"... The merit of Holmes' contribution is that he has drawn attention to the fact that many divergencies between LXX and the Masoretic text are interrelated and therefore point to a systematic revision, rather than to a chain of textual corruptions. Nevertheless, whereas Dillmann placed exclusive emphasis on the priority of the Masoretic, Holmes seems to have taken the other extreme position in order to defend the priority of the postulated Hebrew Vorlage of LXX Joshua ... Holmes did not make any explicit statements about the textual history of Joshua. His thesis presupposes a two-stage model: an original Hebrew text, still represented by the Greek text, was later revised by a Hebrew scribe. G. A. Cooke in his small 1918 Cambridge commentary on Joshua formulated this model as follows: 'Apparently the Hebrew text of Joshua was once current in two forms: the one which lay before the Greek translators, and perhaps was generally accepted in Egypt; the other which is represented by the Masoretic Text, and perhaps was best known in Palestine. The latter form of the text remained open, possibly as late as 200 BC to additions which never found a place in the text used by the LXX<' ..."

Meer, Michaël N. van der Formation and Reformulation: The Redaction of the Book of Joshua in the Light of the Oldest Textual Witnesses (pp. 40-41) Brill, 2004