Computer Linguist Daniil Skorinkin Answers the Question: What is Stylometry?

Stylometry is … counting words. That sounds silly, doesn’t it? And not even the interesting words! We count words that seem meaningless to the ordinary listener: conjunctions, articles, or prepositions such as “and,” “the,” “a,” “of,” “or” ... Sounds even sillier, I know. But what if I told you that by counting these words, we are able to uncover authors who hide their names behind a pseudonym? The writer J.K. Rowling is such a famous example, but not the only one.

It has long been suspected that an author’s individual style – their so-called authorial fingerprint – is exposed in measurable text statistics. The Polish philosopher W. Lutosławski was one of the first to realize the significance of seemingly unimportant words such as interjections and prepositions. He was also the one who suggested the word “stylometry” in 1890. The breakthrough came with the spread of computers, which are able to count all the words in even the thickest book in milliseconds instead of weeks or months. Nevertheless, even this method has its limitations. I will never be able to identify the author of a tweet or even a “long” Facebook post, for example. But if you publish a novel under a pseudonym and there are other extensive texts of more than 10,000 words that you have written under your own name, stylometry will unmask you.

It works because we do not choose every word, phrase, or grammar structure deliberately. We acquire our language skills over many years through communication and reading and develop many habits. Some people say “typical of” where others would use “typical for.” Some love subordinate clauses, while others tend to use participial phrases... Such habits influence the frequency of function words in their texts and make them somewhat similar regardless of the specific plot or topic. By the way, so far stylometry has worked in every language it has been tested in, from Ancient Greek to Chinese. I tried it with Armenian, and it worked just as well.

The most famous “victim” of stylometry is J.K. Rowling. The author of Harry Potter wrote murder mystery stories under the pseudonym Robert Galbraith. Some journalists contacted the American stylometry expert P. Juola, who proved that the book The Cuckoo’s Calling is very similar to Rowling’s works. Another case is Elena Ferrante, an extremely popular writer who has sold millions of books but does not actually exist. The real person behind Elena Ferrante is the Italian writer Domenico Starnone, who may have written the novels with some help from his wife, the translator Anita Raja.

This text was published in the university magazine Portal - Zwei 2025 „Demokratie“. (in German)

Here You can find all articles in English at a glance: https://www.uni-potsdam.de/en/explore-the-up/up-to-date/university-magazine/portal-two-2025-democracy

Back