Bowdlerised "Proof of Evidence"


Introduction Text Results Index

Analyse it yourself.

This proof of evidence only came to light after the person who was supposed to have written it had died. It was produced some weeks after the first day of evidence only a few minutes before it was to be used as evidence on the second day of evidence - the person having died just over a week before. There was no opportunity to check it in any way.  
The solicitors were unable to supply any version with a signature and eventually admitted that there had not been a signed version. Further, this proof of evidence contains statements that the person who was supposed to have written it would have known at the time of writing to have been false - these statements going on to contradict evidence already given both verbally and in writing during the first day by the company.  
The person who was supposed to have written it did not have legal training and certain words crop up here that analysis of all other texts written by him and records of words said by him show occurred nowhere else.  
The text remains complete in all respects except for the names of those concerned. All of the names have been changed - first names to names selected randomly from a babies' names book and last names selected randomly from an English place-names dictionary.  
The question is how many people wrote it? This is where sophisticated analytical techniques come into play.

These are up to you.

Corel WordPerfect is freely available and has the Grammatik grammer checker which gives statistics - many do not. In addition, simpler methods may be used such as measuring the proprtion of given words.

Automatic Grammar checking and word ratios are computer based and are mechanical and rely upon there being enough of a sample to make them reliable - whole texts can be shown to have the same or different ratios as other texts and lead to clues about their authorship. Another thing to consider is that people's texts do vary to some extent - the question being; Does this text show signs of varying that a known consistent text does not? In addition, these mechanical methods do not take into account the choice of words to express anything.

Choice of words can distinguish between authors and, where a text is on the short side may have the edge over the mechanical methods. Styles do change over a long period of time but it must be remembered that in the case of this document, we know that it was written only in a few weeks - not enough time for anything to change significantly.

Try anything you like and e-mail me with your answers stating what you do for a living (student, writer, university professor, forensic scientist, car mechanic and so on) and where in the world you are from (town, district and country will do). I will publish the results - not any names. I look forward to hearing from you.

To see analysis already done, look at the Results Page.

  Note - there are two people called Ledston.

In addition, the fifteenth paragraph is not numbered, the sixteenth paragraph is labelled 15 and so on. To be consistent, call the paragraph after paragraph 14, "14a".

If you want to download the text file - poe.txt (16,188 Bytes) - click here (you may have to right-click and choose to save the page as . . .), otherwise just save the page or highlight it and copy it to wherever you want for analysis.

Site Map
Back to the Index
Copyright 1998 - 2003 P.A.Grosse.
All Rights Reserved
Recipe Index Computer Security