Reflections on YorVoice: a radical, interdisciplinary project bringing together researchers working on the human voice
Your voice is unique. It contains information about your identity, mood, health - even your humanity. It also affects how you navigate the world; people judge us based on how we sound, while the growth in speech AI means that the voice is now a key way we engage with technology. A vast range of academic work seeks to understand the human voice across many disciplines: linguistics, computer science, health, singing and performance, and education. While many of us work on similar questions, the nature of academic research, grouped by departments and faculties, means we rarely talk to each other. Two years ago at the University of York, we got the chance to change this.
YorVoice is a University of York-funded project supported by the SPARKS initiative. YorVoice was based on one simple observation: understanding the human voice is an interdisciplinary problem that needs an interdisciplinary approach. With a budget of £300k over two years, we were given free reign to fund projects that asked questions about the human voice. We knew from the outset that we didn’t want people to just continue working on their usual programme of research with their usual collaborators, but we also didn’t want to dictate what their research questions should be. So we held a multi-day ‘sandpit’ event where we brought together people from across the university, made them talk to each other, and come up with a novel research proposal. We had just a few rules: teams had to be interdisciplinary (from at least two different departments), the project had to be about some aspect of the physical human voice (as opposed to a metaphorical ‘voice’), and we encouraged teams to include early career researchers in leadership positions.
From this we got six sub-projects involving 15 academics across six departments, over half of whom are early career researchers; most had never worked together previously. The sub-projects investigated a wide range of themes, using an array of methods and approaches:
- Breathappy: developing an app to administer breathing exercises to support speech and language therapy, in partnership with York Hospital;
- Vocal Authenticity: producing a documentary (My Voice Journey) following marginalised members of the community in Middlesbrough as they engage with singing projects led by North East Opera;
- Setting the Legal Tone: developing a framework to legally protect characteristics of the voice against threats such as deepfakes;
- Atypical voices: exploring listener perceptions of people who stutter, and those speaking in a second language via an interpreter, in high-stakes situations such as police interviews;
- Love your Voice: developing an intervention for people who have been told they can’t sing, to build confidence and self-esteem;
- Deepfakes as forensic evidence: working on a framework for analysing whether voices are real or deepfakes, using techniques from linguistics, phonetics and audio authentication.
The researchers on the sub-projects enjoyed the process. Dr Peter Harrison, from the York Law School and member of the Setting the Legal Tone sub-project, said:
“We found the entire YorVoice process entirely refreshing. The original networking and project development "sand-pit" threw a diverse band of folk together to explore their common interests and potential inter-disciplinary projects in a way which allowed all voices to have a say.”
YorVoice has also built a wider community of around 40 members at York who engage with regular events supporting interdisciplinary working. Most of these events have been focused on practical aspects of interdisciplinarity, such as intellectual property and public engagement. Another major aim was to build infrastructure to support interdisciplinary voice research, both at York and beyond. To do this, we created the YorVoice Data Catalogue which signposts to a range of curated resources, and is continually updated by the voice research community.
As YorVoice enters its final year, we’re now looking outwards - building wider networks and developing new non-academic partnerships to support impactful work. We hosted the UK and Ireland Speech (UKIS) Conference in June 2025, which brought together over 150 researchers from academia and industry working on speech technology. The day after UKIS we held the inaugural YorVoice conference where we shared our reflections and the work of our sub-projects with a new audience. It’s clear that there’s an appetite to build a wider national and international community of interdisciplinary voice researchers, and it's something others are thinking about as well - for example, in 2024 and 2025 colleagues at Cambridge hosted the Interdisciplinary Voice Workshop, which intersects with the aims of YorVoice.
As we reflect on the aims of YorVoice, we can be proud that we achieved most of what we set out to do. A crucial part of our success has been the ability to incentivise engagement. Academics are busy, and with the best will in the world it’s hard to commit time for all of the things that may be interesting, but not urgent or essential. Having access to funding allowed us to run the sandpit event and fund significant pieces of work, which was the only way that we could realistically have expected buy-in from the community. By not being prescriptive about the topics we wanted to support, colleagues felt they had the flexibility and freedom to pursue only the ideas and collaborations that were most interesting to them, which led to a much richer range of sub-projects than we ever could have imagined. In the longer term, communities need to be nurtured. Having regular events and newsletters can help with that sense of community and we’re now trying to build a strategy for sustainability.
It hasn’t always been easy. One of our biggest challenges has been overcoming the administrative hurdles of working across university departments. Universities talk a good game about interdisciplinarity, but institutional structures often don’t support this, especially when money is involved - Which department manages the budget? Who is in charge of ethics? What happens in terms of staff buy-out? Who gets to claim things for REF? These are things that universities can do better, and will be essential to building successful, truly interdisciplinary communities, to support the next generation of academics and researchers.
Image caption: Audience participation at the YorVoice conference. Credit: Paul Shields