The inks and paper used by the Scottish poet Sir Robert Burns have been analysed allowing the original manuscripts to be accurately distinguished from forgeries. The findings, which could have applications in the authentication of historical manuscripts, are presented in Scientific Reports.

Robert Burns’ writings have been subject to numerous forgeries over the years and although an original work might sell for between ￡6,000 to ￡90,000 manuscripts of dubious authenticity continue to appear at auctions.

Using mass spectrometry, Karl Burgess and colleagues compared a sample of the author’s authentic manuscripts with a number written by the skilled, contemporary forger Alexander Howland Smith. Unlike traditional mass spectrometry, which must be performed in a laboratory, the sample preparation method used in this study allows for samples to be taken where the manuscripts are stored. The authors extracted samples from the manuscripts’ surface, in a minimally invasive process, enabling them to analyse the ink used by Burns, and distinguish it from manuscripts written by Alexander Howland Smith. The authors then used the spectrometry data to produce a machine learning method that accurately distinguishes manuscripts by Burns from Smith’s forgeries.

The authors suggest that the sampling, analysis and machine learning methods described in this study could have a significant impact on determining literary authenticity.