Monday, 08 Sep, 2008 Science

Police Calls for Language Experts to Help Find Criminals


Police looks forward to catch more criminals who send anonymous text messages. Now police seeks help from ecologists and language experts, who are using techniques of studying ocean life in order to identify authors of text messages by their grammar, syntax and the way they use words in a message.

The method had already been used in several cases of murder but it still needs to pass some tests so the court would accept it as routine proof. In 2005 police applied to linguistic analysis in order to find Yorkshire teenager Jenny Nicholl.

Language experts found that text messages sent from the teenager's phone have probably been written by David Hodgson, a murder suspect. By analyzing the characters of abbreviation scientists found that Ms Nicholl usually wrote "I am" and "Myself", however, after the teen's disappearance, these words turned into "im" and "meself", which was quite common for Hodgson's text writing. Thus evidence was able to secure Hodgson's conviction for murder.

Tim Grant, deputy director of the Centre for Forensic Linguistics at Aston University, applied to the help of ecologists, experts able to differentiate between marine mediums by identifying how close or far apart the environments are according to species. By using the same techniques experts can identify how close or distant various species of writing are.

Currently scientists are working on developing a database of text messages provided by volunteers to be able to identify variations in style and language among individuals, which can then be used to investigate anonymous messages from suspects.

Till now scientists managed to gather about 8,000 text messages from about 1,000 members of the public.

"We can measure similarity and difference by means of the stylistic distance between any two text messages. Pairs of text messages by the same author are going to be more alike. And where you get groups of people who text one another a lot, or pairs of people, their language becomes more similar. We're moving away from human-based expertise and opinion to a method which will make it easier to discuss things like error rates," said Dr Grant.


Powered by

Add your comment:

antispam code