I've finished writing an article about using multivariate analysis to map the space occupied by witnesses of the New Testament, particularly those of the Letter to the Hebrews.
The first half is concerned with statistical considerations. If not interested in that, skip to the final parts which consider the picture of what I call "textual space" developed from multidimensional scaling and divisive clustering results.
I think that a case can be made for reading the results geographically, as per Streeter's theory of local texts.
The article is available here: