Science News

Read my words

Sun, 13th Dec 2009

Listen Now    Download as mp3 from the show Was Swine 'Flu Man-Made?

New computer software can read a book’s literary fingerprint that is unique to the author who wrote it.

(c) unknown

" alt="Herman Melville, American author. Reproduction of photograph, frontispiece to Journal Up the Straits." />It turns out that as books get longer, even the very best writers eventually start to run out of new words to use. But, the rate at which new words drop off depends on the skill of the author and is always the same for each author, giving them a unique word-frequency fingerprint.

Based on analysis of books and stories by three great writers, Herman Melville, Thomas Hardy and DH Lawrence, a team of researchers from Sweden, led by Sebastian Bernhardsson, have come up with the idea that inside every writer’s head is a ‘meta-book’. Writers reach into their own personal meta-book and pull out words to put down on paper or computer screen.

The study published in the New Journal of Physics provides insights into the way language is used especially by the greatest writers.

The findings go against a 75 year old theory, called Zipf’s law, which suggested that for any written work, the frequency of any word is inversely proportion to its rank. So, if the word “the” is most frequently used and accounts for 7% of all the words used, the next most frequent word “and” accounts for around 3.5% and so on.

But, rather than a universal literary rule, this new study shows that linguistic ability of individual authors is more important in determining their new-word frequency.

The team is planning to conduct similar research on other literary works, and eventually think it should be possible to pick out an unidentified author from the fingerprint left in their words.



Subscribe Free

Related Content

Not working please enable javascript
Powered by UKfast
Genetics Society