TedM

Sure they compare numbers. My point was that all one can say from that is "yup, their different, and it looks significant". I said that much in the OP. What they COULDN'T say was "no author could possibly have written both works", because they didn't have the data to make such a claim: ie, they didn't have OTHER such comparisons which they could then compare to the differences they were seeing.

That's what I mean by saying there is no "reasonable comparison basis". They ASSUMED that the differences were supportive of another author instead of the many different ALTERNATIVE explanations that exist and are consistent with the same authorship.


What matters to me is the quality of the statistical arguments considered generally to be meaningful. This site appeared to address a few of them, and I thought quite well, providing numbers that seemed meaningful. I get your point. You don't trust them and want me to find better sources. Feel free to direct me to some.

ted

Vorkosigan

No, TedM, the only one with assumptions here is you. You see, real scholars work from the premise that the text in front of them has no particular authorship until authorship is demonstrated. The use of the name Paul is only prima facie evidence, since it is well known that many texts in antiquity falsely use Paul's name. In fact, that knowledge is what compels us to search for evidence on authorship using accepted methodological stances. Once stylistic and other analyses all converge on a single conclusion, namely that the Pastorals are different from the other letters with Paul's name on them, then conclusions are made.

Explorations of alternatives are made when competing alternatives are offered using POSITIVE arguments. For example, when scientists argued about whether dinosaurs are warm-blooded and how they are, the alternatives were generating by reading data through methodology to produce a POSITIVE ARGUMENT -- bone structure indicates one kind of warm-bloodedness, volume/mass ratios indicate another. The DISCREDIT position you take here is not a POSITIVE argument. It is not an alternative generated by methodology working on a dataset. It is a defense of an a priori position and as such is not worthy of serious intellectual engagement. Why?

To understand what you are doing, let's suppose you are right and you can comprehensively discredit each method and conclusion underpinning the current overwhelming consensus on the Paulines. What is the correct position then?

A. Paul wrote the Pastorals
B. Paul did not write the Pastorals
C. We don't know who wrote the Pastorals.

For apologists, the correct answer is A. For scholars, the correct answer would then be C -- back to the drawing board and more testing to see if the same guy who wrote Romans wrote 1 Tim. Are you an apologist or a scholar?

Vorkosigan

Steven Carr

Is it possible to look at the words in very small parts of the Pauline letters (eg the words in 1 Corinthians 15 or Romans) and conclude that they differ from Paul's normal vocabulary so much that they must come from very early Christian teaching?

I think Christian apologists can do that, while also maintaining that even examining the vocabulary of whole letters will not let you conclude that the vocabulary is non-Pauline.

http://old.anglicanmedia.com.au/old/pwb/romans1.htm 'In the opening lines of Romans where Paul is giving an apologia for his apostolate to the Gentiles he adapts an existing credal statement about Christ. Its vocabulary betrays its un-Pauline origins, though we do not know precisely the time and place of such origins.'

So I think we should hear less talk about how looking at vocabulary cannot tell us whether Paul did or did not write something, as Christians will happily argue out of both sides of their mouth.

These techniques can be used to determine authorship. They just have to be applied properly.

TedM

I have no problem with that.

Sure. I'm saying that I doubt that the evidence allows ANY conclusion with regard to authorship of the Pastorals, at least as it pertains to the two issues in the OP.

I choose C. That is my position--I'm pointing out what appears to be the flawed conclusion that linguistic evidence weighs against Paul as the author of the Pastorals. IF it could be shown that words in the Pastorals didn't exist in the 1st century or that the style is dramatically more different than what is found for other similar writings by just one author, then that is evidence that can be used. IF that can't be shown, I see no reason to rely on these stats.

I find it interesting that you have not made any comments on the two issues I raised in the OP--comparisons of hapax legomena with Shakespeare, and the similar (even greater) frequency of "Pauline" function words in the Pastorals as compared with function words in works like Romans and 2 Corinthians.. IF, IF , IF the numbers I quoted are correct (regardless of their source), what meaning would you give to them?

ted

Vorkosigan

None. Because they are nonsense. The count given is PER PAGE. no ratios of HLs to total text is given -- the page count is meaningless as there is no definition of page consistent across either Shakespeare and the Pastorals. No definition of a Shakespearean HL is given, so no meaningful discussion can be had. In fact the 19th century style source you are using -- the Catholic Encyclopedia -- has long been displaced, as stylometrics are far advanced. See this article:

http://shakespeareauthorship.com/ox6.html

and this, just because it is so fucking cool

http://shakespeareauthorship.com/ox7.html

As for the function words, your site madly compares the "ten pauline letters" with the pastorals. There are no "ten" paulines, so the database for comparison is all wrong. So nothing can be said until you get a methodology that correctly chops up the Paulines.

Vorkosigan

andrewcriddle

It's a long time since I read the book so the following is IMS.

Kenny's analysis is book by book not chapter by chapter.

He says of his results something like them being compatible with all 13 letters being by one person with a rather variable style with the possible exception of Titus which is so short a book as to raise questions about the significance of the statistical analysis.

I would prefer to interpret the results as saying that at face value they show Titus to be non-Pauline and 1 Timothy to be close enough to Titus and far enough away from the typical Paulines as to strongly support common (non-Pauline) authorship of 1 Timothy and Titus.

2 Timothy being stylistically intermediate between (1 Timothy and Titus) and the other Paulines.

Andrew Criddle

Gamera

I beleive Malcolm Gladwell's book Blink goes into this issue in detail, and may even cite this study.

It's bogus and discreditted. The problem with this privileging of noncognitive knowledge is confirmation bias (better named the "clock radio effect" from the experince of waking up in the morning with a song on the radio which you've just dreamt about).

All these noncognitive studies simply ignore confirmation bias: the fact that you notice when your "intuitive" thinking gets it right, but ignore or discount when your intuitive thinking gets it wrong.

Thus, wake up in the morning with a song on your clock radio that you just dreamed about, and you notice. Wake up with a different song, and you don't even think about it. So you think you're psychic or something when you get it right, not noticing the thousands of times you got it wrong.

Scholars who claim the validity of noncognitive knowledge of the thing at issue (like the famous fake Kouros at the Getty), never notice when their noncognitive knowledge leads them astray. Thus this type of knowledge is useless.

Gamera

I don't know what this means, but I do know Ted and I have raised an important methological fault in stylimetrics. The fact that peer review has ignored this flaw doesn't make it go away.

I noticed you haven't been able to address it at all. So I ask again: if there is no baseline for determing what level and types of differences are significant in determining authorship, then what good are the conclusions of stylometrics in analysing the authorship of the Pastorals.

TedM

(above)Thanks for your thoughts on this Andrew.


Good observations. They are similar to the ones I made of the quotes you provided by Schnell and Erhman. It can be frustrating to try and make sense out of the stats when the definitions aren't clear and much info is missing. I would assume that a Shakespearean HL is a word that occurs in only one of the alleged Shakespeare plays, and that a Pastoral HL is one that occurs in only one of the alleged Pauline letters, but without more information we don't know if that assumption is correct. IF that assumption is correct the numbers show that great variability exist within both sets of works. I agree that the size of the sets of works is a factor that may be significant, and can affect our interpretation of such variability. In this case, if the above assumptions regarding the definition of HL are correct, then the variability found in Shakespeare plays is perhaps surprising since his works comprise a lot of words and many more than found in the alleged Paulines.


I wonder if this more advanced analysis has been made to the alleged Pauline works? It is interesting that as far as 'rare' words are concerned, the article states:

Might this apply to alleged Pauline works, such as the Pastorals also? If so, "rare words" are not useful for determining Pastoral authorship.


There may be ten paulines. We don't know, but I agree that function words are of NO VALUE if one can't determine a proper Pauline baseline to which another work can be compared.

I look forward to your answers to Gamera's questions.

ted

Ben C Smith

Ted, what do you think of 1 Timothy 2.15 and 5.14 as points against Pauline authorship? After reading 1 Corinthians 7, I have a hard time imagining Paul encouraging marriage and childbirth as a matter of course.

Ben.