Failure of the Logic of History in Christian Origins Studies

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

by Neil Godfrey

Screen Shot 2015-06-16 at 4.37.11 pmI have finally found two books on the practice of history, each by a scholar (other than Richard Carrier), that address the core questions I have often raised with respect to flawed methods of New Testament historians dealing with Christianity’s origins. Both works address historical studies in general and only one from time to time casts a glance at what certain biblical historians are doing.

One is The Logic of History by C. Behan McCullagh (2004). McCullagh is a philosopher of history responding primarily to the postmodernist challenges to traditional historical practices in the field of history generally. Some of his arguments apply not only to postmodernist approaches, however, but equally to a number of flawed arguments by more traditional biblical scholars.

The other is Historical Evidence and Argument by David Henige (2005). In my next post I will address his fourth chapter titled “Unraveling Gordian Knots” where he applies his criticism to sentiments we find expressed repeatedly throughout New Testament historical works — and especially in regard to many New Testament scholars’ attacks on the Christ Myth hypothesis.

This post addresses a few excerpts from C. Behan McCullagh’s The Logic of History. 

Why has no-one else argued these points before?

The points have been argued before but apparently rarely applied to the methods of scholars specializing in the history of Christianity’s origins and early growth. Nonetheless, when I first tried to think through how we came believe certain persons and events in the ancient past were historical and others not I was a little surprised that so little appeared to have been directly addressing this question.

Happily I have now found an explanation for my inability to find what I was looking for back then. On page one McCullagh writes:

Historians often learn how to assess their hypotheses by studying debates in history in the course of their education. They acquire a capacity to evaluate their hypotheses critically, without always being aware of the standards of rationality they are applying. Awareness of those standards, however, will make it easier for historians to ensure that their work is rationally defensible.

There are many good books which explain how students of history should undertake their inquiries, but they contain very little guidance as to the logic of historical reasoning. They are almost entirely about searching for answers to one’s questions, and writing up the results. Yet the point of all the good practical advice is to gather information from which sound inferences about the past can be formed. Those inferences and arguments are at the heart of historical practice. (my own formatting and bolding in all quotations)

And in the conclusion of his Introduction on page 4:

I hope that this introduction to the logic of history will quicken historians’ interest in the rational justification of their accounts of the past. It should help guide historians in the rational assessment of their own work and that of others.

So McCullough appears to be acknowledging that most of the current works on the practice of history have overlooked and taken for granted “the standards of rationality” being applied and “logic of historical reasoning”. 

How to be sure we are reading a text the right way

One assumption found throughout most works on Christian origins has been that we should assume that the narratives in the gospels were ultimately derived from historical persons and events. The question facing historians is to decide what those historical events and persons may have looked like beneath all their gospel mythical and theological overlay. It is taken for granted that the gospels were written for the purpose of conveying certain meanings about historical events and persons.

McCullough does not address historians of the Gospels but notice what he does say about the way we historians approach documents:

Normally the conventional meaning of a text is the one which the author intended to convey, but very occasionally an author writes ironically, satirically, humorously or mistakenly, and intends to say something else. Jonathan Swift, for example, intended his book Gulliver’s Travels to be a political satire; and Cervantes intended Don Quixote to be a parody of knightly chivalry (see McCullagh, 1998, pp. 154–5).

One can add to “ironically, satirically,” etc. “theologically, metaphorically, fictitiously….” as we find in several narrative books of the Old Testament and non-canonical Jewish literature of the Second Temple era. Continuing . . . .

To be sure of such interpretations, historians must often discover quite a lot about the circumstances of the text’s composition from other sources. At all events, the conventional meaning of a text is not necessarily its intended meaning.

Is McCullough really saying that we need independent evidence to enable us to be sure of a certain interpretation of a text? So to be sure if a narrative about King Arthur or Julius Caesar or Jesus Christ is historical we need to find “quite a lot about the circumstances of the text’s composition from other sources“? The problem with using the Gospels as sources, however flawed, for the historical Jesus is that we have few “other sources” to help us interpret them. No, that does not mean that we must treat them as entirely fictitious. It means that we have a responsibility to suspend judgment pending other information.

It is almost impossible to guess how people in the past read the texts before them. Even when they are fluent in the language of the text, and familiar with its context, it is common for people to find things in texts which were not there, or to distort what was said in some way. It is only when readers describe a text that an historian can be certain of their reading of it. If a person’s reading is inaccurate, then historians sometimes investigate the causes of the misreading, using it as a way into the mind and personality of the reader.

An historian’s own reading of a text is not always accurate, because historians usually read texts hoping to find evidence for or against a certain hypothesis about the past, and are prone to overlook what does not suit them.(p. 20)

How to know an author’s intentions

All historians know the importance of understanding the perspective and bias of their sources. Nothing new here.

Traditionally it has been very important to know writers’ intentions in order to understand their texts. As well as judging the conventional meaning of a text as evidence of what the author wished to convey, historians have always been interested in the author’s intention in publishing the text. This often provides an important indication of possible bias in the text. Was it written to win favour with someone, or to justify some action, or to destroy a reputation? A text is likely to be biased according to the purpose for which it was written. (p.  28)

I omit McCullough’s argument addressing postmodernist challenges to the very idea that authors even write with personal intentions. Page 28:

. . . How does one discover an author’s intention?

It was once thought that this was done through an act of empathetic imagination. This involved historians imagining the circumstances in which authors found themselves when writing, taking into account their beliefs, values and attitudes, and then imagining how they would have responded in precisely those conditions. We commonly use this technique to interpret what is going on in the minds of those around us, and it seemed natural to apply it to history. (See, for example, W. Dilthey’s theory of understanding in Gardiner, 1959, pp. 211–25; and Collingwood, 1946, Pt V., sect. 4.)

There are a couple of problems with this procedure, however.

The first involves the assumptions being made about an author’s beliefs, values and attitudes. How are these arrived at in the first place? Without specialized knowledge of their state of mind, one could only imagine that their views of the world were like ours, but that is bound to be wrong. So one cannot really discover the mental state of others by empathetic imagination, though often this technique is useful for suggesting possibilities.

The second problem is that this method is quite uncritical. Once the historian imagines what someone intended to say, so long as this hypothesis seems to cohere quite well with what is known about the subject and their situation, it is accepted. But there could be other hypotheses even better supported by all the relevant evidence. Coherence is not enough.

Here McCullough is addressing the common fallacies of correlation or confirmation bias. Rigorous methods require us to look for evidence that contradicts our hypotheses and to test alternative hypotheses. Anyone can find evidence to support a hypothesis that one belongs to any chosen star-sign. Correlations can have any number of explanations: causal, consequential, third party, chance . . . .

Or to apply the point to our immediate interest, just because the Bible says it’s so and entertains no doubts does not necessarily mean that it must at least be a little bit so.

And again, no, that does not mean we must reject a claim for historicity, either. (In my next post I will address what Henige has to say about so-called “hyperscepticism”.) But it does mean that pending further information we must respect uncertainty:

The appropriate method for discovering authors’ intentions is, once again, that of finding the best explanation of all the relevant evidence.

An informed imagination can produce plausible hypotheses, but these must be checked against alternatives to discover which is most worthy of belief.

People’s intentions are often very difficult to discern, and responsible historians will admit to uncertainty when lack of evidence requires it. (For example, see several quite different interpretations of the intended point of Plato’s Lysis and Phaedo in McCullagh, 1998, pp. 152–4.)  (pp. 28-29)

Coherence is never enough; positive evidence not always enough

So many hypotheses about the historical Jesus are built upon arguments of coherence. Placing enough gospel verses (e.g. about poverty) together against other information (e.g. about economic hardship of peasants) does not establish a bed-rock argument for Christianity beginning as a response to economic stresses as some scholars argue.

Coherence is never enough to establish the credibility of an interpretation of a text. One way in which an interpretation of a text can cohere with other credible information about it, is by not contradicting that information, or rendering any of it improbable.

This kind of coherence is clearly not enough to establish the credibility of an interpretation.

An interpretation needs positive support from credible information, as well as this kind of consistency, to be credible. (p. 34)

Positive support is necessary but not necessarily sufficient

Positive support for an interpretation of a text, however, is not always enough. If there is evidence that strongly supports a belief inconsistent with it, then the interpretation remains uncertain.

Historians can display much ingenuity in explaining away inconsistent evidence to render their interpretation consistent with it, but if the explanation is ad hoc, that is without independent evidential support, then it is not credible. . . .

That passage ought to be printed in bold font, framed, and placed on a good many office desks of New Testament scholars writing about the historical Jesus and Christian origins. It’s application would surely have resulted in far fewer books having been written on these themes — or at least have ensured a higher ratio of quality books. Thus,

If no explanation is clearly superior to all others, responsible historians will admit as much and not insist on the interpretation they prefer. (p.34)

I am reminded of one prominent biblical scholar (prominent in the internet and online discussion forums, at least) who flatly insisted that a particular papyrus manuscript should be dated to around 125 CE “because” all the specialist conclusions dated it between 125 and 175 CE. That meant, according to him, that his hypothesis that depended on the earlier date was fully justified! (His audience consisted of a good number of amateurs, I recall, so I can’t be sure he would be so reckless in any formal publication reserved for his peers.)

Independent controls once again

One postmodernist objection that historians inevitably argue in circles is summed up by McCullough:

Historians discover what happened in the past by interpreting documents. But if historians need to know the context of a document in order to interpret it appropriately, how can the document provide evidence of that context? . . . (p. 36)

The postmodernist argument continues by claiming that the historian is merely falling back on choosing whatever story he subjectively assesses to be the more coherent.

In summary form the criticism is not all that different from my own point about the circularity of many NT scholarly arguments from the Gospels. I am not arguing for the postmodernist approach however . . .

McCullough rightly responds to this point:

Sometimes an ambiguity about the historical significance of a document can be resolved only by studying the context in which the document was produced. We saw a case of this above in the debate over the date of Hastings’s execution.

But, as in that instance, it is almost never the case that the document itself is the only evidence of that context. If it were, historians might have a problem.

Usually other documents yield much information about the context so that the uncertainty about the historical significance of a document is not difficult to resolve.

Yes, the Gospels are indeed the “only evidence” of the context assumed by NT scholars so they really do “have a problem”.

Of course the same NT scholars are aware of this problem and that is why someone like Bart Ehrman will insist that they really have multiple independent sources. But these “independent sources” are not the sorts of tangible documents McCullough is explaining are necessary: they are entirely hypothetical.

Another argument McCullough addresses is as follows:

Or perhaps the interpretation of the meaning of a document depends upon an historian’s judgement of the author’s purpose in producing it. Then once again, how can the document provide evidence of that purpose? Historians must interpret the meaning and purpose of documents together, accepting the most coherent account of them.

McCullough’s response:

The same is true when historians search for the author’s intention in writing a document. Often there is a lot of independent information about the author and the circumstances in which the document was written that enables historians to get a fair idea of the author’s intention.

Independent information about the authors (indeed, any information at all about them!) is just what the Gospels lack.

Often one interpretation of the author’s intention is vastly superior to all others, as none other fits the context, both textual and historical, nearly as well. But, when historians have so little evidence that the only thing going for a hypothesis is its coherence, responsible historians are very wary about attributing much credence to that hypothesis.

Postmodernist approaches to Christian origins studies have left a number of New Testament scholars open to the same criticisms applicable to their postmodernist counterparts in other historical fields. My impression is that New Testament scholars have to some extent embraced the postmodernist arguments of their historian peers in an attempt to “paper over” the logical flaws in their traditional arguments. These flaws have long been the circularity at the heart of all arguments relying upon the Gospels and Acts as the sole sources of information about the Christian origins.

Next post I hope to take up a number of Henige’s points from Historical Evidence and Argument.


  • 2015-06-16 16:36:28 UTC - 16:36 | Permalink

    The arguments here are just so… Bayesian! For instance:

    “Coherence is never enough to establish the credibility of an interpretation of a text. One way in which an interpretation of a text can cohere with other credible information about it, is by not contradicting that information, or rendering any of it improbable. ”

    As any Bayesian knows, the strength of an explanation isn’t in what it confirms or coheres with, but what it says shouldn’t be. For instance, if I were rolling dice and I roll a five, this coheres with both a normal six sided die and a 1 million sided die. Indeed, a 1 million sided die doesn’t contradict rolling a five! But the six sided die is more probable because it rules out almost 1 million other possibilities (this, combined with the prior probability of me having a 1 million sided die in the first place…).

    The most obvious example of this failure of explanatory power is in religious apologetics. But barring the extreme cases like 1 million sided die or the existence of god (which would be represented by a die with much more than 1 million possibilities since with god all things are possible…) it becomes a much easier mistake to overlook or gloss over when comparing more plausible hypotheses.

    “Of course the same NT scholars are aware of this problem and that is why someone like Bart Ehrman will insist that they really have multiple independent sources. But these “independent sources” are not the sorts of tangible documents McCullough is explaining are necessary: they are entirely hypothetical. ”

    Yes. All we have to do is ask e.g. Ehrman what is the probability of us knowing about, say M, without the existence of our Matthew text. Our knowledge of the so-called M depends on the existence of Matthew and as such it is not independent. Truly independent sources mean that we would know about them without some other source. Deriving M, L, Q, etc. from our received gospels is one of the instances I see in academia of “clever argument” per se being lauded in the stead of an actual discovery. I think this is one of the reasons why historical analysis that confirms the basic supernatural (conservative) Christian view is so welcome in NT scholarship.

    But maybe it’s just confusion due to our language and irresponsibly moving back and forth between epistemology and ontology. Ontologically, if M exist, then Matthew depends on it. But epistemically, M depends on Matthew. The only magisterium we work in as academics is the epistemic, and we use that as a means to approximate the ontological.

    (I just came across an article in the Oxford Journal of Legal Studies called “Probability – The Logic of the Law”. So not only do we have arguments in favor of probability being the logic of science (Jaynes), probability being the logic of historical analysis (Tucker, Carrier), but we now have an argument that probability is the logic of the legal world, too.)

    • Scot Griffin
      2015-06-17 05:17:56 UTC - 05:17 | Permalink

      “The arguments here are just so… Bayesian!”

      Logical, yes. Bayesian, not necessarily.

  • Tim Widowfield
    2015-06-16 19:38:16 UTC - 19:38 | Permalink

    McCullough’s statement on coherence is particularly apt. Our friends, the Memory Mavens, have to rely heavily on coherence, since in the main they reject the criteria approach. So we end up with a thesis that they insist is the most plausible because it makes sense of all the available evidence.

    In a nutshell, McCullough is telling us that “See? It all fits!” is a pretty lame argument.

  • Neil Godfrey
    2015-06-16 21:27:02 UTC - 21:27 | Permalink

    The point about coherence also strikes at the heart of a number of popular mythicist arguments: think Murdock’s astrotheology and Harpur’s revival of Christianity as another form of ancient Egyptian myths.

  • Reader
    2015-06-21 19:11:03 UTC - 19:11 | Permalink

    A somewhat related book:

    Writing the History of Israel – Diane Nunn Banks

    *From the Introduction, p. 1 (from the book itself)*

    “It is a fact worthy of remark that no biblical historian is included in the standard dictionaries of historians. A search of the indices will show Herbert Butterfield, famous for his Whig Interpretation of History, has written on the subject of biblical history. No one trained in departments of religion or in Semitic languages, however, has an entry. Specifically, historians of ancient Israel, well known to generations of students—such as Kittel, Eissfeldt, Oesterley, Lods, and Olmstead—find no place in these dictionaries… ”

    *From the Amazon blurb*

    “No biblical historian is included in the standard dictionaries of historians. Banks’ study examines the boundaries as well as the links that exists between history writing in biblical studies and the practice of history in university departments of history. She argues that while the influence of the profession of writing history is apparent, there are countervailing forces as well. The presupposition that the Bible is a book of history conditions the outcome of historical research in biblical studies. Banks argues that Julius Wellhausen’s history of Israel set in motion the general tendency toward ever greater congruence between historiography in biblical studies and in academic departments of history; that the initial tension caused by Wellhausen’s work produced a reaction which effectively stalled the movement toward accommodation between secular, academic history and biblical studies; and that a new generation of scholars applying the methods used by secular historians has revived and continued the tendency to promote the practice of secular, academic historiography in biblical studies. Banks applies her method to Wellhausen, Martin Noth, John Bright, and Thomas Thompson.”

    It would also seem(if the claim in the following article is true) that another dictionary (Oxford Classical Dictionary 4th) do not have as an historical entry, the alleged founder of Christianity – Jesus

    I am attempting to get my hands on a copy.

  • 2016-07-21 22:41:56 UTC - 22:41 | Permalink

    Thanks. This was a great article. I’m using McCullagh’s analytic-descriptive approach, and his inference to the best explanation model in my dissertation on Jesus’ Literacy. I think that the biblical historical-critical method (particularly the elusive and contrived Criteria of Authenticity) is long overdue to be retired from use. Finding the pre-literary authentic traditions is less likely than finding the flying spaghetti monster.
    Thanks again.

    • Neil Godfrey
      2016-07-30 01:46:22 UTC - 01:46 | Permalink

      Thanks. I must re-read this post to remind myself of what I wrote and read back then!

  • Leave a Reply

    Your email address will not be published. Required fields are marked *