Updated: 6 days ago
This post is the fourth in a series of interviews with experts in sonic branding and music psychology. And we’re delighted that Peter Harrison, another alumnus of Music, Mind and Brain at Goldsmiths, University of London, shares his thoughts on the psychology of music and how it will have an increasing role in our lives. Peter is the new Director of the Centre for Music and Science at the University of Cambridge, which he joined following a postdoctoral research position at the Max Planck Institute for Empirical Aesthetics in Frankfurt, Germany.
What got you interested in the psychology of music?
When I arrived as an undergraduate student at the University of Cambridge, I saw myself mostly as a scientist/mathematician who happened to pursue music performance as a hobby. Cambridge has a very rich musical scene, though – both in terms of performance and in terms of academic studies – and this rich scene eventually persuaded me to switch to studying Music for my undergraduate degree. I was particularly excited by the work going on at the Faculty’s Centre for Music and Science, which showed me that my musical and scientific interests could be combined in a way I hadn’t considered before. After taking a couple of inspirational classes in the topic I decided to pursue music psychology long-term, and so moved to London for the Music, Mind & Brain MSc programme. The topic has kept me captivated ever since.
How did you get to know SoundOut?
Daniel Mullensiefen recommended I speak to SoundOut because of their extensive work with music testing as well as their blue-sky thinking and commitment to research. They invited me for a three-month internship during my PhD studies in London. It was my first experience working in a real office and I really enjoyed getting to know the friendly team.
David and Grace wanted to make sure that the reviews provided by their panel members were of high quality and asked me to help create a way to identify and reject any that did not meet the required standards. This allowed me to explore some new machine-learning ideas that I'd been working on during my PhD, resulting in software that still runs every day behind the scenes to process incoming reviews. Alongside this I also did some interesting work on the predictive analytics side, studying how we can predict the sales success of consumer items on the basis of questionnaires administered to SoundOut’s reviewer panel.
What excites you about being back at Cambridge?
There is something particularly meaningful about returning to the place that first inspired me into music psychology. I’m deeply invested in the idea of helping today’s students to discover this fascinating field and giving them the practical skills to be proficient scientists in their own rights. My position also gives me a very special opportunity to develop truly long-term research programmes of my own, with the goal of tackling the big remaining unsolved questions in music psychology. SoundOut is proving already to be a valuable partner in both regards, hosting students for data-science internships, and collaborating to develop exciting new technologies for audio branding.
What do you think AI will offer brands in terms of their audio-branding choices?
We're in an exciting new era of deep learning where computers are really beginning to understand vision, sound, and language in ways that were hard to anticipate before. For example, one might design a machine-learning model to predict the emotions evoked by a piece of music. Once trained, the model doesn’t require any humans in the loop, and can be applied in moments to very large databases of music. Automated analysis techniques such as these should ultimately be very valuable for helping companies to find their ideal audio brand matches within the massive pool of available music.
Why do you think brands should use rigorous data science to inform their investment in brand assets?
The world is constantly changing, and we need to stay informed and up to date with how our customers' perceptions are changing. Data science techniques allow us to 'plug in' to this massive population of consumers and keep our finger on the pulse in a way that would never be practical if we had to run through everything manually.