about

sk

I am Senior Lecturer in Phonetics at Lancaster University, where I co-direct the Phonetics Laboratory and lead the Speech Physics Group. I am also affiliated with the Data Science and AI Institute.

I was awarded my PhD in 2014 from the University of Sheffield. I currently serve as associate editor at Journal of Phonetics and was awarded the 2026 ISCA Mid-Career Excellence Award in recognition of advancing understanding of speech dynamics through integration of articulatory imaging, computational modelling, and dynamical systems theory.

My research aims to discover the physical and cognitive principles behind spoken language. Speaking depends on precisely coordinated movements of the tongue, lips, jaw and larynx that shape moving air into audible language. I measure these movements using vocal tract imaging (MRI, ultrasound, electromagnetic articulography) and build mathematical models of how they are planned and controlled. This work draws heavily on dynamical systems theory and sits at the intersection of linguistics, speech motor control, and the physics of living systems. My research is funded by UKRI (2019-25) and The Royal Society (2025-27) and currently focuses on the following areas:

  • Dynamical principles of speech movements. What principles describe how the vocal tract moves when we speak? And how much of speech control is handled by the body’s own mechanics? I pursue these questions by analysing vocal tract dynamics in tandem with data-driven equation discovery and computational modelling. A major current strand, with Patrycja Strycharczuk, looks at variation in articulatory strategies and asks how much of it comes from anatomy rather than learned behaviour.

  • Time-varying dynamics of speech. As in many living systems, the dynamics of speech are neither fixed nor isolated: they drift and interact with rhythms in the brain, body, and environment. With physicist Aneta Stefanovska, I am exploring how methods developed for time-varying systems, such as the cardiovascular system, can reveal how the control of speech involves coupling between subsystems, where that coupling itself changes over time.

  • Interpretable acoustic-articulatory inversion. Can we predict how the tongue and lips are moving from audio recordings of speech? With Anton Ragni and Vishal Kumar, I am developing physics-informed machine learning models that link speech acoustics to vocal tract movements and their underlying control dynamics. We aim to do this while keeping those models interpretable, so we can also advance progress on how vocal tract movement generates sound.

Upcoming talks & events