GE Notes:

Howe & Windram treat words as their units, classifying them as either essentially the same ("equivalent", Figure 2 caption) in two texts or essentially different. They regularized spelling as a step towards making this call. We want the flexibility to do that as well as to treat letters rather than words as our units.

Howe & Windram and Spencer et al. are working with manuscripts in which there are many stages of copying and recopying (during which changes may occur) between the authorial original and the documents being tested. With our Shakespeare plays, there are almost certainly fewer layers of copying: the printed editions are probably much closer to the authorial original than is the case with medieval manucripts. The 'trees' of descent are shallow.

Spencer et al. mention recombination ("the production of entities containing information from more than one source") as something that current phylogenetic research is not well equipped to handle. We think that is some cases this is likely to be present in the early editions of Shakespeare. See the stemmata in the Textual Companion to the 1986-76 Oxford Complete Works for a sense of the worse cases(eg Richard 3 p. 230, Richard 2 p. 307, and Henry 5 p. 375)