# Parallels or Parallelomania (Josephus and Mark) How can we tell the difference?

• #### Tammuz

Can a Bayes equation demonstrate that these parallels establish something approaching a "historical certainty" that Mark depended upon Josephus?
A statistical approach cannot possibly provide any certainty in this case but at best a rather high probability, say 98,7 %, of the correctness of Carrier´s hypothesis. So there remains a rest of probability that the hypothesis is wrong. There cannot possibly be proven a, as you put it, "historical certainty".

But the main problem with such an approach is that a Bayes calculation needs a certain amount of definite figures with which it calculates. However, 22 events in text A and similar 22 events in text B is not enough for calculating the probability that they are causally related. You have exactly to know, for example, if and how many other similar cases of persons, who were called Jesus and who lamented over the temple and were interrogated by authorities and were scourged and so on, were given in the decades before the Mark gospel was written, and which could have served a a model for the Markan Jesus, instead of Jesus ben Ananus. If you say it is very improbable that more such cases were given, you have to calculate this improbability before you calculate Carrier´s hypothesis. But how to calculate this? You have exactly to know, for example, how many Jeshuas lived in Jerusalem in the decades before the Mark gospel was written. This is impossible. So you cannot rule out the possibility that there actually were more such cases like the Jesus ben Ananus case, even including the name forename.

Therefore a statistical approach turns out to be a phantom calculation since there are too many unknown factors to take into account, so that the whole enterprise of calculating the probability of the hypothesis proves to be a grotesque sleight of hand.

Look at some examples of Bayesian calculations on the web and you will find that they are all working with exact figures within an exactly defined frame of conditions, what is impossible in the case of the Carrier hypothesis.

In a situation like that, the first thing to determine is who could have copied who, that is to say the chronology of the two "portraits".
The second step would be to try and understand if the copycat acted to interpolate [to feel in the blanks of an incomplete tale of a personage] or simply transposed a "historical" figure on an other one [historical or not].
Clearly in the second case the real identity of the later "Jesus" would be not above suspicion.
This is similar to what I proposed (contextualization). As to the chronology, Carrier of course assumes that Mark is later than Josephus. It´s at any rate extremely improbable that Josephus copied from Mark.

Last edited:

• #### Kookaburra Jack

Returning for a moment to: the absence of a supporting probability argument to the claim that the parallels are (historically) sigificant.

Charlie Eppes from the series ´Numb3rs´ might be able to show us how very low the probability is that the parallels are random. So there are 20 or so parallels, each of which - let's say very roughly - is random on average with a probability of 50%. One starts with one parallel and then calculates successively how likely it is that more and more parallels occur together until all 20 parallels have been calculated. The initial probability of 50 % is thus halved 19 times. In the end, there remains such a low probability that the 20 parallels are random in their sum that the probability that they are NOT random is practically 100%. And yet all parallels could be coincidences. So the probability argument is not compelling.
This probability argument is at least a start in dealing with this issue. But it is not compelling because it looks to be equivalent to flipping a coin 20 times and recording heads every time. This is an astounding large improbable event. It is unlikely to be random. It is more likely that e have a coin with two heads. As a result I dont think this specific probability argument is valid. However there may be other forms of probability arguments which could throw some light on the question "How do we tell the difference" between valid parallels and invalid parallels of this kind? I cant see a simple solution to this atm. Any solution will need to be able to define what constitutes a "parallel", and may have to define how likely each is.

ETA: I see that you have arrived at a similar conclusion in the above post.

• #### Tammuz

This probability argument is at least a start in dealing with this issue.
Since I noticed the significance of parallel 22 (same unusual term for ´temple´ in Josephus and in a relevant place in Mark), I now tend to see Carrier´s hypothesis in a more favorable light. Parallels 1-21 might be coincidental but in connection with parallel 22 (which is quite astonishing) the hypothesis clearly makes sense.

(22) In both the “Jesus” stories, the term for Temple used in the syntax of their reputed pronouncements against the Temple is NAOS (_J.W._, VI. 301, 309 and Mk.14:58). This is a particularly interesting linguistic features the texts share in view of the fact that, except for this particular Markan linguistic parallel with the Josephus story, with respect to the use of NAOS as a term for Temple, Mark, otherwise routinely chooses to use the term hIEROS when he makes reference to the Temple in his narrative (11:11, 15, 16, 27; 12:35; 13:1,3; 14:49). Only in 14:58 (when the charge is presented— falsely so, according to Mark— that Jesus vowed that he would destroy the Temple), in 15:29, (where Jesus is mocked with that charge in the taunt at the cross: 15:29), and in 15:38 (where Mark seeks to vindicate Jesus’ attack on the Temple cult through the narration of the apparent divine rending of the Temple veil) does Mark choose to use NAOS as a term for the Temple rather than hIEROS.

As for a probabilistic approach:

Of course I mentioned Charlie Eppes just as a joke, the described "probability calculation" is my idea.

Now, as for the Bayesian calculation you mentioned, it´s working with fixed probability figures, say "97 %", what has nothing to do with flipping a coin, but rests on empirical measurements. In the case of my "method", I simply assumed - purely heuristically - that the probability of each parallel (separately viewed, that is, independent from the other parallels) to be coincidental is 50 %. Since that figure seems much too low, let us say the probability is 99 %.

So there is, heuristically, 99 % probability for each parallel found in Mark that it is not copied from Josephus (or an orally spread story about Jesus ben Ananus) but purely random.

Now we have to calculate the probability that ALL 22 parallels are random. Of course this probability should be significantly lower than 99 %, since the probability that they are not random but copied by "Mark" increases with the number of parallels.

Unfortunately I have no sufficient mathematical knowledge for an exact statistical calculation of that problem, it requires a certain and rather simple algorithm which I will find maybe later on the web.

So I can only very roughly proceed with my "method".

1) That the 2 Jeshuas are publically rebelling against the Jewish priesthood is 99 % random.

2) That they are doing so and cite from Jeremiah is, say, 95 % random.

From now on, I lower the probability with each step by 5 %.

3) That they do so and are interrogated by the Sanhedrin is 90 % random.

4) That all this happens and they are additionally scourged is 85 % random.

5) That all this happens and both of them are called "Jesus" (Jeshua) is 80 % random.

And so on.

In the end, you get a probability of 0 % that the synchronous occurrence of all parallels is random, that is, there is 100 % probability that they are copied by the author of the Mark gospel.

But of course this is the result of my very unprofessional method. Maybe the correct algorithm would provide quite another result.

And, of course, the result would be very different if we assume a probability of 99,9 % instead of 99 % that each parallel, individually viewed, is random. It would rather turn out that Carrier´s hypothesis is quite improbable.

However, in this case a probabilistic method even if professionally calculating can only work with estimated figures (probability of each parallel) so that the whole thing cannot be seen as a serious method for finding out to what degree Carrier´s hypothesis is correct.

Last edited:

• #### Kotromanic

However, in this case a probabilistic method even if professionally calculating can only work with estimated figures (probability of each parallel) so that the whole thing cannot be seen as a serious method for finding out to what degree Carrier´s hypothesis is correct.
Agree. (!)

Bayesian probability calculations are ONLY useful when the likelihoods of component probabilities have high confidence. They really should be based on observations.

Although the Hegelians will disagree with this, history is not a science. Rather it is by nature closer to theology or philosophy than a physical science. Carrier is not improving historical research methods with the muddying of the waters. My opinion.

Linschoten

• #### AlpinLuke

If we want to enlarge the perspective to the nature of history ... history is among the "human sciences", that is to say the not experimental scientific doctrines. The point is simple: how can you make a historical experiment? Can you recreate exactly the historical context of the D-Day making the Germans react more quicly to see what happens? Obviously you cannot.

History has go this limitation in common with psychology, sociology, economics, political science ... [it wouldn't be ethical to kill the pet of 100 children to observe their emotional reaction!].

Not only this, the action of the human will appears to be at least partially random [we say emotional], not totally rational, so that we cannot state expected values for our equations when we study the story of a set of human individuals. This is why who suggests a quantum approach to history is not exactly correct: QM, as odd it can look at first sight, is anyway a scientific experimental theory about mechanics. If the system is enough "big" [let's make it simple] QM is reduced to classical mecanics and a ball wan't pass through a closed window... if not breaking it!

Once I had occasion to hear a reasoning about history as a field and present as the normalization of the "Time Field". Philosophically, stretching a bit the "parameters", we could sustain such a perspective: present would be the sum of all potential pasts and when we observe what there's around us, stating a present, we normalize the Time Field stating also the past. But this doesn't move the problem: the will of the observer normalizes the Time Field.

Evidently, such an approach is out of context.

To go back to parallelomania, there is a detail we should keep in mind: in case of parallels and similarities, we could even not exclude a will to copy something previous. In History there have been personages with a model in mind.

• #### Kotromanic

Not only this, the action of the human will appears to be at least partially random [we say emotional], not totally rational, so that we cannot state expected values for our equations when we study the story of a set of human individuals. This is why who suggests a quantum approach to history is not exactly correct: QM, as odd it can look at first sight, is anyway a scientific experimental theory about mechanics. If the system is enough "big" [let's make it simple] QM is reduced to classical mecanics and a ball wan't pass through a closed window... if not breaking it!

.
Scientists present natural selection as the context to think about human behavior, human nature.

We see that as a rule the mammals with a three generation family or clan unit are emotionally driven in their lives. To attempt to model human behavior as primarily rational (economics of the 1960s/70s "Chicago school" for instance) is inviting poor match to human behavior "in the wild".

• #### AlpinLuke

Scientists present natural selection as the context to think about human behavior, human nature.

We see that as a rule the mammals with a three generation family or clan unit are emotionally driven in their lives. To attempt to model human behavior as primarily rational (economics of the 1960s/70s "Chicago school" for instance) is inviting poor match to human behavior "in the wild".
Economics, as any other scientific doctrine, has got some assumptions and principles which generally works. But even these A & P vary a bit when contextualized.

Among the A & P we can put, for example, the general mechanism which forms prices: the meeting of demand and offer.

But when we reason about "political economy" [or, if we prefer, general economy] there are well more complicated [and absolutely not linear] processes running in the system. I could remind here the famous "animal spirits" mentioned by Keynes to give an idea. And that's reality.

A nice example was Reagan ... his economic doctrine [usually called "Reaganomics"] was simple and linear, but at the end he had to expand public expenses to sustain it [I mean that the public dept increased its percentual weight regarding GDP].

• #### Tammuz

Rather it is by nature closer to theology or philosophy than a physical science.
It´s not closer to the first than to the latter; the science of history is a science of its own. Its material consists of documents and cultural and technical objects, of which many are archeological. Moreover, there are the remains of corpses and some very special objects, the mummies. Due to the fact that everything that is produced or arranged by humans has a meaning resp. a purpose, these objects - documents as well as material objects - have to be interpreted and placed into contexts. The discipline which deals with contextual interpretation is called Hermeneutics (originally a theological discipline but since the 19th century also applicable to everything that is the object of interpretation). The inventor of Hermeneutics in the field of human science is the German Wilhelm Dilthey.

Look, for example, how the art/science of interpretation is applied to the problems with which historians are faced with regard to Amenophis IV / Akhenaten. A few examples: Has there been a coregency with his father Amenophis III, and how long did it last? How old was Akhenaten when he decided to build the new town Akhetaton? Maybe only 14? Depending from the answers, the extent of Akhenaten´s share in the development of Atenism can be differently estimated. As to the seemingly distorted proportions of his body in depictions, it´s also not fully clear if it is due to an illness or the result of an aesthetical styling, while his body itself was normal. To come to conclusions about all that, Egyptologists have to interpret depictions and texts and to investigate mummies, what has nothing to do with theology or philosophy, but is Hermeneutics and natural science and the use of technology (C14 analysis, satellite photography research etc.).

Carrier is not improving historical research methods with the muddying of the waters.
Since this statement is in itself rather unclear, please clarify what you mean.

To go back to parallelomania, there is a detail we should keep in mind: in case of parallels and similarities, we could even not exclude a will to copy something previous. In History there have been personages with a model in mind
Someone said art is 10 % inspiration and 90 % transpiration.

This applied to our issue, I would say, religion is 10 % inspiration and 90 % imitation.

Last edited:

• #### Kotromanic

It´s not closer to the first than to the latter; the science of history is a science of its own.
.
I suppose my point of view won't allow me to find much common ground with you as I worked in applied sciences for my income for a little over 25 years.

I'll answer some of your comments later this evening. I'm not as verbally gifted as AlpinLuke, however I am a bit of an outlier as someone who practiced physical science, ie various disciplines of communications and semiconductor development engineering, for a long time.

For you to assert that history is nearer to physical science than it is to theology as a discipline doesn't seem at all justified, given my experience in working in the sciences and advancing the capabilities of humble (measured by price point) cell phones and wireless networking gear. But look forward to learning more of your point of view.

Carrier and also Jared Diamond, to me, are great examples of scientists who perform very poorly as historians.

• #### Kookaburra Jack

However, in this case a probabilistic method even if professionally calculating can only work with estimated figures (probability of each parallel) so that the whole thing cannot be seen as a serious method for finding out to what degree Carrier´s hypothesis is correct.
Agree. (!)

Bayesian probability calculations are ONLY useful when the likelihoods of component probabilities have high confidence. They really should be based on observations.

Although the Hegelians will disagree with this, history is not a science. Rather it is by nature closer to theology or philosophy than a physical science. Carrier is not improving historical research methods with the muddying of the waters. My opinion.
I tend to agree with the view that the success of any analysis, Bayesian or otherwise, is more effective when there is a great amount of known data available. As the amount of data avaliable decreases, as in the case of ancient historical events, the great shadow of the unknown gets in the road of any great certainties more often.

I agree that history is not a science, but it can use science (and statistical mathematics) as an advisory reference. The scientific tool of C14 dating for example provides a great advisory reference for historians, and far more use of it should be made to stress test long held church dogma surrounding key manuscripts.

Kotromanic