Mass spectrometry technique helps identify forged Robert Burns manuscripts

July 27, 2018 by Bob Yirka, report
Mass spectrometry technique helps identify forged Robert Burns manuscripts
Robert Burns by Alexander Nasmyth (1758–1840). Credit: Public Domain

A team of researchers at the University of Glasgow has developed a mass spectrometry technique to identify forged manuscripts. In their paper published in the journal Scientific Reports, the group describes their technique and how well it worked when testing known forged Robert Burns manuscripts.

Robert Burns was an influential Scottish poet and is best known outside his home country for a poem he wrote that was turned into the song "Auld Ang Syne"—which is traditionally sung on New Year's Eve around the world. Burns lived from 1759 to 1796, and wrote poems, songs and other works on paper held in binders, which subsequently became manuscripts. He also used a variety of inks which were also typical of the time. Burns eventually became so well known that his original manuscripts became valuable—so valuable that others attempted to cash in on his success by forging manuscripts, which they attempted to sell as originals. One forger, Alexander Smith, was particularly good at it, and reached a degree of notoriety himself when he was finally exposed and sent to prison. In this new effort, the researchers have used the work of both men to test a forgery detecting technique they developed based on both mass spectrometry and a network—a technique that does not damage the material being tested.

The technique involves spraying a very small amount of a methanol and water solvent (2 µL) onto a section of the manuscript where ink is present. Next, the solvent and ink sample are pulled from the manuscript. After that, a microfluidic is used to spray just nanoliters of the sample into a . The data from the mass spectrometer is then fed into a deep learning network—one that has been trained to identify components in ink that were used during the time when Burns was writing his . The system then gives a report on what it found.

The researchers tested their system by applying it to three works known to be authentic and nine that were done by Smith. They report that the system found 16 differences between them which were described as significant—an indication that the ones by Smith were forgeries.

Explore further: Research uncovers seven lost Burns manuscripts

More information: James Newton et al. Minimally-destructive atmospheric ionisation mass spectrometry authenticates authorship of historical manuscripts, Scientific Reports (2018). DOI: 10.1038/s41598-018-28810-2

Related Stories

Research uncovers seven lost Burns manuscripts

January 14, 2013

A Scottish researcher has unearthed seven long lost manuscripts, including correspondence between Robert Burns and his close friends, which throw significant new light on the life and work of the poet.

Japanese IT firm to digitize Vatican manuscripts

March 20, 2014

A Japanese information technology company has agreed to digitize 3,000 Vatican manuscripts in a deal to make some of the Catholic Church's most historic documents available online.

Israel-British project makes Hebrew texts available online

August 3, 2015

One of the oldest surviving Hebrew manuscripts, a bible dating back more than 1,000 years, will soon be available online in a joint project with The British Library in London, the National Library of Israel said Monday.

Recommended for you

Carbon fuels go green for renewable energy

December 18, 2018

For decades, scientists have searched for effective ways to remove excess carbon dioxide emissions from the air, and recycle them into products such as renewable fuels. But the process of converting carbon dioxide into useful ...

Data storage using individual molecules

December 17, 2018

Researchers from the University of Basel have reported a new method that allows the physical state of just a few atoms or molecules within a network to be controlled. It is based on the spontaneous self-organization of molecules ...


Adjust slider to filter visible comments by rank

Display comments: newest first

1 / 5 (1) Jul 27, 2018
A deep learning network trained on a total of 12 samples (3 authentic and 9 known forgeries). Sorry, this sounds like someone desperately wanted a "deep learning" paper on their CV.
not rated yet Jul 28, 2018

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.