“It’s very important, for me, to put science into perspective. Science is one way to view reality. And art is another way to view reality. But they’re both equal.”

At once a mathematician and a musician, Vermeulen finds himself at the nexus of where art complicates--or should be complicating--science. In particular, he observes a mainstream habit, in our society, of relying on statistical models to the point of dangerous over-simplicity; and this is where, perhaps, artistic voices could broaden the narrative.

“It worries me a lot,” he says, referring to the misuse of data and particularly of mathematical models. He is spontaneously conversant on the underlying causes of the 2008 financial crisis, elaborating on the 1973 paper which introduced the core model for a secondary market that, its authors promised, wouldn’t crash. “But it’s a super simple version of reality,” explains the artist, who sonified this model and its simulations for an electronic music project a few years ago.

Our over-reliance on such models continues today, he says. “You see in our society… that everything is becoming more and more rationalised… more and more datafied.” As examples he cites PredPol, a software designed to predict where crimes are likely to occur, as well as China’s social credit system, which tracks the “trustworthiness” of entities and individuals. He also mentions the coronavirus pandemic, during which everybody had their own data and interpretations. “We were more and more made to believe that, if the numbers are correct, then it should be correct.”

By “it” he seems to mean truth, reality itself--the situation as an unassailable certainty.

But, he cautions, data is inordinately easy to manipulate. And this is what he wants to expose: “If you can put a scientific model or a prediction into perspective,” he says, “then I hope that the absolute value and sometimes perverse power of such models can be broken.”

Performance in Belval

Vermeulen has a PhD in pure mathematics and works both as a maths researcher at the University of Antwerp and as a data scientist. But he is simultaneously an established sound artist, music researcher and electronic musician whose 2021 album Mikromedas AdS/CFT 001 garnered worldwide attention, with press in 28 countries having picked it up.

In his upcoming lecture/performance at the Rockhal on 1 May--part of the Sound of Data project, an Esch2022 initiative--Vermeulen hopes to problematise wrongly used statistical models and their dominance by, in a way, rehumanising them and the people who make them. “The goal with the lecture/performance,” he says, “is to convince people about the intuitive, creative nature of science.”

For Vermeulen, creativity is at the heart of science. During his PhD, he spent six months solving a single large problem. “But the way I [solved it]… was not completely rational. And that’s a misconception I think a lot of people have,” i.e. that scientists must be rational all the time. He draws a link to music: “You find the same thing if you’re making a piece of music. Why this melody? Why these chords?” True, he adds, some rules should be followed, but curiosity and intuition are at the root of it.

The lecture/performance will include excerpts from Vermeulen’s Mikromedas project, in which he uses models from theoretical physics and data from deep space to make music. One piece is a sonic representation of a voyage from Earth to an exoplanet in the centre of the Milky Way, which is just the type of deeply speculative scenario that is well-suited to the medium, according to the artist. “Sound is much more easily connected to the imagination,” he comments.

“If you’re talking about black hole astrophysics, very easily you can end up in a ten-dimensional version of our reality,” he adds. The suggestion is that, in such a profoundly weird context, sound is as good a method as any for parsing the facts. And in some ways a better one.

Astronomy seems to be a particularly fruitful area for creative thinking. People commonly romanticise the work of astronomers, picturing somebody who gazes at the night sky and contemplates the cosmos, but in reality this profession largely consists of analysing abstract datasets and developing mathematical models because more conventional methods of observation are simply impossible--you cannot visit a black hole and see what it’s all about.

Thus, imagination comes deeply into play. And indeed, as a musician exploring worlds typically under the remit of scientists, Vermeulen gains access in surprising or at least unusual ways. His musical output “stays close” to the science, he says, but it isn’t meant to be, itself, science. “I can freewheel,” he explains. “I see myself sometimes as a pirate, just navigating the space of black hole astrophysics, taking a wire and connecting this to that, and maybe that to this, and maybe something new comes out or not.”

His output hasn’t led directly to any scientific discoveries, he says, though it has augmented his own understanding of the field. And it certainly enters into a wider societal dialogue about what science is, what information is, how it can be processed, what types of authority it has.

Sonifying data

The event on 1 May (entrance is free and no tickets are required), where Vermeulen will speak, will officially kick off the Sound of Data project. Vermeulen will then run a series of workshops for local musicians, , to help them learn how to sonify data. Their compositions, along with an original piece by Vermeulen, will be performed at the culmination of the project: a concert at the Rockhal on 3 December 2022.

“Sonifying” data can take several forms. It can be as simple as matching a numeric value or other data point to a tone on the Western 12-note scale, by which you could (for example) map the global rise in temperature over the last centuries with a cascade of notes. Vermeulen calls this the “classical” approach, adding that he is less interested in it because of its limitations: with only a dozen notes, the data cannot be very closely reproduced in sonic form.

He is more taken with newer methods such as digital signal processing, which uses technology to mirror datasets far more accurately. With the global warming example, you might use a pitch oscillator to hear the temperature changes in minute detail. Another method is to use algorithmic music composition techniques, where data feeds algorithms that can generate rhythms, textures or timbres.

During the Sound of Data project, local musicians will apply these techniques to various datasets, including 3D body scans, historical data (numbers but also words), crowdsourced geometrical drawings, traffic data and covid-19 statistics.

“From the participants in the workshops, a lot of new things could come out,” Vermeulen comments. “I think it will be very interesting to see how close or how far the artists will go from their everyday practice using the data. I mean, either direction is good--to use it a lot or not.”

“But for me, it can also be inspiring to see that,” he adds.

The Sound of Data is an Esch2022 project run by the Rockhal/Rocklab, the National Research Fund (FNR), the University of Luxembourg, and the Luxembourg Institute of Science and Technology (LIST).