“When I’m working with scientists in a collaboration, I would say there’s always an element of the human narrative that is influencing me. That’s how I’m connecting with their ideas, following their scientific pathways — they’re being told to me through human connection, and human conversation.”
Zakiya Leeming
Manchester-based composer Zakiya Leeming is currently Artist and Producer in Residence at the Royal Northern College of Music’s Centre for Practice & Research in Science & Music (PRiSM). Zakiya frequently engages in interdisciplinary collaborations with health data scientists and doctors, with recent projects including collaborations with health data scientists such as ISARIC4C (‘Dawn, or the Morning After the Storm’), Connected Health Cities (#MusicDataSavesLives), and Oxford University Professor Paul Klenerman exploring immune memory and music, funded by the Wellcome Trust. Zakiya has been commissioned by ensembles such as Riot Ensemble, Ensemble Recherche, Future Music Festival, Explore Ensemble, and Psappha; she was a Royal Philharmonic Society Composer 2021-22 and a London Philharmonic Orchestra Young Composer 2022-23. Zakiya studied her doctorate at the Royal Northern College of Music, and previously studied at the University of Tasmania.
We chatted with Zakiya over Zoom, discussing collaborating with scientists, generative AI in composition, centring human stories, and “thinking through music”…
–
Zyggy/PRXLUDES: Hey Zakiya! Thanks so much for joining me today. You’ve started making a name for yourself through your collaborative practice combining composition and science — tell me a bit about how your interest in science began?
Zakiya Leeming: It was always there, but something I learned I had to suppress so as not to drive everyone mad, including me. One moment I think sums it up well. In a primary school library session, we were given the task to pose a question, then use the resources to find the answer. My question, that nobody had yet sufficiently answered, was: “if the universe is everything, and everything has to be somewhere — where is the universe?” After not finding the answer in the modest primary collection, I was given a calculated look by the teacher, and told to ask a different question. That moment stuck with me. Not just the fact I didn’t get an answer, but how it felt to look around the class and see everyone looking up what suddenly felt like bits of tiny, useless facts that moments ago I would have been thrilled to know, when nobody apparently knew or cared to explain where we even were.
I mostly managed to suppress my existential dread over the next decade or so; studying both the sciences and arts equally, with an unarticulated feeling that somewhere between the two there must be some answers. There was a moment of serious consideration of whether to pursue physics or music at university, and music won on the basis that if it failed (as I thought it would) I could continue exploring the sciences like nothing happened. Fortunately, something happened, and that was the realisation that I could, in fact, keep both in my life.
It wasn’t intentional at first. I would just be reading a new bit of astrophysics, and what would happen was: in trying to understand what I was reading about, I would put it into musical terms, or musical structures. That led to me writing pieces — because I was really just fundamentally inspired by these ideas that had come from the science. That’s something I was doing naturally in my practice. When I came to do my PhD, that’s when I thought “well, I really want to investigate why it is that I do this — why do I think about science in terms of music?” — and then, if I looked at that more closely, what could it become? That’s what started me down this pathway.
Of course — were there any particular collaborations that served as a catalyst for your practice?
I did a project with PRiSM [Centre for Practice and Research in Science and Music]. It was the first ever PRiSM project; it’s this new centre for research at the Royal Northern College of Music. One of the projects that they put on early on was this project called 8³ [8 Cubed], and the idea of it was to pair composers with scientists. The idea was to start a collaboration, and anything could occur from that; the point was to collaborate, to share ideas.
I was fortunate to work with a wonderful scientist — Dr Kirstie Andrews — who is a biochemical engineer. This was a huge turning point in my practice, because: whilst I was very happily being inspired on my own, finding these interesting things in physics, and astronomy… What happened when I started working with a real scientist — someone who’s actually doing science out there in the world — it grounded me in reality, and connected me with that lightning-bolt feeling that I wasn’t in control. First of all, I couldn’t choose the topic! I therefore wasn’t being led by what my brain was naturally finding patterns and affordances with; suddenly, I was having to immerse myself in areas that I was totally unfamiliar with. Biochemical engineering was a new side of science to me! It really made me think in different ways; it made me go back to the drawing board in terms of how I’m interacting with this knowledge, how I’m understanding it. Developing a whole new framework to pin those things on.
Tell me more about your collaboration with Dr Andrews. How did the initial concepts for your work take place, considering the differences in your disciplines?
We were having conversations about Kirstie’s work, where it came from, her history. All those sorts of things. It all started to become part of the package — part of what I was being informed by. So the very first piece that I did was a narrative piece; because in those initial stages, while my brain was still creating these new pathways, I was able to understand more immediately the human story she was telling me. It became more narrative in that way — of her journey through her research — using the musical analogies to things such as directional axon growth as a framework. Through which I was telling her story: of finding a rigid theoretical base that instinctively seemed wrong because it ignored the more complex patterns of nature, and ultimately being proven right through experiment.
Tell me a bit about how that collaborative process looks and feels, to you — through humanising those experiences?
For me, it’s always about the collaboration itself — and not about an outcome I have in mind. When I begin a collaboration, I don’t put any goals on that initially, other than: to engage with the person, to find out about their work, to try and get an understanding of what it is that they do, to the best that I can. Ultimately, I chose music as my main mode of study and expression. I am a composer; I’m not a biochemical engineer, or an immunologist. But I’m always interested in exploring that alternate universe, and I can experience that through connection with those who did chose the other door.
But also, thinking about the process… One of the things I have noticed in these collaborations is that either though we’re coming from very different backgrounds, genres, topics — in some cases, they’re topics I’m completely unfamiliar with — I’m always surprised by [the fact that] when we get talking about the work that we do, we always find similarities in our processes. It’s that “ah, we do work in similar ways!” — even if the thing[s] we are working on [are] seemingly very different. It’s interesting to find where you then become inspired by somebody else’s practice; the insight that gives you. They might share something about their method of working, and that makes you think about something that you do in a different way.
For me, the collaboration really is about a human connection, first and foremost. What is the thing that is so interesting that you can’t put it down? [Because] for me, that’s also what composition is.
Exactly — I think between any kind of practitioner in the arts or sciences, there are way more similarities than there are differences. What is it about scientific process, or scientific method, that resonates with you?
I think that it comes first from the fascination of “what is this place, what is this life?” — you know? For me, it does start with those questions. That’s why science is interesting to me. The methods that you use to determine those also become interesting, because (and this is the beauty of science) it recognises that you can’t really say anything for certain. The scientific method is the hypothesis, then the test… You get some results back that tell you something about what you’re looking at, but it’s never done, it’s never finished. You never have the complete and total answers. You never get to 42. -laughs-
Equally, I think we are all familiar with bad science — how you can come to conclusions based on poor methods. So you always have to interrogate that as part of your practice; not only the thing that you’re looking at, but also the way that you’re doing it. I think that’s why, for me, there are some parallels with artistic methods. It’s not because I think that there are empirical answers in arts — it’s not that you’re trying to find an answer, per se — but it’s the looking that is interesting.
How do ideas of scientific methods, and processes, tend to manifest in your work?
The way that I tend to work is through metaphorical analogy; sort of an extrapolation. I’m never claiming to be “doing science” in music, or anything remotely close to it. I think it’s possible to organize ideas in a strictly rule-based system, and that can produce interesting musical compositions, but since that is still composition, and not science, I don’t tend to lean toward that as an outlet. I’m also not trying to “translate” science either. It really comes down to this: that concept, that framework, prompted me to think of this musical idea. It’s a set of connections; some that I am aware of and can identify the pathways between the musical and non-musical ideas, and others that I leave uninterrogated.
Of course — I understand you recently wrote a piece for the London Philharmonic Orchestra based on the movements of a trapeze artist…
When they announced the topic would be music inspired by art, I immediately knew what my art form would be — circus arts. It was an art I had found to be beneficial for my condition, but equally also due to it, had been unable to practice. But in considering it as a topic of the work, I did naturally revert to the basics with questions like: what does this art form “come down to”, if you want to scrutinise it at its most basic level? Of course, this is not the end point, because it’s an art form; but it essentially starts with the mechanics of movement. So I was looking at ideas of: what is balance? If you’re balancing on a trapeze in the air, where does that come from; what is happening in the body? It’s muscles firing in opposite directions — a miniscule, muted battle of thousands of opposing cannons firing shots across a massive distance. That is where the piece starts; this microtonal fluctuation in the trombones, tiny firings in opposite directions, keeping you in the middle.
That makes sense — when it comes to this kind of process, would you say you have a scientific approach to composition?
Whilst I was really interrogating what it meant to write music informed by science, I experimented with the question: what I were to take a more systematic approach? What if I were to make it more rule-driven, and try to map every single parameter of music onto the parameters of the subject I’m looking at? So I did that, and I feel happy at having done it, and can move on. It came from a question I had to ask myself, which is: if I am being inspired, or informed, by science, but I am not using anything close to an empirical method, how can I say it has anything whatsoever to do with science? And I think ultimately, it’s because I’m not trying to “do science”, or translate it. It’s my response, as a composer, to these ideas. There’s a Xenakis quote that goes along the lines of “thinking through music”, and that’s what I relate to the most; I’m thinking about the world, but I’m thinking about it in music.
Yeah — the subjectivity of composition, that’s something you can’t really get from anywhere else besides yourself. Even in things like process music or spectralism, you’ll still have elements of subjectivity weave their way in…
Yes, and I highly respect composers who have worked in a more rule-driven way, some of the most ground-breaking and frankly enjoyable music has been written with highly systematised methods. But the question I was asking was really an interrogation of my own self and methods. What I learned is this metaphorical approach is just the way I am driven to compose; what in most cases I would have to actively reject or suppress because that’s what’s coming from the fountain. There has been huge growth in the times I have forced myself to work against my natural tendencies, that’s also an important part of the craft, but also, I think it’s okay to embrace the natural way that I work.
How important then is it for you that the audience understands the conceptual material of your work?
Not that you really can, but I’m always trying to have several different ears listening to the composition as I’m writing it. Some of them deal with the concept, and how it’s being told through the language of music. Another is whether the music is developing (or not) in an interesting compositional way. But the one I’m usually most concerned about is: taking away all else, what does it sound like? It can get like a bit like a hermeneutical spiral that never ends, because you’re always revisiting each point — and then of course it’s a temporal art form, but one that can be revisited, so you can’t ever get to the end… -laughs- It can get a bit mad!
Maybe the natural order of things is to not have the answer, or to not know the answer.
Sometimes I feel that you have to allow — or even force yourself — to stop asking these questions at some point, just for a little bit. So at least you can get over the next hurdle.
That’s actually a great segue into another facet of your work — particularly as you’ve done some compositional work with PRiSM researching AI and machine learning. How has your work in machine learning and AI informed your compositions?
I’ve done a number of works using AI. Primarily, I have used one tool, which is PRiSM’s SampleRNN; so a lot of my understanding is really specific to that type of AI. Currently the least understood about thing about AI, I think, is that it isn’t one thing. It really depends on what tool you’re using, what data you’re working with, those things in combination produce entirely different possibilities.
For PRiSM SampleRNN… it’s sample-based, meaning it works one sample at a time. It’s just making predictions, essentially. You give it a dataset, and it will say “in the dataset, if I had this kind of sample here, then the next sample is likely to be this.” That’s all it is — it’s not that deep, in a way. -laughs- The other important thing to say about this particular tool is that there’s no deep learning associated with it. I think the tools that people are more familiar with have been trained on enormous corpuses — things that have been scraping the internet. But this particular tool only knows what you give it. It doesn’t understand music, it doesn’t even know what a sample is really, it’s just running calculations.
Tell me about how you used SampleRNN — what kind of material did you give it, and how does that differ from the deep learning AI software most people are (becoming) familiar with?
I’ve always given it material that originated through my compositional practice. So things I’ve been working on together with performers; it doesn’t include anything I didn’t originally prompt, or had a composer role in. So immediately, there are a lot of ethical questions that are not relevant.
The purpose of AI is to generalise. What it’s built to do is take a very large amount of data, and then give you something that could be mistaken for some of the source data. That’s what its purpose is. That can tell you certain things, it can shine a light on generalisations you didn’t realise were happening. But it was also quite early days when we started working on it. Things are moving fast, and this tool is technically very underpowered based on what you can use these days, but I find the clunkiness of it quite useful. I don’t really need a tool that can do what I originally did: create one more version of the source data. I am looking for new questions to ask: like what after-image-like imprint do I get when I give it substantively different datasets? It gives interesting answers to that. And equally interestingly, I won’t know what sounds will result from the conditions I’ve given it: that’s an unusual position to be in as a composer today.
How do you view your approach to using AI, as well as AI in composition more generally? Do you see it more as a tool, or set of parameters, rather than as a “monolithic” thing?
For me it’s definitely a tool in the toolshed; not a co-creator, or disrupter. Maybe understanding how it works, especially through practice, is what dispels the illusion of something less tangible. Of course, tracing threads might become more difficult as deep learning and more complex neural networks develop; but at the same time, as soon as you remember that it is built on what is already in the world, the less profound, and more derivative and identifiable in origin those predictions appear. It’s important to remember those origins are still us.
In terms of application, I’m sure there will come a day when you can hit a button, and it can produce something that you might have written on your own — without having to actually do that. It was Jennifer Walshe who first said it — there will be composers who do that, and that will be okay. And I agree with that. Because we’ve already had this argument in art! We’ve already discussed framing. And collage, distributed creativity, sampling, quotation, pastiche, aleatoricism, “computer music”… I could go on. I have yet to hear an argument about AI that situates it in a truly new space in terms of artistic merit and authorship.
That said, I don’t particularly see myself using “hit a button to generate x” as the one and only stage of composition. The point is that it can be used like any other tool to generate, challenge or refine ideas as part of a larger process that is ultimately puppeteered by human hands, hearts and ears.
Of course — I do think that’s something that a lot of conversations around AI tend to neglect. -laughs-
I think it’s important to remember that everything AI is going to spit out comes from humans. However — there is an interesting event horizon that a lot of people are looking out for now, and that is: if you have these scrapers that are running constantly on the internet, you will hit a critical mass where the majority of what it is pulling down is no longer solely human-generated, but now also mixed with results of prior AI processes. Everything still points back to humans in this chain; but what happens to the dataset when it’s been through the mixer that many times? We don’t know yet.
To bring things back to your collaborations with scientists: one striking piece you’ve recently worked on was with Prof. Paul Klenerman as part of a collaboration with ISARIC4C1, looking at the experiences of health data scientists during the pandemic. When you’ve done more hands-on collaborations with scientists, in what ways do either the human experiences or research manifest for you?
When I’m working with scientists in a collaboration, I would say there’s always an element of the human narrative that is influencing me. That’s how I’m connecting with their ideas, following their scientific pathways — they’re being told to me through human connection, and human conversation.
For some works, I’ve allowed a heightened focus on that human element such as in ‘Dawn, on the Morning After the Storm’. The writing process involved talking to each member of the team about their experience during the pandemic, but starting first with their musical histories, and why they (unlike myself) chose to pursue science in their career. One common factor between the researcher/performers was a history in Scottish folk music, so started thinking about movements as “airs” and “jigs”. There are some uses of the strings in one section that was a response to the experiences being relayed to me about the stress, the difficulty, the complex emotions in the immediate sense. It was my response to that. So that piece came more specifically from those personal, and shared human experiences.
But there can also be a connection between the human story and the musical scientific analogies. That’s something I’m exploring now with Prof. Paul Klenerman, who I met through ISARIC4C. We are telling a story about immune memory, using human memory as a vehicle through which to explore it.
That’s a really interesting way of looking at the treatment of material — have there been any pieces that have drawn that connection in a more literal sense?
To bring it back to my collaboration with biochemical engineer Dr Kirstie Andrews — [her] work involved repairing neural pathways when there’s been damage or disease in the human nervous system. Her role in this research was to look at how to create a synthetic scaffold that could be used in the body, to allow the axons to grow toward each other again so that those connections can be re-established.
The literature proposed rigidly straight, road-like channels as the best way to achieve this directional growth. She said she tried that, and immediately discovered that they did not follow the structures… -laughs- If you think about it, very few things in nature travel in a straight direction. So her theory was “what if we made it more loopy, a spaghetti-like structure?” — and found that yes, in practice, that worked. Musically, I had these ideas of dominance and control that I was representing in various ways; and then also, this idea of freedom. I was using composition methods as well as the music itself to make this analogy. But also, I was writing the music to tell this narrative; to have these channel-like musical structures — blocks of sound with gaps in-between — and then I’m using pitch to convey this idea of looping. Even though I’m telling a “human” story, I’m still using ideas from the science itself in [the] musical language — to inform my musical choices.
I understand you’re currently working on another piece with Paul Klenerman — can you tell me a bit about how both scientific research and human experiences are coming together in this upcoming work?
I’m working with Paul on an opera for singer Rosie Middleton, which will be premiered on 24th June 2025 at the RNCM. We’re looking at the incredible story of Lady Mary Wortley Montagu, an English noblewoman who brought the practice of variolation from Turkey to England and demonstrated it in 1721. If you know anything about immunology, you know about Edward Jenner; but less is known about those who provided the steps along that path that led to the next innovation.
There’s a direct analogy we are making between resistance put up by the medical profession at the time against Montagu’s work to introduce and multiply this foreign idea, and the battle between viruses and the immune system. I find the immunological concepts are very easily communicable with musical ideas. I mean, the idea of a cell — that has an obvious metaphorical use, you could say! You have the idea of division, of multiplication. Things spreading. These connections, whilst perhaps different for each composer or listener, can be easily made between the two topics. So the dramatic narrative is framing the scientific idea in a metaphorical way, and the music itself then goes a little further in exploring Paul’s research — the behavior of cells and the role of memory in each unique, human body.
–
Learn more about Zakiya and her practice at:
- https://www.zakiyaleeming.com/
- https://www.instagram.com/zakiya_leeming/
- https://soundcloud.com/user-416324597/
Links:
- Who was Lady Mary Wortley Montagu?, National Trust
- International Severe Acute Respiratory Infection Consortium ↩︎
