Response to Sanford, Aked, Moxley and Mullin
Reverend A. Q. Morton · International Journal of Speech Language and the Law · 1995
In a recent issue of Forensic Linguistics, Volume 1 Number 2, a paper by Sanford, Aked, Moxley and Mullin purports to examine my cusum technique of authorship attribution and concludes that it is ill-founded and unreliable, unsuited to any forensic application.The paper makes no such examination and its conclusions are invalid.The Aunt Sally knocked down is one of their erection.The facts are as follows.Lecturing to Manchester Statistical Society in March 1965, I used cumulative sum charts to illustrate the periodic nature of sentence length distributions.I ended the lecture by suggesting that these charts could be used to show habits of authors and the coincidence of two or more charts would powerfully reinforce an argument.It was later that the first computer program to construct such charts appeared and the ICI research department who made it ran off some examples at the request of W C. Wake and myself.These covered Plato's 7th epistle and led to the paper 'The