A Reply to Mikeal Parsons and Heather Gorman

Abstract

A Reply to Mikeal Parsons and Heather Gorman Patricia Walters 1 Background In edition 46.1 of Neotestamentica, Mikeal Parsons and Heather Gorman wrote a timely and thorough review essay of my book, The Assumed Authorial Unity of Luke and Acts: A Reassessment of the Evidence (Walters 2009). The review essay assessed the strength of the argument in my book that, based on a five-part statistical analysis of Greek prose compositional style in the seams and summaries of Luke and Acts, the common authorship hypothesis must be abandoned (Parsons and Gorman 2012, 139–152). Given the conviction that my work has the strength and robustness to stand on its own, a reply to their review did not seem necessary until now. Because my results have received noteworthy independent corroboration (Mealand 2016), however, a reply to Parsons and Gorman is in order. Briefly, the analysis in my book proceeds in three stages. First, it uncovers the least contested authorial passages in Luke and Acts, which turn out to be seams and summaries: the literary sutures that connect pericopes (Walters 2009, 43–89). Second, it culls out the often subtle and secondarily-studied Greek prose compositional style conventions popular during the first century CE, of which five were most pertinent: hiatus, dissonance, prose rhythm, final syntax and clausal connectives (Walters 2009, 90–136). Third, by statistical testing, it analyses whether the five conventions, present or not, in the Luke and Acts seams and summaries show verifiable similarities—expected with single authorship—or significant differences (Walters 2009, 137–189). In all five cases, the analysis shows statistically significant differences, which I argue must lead to the conclusion of different authorship (Walters 2009, 140–149). 2 Response to the Review Essay Parsons and Gorman (2012, 139–152) focus on four areas that they regard as problematic to my methodology: (1) textual variants; (2) clausal segmentation; (3) data selection; and (4) stylistic criteria. Each area is [End Page 489] treated separately below. What Parsons and Gorman appear to misunderstand in general is the subtle nature of the chosen prose style conventions, which for all intents and purposes philologically point to identifying characteristics of a writer (Walters 2009, 94–96, 145–147). 2.1 Textual variants In this, the shortest discussion of the four, Parsons and Gorman (2012, 140–141) claim that one must account for manuscript variants. They cite the example of hiatus found in the uncontracted τετραάρχ-stem (Luke 3:1 [x3], 19; 9:7; Acts 13:11) compared with the contracted τετράρχ-stem, which is a manuscript variant with no hiatus (Parsons and Gorman 2012, 140–141; cf. Metzger 1994, 353). They suggest that by tabulating the contracted stem here and by evaluating germane textual variants elsewhere, my statistical results would change (Parsons and Gorman 2012, 141). They are correct, but not for the reason they suggest. It was a predetermined decision to rely on the text of the Nestle-Aland Novum Testamentum Graece 27th critical edition (1993; hereafter NA27), reconstructed by a committee of expert text-critical scholars. To cherry-pick among textual variants, possibly even significant ones, whether in the critical apparatus of the NA27, the Metzger Textual Commentary (1994), or another published version of the Greek NT, would impute to the data a stratum of interference. The result would be a lack of data integrity and a flawed methodology. Had I cherry-picked among textual variants, a counter-charge would undoubtedly be "stacking the statistical deck" in my favour. The most judicious approach to textual variants is to eliminate idiosyncrasy by letting the experts decide. 2.2 Clausal segmentation In this part of the review essay, Parsons and Gorman take issue with the way seams and summaries are segmented or divided into clauses for use in the statistical analyses, especially the final syntax of a clause. As with textual variants, clausal segmentation must avoid idiosyncrasy to the greatest extent possible. Therefore, segmentation guidelines were generally derived from a fourfold set of references: first, the ancient critics themselves (e.g., Dionysius of Halicarnassus, Comp. 18; cf. Thucydides, Hist. 2.35.1); second, H. W. Smyth's Greek Grammar (1956); (3) third, [End Page 490] the NA27; and fourth...