Tuesday, May 17, 2011

Sharon Mattila's Statistics and the Synoptic Problem: Some Quibbles

One of my favourite spare time activities is to dip into statistical studies of the Synoptic Problem. One day I might even write something about it. In going back over an article by Sharon Lea Mattila recently, I noticed that I had scribbled all over my print copy and that some of my scribblings might usefully be transferred to a blog post lest they get lost.

The article I wish to comment on is Sharon Lea Mattila, "Negotiating the Clouds Around Statistics and 'Q': A Rejoinder and Independent Analysis," Novum Testamentum 46/2 (2004): 105-31.  The article is a detailed, lengthy and frequently polemical response to Charles E. Carlston and Dennis Norlin, "Statistics and Q -- Some Further Observations," NovT 41 (1999), itself largely a response to an earlier article by Mattila.  For an excellent summary of  these and other works on statistical studies of the Synoptic Problem, and for full bibliography, see John Poirier, "Statistical Studies of the Verbal Agreements and their Impact on the Synoptic Problem," Currents in Biblical Research 7 (2008): 68-123.

The gist of the article is that Mattila thinks that Carlston and Norlin have overstated the case that Matthew and Luke preserved the Q sayings more faithfully than the Marcan sayings.  One of the key elements in the article is "an independent statistical analysis" that attempts to avoid what she sees as the flaws in their earlier study.

My comments here relate to difficulties with Mattila's figures tabulated on pp. 125-6 of the article:

(1) Mattila claims to avoid "such pitfalls as including sayings containing words from Scripture" (p. 121) but two of her parallels feature quotations from Scripture:
  • Matt. 10.34-6 // Luke 12.51-53 quotes Micah 7.6
  • Matt. 21.13 // Luke 19.46 is a famous composite quotation of Isa. 56.7 and Jer. 7.11.
(2) Mattila makes the interesting methodological move to count passages normally attributed to Mark-Q overlap as Triple Tradition (p. 122). In some contexts, this could be a strong and defensible move, e.g. it could help us to take seriously the existence of major agreements in triple tradition material, something that has a bearing on arguments for and against the 2ST. However, in this context, in which she is discussing Carlston and Norlin, and dealing specifically with claims about a written Q in comparison with a written Mark, she is potentially transferring material that should be in one table to the other. Four of her Triple Tradition pericopes are in Mark / Q overlap
  • Matt. 10.19-20 // Luke 12.11-12
  • Matt. 13.31b-2 // Luke 13.18b-19
  • Matt. 16.6b // Luke 12.1b
  • Matt. 18.6-7 // Luke 17.1b-3a.
It's an important difficulty with the analysis when evidence that belongs on one side of the experiment ends up on the other.

(3) Mattila includes the Lucan woes in her analysis in spite of the fact that they have no parallel in Matthew, and against her stated objective of treating Aland's synopsis divisions as the standard.  She is conscious that this is problematic (p. 121).

(4) This is a point that is rarely treated seriously in statistical studies of the Synoptics.  Parable material occupies an interesting position here in that it is both sayings  material (in attribution to Jesus) and narrative (in form).  Parallel parables appear often to be much less close to one another in wording than is other sayings material. Indeed, they sometimes imitate the way that narrative material appears. Mattila's double tradition table includes four parables across 47 verses of Matthew:
  • Matt. 7.24-27 // Luke 6.47-9
  • Matt. 18.12-14 // Luke 15.4-7
  • Matt. 22.2-14 // Luke 14.16b-24
  • Matt. 25.14-30 // Luke 19.12b-27
Her triple tradition table includes three parables across only 23 verses in Matthew (the second is the Mark-Q overlap parable of the Mustard Seed listed also above):
  • Matt. 13.3b-9 and 18-23 // Luke 8.5-8a and 11-15
  • Matt. 13.31b-32 // Luke 13.18-19
  • Matt. 21.33-40 // Luke 20.9b-15
(5) Unlike Carlston and Norlin, Mattila excludes from consideration speech from characters other than Jesus (John the Baptist, the Centurion, etc.).  I am not sure about this move, especially as they include famous examples of very high verbatim agreement, thereby potentially skewing the figures for the double tradition.

No comments: