Monday, March 23, 2009

Legends, Lüdemann, and Inductive Inference

In my discussions last year with David Parker regarding the legend theory (the theory that the New Testament documents contain legendary material), his counter was to defend the view that the earliest documents, namely the Pauline letters (in particular 1 Corinthians 15:3-8), were “too early” to be the product of legendary development. In the sources which he introduced in his defense, David emphasized the fact that historians generally affirm their conclusions as a matter of probability rather than as incontestable certainties. I of course acknowledge this, but pointed to numerous examples in the Christian apologetic literature which affirm that passages like 1 Corinthians 15:3-8 are “too early” to contain legendary material, with no suggestion that such positions have only a probable nature. In fact, often quite the contrary is the case.

A prime example comes from the very passage in Geisler and Turek’s book I Don’t Have Enough Faith to Be an Atheist which started our lengthy discussion in the first place, where they say of the passage in question that “there’s no possible way that such testimony could describe a legend, because it goes right back to the time and place of the event itself” (p. 242, emphasis added). There is a strong tendency in the apologetic literature to portray apologetic positions and arguments as aligned with, informed and corroborated by the latest, most rigorous historical scholarship. And yet here are Geisler and Turek telling us that “there’s no possible way” that 1 Corinthians 15:3-8 could contain legendary material, even though we are repeatedly told that historical conclusions are at best probable in nature.

In my analysis of Geisler and Turek’s misleading statements about 1 Corinthians 15:3-8, I pointed out that such declarations take the timeline for Jesus’ life, crucifixion and resurrection as found in the NT gospels for granted, which of course begs the question against the legend theory, the very position to which statements like the one quoted from Geisler and Turek are apparently trying to answer. Specifically, Geisler and Turek’s view assumes that Paul was talking of a man who recently lived, who was crucified and resurrected in his own lifetime. Unfortunately, nothing in Paul’s letters explicitly confirm any of this, and statements from Paul’s letters which are taken to imply that Paul’s Jesus was a recent historical figure are questionable at best, and difficult to reconcile with the overall treatment of Jesus found throughout Paul’s letters (see for instance here, here, and here). Paul nowhere gives any explicit timeframe for Jesus’ life on earth, never states when or where he was crucified and resurrected, and only vaguely (without any details pertaining to place, time or situation) mentions Jesus’ post-resurrection appearances, and even then only in passing, and only once in all his letters, namely in 1 Corinthians 15:3-8. This is in stark contrast to the detailed stories we find in the later NT strata, namely the gospel narratives.

In substantiating the position that it is improbable that legends in the first century AD developed in less than a generation, David referenced the work of Gerd Lüdemann, Professor of History and Literature of Early Christian at Georg-August-University Göttingen, Germany. According to David, Lüdemann is responsible for breakthrough research on legends in the first century, and found in that research that no legends outside Christianity developed in less than a generation, and may have even taken longer for them to develop. When asked what is meant by “develop” in this context, David clarified as follows: “the idea is that within a generation one would not see much changing of the story” (David Parker, 28 Nov. 2008).

Keep in mind that, prior to having been written down by Paul in his letter to the Corinthian church, this “creed” in I Corinthians 15:3-8 was supposedly an oral tradition that was passed around among Christians. That’s one of the operating assumptions lurking beneath claims, like Geisler and Turek’s, that the content of what we’re reading in this passage “goes right back to the time and place of the event itself.” I have sought from David validation for this assumption, and unfortunately it was rather fleeting, and in his 4 Aug. 2008 comment he admitted that he “cannot personally date the creed,” and “guesses” that I am correct that those who do assign a date of the early 30s AD to what is assumed to be a creed in 1 Corinthians 15:3-8, like Gary Habermas, cannot figure such a dating without relying on the gospels. Without the gospels, there is no indication of a date for Jesus’ crucifixion and no basis for assigning the date of what is claimed to be a creed in the passage in question to the early 30s. Consequently, if the content of the gospels has a legendary character, as the legend theory holds, then they are unreliable for purposes of establishing the date of Jesus' crucifixion. We are then left with the early epistles, which not only fail to identify a timeframe for the crucifixion, but in fact treat it as if it happened in some remote past, with no interest paid to time, place or circumstances of the event in question.

If I Corinthians 15:3-8 does contain an “early creed,” and Paul did not formulate it himself, who did? Of course, there seems to be no clear answer here. If it is a creed, it appears to be anonymous. This apparently does not raise any concerns for Christians insisting that it does contain a creed, for they also typically insist that Paul would have researched it and verified every element of what it stated, otherwise he would not have recited it in his letter. They take statements from Paul’s letter to the Galatians, where Paul mentions that he met with Peter and James on his two visits to Jerusalem, as license to make this kind of assumption. We are to believe, then, that Paul was not just out missionizing various gentile locales, but also an avid fact-checker, chasing down any possible shred of evidence to back up every claim he makes in his letters. We are thus assured that Paul did all the necessary homework, and we’re expected to simply believe what he says on this assumption. If a defense could be weaker, I wouldn’t want to see it.

As for the original content of an oral tradition (that is, when it is first formulated), when asked how we can know whether or not the tradition that finally got penned to a piece of paper years later underwent change or remained entirely intact, if it were embellished or modified along the way at some point, David conceded, in a 1 Dec. 2008 comment, that “historians can't know for sure,” but added that “they can formulate what the most probable conclusion [in] that best explains the data.” So we’re back to what someone thinks is “most probable,” and what apparently governs the assessment that a conclusion is “most probable” is how well “it explains the data.” This gives little cause for supposing that what we read in 1 Corinthians 15:3-8 in fact reflects an intact oral tradition which originated in the early 30s. Moreover, after ample attempts to criticize my position, I see no reason why the legend theory fails to "best explain... the data," or why it is less probable than the miracle stories contained in the literature.

At any rate we have, on this view, an oral tradition, which gives no details as to time, place or circumstances of the post-resurrection appearances of Jesus listed therein, whose author(s) are unknown, whose date of origin is at best speculative and based on later sources which clearly exhibit the telltale marks of legendary development, and which cannot with any confidence be verified not to have undergone change or modification since its original formulation. Paul then picks it up somewhere (though he never calls it a creed, and never identifies any human being(s) as its source), and recites it only one time in all his letters, tagging onto its end his own claim that the risen Christ “appeared” to him, and never elaborating on the experience he claims here to have had in any of his letters. There are numerous other problems with this passage which I will bring out in my next blog. For now, what is interesting at this point is the heavy reliance on inductive methodology in concluding that the New Testament, especially early letters like Paul’s, could not have been tarnished with legendary development.

David made the reliance of the “too early to be legend” thesis on inductive reasoning explicitly clear in several statements he made in the comments section of this blog.

For instance, in his 28 Nov., 2008 comment, David wrote:

The argument is a rather simply inductive move, taking the trend in first century legend develop and extrapolating to a probability about Jesus' story.

I had asked:

is your position akin to the view that, since Ludemann has (purportedly) shown that no other legend in first century Palestine developed in less than a generation, therefore the Christian legend did not develop in less than a generation?

In the same 28 Nov., 2008 comment, David responded:

Insert probably before "did not develop" and it looks good. Its called extrapolating from the sample to the general population. The heart of the inductive method.

And to confirm David’s position, I asked (in my 28 Nov., 2008 response to his above comment):

Okay, so the basic reasoning is: Since it is believed that no other legend in first century Palestine developed in less than a generation, it is supposed that the Christian legend therefore probably did not develop in less than a generation. How’s that?

To this David responded:

Yup, just like the old "all observed polar bears are white, therefore inductively we can assume that probably the polar bear in that room is white." Don't remember where I heard that example, but its common in books which introduce the inductive method.

Furthermore, in his 1 Dec., 2008 comment, David wrote:

it is my position that classical deductive/inductive arguments in addition to historical evidence are sufficient to make a rational case for Christianity.

These comments make it clear that David clearly assumes the validity of induction, and even states for the record that inductive arguments have a primary role in “mak[ing] a rational case for Christianity.”

Now recall that, when asked what his ultimate starting point is, David David gavethe statement “the Bible is the Word of God.” In that same exchange I had given my own critique of this statement as a starting point, and at that point David took it to a friend of his, a Christian named Dominic “Bnonn” Tennant, who posted his own response to my points on his own blog. (In turn I posted my own rejoinder to Tennant’s points, and so far I’ve seen no response to it from any Christians.) In going to Tennant for guidance on such matters, David clearly demonstrated that he considers Tennant to be an authority on things philosophical.

The problem is that Tennant’s position dramatically undermines David’s defense of the “too early to be legend” thesis. In his 18 Jan. 2009 comment to this Triablogue post, Tennant declared that “induction is an informal fallacy.” When challenged on this (by John Donohue, an occasional visitor to my blog), Tennant, in the same blog comments section, quoted an encyclopedia no less, no doubt a stalwart source of philosophical acumen:

In logic, a type of nonvalid inference or argument in which the premises provide some reason for believing that the conclusion is true. Typical forms of inductive argument include reasoning from a part to a whole, from the particular to the general, and from a sample to an entire population. Induction is traditionally contrasted with deduction. Many of the problems of inductive logic, including what is known as the problem of induction, have been treated in studies of the methodology of the natural sciences. (Britannica Concise Encyclopedia, 'induction'.)

If it is the case that “induction is an informal fallacy,” that it is “a type of nonvalid inference or argument,” what then can be said on behalf of David’s defense of the “too early to be legend” thesis? Wouldn’t this mean that the conclusion that, since no other legends in the first century developed in less than a generation, therefore Christianity probably did not develop in less than a generation, is based on a methodology which is informally fallacious?

I’m glad these aren’t my problems!

by Dawson Bethrick

12 comments:

David Parker said...

Inductive arguments are formally invalid. Formally valid means there is 100% certainty of the conclusion given the premises. You can't get that with induction...therefore induction is formally invalid.

Thus Tennant said:
"Then there is no certainty in it—induction is an informal fallacy."

Also:
"Induction is an informal fallacy—it is a kind of logically invalid inference which may nonetheless yield true results."

I don't know about his use of "informal fallacy" as a term, but I agree with him completely that you can't have certainty by induction.

Bahnsen Burner said...

Hi David,

Good to hear from you. How's Denver? Did the move go okay? Is it warming up??

You wrote: "I agree with him completely that you can't have certainty by induction."

Are you certain that this is the case for all instances of inductive reasoning? If so, how does one come to this conclusion without using induction?

Also, if I burn my finger on a hot stove 10 times, why shouldn’t I inductively infer from this that if I touch my finger to a hot stove an eleventh time, it will certainly burn me again?

Regards,
Dawson

madmax said...

Dawson,

Induction is universally attacked by almost all thinkers today. The so called "problem of induction" is often sited. I always have the example of the "white swan" thrown at me. Anti-induction types will say that induction by enumeration is flawed; ie what if one day we find a non-white swan? I think Objectivism rejects induction by enumeration but I am not certain of what exactly is Objectivism's approach to induction. I think Peikoff has addressed this recently but I haven't listened to his more recent lectures. I'm hoping that at some point in the future you could blog on induction and the Kantian / Popperian criticisms of it.

David Parker said...

Denver is going well so far. I'm still getting adjusted to the bigger city lifestyle, but have enjoyed being able to take the lightrail to work in the morning instead of driving. Definitely warmed up last week, but we're supposed to be getting snow this week! Glad to hear your travels abroad were eventful.

Are you certain that this is the case for all instances of inductive reasoning? If so, how does one come to this conclusion
without using induction?


When extrapolating from a sample to the general population, there is always the possibility of encountering an instance to the
contrary (more about that below). If you have the entire population as the sample, then you aren't extrapolating, and also you aren't using induction.

P1. Every person in this room is male
P2. John is a person in this room
C. John is a male

We are 100% certain that if the premises are true, the conclusion must be true.

You said:Also, if I burn my finger on a hot stove 10 times, why shouldn’t I inductively infer from this that if I touch my finger to a hot stove an eleventh time, it will certainly burn me again?

There is a deductive argument with a premise that requires induction:

P1. All hot stoves will burn your finger
P2. The object before me is a hot stove
C. Therefore, the object before me will burn my finger

But to justify P1 one must use induction.

Have we examined "all hot stoves" to verify that they will burn our finger? If we haven't, then really P1 doesn't correspond to
reality. What P1 should say is that all observed hot stoves have burned fingers, therefore we should probably expect the next one to do likewise.

You cannot be certain that the next stove will behave likewise, which is a limitation Hume coined as the "problem of induction."

Bahnsen Burner said...

David: “I'm still getting adjusted to the bigger city lifestyle, but have enjoyed being able to take the lightrail to work in the morning instead of driving.”

Ah, lightrail! I know it well. I imagine Denver’s system is much newer than San Francisco’s. Regardless, be careful of any sudden brake-stops. I’ve been hurled a few times myself. So far, so good. But I’ve seen some broken bones here and there over the years. Not fun!

David: “If you have the entire population as the sample, then you aren't extrapolating, and also you aren't using induction.”

If “the entire population” in question is *all instances of inductive reasoning* (which is in question, given the general prognosis of your/Tennant's position), are you confident that you have that as your sample? If not, then we’re back to my original question. If so, how did you acquire knowledge of this population (it's open-ended)? It certainly does not seem analogous to inquiring on the sex of all persons located within a single room. That’s quite confined. But “all instances of inductive reasoning” is not so confined. How does your analogy relate?

David: “Have we examined "all hot stoves" to verify that they will burn our finger? If we haven't, then really P1 doesn't correspond to reality.”

Really? What does “hot stove” mean to you then? Does it not refer to things known from your experience with reality?

There is a much better way to understand induction, David. But you won’t find it in Christianity or in the Christian apologetic literature. You might want to consult David Kelley on the matter. Then again, he’s an Objectivist, and as such he rejects all forms of mysticism (including Christianity), so you might have a problem with that. But he avoids the pitfalls that appear to be hindering your understanding of induction.

David: “You cannot be certain that the next stove will behave likewise, which is a limitation Hume coined as the ‘problem of induction’."

Kelley addressed this as well. He explains where Hume is mistaken (on several issues, including the starting point of knowledge, the nature of conceptualization, the nature of causality, etc.), and shows how the Aristotelian understanding of causality and the Objectivist theory of concepts together avoid and correct the problems which saddled Hume’s understanding of these issues. Quite an enlightening solution if you ask me. Now I understand how certainty can, in at least some instances, be acquired through induction, such as when I approach a hot stove. You won’t see me touching one of those any time soon. I bet you avoid touching them as well, no?

Regards,
Dawson

madmax said...

"What P1 should say is that all observed hot stoves have burned fingers, therefore we should probably expect the next one to do likewise.
You cannot be certain that the next stove will behave likewise, which is a limitation Hume coined as the "problem of induction.""

And there it is, right on que.

David has given the textbook argument against induction and for Humean skepticism. I think I know how Objectivism answers that but technical epistemology is not my strongest suit. I'll let you take the first crack at it Dawson.

madmax said...

Dawson,

Where did Kelley deal with induction? In "The Evidence Of The Senses?" I ask because I am going to read that book next. I really want to be able to answer the Humean skeptics who are everywhere especially amongst atheists.

Bahnsen Burner said...

Madmax: "I think Objectivism rejects induction by enumeration but I am not certain of what exactly is Objectivism's approach to induction."

One of the keys to understanding induction properly is identifying the causality behind the phenomenon in question (e.g., touching a hot stove top and your finger feeling the pain of a burn). In my response to David I deliberately played to his implicit assumptions about induction, such as that repetition plays an essential role in inductive generalization, by speaking of ten occurrences of touching a hot stove and extrapolating from those ten occurrences to an eleventh which has not yet taken place. I did this to see if David would catch on. He didn’t. Thus I’m pretty confident that he is unfamiliar with the Objectivist view of induction (at least as explicated by David Kelley in his lecture Universals and Induction) (still apparently available only on audio cassette – for instance on Amazon.com; I burned mine long ago to CD and uploaded it to my iPod – great listening pleasure!). Kelley explains how Mill’s methods of induction were a major contribution on the topic, in spite of Mill himself being an empiricist in Hume’s tradition, particularly in establishing the causality behind a particular kind of phenomenon (such as touching a hot stove top and burning your finger). Kelley focuses on the first three of Mill’s methods, and notes that

in all three methods we had to examine more than one case. These cases differ, however, and the differences between them… are crucial to the conclusions that we draw. So we are not simply gathering confirming instances, as in Hume’s approach. We are looking at a very structured way at a single set of instances. Now remember Hume’s argument… that if there were a necessary connection between cause and effect, we could generalize from a single instance. We wouldn’t need to repeat the observation or the experiment. Well that’s true. There is a necessary connection between cause and effect, and since the cases that we compare in any of the methods, constitute a single set of connected observations, we do in a sense generalize from a single instance. At any rate, we do not rely on repetition. We don’t have to collect identical confirming instances…. Repetition plays no essential role in knowledge at all: not in induction, not in concept-formation, not in any reasoning process. Whenever we reach certain conclusions about a given phenomenon – by observation, or by inference – the occurrence of an exact repetition of that phenomenon does not allow us to draw any *new* conclusions, except the obvious conclusion that this has happened before. Repetition as such, the sheer fact of repetition, is epistemologically barren. In concept-formation, for example, we could not form the concept ‘red’ by observing two identical shades of red. We form the concept by omitting measurements. But in order to do that we have to have some measurements to omit. So we have to have shades that differ quantitatively so that we can grasp the measurement relationship between them. And once we have two different shades that differ quantitatively, and we notice that quantitative relationship, we have all that we need to form the concept. Further instances add nothing essential (although they may facilitate the process psychologically). And the same is true of induction. Observing the same action occurring in two or more identical situations would not allow us to isolate the relevant causal factor. We need some variation among cases in order to use any of Mill’s methods. But once we have the kind of variation required by a given method, we have all we need in order to generalize. Further experiments that simply reproduce the cases, add nothing essential.

Overall, Kelley presents a radical departure from the mainstream (i.e., Hume-influenced) view of induction, which takes for granted many of Hume’s more fundamental errors. Notice though that it is not Mill’s methods alone which overcome Hume’s errors. Also notice that any time Bahnsen (or his ilk) raises the problem of induction, he does not question any of Hume's premises which led him to an impasse on the issue in the first place. That alone should raise one's eyebrows I'd think.


My view is that induction is actually an extension of concept-formation. Where concept-formation allows us to form integrations of entity-classes, for instance, induction takes this process as a model and allows us to form integrations of causal connections by applying the law of causality to entity classes so formed. For some further background on induction, see my responses to a commenter calling himself “Apologia4JC19” in my blog Presuppositionalism and the Argument from Ignorance.

Madmax: "Where did Kelley deal with induction?"

See above. So far as I know, it's available only in audio format. But it’s the best treatment of induction I’ve ever come across.

Madmax: “In ‘The Evidence Of The Senses?’ I ask because I am going to read that book next. I really want to be able to answer the Humean skeptics who are everywhere especially amongst atheists."

Kelley’s book The Evidence of the Senses is a great read, though it does get a bit technical in some places. If you absorb this, Kelley’s lecture on induction that I mentioned will go down like a smooth aperitif!


Now for David, I’m still wondering why, if “induction is an informal fallacy,” Lüdemann’s conclusion has any reliability.

Regards,
Dawson

David Parker said...

Now for David, I’m still wondering why, if “induction is an informal fallacy,” Lüdemann’s conclusion has any reliability.

Formal validity isn't a necessary precondition for "reliability."

Bahnsen Burner said...

David: "Formal validity isn't a necessary precondition for 'reliability'."

Are you saying that Lüdemann's argument is not formally valid, but it's reliable anyway?

Regards,
Dawson

David Parker said...

I'm suprised that my previous comments don't answer your question, so perhaps you are playing tricks again.

Is there some Objectivist definition of reliable that you're waiting to bring forth?

As an example: we can accurately predict the amount of time it takes a 1 gram steel ball to fall 1,000 feet in a vacuum chamber. Is this reliable? Yes, unless by reliable we mean "deductively certain." Is there a deductive argument to prove the sun will rise tomorrow? Nope, but I still call that reliable. Maybe you don't.

Perhaps you explain how one knows with certainty that a 1g steel ball will fall at the same rate given the same conditions every single time? (given Objectivist's definition of induction).

Bahnsen Burner said...

David: "Is there a deductive argument to prove the sun will rise tomorrow? Nope, but I still call that reliable. Maybe you don't."

David, it's not my view that "induction is an informal fallacy." Nor is it my view that certainty is never possible in inductive inferences. I'm certain that I can burn my hand on every hot stove I come in contact with. That's why I'm careful around them. Of course, I subscribe to the contextual view of certainty, which is expounded in Peikoff's Objectivism: The Philosophy of Ayn Rand, pp. 171-175.

Regards,
Dawson