The Extent of the Pauline Corpus: A Multivariate Approach

Abstract

Samples of 1000 words were taken from the New Testament Epistles. From these 25 stylistic variables were analysed. After testing for normality 19 of these were used, first in cluster analysis, then in discriminant analysis. Where there were several samples from one epistle these usually grouped together. Epistles known to be by different authors (Paul, Hebrews, 1 Peter, 1 John) usually clustered in separate areas. Within the Pauline corpus the major and minor Paulines grouped together, but Colossians and Ephesians separated in one direction and the Pastorals in another. The cluster analysis identified these main groups. This was followed by discriminant analysis in which the data were partitioned into calibration and test sets. The results clearly assigned some initially marginal items to the groups mentioned above. Key probability factors confirmed the discrimination, and tables of distances corroborated the greater distance of the deutero-Paulines from the four major and three minor Paulines.

Who Wrote 2 Thessalonians? A Fresh Look at an Old Problem

“…Interestingly, on the stylistic map this cluster occupied the central position with the other clusters showing varying degrees of divergence from this medial group (Mealand 1995: 68). Also on some of the statistical measures, Colossians separated away from Ephesians and tended towards this central grouping (1995: 78). From his extensive analysis, Mealand found that the multivariate analysis supported the following findings in relation to 2 Thessalonians.…”

“…Interestingly, but perhaps reassuringly, Colossians and Ephesians formed another closely related cluster. However, Mealand’s analysis revealed that ‘Colossians and Ephesians separated in one direction and the Pastorals in another’ (1995: 92). This Colossians and Ephesians grouping was not as distant on the stylistic map from the Romans, 1 and 2 Corinthians and Galatians group as the Pastorals.…”

“…It may be worth recording that of these three letters it was 1 Thessalonians rather than 2 Thessalonians which seemed to be slightly more distant on the criteria used. When 2 Thessalonians was specifically subjected to discriminant analysis as a doubtful sample it was decisively classed with Paul (1995: 86).…”

“…The two exceptions are that 2 Thessalonians only weighs 823 words, and Titus is a mere 659 words long. In these cases I used almost the whole epistle as a sample but made the total up to 1000 by multiplication’ (1995: 64-65).…”