who consider consensus to be philosophically valuable to attempt to prescribe whose opinions matter for determining whether or not there is a consensus on beliefs. Reliance on professional organizations or academic institutions, affiliations, and certifications is a tempting approach. If all professional astronomers agree that the earth is round and revolves around the sun, it must be because they have knowledge of the heavens and those who honestly dissent must be either ignorant or irrational. At their best, professional and academic institutions should indeed reflect the qualities that make consensus among their members significant as a possible indicator of knowledge. Still, this appeal to authority failed too often in history. Academic and professional institutions have proved themselves to be too susceptible to political threats and manipulation, economic dependence and graft, or the common biases of a class of people who share professional interests and often social background. The corruption of the German academic system under the Nazis, the rejection of relativity in physics, the destruction of psychology, anthropology, and historiography in favor of dark racist fantasies is just the most extreme example of the ever-present temptation of academic and professional institutions, even in previously excellent academic systems, to prefer their institutional, class, and professional interests to intellectual integrity. (33)
So what does Tucker advise given the fact that the consensus of authorities may not always be a reliable guide to the truth of a matter?
The alternative approach I present here does not have to rely on authority, nor does it require universal consensus. It merely requires unique noncoerced heterogeneity: It does not matter if some do not agree to a set of beliefs, as long as the people who do, are sufficiently different from each other to reject alternative hypotheses to the knowledge hypothesis, and those who dissent are sufficiently homogenous to support hypotheses that explain their dissent by particular biases. For example, the group that reached consensus on Darwinian evolution is uniquely heterogeneous, it includes people who are secular and religious, and of many different faiths. The community that upholds the alternative “creationism” is quite homogenous, composed exclusively of biblical fundamentalists, almost all of whom are American Protestants, though many American Protestants believe in Darwinian evolution. Their bias in favor of an anachronistic, historically insensitive interpretation of Genesis is the best explanation of their beliefs. (34)
I know many readers of this blog will be thinking of the consensus on the historicity of Jesus. In the light of the above, we may see a reason certain critics of the Christ myth theory attempt to portray “mythicists” as a singular group of Christian-hating atheists. By ignoring the broad spectrum of “mythicists” — atheists, Buddhists, prominent public intellectuals, and even faithful Christians and others who have expressed a high regard for Christianity — the mythicists are portrayed as a narrow, homogenous group with a unique hatred of Christianity and thus able to be comfortably dismissed.
Tucker says a reliable consensus must be uncoerced and sufficiently large. How large?
The heterogeneous group that reaches consensus must be sufficiently large to avoid accidental results. Small groups can never be sufficiently heterogeneous to exclude hidden biases. . . . If only four scientists work on a particular problem and they agree on a set of related beliefs, it does not imply that their agreement reflects common knowledge. The four may be a professor, her assistant, a former student, and an untenured member of faculty who needs her vote on the tenure committee. The minimal size of a significant consensus depends on local circumstances such as whether the people who develop a consensus are related socially, the nature of their relations, and whether they attempted to replicate the process that generated the beliefs or merely accepted the conclusions of others on faith or authority (Sarkar, 1997, p. 510). Usually, when the consensus involves hundreds of people who are geographically, institutionally, and professionally dispersed, it is safe to assume that it is large enough. If the consensus is on an esoteric topic, and only a handful of experts are competent or interested enough to reach the consensus, they may possess knowledge, but their consensus cannot function as an indicator of knowledge. It is necessary to follow their reasoning to evaluate the status of their beliefs. (34)
Those are three key features for “mere outsiders” to look out for when deciding whether they have a right to hold any reservations about a consensus among the experts:
The consensus is uncoerced;
The consensus is found among a “sufficiently large” and heterogenous group;
Dissenting voices are from a “sufficiently homogenous” group such that their biases are readily apparent.
They look like a neat rule of thumb to me.
Goldman, Alvin I. “Foundations of Social Epistemics.” Synthese 73, no. 1 (October 1987): 109–44. https://doi.org/10.1007/BF00485444.
Tucker, Aviezer. Our Knowledge of the Past: A Philosophy of Historiography. Cambridge University Press, 2009.
Since I began drafting this post, Richard Carrier has responded specifically to some of my subsequent comments that were made in an exchange over what I, rightly or wrongly, understood to be some confusion about my view of Bayes’ theorem. He has not, as far as I see — again, I am open to correction and reminding — responded to the central argument of these posts. Further, I began posting a detailed review (scroll to bottom of the page for the reviews) of Carrier’s Proving History in 2012 but other questions arose that distracted me from that project after three posts. This series might be seen as an update on my views of Carrier’s application of Bayes’ theorem to history generally and Jesus in particular.
My position on the historical Jesus
To begin with, I think the only figures of Jesus of any relevance to the historian are
the Jesus in our early sources (especially the New Testament writings)
Attempting to “discover” a Jesus “behind” the sources through memory theory or other means (criteria of authenticity, form criticism) necessarily begins with the assumption that the stories told in the gospels have some kind of relationship with a historical Jesus. In other words, they assume a historical figure was the starting point of everything. (A passage found in a work by the Jewish historian Josephus, even if only partially authentic, can tell us nothing more than what was being said about Jesus some sixty years after he was supposed to have lived.)
My position regarding Bayes’ Theorem
One more point I should reiterate. I have said many times now that Bayes’ theorem is a fine tool to apply to many hypotheses. My point, though, is that I see little historical value in hypothesizing the existence or non-existence of Jesus per se. What is of historical interest is how Christianity emerged. A hypothetical Jesus or hypothetical non-Jesus alone doesn’t help us with that question. We simply don’t know if there was a figure identifiable as Jesus at or near the start of Christianity. The reason we do not know arises from the lack of independent and empirical data to establish his presence. Historical explanations can draw on hypothetical scenarios but when they do they can never be more than hypothetical proposals. I prefer that a historian works more modestly with what can be securely known and seeks to explain that much.
My position with respect to Richard Carrier’s historical methods
When Richard Carrier’s books Proving History and On the Historicity of Jesus first appeared I was intrigued by the Bayesian approach and in large measure rode with it. But what especially attracted me was the comprehensiveness of Carrier’s approach to the question that had at that time been a “hot topic” ever since Earl Doherty’s contributions. At the time I attempted to shelve some discomfort I felt over Carrier’s portrayal of “what historians do” and “how they do things” more generally. He seemed to me to be returning to a positivist view of history, a view that had largely been left in the margins especially since the mid twentieth century. One other discomfort I had was that I thought he was weakening his position and making himself too-easy-a-target for critics by adding new speculative arguments to those of Earl Doherty. I felt a stronger case and smaller target would have been made with less rather than more — with zeroing in on a selection of core arguments from Doherty rather than trying to cover everything that had been argued and adding even to that.
Though I have written many posts in favour of the application of Bayes’ Theorem to questions arising in biblical studies (and I have Richard Carrier to thank for introducing me to the usefulness of Bayes) I have also found myself in disagreement with some of Carrier’s views:
Ironically, in most of those cases, I think that it is Carrier who has dropped the Bayesian ball along with “rational-empirical” argument and it is yours truly who is using Bayesian reasoning to demonstrate where some of Carrier’s views are amiss.
At the time I held back my criticisms mainly because I did not want to be seen as part of what was then a hostile internet backlash against Carrier. But since I have recently been deeply re-engaging with the nature of historical knowledge and history itself I have felt the time is right to try to resolve some issues raised by Carrier’s work that originally left me a little uncomfortable.
Misapplied Conclusions from Bayesian Analysis
If we conclude from Bayesian reasoning that a historical Jesus is not likely to have existed, it tells us nothing useful. All it would mean is that if Jesus did exist there were many views expressed about him that gave rise to suspicions about his existence among later readers. He would not be the first.
No historical event is the same as another and no historical person is the same as another. Each and every historical event and historical person and circumstance is unique in some way. Hypotheses about the existence or non-existence of Jesus are hypotheses about a unique event.
But it is impossible to compute the frequencies of events that are unique. (Tucker, 136 — Tucker further notes, without comment, Carrier’s response to this problem, which is to assign a range of probabilities including subjective but informed probability estimates of experts — that is, measuring )
A Case Study
Carrier argues that even subjective expectations are ultimately (though perhaps hidden from one’s immediate consciousness) based on calculable frequencies of the same kinds of events:
Any time you talk about degrees of belief or certainty, just think about what you base that judgment on, and what facts would change your mind. Always at root you will find some sort of physical frequency that you were measuring or estimating all along. (272)
I am not so sure. Are “degrees of belief” or subjective “certainty” always or necessarily based on conscious or subconscious rational calculation? If so, there can be no such thing as “subjective” belief: every belief would be at some level a rational calculation based on relevant frequencies. I am not as confident as Carrier in the fundamental rationality of all subjective beliefs and expectations.
Alexander David Cooper – THE DEATH OF KING WILLIAM II. Image from https://clasmerdin.blogspot.com/2020/08/the-death-of-william-rufus-accident-or.html
Let’s take a unique historical event, one that Carrier discusses in Proving History, and examine his argument. The historical hypothesis here is that Henry I plotted to kill William II.
In our personal correspondence, C. B. McCullagh observed that to apply BT to questions in history
the hypothetical event has to be considered as a generic type, similar in some respect to others. That might worry historians, whose hypotheses are so often quite particular. For instance, consider how the hypothesis that Henry planned to kill William II in order to seize his throne explains the fact that after his death Henry quickly seized the royal treasure. The relation between these events is rational, not a matter of frequency. . . . (The example being referred to is discussed in C. Behan McCullagh, Justifying Historical Descriptions (New York: Cambridge University Press, 1984), p. 22.)
But, in fact, if the connection alleged is rational, then by definition it is a matter of frequency, entailed by a hypothetical reference class of comparable scenarios. To say it is rational is thus identical to saying that in any set of relevantly similar circumstances, most by far will exhibit the same relation. If we didn’t believe that (if we had no certainty that that relation would frequently obtain in any other relevantly similar circumstances), then the proposed inference wouldn’t be rational. Explaining why confirms the point that all epistemic probabilities are approximations of physical frequencies. The evidence in this case is that Henry not only seized the royal treasure with unusual rapidity, but that his succeeding at this would have required considerable preparations before William’s death, and such preparations entail foreknowledge of that death. Already to say Henry seized the royal treasure “with unusual rapidity” is a plain statement of frequency, for unusual = infrequent, and this statement of frequency is either well-founded or else irrational to maintain. And if that frequency is irrational to maintain, we are not warranted in saying anything was unusual about it. Likewise, saying “it would have required considerable preparations” amounts to saying that in any hypothetical set of scenarios in all other respects identical, successful acquisition of the treasure so quickly will be infrequent, and thus improbable, unless prior preparations had been made (in fact, if it is claimed such success would have been impossible without those preparations, that amounts to saying no member of the reference class will contain a successful outcome except members that include preparations). Again, the result is said to be unusual without such preparations, or even impossible; and unusual = infrequent, while impossible = a frequency of zero. Hence such a claim to frequency must already be defensible or it must be abandoned. Similarly for every other inference: making preparations in advance of an unexpected death is inherently improbable for anyone not privy to a conspiracy to arrange that death, and being privy to such a conspiracy is improbable for anyone not actually part of that conspiracy, and in each case we have again a frequency: we are literally saying that in all cases of foreknowing an otherwise unpredicted death, most of those cases will involve prior knowledge of a planned murder, and in all cases of having foreknowledge of a planned murder, few will involve people not part of that plan. If those frequency statements are unsustainable, so are the inferences that depend on them. And so on down the line.
Thus even so particular a case as this reduces to a network of generalized frequencies. And all our judgments in this case necessarily assume we know what those frequencies are (with at least enough accuracy to warrant confidence in the conclusion). We won’t know exactly the frequencies involved, but we know they must be generally in the ballpark stated, otherwise we wouldn’t be making a rational inference at all. (273f – my highlighting in all quotations)
Here Carrier notes as background knowledge that “considerable preparations” would have been required “before William’s death” for Henry’s actions to have succeeded. He confuses this background knowledge with “evidence” but understanding the complexities involved in asserting full control of the royal treasure is really background knowledge. That background knowledge forms the basis of the hypothesis that Henry murdered William. Carrier leaves this aside. He suggests that Henry’s ‘unusual rapidity’ in taking control reflects subconscious knowledge of how infrequently such events occur under normal circumstances. But I do not know how Carrier could verify that historians really do reflect on how many times a royal treasure has been taken over with such speed, or how often comparable “classes of events” have occurred. Carrier does not give examples of similar “reference class acts” with which to compare and I suspect most historians will need time to think before they could offer instances. Even if one did compile a list of comparable successions in an attempt to establish a reference class with which to compare Henry’s succession of William, historians would be hard pressed to tease out all factors that made each situation unique and to justify its relevance to the particular event of Henry’s replacement of William. Rather, it is simpler and more likely that historians who are informed of the political structures and scale of England at the time use that background knowledge to infer that the speed of Henry’s acts points to the likelihood of murder.
While typing up this post I was distracted by a news item about a woman being interviewed who said that she was told she had a “only a 10%” chance of contracting a certain terminal illness but now she had it. There was no longer any 10% business about it. Statistics and probabilities are relevant when dealing with effectively infinite numbers of factors. But historical contingency is not a probability event. It happens to a particular person with a certainty of 1 regardless of what the odds are in an infinite universe and there was no way to estimate in advance that that particular woman was going to get the illness. That that unfortunate person was part of a 10% subset within a population of many thousands was meaningless to her and her loved ones. A science body had produced statistics. This person experienced an historical event. Most historical events are unforeseen — except, as I keep saying, in hindsight.
Confusing History with Science
Geology and paleontology, for instance, are largely occupied with determining the past history of life on earth and of the earth itself, just as cosmology is mainly concerned with the past history of the universe as a whole. . . .
History is the same. The historian looks at all the evidence that exists now and asks what could have brought that evidence into existence. . . .
And just as a geologist can make valid predictions about the future of the Mississippi River, so a historian can make valid (but still general) predictions about the future course of history, if the same relevant conditions are repeated (such prediction will be statistical, of course, and thus more akin to prediction in the sciences of meteorology and seismology, but such inexact predictions are still much better than random guessing). Hence, historical explanations of evidence and events are directly equivalent to scientific theories, and as such are testable against the evidence, precisely because they make predictions about that evidence. . . .
[T]he logic of their respective methods is also the same. The fact that historical theories rest on far weaker evidence relative to scientific theories, and as a result achieve far lower degrees of certainty, is a difference only in degree, not in kind. Historical theories otherwise operate the same way as scientific theories, inferring predictions from empirical evidence—both actual predictions as well as hypothetical. Because actual predictions (such as that the content of Julius Caesar’s Civil War represents Caesar’s own personal efforts at political propaganda) and hypothetical predictions (such as that if we discover in the future any lost writings from the age of Julius Caesar, they will confirm or corroborate our predictions about how the content of the Civil War came about) both follow from historical theories. This is disguised by the fact that these are more commonly called ‘explanations.’ But theories are what they are. (46ff)
I have recently addressed historical positivism at
What Carrier is describing here is a “positivist” view of history. This is a notion of history that was more widespread up to the middle of the last century. One of its leading exponents was Hempel who argued that historians should be seeking to discover predictable cause-effect relationships. (Hempel took positivism a step further than Carrier by claiming actual “laws” in history could be found.) His colleague, Carnap, stressed the importance of probabilistic reasoning in such an endeavour. The view that history could aspire to be akin to the natural sciences in method grew out of the Enlightenment when there was burgeoning confidence that Reason and Empiricism could liberate humanity from the shackles of superstition and dogma. But positivist history has long since been under strong attack from many quarters.
It is this positivist approach to history that explains the relevance of Carrier’s use of Reference Class. The idea of a reference class is to generalize historical events or incidents so that they can be compared with one another as a common type. That means they are temporarily removed from their historical contingency and treated as sharing common features for the sake of comparison. The point is to isolate generalized cause-effect principles.
Strictly speaking, prior probability is the probability of getting a specific kind of h when you draw at random from a reference class of all possible h → e [hypothesis to evidence] correlations. Those correlations don’t have to be causal, although in history they usually are. Because, in history, we are almost always asking what caused e and proposing h as the answer (see chapters 2 and 3). I’ll thus focus mainly on causal hypotheses and explain how to ascertain prior probabilities in a way that can produce intersubjective agreement among expert historians, and when and why such a process is logically valid. Some critics of BT are skeptical of causal language in applying the theorem, but that’s fundamental to many theories, especially historical ones, since any statement about what happened in history reduces to a statement about what caused the evidence we have. And you can’t propose historical explanations without proposing causes. Historians do distinguish claims about what happened (or once existed) from claims about why it happened (or why it existed). But ultimately all claims about ‘what’ entail claims about ‘why.’ (229)
I pause and ask if that is so. Many historians may agree with the above, but even among those who do, I think most would be sceptical about any attempt to assess varying degrees of causal probability to any of the factors associated with an event. Understanding human behaviour is not so mechanical an enterprise. The example Carrier offers does to me come across as unrealistically mechanically causal and even positivist with a vengeance:
. . . a hypothesis that a religious riot was caused by prior beliefs of that community (such as an ancient prophecy) in conjunction with new events (such as the appearance of a comet) obviously proposes a causal relationship between those prior beliefs and the riot . . . (230)
Such a view of human nature in general and historical events in particular is not one I share. I doubt that many historians have ever concluded that there can ever be such a simplistic one-to-one cause-effect of a riot as “a belief” of some kind. I propose that where riots occur a range of conditions will normally be found to help us understand the what and the why.
I think the principle applies to most works of historians today. Few, I believe, would think they can reduce historical events to isolated or particular combinations of specific causes each bearing a certain probability factor in the final equation.
In another instance, this time in On the Historicity of Jesus, Carrier continues the same refrain: the existence of prophecies would in effect have caused would-be messiahs to seek martyrdom, so in such a context, Christianity “almost becomes predictable”.
God had promised that the Jews would rule the universe (Zech. 14), but their sins kept forestalling his promise (Jer. 29; Dan. 9), which would also create a motive for would-be messiahs to perform atonement acts, which could include substitutionary self-sacrifice (see Element 43), out of increasing desperation (Elements 23-26). Christianity almost becomes predictable in this context. (OHJ 71)
Admittedly Carrier relegates this statement to a footnote but it does further illustrate the simplistic cause-effect positivism approach he has to the question of Christian origins: prophecy — would have inspired (caused) — would be prophets — to do an atonement act like Jesus — Christianity conceptually predictable (law of cause and effect) in such a scenario. If there were historically verifiable prophets acting that way, most historians would prefer to seek a deeper understanding about why such behaviour emerged at that time and place than the mere existence of a prophecy rolled away in the scrolls.
A recent work of history that I read is Killing for Country by David Marr. Along with Tom Petrie’s Reminiscences of Early Queensland and Libby Connor’s Warrior, I have been left with a deep sense of shame about white treatment of the indigenous population of my state and a strong political and social conviction of what we owe their survivors. Those historical works were not about “cause and effect” but about understanding and awareness. There will always be causal elements in any explanation but causes per se are not always what history is about.
Reference Class Revisited
As far as I understand Carrier’s approach, he introduces Reference Class in the question of the historicity of Jesus in order to establish a prior notion of how likely a certain idea of Jesus is the result of a generalized cause-effect class of events. This is an attempt, as I understand it, to introduce some kind of “scientific” validity to the study of history. If we understand the “scientific” approach as one that seeks to establish the general from the particular, this is the intended function of assigning Jesus to the Rank-Raglan hero class and drawing probabilistic inferences based on cause-effect principles found in that class.
The idea is that among figures found in a subset of the Rank-Raglan class few or none are known to be historical. The principle Carrier wants us to conclude from this is that stories of a certain type are “caused” by something other than a historical figure behind them.
If therefore we find Jesus within this subset of story types, then it logically follows that those stories about him likewise owe their existence to something other than an actual historical figure of Jesus.
I agree that in principle — and it is the principle that counts — that is a correct conclusion. Lord Raglan himself expressed the same point:
If, however, we take any really historical person, and make a clear distinction between what history tells us of him and what tradition tells us, we shall find that tradition, far from being supplementary to history, is totally unconnected with it, and that the hero of history and the hero of tradition are really two quite different persons, though they may bear the same name. (The Hero, 165)
Further, I think a good many biblical scholars will also agree that what we read in the gospels about Jesus is in large measure unconnected with a historical Jesus. Many argue that the stories that arose about Jesus were fabricated to meet the needs of later generations.
In other words, the reference class itself is irrelevant to the question of the historicity of Jesus. It is a misguided attempt to establish a quasi-scientific or positivist approach to history by establishing a principle that transcends the uniqueness of each historically contingent event and person.
The mythical stories about Jesus tell the historian something important to the interests of early Christianities but as Lord Raglan pointed out by implication — those stories of themselves cannot have any relevance to the question of whether there was some kind of historical Jesus at the start of it all. If we think otherwise we would need to argue the case with evidence.
Of course, many other biblical scholars are quick to deny this point and will claim “memory theory” and “triangulating” “gists” of gospel stories and sayings can help them see “through a glass darkly” some outline of the historical Jesus. But such notions are founded entirely on the assumption that Lord Raglan was wrong and that the stories did evolve from a historical person.
We simply have no way of knowing if “a historical Jesus” existed. There are many interesting studies that explain the New Testament sources emerging from within the historical, philosophical and literary milieu of the day without appealing to a hypothetical role for a historical Jesus. We don’t need to over-reach and try to “prove” anything within any margin of probability. Hypothetical notions relating to the existence or nonexistence of Jesus cannot help the historian produce any serious reconstruction or understanding of Christian origins. Let’s be content with what we cannot know and focus on what we do know. Carrier’s On the Historicity of Jesus, especially its Backgound/Context section, offers many areas for further study. As I pointed out above, I think there are some areas where even Carrier can more consistently and profitably apply Bayesian analysis.
Carrier, Richard. Proving History: Bayes’s Theorem and the Quest for the Historical Jesus. Amherst, N.Y: Prometheus Books, 2012.
Carrier, Richard. On the Historicity of Jesus: Why We Might Have Reason for Doubt. Sheffield: Sheffield Phoenix Press, 2014.
Tucker, Aviezer. “The Reverend Bayes vs. Jesus Christ,” History and Theory 55, no. 1 (February 1, 2016): 129–40.
