The application of mathematics to the New Testament

Mathematics has been applied to the New Testament to assist in determining the Greek text, investigating document authorship, delineating sources, and (in John's gospel) supporting dislocation hypotheses. The present project uses it help identify interpolations, and in finding the structures which NT authors had in mind.

Textual Criticism

A relatively simple way to decide which manuscript reading appears to be the most original is to allocate one vote to each manuscript and count the votes for each alternative reading. This produces what is called the "Majority Text". However the method is simplistic, for the most ancient reading might happen by chance to have been copied the least. For an expert scholarly criticism of the method, see D.C.Parker, "The living text of the gospels" (CUP, 1997, pp.129-131).


In "The problem of the Pastoral Epistles" (1921), P.N.Harrison presented his stylistic studies which involved, among other things, counting the number of words in the Pastorals which did not occur elsewhere in the New Testament. The unusually large number of these 'hapax legomena' helped to convince him that the Pastorals were not written by Paul, but were produced in the second century.

In "Paul, the Man and the Myth" (Hodder and Stoughton, 1966), A.Q.Morton and J.McLeman counted sentence lengths and occurrences of the Greek word kai ("and"), concluding (p.13) that "... no more than five of the fourteen Epistles attributed to Paul can safely be regarded as his".

More recently, D.L.Mealand, "The Extent of the Pauline Corpus: A Multivariate Approach" [JSNT 59 (1995) 61-92] did a much more thorough statistical analysis using more variables. His research lends support to the view that Colossians, Ephesians and the Pastorals are deutero-Pauline.

Source criticism

A.Q.Morton has tried to use mathematics as an aid to source criticism. His basic premise is that the NT documents were penned in columns, and that there would have been some advantage to the writer in ending a section at the end of a column. If he were using multiple sources, some column boundaries would have coincided with source boundaries, thus helping us to discover the latter. [A.Q.Morton & J.McLeman, "The Genesis of John" (Saint Andrew Press, Edinburgh, 1980) p.25]

But Morton is surely too ambitious here. Firstly it is doubtful whether an author would start a new section when he encountered a new source. Certainly a good author would be expected to digest the source and create sections with boundaries which do not reflect the source boundaries. Secondly any correlation between column and text boundaries would be expected to reflect primarily the author's own sections and only secondarily (if at all) the source boundaries. As the primary effect is quite difficult to detect, any secondary effect would be virtually impossible to detect by this means.


Dislocations in the fourth gospel have been suspected for over a hundred years. [c.f. F.Spitta, "Zur Geschichte und Literatur des Urchristentums" (1893)] C.K.Barrett pointed out the most common passages which are candidates for dislocation (with lengths in Greek letters added here in parentheses): 3:22-30 (740), 6 (5663), 7:15-24 (765), 10:19-29 (776), and 15-16 (4781). [Peake's Commentary on the Bible, (Ed. Black & Rowley, Nelson, 1963) p.845] These numbers are all quite close to whole multiples of 800 and are undoubtedly significant.

The logical conclusion of this idea can be seen elsewhere on this site in the section entitled "the formation of John's gospel".

go to "the formation of John's gospel"


So far as I know, I may be the first to use (simple) mathematics to help to locate interpolations. The technique has been applied to John's gospel.

go to "Picking out the later interpolations"

It has also been applied elsewhere in the New Testament (full results yet to be published).

Book structure

I have also developed a mathematical formula (yet to be published) which sheds light on book structures.