Linguistic forensic analysis is the process by which a text undergoes statistical examination to extract information that is not explicitly contained in the text. The analysis is often employed to determine the identity of the author of an anonymous work. It has been convincingly used in cases from the _Federalist Papers_ of the 18th century to the novel _Primary Colors_ of about a decade ago.
Since the death of Rutherford, nearly all articles and books being emitted from the orifices of the WTBTS do not carry the names of the individual authors. While in some cases the true authors' names have become public, it would be interesting to find out the identities of the authors of the remaining texts.
All that is needed to do this is a source of the e-texts and the appropriate analysis software. The software is fairly easy to write and the techniques for doing so are well known.
Think of the cool things we could do with this. We could generate a ranking of sorts to see which authors had a penchant for convoluted syntax and winding sentences (kind of like how I write), which authors have the most limited volcabulary, which authors were most dependent on borrowing phrases from other authors, which authors referenced which scriptures, which authors were more strident and which were more gentle. By connecting the author to the e-text source, we could find out which ones were more favored for wrting study articles, for writing pamplets, and for writing entire books.
Topical word subsets could be used for further entertainment. I'll wager that once we know most of the identities, it will turn out that there are likely just one or two that write about fornication, adultery, porneia, and the like. Another two or three will show up has heavy users of FD&S, loyalty, organization, apostacy, purity, and, of course, disfellowshipping. Then there will be the group that uses Babylon the Great, Christendom, worldly, and outsiders.