Rule #1 of Historical Reasoning

[MrMacSon]

This post is an addendum to my earlier Rules of Historical Reasoning. It addresses a particular instance where a historian uses a very late document as a repository of a very early "primary source" from a much earlier period.

Rule #1 (from Mark Day's The Philosophy of History, 2008, pp. 20-21) addressed the necessity of prioritising primary sources.

Primary sources here are understood to be the documents and other material artefacts that belong to the period being researched.

So coins minted by a king are primary sources for the reign of that king; a written account of that king that looks back on his reign subsequent to his death is a secondary source¹. A monument or decree written by order of the king that survives today is a primary source. A later historical work asserting claims about what the king wrote or decreed is a secondary source¹.

• (These are common definitions of the terms that have been in use since the nineteenth century. There is some fluidity among historians about how they use the terms but I have set out how they are used by Mark Day in his rules of historical reasoning.)

But what happens when the historian has no primary sources? That is, when no documents from the person/period being studied survive although the historian does have much later purported copies of primary sources²?

For example. Josephus, writing in Roman times, quotes what he claims is correspondence between the second century BCE Seleucid persecutor of the Jews, Antiochus IV (Epiphanes), and Samaritans. Is it legitimate for a historian to use Josephus's "record" of this correspondence as primary source material^3a for the actual events of the second century BCE?

The answer, I believe, is found in Elias Bickerman's analysis of Josephus's narrative, 'A Document Concerning the Persecution by Antiochus IV Epiphanes' in Studies in Jewish and Christian History, 2007, pp. 376-407.

Bickerman goes into very detailed argument to establish a reasonable argument that the historian is indeed justified in using Josephus's record as a genuine copy^3b of original correspondence dating from the time of Antiochus IV.

His arguments is based on several lines of evidence:

• archaeological evidence supporting originality of the correspondence in Josephus's work and providing details highly unlikely to have been known in the time of Josephus;

• misunderstandings by Josephus in his use of the letters that demonstrate an ignorance of practices alluded to in the letters that passed from usage in the Roman era;

• anachronistic references by Josephus that demonstrate a failure to understand the original context of the correspondence;

• other examples of genuine and forged correspondence^3c used as controls in Bickerman's argument;

• the extraordinary difficulties a forger would have had in getting specific details correct -- formulae appropriate to a narrow geographical and chronological range; accurate dating despite many potential chronological traps such as years beginning differently from one city to another, -- as they are in the correspondence cited by Josephus.

Bickerman concludes his argument for authenticity of the correspondence by addressing the possibility that

a forger was skilful enough to fabricate, one or two hundred years later, an impeccable document dated to 166 BCE. His diligence would not have done him any good; indeed, it would actually have detracted from the plausibility of his work because, if his readers were to be tricked into accepting it, they needed a document drawn up in the terms with which they were familiar, -ie. in the style of their own historical period. This explains the remarkable fact that forgers in antiquity normally employed the official formulae of their own period when they produced their texts.

In other words, Bickerman is very aware of the absolute necessity to establish a source⁴ as a "primary source"⁴ in order to use it as a basis for a historical reconstruction of the period being investigated.

I confess I was at first very suspicious of Bickerman's historical methods. The first work of his I read was God of the Maccabees in which he baldly stated, at one point, that we have various sources from the Seleucids pertaining to the Maccabean revolt. It was only after reading his justifications for this claim (as in the article discussed in this post) that I backed down and gave his argument some credence.

What Bickerman has given historians is a very solid argument. He has not given them primary sources. But he has given historians reasons to have some degree of confidence that they do have access to primary sources⁵. That means any argument based on these primary sources must necessarily remain hypothetical, always with awareness that the "sources" upon which the argument is based are conclusions of argument, hypotheses⁶, and not the "hard facts" as we have with coins or stone monuments or preserved clay tablets, etc.

Bickerman's primary "sources" will, like any and all primary sources, remain open to question and challenge. After all, that's what Mark Day (like many other historians) calls for: a constant testing and evaluation of the historian's source material.)

Bickerman's use of the documents as cited in Josephus are not a shoddy licence to make easy excuses for "making do with what we have or else we cannot do the history we want to do" type of unprofessional, unscholarly approach.

1 It would depend on what sources the writers of those 'accounts or 'works' used. For their accounts or works to be good secondary sources, their sources should, likewise, be good primary contemporaneous sources and their accounts of them good. Such an account or work may only be an unverified narrative.

2 copies? or other's accounts of previous documents [or purported previous documents]?

---

Neil, I appreciate you are posting here to clarify a lot of this. My commentary on your commentary is not to diss or criticise you per se or personally, but to facilitate this; particularly as you have an interest in this area and because you have an influence by your reputation and through vridar.

---

3 I haven't read Bickerman's account but a few thinks strike me about the terminology here -

• Is Josephus' account really a copy? or is just an account of 'correspondence between Antiochus IV (Epiphanes; the 2nd century BCE Seleucid persecutor of Jews) and Samaritans' ??

1. it would seem Josephus's account would best be termedas a secondary source.
2. a purported copy?
3. is the implication that Bickerman discusses whether
   1. Josephus had forged the correspondence [between Antiochus IV (Epiphanes) and Samaritans]? or
   2. Jospehus might have been using a forgery?

4 well, access to a secondary source that likely had access to primary sources

5 -ie. "access to secondary sources that likely had access to suitable primary sources."

6 or may be, and often are, accounts of accounts

[neilgodfrey]

Bickerman goes into very detailed argument to establish a reasonable argument that the historian is indeed justified in using Josephus's record as a genuine copy^3b of original correspondence dating from the time of Antiochus IV.

Too many arguments there. Thanks for drawing the sloppy wording to my attention. Fixed it in the OP.

(As for your other points, I'd like to think a second reading of the OP would clarify them.)

[Paul the Uncertain]

Neil

I am in general agreement with you that Bickerman appropriately admitted and weighed the evidence he discussed.

Usually, the same overt act can be described more than one way. In your telling, Bickerman defines primary in a way he finds to be situationally appropriate, but differently than others might. Having done so, he can preserve the form and apply a heuristic which could not otherwise be applied, the "Rule #1" of the thread title.

There is a different way of describing what Bickerman did. Seeing that one popular heuristic, Rule #1, could not be applied as it has been written, Bickerman applied a different heuristic instead. Since the two heuristics have similar "spirit" and concern similar attributes of the evidence, there are loose mappings between them. In this description, the Bickerman heuristic is a relaxation of Rule #1, and functions as a near substitute for it.

As near as I can make out, what's at stake in the preference for one reasonable and useful description over another of the same overt act is found in your OP's last lines,

Bickerman's use of the documents as cited in Josephus are not a shoddy licence to make easy excuses for "making do with what we have or else we cannot do the history we want to do" type of unprofessional, unscholarly approach.

Preserving the form of Rule #1, as your description of Bickerman's analysis does, is also desirable because you profess that there exists, in the words of a noun phrase you used:

the absolute necessity to establish a source as a "primary source" in order to use it as a basis for a historical reconstruction of the period being investigated.

Taking the two points together, I infer that your view is that if the form of Rule #1 isn't preserved, then that would open the door to shoddiness, easy excuses for adopting a slogan that you disapprove of, and an unprofessional or unscholarly approach to investigating the human past.

How am I doing so far?

[neilgodfrey]

Paul, I find your posts very difficult to follow -- your terms like "situationally appropriate", "preserve the form", "apply a heurstic" ... are too nebulous for me in the way you express them and I find my eyes glazing over with very little concrete substance registering in my brain.

To the extent that I think I can grasp something of what you are trying to say, . . . .

No, there is no relaxation of Rule #1. That is not what I have suggested happens in my OP. Nor is there anything in Bickerman's discussion as I read it that suggests anything like a relaxation of Rule #1. Quite the opposite, in fact.

No, I do not agree that Rule #1 "could not be applied as it has been written". The rule "as written" is that primacy must be given to primary sources. That's what Bickerman's work clearly agrees with. It is what my own post assumes throughout.

All the OP is addressing is that sometimes a primary source might be located in an unexpected place. There is no "weakening" or "relaxation" of Rule #1 or failure to sustain the rule "as written" at all.

I am simply following every discussion of historical method that I have ever read by professional historians that primary sources always take paramount position in any historical research. I don't know of any historian who would disagree. Do you know of any historian who does not follow that principle?

Even biblical scholars addressing Christian origins give tacit acknowledgement to that same principle every time they choose to date the gospels as closely as theoretically possible to 30 CE.

[Paul the Uncertain]

Neil

That is not what I have suggested happens in my OP.

Whether or not you suggested it (we agree that you didn't), it is a fair description of what Bickerman did, according to your account. You established the situation with the question

But what happens when the historian has no primary sources?

and that question is answered by your desciption of what Bickerman did in a specific instance, with your approval (= "situationally appropriate").

The rule as you gave it in the previous thread's OP was

..... the historian should prioritize primary sources, though should nonetheless be critical of these sources.

If the historian has no primary sources, then this heuristic literally offers no counsel whatsoever. It cannot be "violated" in the situation you posed, since it doesn't apply.

Ironically, the sense of primary given in the previous thread referred to observables that

transport the historian directly back to the past that the documents describe and of which they were a part, permitting the historian knowledge of that past without the accretion of subsequent interpretation and tradition.

That would seem to license Bickerman's maneuver almost directly, by allowing him to answer "but Josephus does permit me to look at something without the accretion of subsequent interpretation and tradition." That is, "I am not in the situation of a historian who has no primary sources in the sense just quoted," even though he is without primary sources in the sense you gave in this thread,

Primary sources here are understood to be the documents and other material artefacts that belong to the period being researched.

[neilgodfrey]

Neil

That is not what I have suggested happens in my OP.

Whether or not you suggested it (we agree that you didn't), it is a fair description of what Bickerman did, according to your account. You established the situation with the question

But what happens when the historian has no primary sources?

and that question is answered by your desciption of what Bickerman did in a specific instance, with your approval (= "situationally appropriate").

No, that is not a fair description of what B did according to my account.

The question applies to the start of the historical research phase. The historian only has at the outset secondary sources -- in this case Josephus.

But on analysis of that secondary source B discovers that he does have a primary source, after all. At least he has the intellectual content, verbatim, of the primary source. The primary source has been stripped of its original material container but its intellectual content is preserved word for word, and its provenance, the provenance of that verbatim intellectual content, can be established as having derived from 168 BCE, Palestine and Syria.

What we have in this case is just as spin more cogently expressed it: we find that we have a primary source nested within a secondary one, something that was not suspected at the start of the research.

What we have is not a paraphrase by Josephus or a creative reconstruction by Josephus of what the parties most likely said to each other. We have the intellectual content, the blood and guts, of a primary source whose provenance can be established to the time of Antiochus IV.

The rule as you gave it in the previous thread's OP was

..... the historian should prioritize primary sources, though should nonetheless be critical of these sources.

If the historian has no primary sources, then this heuristic literally offers no counsel whatsoever. It cannot be "violated" in the situation you posed, since it doesn't apply.

The point is that the historian discovered that he did have a primary source after all! Rule #1 applies. The rule does apply and B is very careful to ensure that it is not violated. That is the point of his detailed critical analysis.

My point was that B began with only a secondary source but on critical analysis of that secondary source concluded that it contained a primary source.

So the historian B discovered a primary source (i.e. its verbatim intellectual content) in the course of his inquiry. He did not begin with a primary source. He began with a secondary one but then he discovered a primary one.

So once he had that primary source, newly discovered, he prioritized that above the secondary sources.

Ironically, the sense of primary given in the previous thread referred to observables that

transport the historian directly back to the past that the documents describe and of which they were a part, permitting the historian knowledge of that past without the accretion of subsequent interpretation and tradition.

That would seem to license Bickerman's maneuver almost directly, by allowing him to answer "but Josephus does permit me to look at something without the accretion of subsequent interpretation and tradition." That is, "I am not in the situation of a historian who has no primary sources in the sense just quoted," even though he is without primary sources in the sense you gave in this thread,

There was no license involved. Nor was there any "maneuver" by B.

Josephus himself, the author, did not produce the primary source. Josephus did not understand much of what it was he was citing and he made a hash of it, raising questions in the mind of the attentive reader.

B was simply doing what the rules state: critical analysis on the secondary source that he had in hand. That's not a "maneuver" as you put it, nor is it a way to acquire a "license" to bend any rules.

Primary sources here are understood to be the documents and other material artefacts that belong to the period being researched.

Yes, that's exactly what B appears to have discovered in Josephus: a transcription of a document that belonged to the period being researched.

The only caveat is the following:

Obviously what Josephus appears to cite is not the material document that was directly composed by the second century BCE scribes on behalf of Antiochus IV Epiphanes and the Samaritans. But B's analysis yields the conclusion that the intellectual content, the words actually written by both Hellenistic parties, was somehow available to Josephus and he transcribed them into his otherwise secondary source.

If there is a dilution of the Rule #1 it lies only in the fact that we appear to have the verbatim intellectual content of the primary source without its material container.

What we have here is an unusual manner of coming by a primary source, at least the verbatim intellectual content of the primary source (which is the most significant component of the source). No rules are broken. In fact B's lengthy discussion is an attempt to demonstrate to his readers that he is very conscious of "rule #1" and demonstrating as thoroughly as he can how his study is in accord with that rule.

The unusual manner of uncovering the primary source (without its physical container) does not mean Rule #1 itself is violated or bent or compromised in any way.

The OP was intended to demonstrate an instance where a historian adheres fast to Rule #1 even in the very unusual situation of locating a primary source in a secondary container.

There is no dilution of the rule. What is being given primacy is the primary source -- the Hellenistic correspondence itself, the words exchanged between the two parties and that can be established as originating in 168 BCE in the Levant.

They do not originate with Josephus. They are preserved within Josephus.

That is the point of B's argument. It is subject to challenge, as all arguments are. And as all primary sources also are, as any source, primary or secondary, is. Always.

[neilgodfrey]

On further reflection, in the wake of responding to Paul I wonder if I would have been more accurate if I had said that what we have in Josephus (according to B's "discovery") is "primary source content". (It's a question I am still thinking through; I am open to suggestions.) Primary sources are materially situated in or established as being productions from the time/place/persons being researched.

Bickerman himself speaks of the Hellenistic correspondence between the Samaritans and Syrian king found in Josephus as "sources" -- implying they are Hellenistic or primary sources. Their provenance, B concludes, is 168 BCE, Palestine.

The question arises, then, whether any original source quoted in a secondary source should be considered a primary source. The answer has to be "No".

If a secondary source claims to quote a primary source, that quotation is still a secondary source. The historian's job is to critically evaluate sources. The only way the historian can establish whether a secondary source is indeed accurately quoting "a primary source" is by checking or testing artefacts, documents, sources outside and independent of the (secondary) text.

A quotation of a primary source in a secondary source does not turn the secondary source, even in part, into a primary source.

Herodotus claims to quote the inscription on the Karabel monument in Asia Minor but historians can see that monument for themselves and see that Herodotus's account is false. Had the actual inscriptions proven Herodotus was correct in all details then the best we can say is that Herodotus provides us with the verbatim intellectual content of a primary source. But Herodotus is not our authority; Herodotus is not the primary source. We have established the primary source by reference to data that is independent of Herodotus.

Bickerman likewise uses data external to and independent of Josephus to test the Hellenistic correspondence found in Josephus. If all Bickerman had to study was Josephus then there is no way he could have concluded that Josephus contains "verbatim intellectual content" of a primary source. Bickerman relies upon archaeological finds; texts of other correspondence from the time of the Seleucids.

Bickerman "discovers" the "primary source content" in Josephus "in spite of" Josephus himself. Josephus fails to understand some of the protocols relating to the correspondence and in subtle ways misrepresents it. Josephus's presentation of the correspondence raises questions that can only be satisfactorily answered by reference to Hellenistic era sources.

So the only way the historian can know if a quotation is a copy of the content of a primary source is by independently checking the primary source itself.

What Bickerman has done is establish that the Hellenistic correspondence cited by Josephus was "bigger than" Josephus himself in that Josephus failed to understand its protocols and in subtle ways misrepresented it. Bickerman was able to establish that the correspondence survived intact in Josephus and could be provenanced to the mid-second century BCE.

I said I was not interested in quibbles over definitions, but now it seems I am getting deep into problems of definition.

What is important is the clarity of the concepts themselves, whatever we call them, so that we can understand what we are relying upon when we are doing historical research.

It is important for the historian to clearly establish what it is he or she is analysing and using as a means of researching historical questions. I don't think this step is always well done by most of us. Many assumptions underlying arguments and analyses of the gospels and letters of the NT would be challenged, I think, if we took more care to think through exactly what it is that we are relying upon whenever we pick up our texts, whether the Gospels or Josephus.

[Peter Kirby]

It seems to me that, by preserving the strict definition of what a 'primary source' is, we tend to undermine the idea that they're strictly necessary for good history. Since the definition of 'primary source' doesn't include anything about being the minimum standard necessary to be usable as evidence, it's only logical that things other than primary sources (including some of those 'secondary sources' that we keep hearing about) could possibly be usable as evidence.

(This conversation is hard to follow. If someone believes they didn't make a point, please don't assume that I attributed it to them.)

[MrMacSon]

On further reflection, in the wake of responding to Paul I wonder if I would have been more accurate if I had said that what we have in Josephus (according to B's "discovery") is "primary source content". (It's a question I am still thinking through; I am open to suggestions.)

. . . . . . < . . Snip . . >

I said I was not interested in quibbles over definitions, but now it seems I am getting deep into problems of definition.

What is important is the clarity of the concepts themselves, whatever we call them, so that we can understand what we are relying upon when we are doing historical research.

It is important for the historian to clearly establish what it is he or she is analysing and using as a means of researching historical questions. I don't think this step is always well done by most of us. Many assumptions underlying arguments and analyses of the gospels and letters of the NT would be challenged, I think, if we took more care to think through exactly what it is that we are relying upon whenever we pick up our texts, whether the Gospels or Josephus.

I think you/we are getting into the language of clarification (for want of a better term): specifically, the nature of adjectives and adverbs, and how many might be needed to describe or clarify the nouns 'source', 'text', 'document', etc. I think we sometimes need two, and careful phraseology.

The Bickerman example is quite a complicated one, but I think you've teased it out quite well in this post.

In terms of "primary source content", a term you used twice - quoted above and later in your post - the 'content' in this case is either an induction or a deduction by Bickerman. And he concludes that Josephus misrepresented the likely original content. The correspondence 'content' that Bickerman has deduced (or induced) is probably still 'secondary' - there is still likely to be a probability as to its veracity.

Bickerman himself speaks of the Hellenistic correspondence between the Samaritans and Syrian king found in Josephus as "sources" -- implying they are Hellenistic or primary sources. Their provenance, B concludes, is 168 BCE, Palestine.

The question arises, then, whether any original source quoted in a secondary source should be considered a primary source. The answer has to be "No".

If a secondary source claims to quote a primary source, that quotation is still a secondary source. The historian's job is to critically evaluate sources. The only way the historian can establish whether a secondary source is indeed accurately quoting "a primary source" is by checking or testing artefacts, documents, sources outside and independent of the (secondary) text.

A quotation of a primary source in a secondary source does not turn the secondary source, even in part, into a primary source.

Herodotus claims to quote the inscription on the Karabel monument in Asia Minor but historians can see that monument for themselves and see that Herodotus's account is false. Had the actual inscriptions proven Herodotus was correct in all details then the best we can say is that Herodotus provides us with the verbatim intellectual content of a primary source. But Herodotus is not our authority; Herodotus is not the primary source. We have established the primary source by reference to data that is independent of Herodotus.

Bickerman likewise uses data external to and independent of Josephus to test the Hellenistic correspondence found in Josephus. If all Bickerman had to study was Josephus then there is no way he could have concluded that Josephus contains "verbatim intellectual content" of a primary source. Bickerman relies upon archaeological finds; texts of other correspondence from the time of the Seleucids.

Bickerman "discovers" the [likely original correspondence] "content" [used by] Josephus "in spite of" Josephus himself. Josephus fails to understand some of the protocols relating to the correspondence and in subtle ways misrepresents it. Josephus's presentation of the correspondence raises questions that can only be satisfactorily answered by reference to Hellenistic era sources.

So the only way the historian can know if a quotation is a copy of the content of a primary source is by independently checking the primary source itself¹.

What Bickerman has done is establish that the Hellenistic correspondence cited by Josephus was "bigger than" Josephus himself in that Josephus failed to understand its protocols and in subtle ways misrepresented it. Bickerman was able to establish that the correspondence survived intact in Josephus and could be provenanced to the mid-second century BCE.

¹ or by checking other contemporaneous information. As you said "Bickerman likewise uses data external to and independent of Josephus to test the Hellenistic correspondence found in Josephus."

Primary sources are [materially] situated in or established as being productions from the time/place/persons being researched.

As far as the Herodotus-Karabel monument goes -

Herodotus claims to quote the inscription on the Karabel monument in Asia Minor but historians can see that monument for themselves and see that Herodotus's account is false. Had the actual inscriptions proven Herodotus was correct in all details then the best we can say is that Herodotus provides us with the verbatim intellectual content of a primary source. But Herodotus is not our authority; Herodotus is not the primary source. We have established the primary source: [the monument itself] by reference to data that is independent of Herodotus.

[Paul E.]

It seems to me that, by preserving the strict definition of what a 'primary source' is, we tend to undermine the idea that they're strictly necessary for good history. Since the definition of 'primary source' doesn't include anything about being the minimum standard necessary to be usable as evidence, it's only logical that things other than primary sources (including some of those 'secondary sources' that we keep hearing about) could possibly be usable as evidence.

(This conversation is hard to follow. If someone believes they didn't make a point, please don't assume that I attributed it to them.)

Regardless of any semantic debate about the definition of "primary source," I don't think many historians (or any?) would be dogmatic that they are "absolutely necessary," would they? You use what is usable as you see it. Chracterize it and prioritize it as you will, making your arguments and qualifying your conclusions, and others will ultimately judge your work. I guess I concur in your confusion about these posts.