My wife says I shouldn’t be, but I am always shocked at how credulous people will act toward anyone who simply writes a book and labels himself an expert. Case in point is a Slashdot book review by Jon Katz of Don Foster’s Author Unknown.
Foster claims that, “no two individuals write exactly the same way, using the same words in the same combinations, or with the same patterns of spelling and punctuation,” and so by comparing an writing sample whose authorship is unknown or questioned with enough known writing samples, that he can infer who the author of the passage is.
I am extremely skeptical of Foster’s claims. My skepticism is only increased by the fact that numerous online reviews of Foster’s book note that he spends little time discussing exactly how he goes about comparing written passages to each other.
Before I’d buy into Foster’s claims, I’d like to see a series of double blind tests which involve Foster matching unknown samples with their authors. This could be easily accomplished by using say long forgotten novels from the 19th century and correspondence by their authors in a genre that is relatively similar in form and style.
This sort of technique has been abused in the past. In 1991 two National Institutes of Health researchers, Ned Feder and Walter Stewart, who were supposed to be looking for fraud in the biomedical field decided to turn their attention to historical biography. They created a computer program which compared Stephen B. Oates’ Lincoln: The Man Behind the Myth with a 1950s-era Lincoln biography by Benjamin Thomas.
The charges created an initially sensational reaction, but when the details filtered out they proved groundless. Here, for example, is one of the strongest pieces of evidence of Oates’ plagiarism.
I strongly suspect that this sort of detective work is right in line with other popular but suspect methods such as the so-called lie detector test.