Science & Technology, New Zealand (Commonwealth Union) – Researchers of a study point out that this year, Ed Sheeran successfully defended himself in a legal case, convincing a jury that he did not plagiarize Marvin Gaye’s “Let’s Get It On.” In contrast, Pharrell Williams and Robin Thicke were unable to prove that their song “Blurred Lines” was not a copy of Gaye’s “Got to Give It Up.”
The question arises: Could automated algorithms introduce a new level of objectivity to decisions concerning music copyright infringement? Such a development might potentially reduce the number, scope, and cost of court cases.
Musicologist Dr. Patrick Savage, in collaboration with Yuchen Yuan from Keio University in Japan, as well as experts in music psychology and copyright law from Goldsmiths, University of London, and George Washington University in the United States, has conducted research on this subject.
“It’s the largest study so far of how the best algorithms compare with humans in judging when music crosses the line into plagiarism,” added Savage, who is a senior research fellow at the Keio University, School of Psychology. “It’s fair to say that algorithms won’t be taking over any time soon.”
Savage’s involvement in this field included providing expert evidence in the form of an amicus curiae brief, which played a role in overturning a decision in a Katy Perry case.
In their study, 51 individuals were tasked with evaluating 40 instances of alleged music plagiarism spanning from 1915 to 2018. This collection included a 2014 New Zealand National Party campaign advertisement reminiscent of Eminem and the 1970s hit “My Sweet Lord” by ex-Beatle George Harrison.
Two prominent publicly available tools for identifying music plagiarism, Algorithms PMI and Musly, were used to assess the same set of songs.
The assessments made by the study participants aligned with court decisions in 83 percent of the cases (33 out of 40), while the algorithms achieved a 75 percent match (30 out of 40).
Researchers pointed out that it’s worth noting that the study is based on the underlying assumption that the court decisions were correct. As Savage pointed out, the “Blurred Lines” case, for instance, generated substantial controversy, with neither the study participants nor the algorithms strongly supporting the legal verdict. This view was shared by many musicians, musicologists, lawyers, and judges.
A fundamental constraint when employing algorithms in copyright cases is that non-musical factors can influence the outcome.
For instance, even if two songs exhibit a high degree of similarity, copyright infringement may not occur if the composer accused of plagiarism can demonstrate that it was impossible for them to have been exposed to the earlier song, as indicated by Savage.
In the end, while trial by algorithm may not supplant trial by jury, the objective assessments provided by algorithms can be a valuable factor to consider in such cases.
“For example, Spotify is already experimenting with a Plagiarism Risk Detector that might help artists automatically catch unintended similarities with existing works before they release new songs,” explained Savage. “Future court cases might also be able to include graphs of how similar two songs are in relation to past cases to give judges and juries more objective data and context to aid their decisions.”
The escalating prevalence of litigation has led to concerns about “unjustified music copyright lawsuits not only inhibit music creativity but also waste millions of taxpayer dollars annually to cover the adjudication of these disputes,” as stated by Savage and his co-authors in their paper published in the journal “Transactions of the International Society for Music Information Retrieval.”
The algorithm “Percent Melodic Identity” was originally developed by Savage and Professor Quentin Atkinson from the School of Psychology to investigate the evolutionary aspects of folk song melodies.
On the other hand, “Musly,” created by Dr. Dominik Schnitzer at the Austrian Research Institute for Artificial Intelligence, incorporates elements such as rhythm, timbre, and melody into its analysis of music compositions.
