Biblical Criticism & History Forum - earlywritings.com

FWIW, my introduction to Stylometry was A Q Morton's Paul, The Man and the Myth (1966) which concentrated on "sentence length" (measured in the number of words between commonly accepted punctuation marks known as full stops, which he called "spons").

The problem is that punctuation in the 1st century CE was not as sophisticated as it became in the late middle age minuscules, so our modern punctuation may not be an especially good criterion. The other criterion he looked at (vocabulary mostly) seemed kind of rudimentary to me even when I read this book back in the 1990s.

Nevertheless, from very extensive statistical tables in the appendices at the end, I created graphs in an Excel spreadsheet that demonstrated some very interesting, and baffling, patterns between all the books of the Pauline corpus and a few others such as 1 Clement, Barnabas, and a couple Greek writers. I think I calculated the coefficients of correlation for all possible combinations, maybe even tried to determine whether results were "significant" or not, all sorts of things like that.

Oh, all my brilliant work was pure crap, but I learned a lot.

DCH

Edit 5/21/22: I have attached an old Excel97 spreadsheet that had the data and coefficients of correlation. There is even a graph on one tab. The correlations seem high, but sample size may require much higher correlations to be considered significant. There may have been a time I had tried to see if these correlations were significant, but I can no longer find it in my computer.

DCHindley wrote:FWIW, my introduction to Stylometry was A Q Morton's Paul, The Man and the Myth (1966) which concentrated on "sentence length" (measured in the number of words between commonly accepted punctuation marks known as full stops, which he called "spons").

The problem is that punctuation in the 1st century CE was not as sophisticated as it became in the late middle age minuscules, so our modern punctuation may not be an especially good criterion. The other criterion he looked at (vocabulary mostly) seemed kind of rudimentary to me even when I read this book back in the 1990s.

Nevertheless, from very extensive statistical tables in the appendices at the end, I created graphs in an Excel spreadsheet that demonstrated some very interesting, and baffling, patterns between all the books of the Pauline corpus and a few others such as 1 Clement, Barnabas, and a couple Greek writers. I think I calculated the coefficients of correlation for all possible combinations, maybe even tried to determine whether results were "significant" or not, all sorts of things like that. Oh, it was pure crap, but I learned a lot.

DCH

There is indeed a very old literature on sentence length (as long ago as 1887, Mendenhall worked with sentence length in English authors).

It is indeed problematic, not least for the reasons you mention (sentence length having as much to do with the editor of an ancient Greek text as it has to do with the author).

As a solitary criterion, sentence length has been frequently analyzed, with a negative conclusion ('not enough evidence,' 'not enough evidence!' -- to steal a paraphrase from Bertrand Russell).

There is also a huge literature on 'vocabulary size', not all of it strictly 'stylometric'. There is some good mathematical work done here, but it's unlikely to yield anything useful for stylometry without extremely large samples (>100,000 words I reckon).

I hope you find the 'basic' program that I put out to be useful. If you can suffer working with Excel spreadsheets for this stuff, it should be a welcome change of pace I think. I'd like to hear what someone else can do with it.

DCHindley wrote:I am not seeing anything about stylometry directly, unless it is in that second asterisked point about the "style [being] radically different from Paul’s".

I believe that is the case.

It has been many, many, many years since I read the book, so of course I may be misremembering.

Ben.

DCHindley wrote:FWIW, my introduction to Stylometry was A Q Morton's Paul, The Man and the Myth (1966)

Incidentally, A. Q. Morton's notoriety is something of a mixed blessing, given that a good number of his methods (even later in his career--the whole "QSUM" thing, which was even occasionally brought into court) have been generally discredited (IMO).

Authors of the Mind: Some Notes on the QSum Attribution Theory
http://www.jstor.org/stable/40372114

An Investigation of the Basis of Morton's Method for the Determination of Authorship
http://www.jstor.org/stable/42945562

Pseudoscience: A Comedy of Statistical Errors
http://www.jstor.org/stable/42945741

CUSUM: a credible method for the determination of authorship?
http://www.ncbi.nlm.nih.gov/pubmed/9206319

Ben C. Smith wrote:I myself am quite interested to see your stylometer used on the Pauline epistles. IIRC, my first exposure to stylometry was Colossians as Pseudepigraphy, by Mark Kiley. I do not think he himself used such methods, but he drew on someone who did (I forget who), and IIRC those little function words, as you call them, were pretty important.

Ben.

Anthony Kenny http://www.amazon.com/A-Stylometric-Stu ... 0198261780 may be of interest.
Kenny argues that it is possible on stylometric grounds that all 13 letters were written by the same person with the Pastorals being the least likely to be written by the main author.

Andrew Criddle

Great post, Peter. I've been on the fence about the authenticity of the passage. You've knocked me off the fence!

Thanks, Roger.