Homepage and blog
by Arjen P. de Vries
The CIDR 2023 conference in Amsterdam started on Jan 8th with a very special opening session: a Memorial session to commemmorate Martin Kersten, who passed away in summer 2022.
Peter Boncz (CWI colleague and Martin’s PhD candidate who did the main implementation work on MonetDB) and Arno Siebes (my boss when I joined CWI as postdoc) spoke first, followed by my contribution below, after which the session continued with EPFL’s Anastasia Ailamaki, University of Athens’ Yannis Ioannidis (also the current ACM president), ETH’s Gustavo Alonso, and, finally, CMU’s Andy Pavlo who closed the session with a video that made everyone smile.
I addressed the audience with the following speech.
Dear colleagues and friends, Martin’s sons Rene & Joost, less than one year old grand-daughter Martha, we’re gathered here at CIDR, in Amsterdam, to share our thoughts and memories. And how appropriate it is, to have this IM here and now!
Martin lived and worked here in Amsterdam, but was much more a part of the world-wide, somewhat US-centric database community, than he was a Dutch scientist. During my PhD, I’ve heard comments such as “Martin with his database system”, expressed in a derogative way, and, to be fair, that feeling was a mutual one! When I as a postdoc attempted to acquire funding from NWO, the Dutch science foundation, Martin was not exactly motivating in his remarks to “forget about it, they only care about past performance”.
We can state without hesitation that Martin was too extraordinary to be bounded by our scientific system. And, I will argue here, among friends, that Martin could not care less - he had a much grander vision for what to do with his remarkable talents. To me, Martin has always been more of an artist than a scientist; and I believe that he would strongly approve of this view. Who has seen the works by Monet in real life immediately understands the grandeur that matches Martin’s art. Art, presented in the form of data management contributions instead of paintings, but really stretching far beyond “just” scientific work.
Really, Martin mainly cared about his own creativity and expressing that in crazy original ideas. He was never impressed by papers that I thought interesting, and he did need the rigour of Peter and Stefan to sell his ideas in the scientific community. I hope that I make it sufficiently clear that I do not say these things to disregard Martin’s contributions to the database field - these are immense! But the key to these contributions are their originality, enabling us to view the same world in a new way, again and again, really as an artist creating different works in a unique and impressive style of their own.
So how would one characterize Martin’s style as a data artist? I’d say that his work resembles more Mondrian than Monet, in the sense that he opened up new perspectives using very basic principles, just like the colours and lines in De Stijl. He’d ask questions like “what would you type when I present you an empty database shell”, or “why would we never expire data”, and then throw in excellent analogies to help steer our thinking. He wrote about data management using terminology including data fungi, fragmented worlds, a data cyclotron, mammals eating dinosaurs, database cracking, your friendly butler Charles - and then I’m only listing a tiny fraction of all the fantastic beasts that Martin would come up with to express his ideas.
You can only try to imagine how incredibly inspiring it has been for us to witness Martin’s thinking process, seeing first-hand how great ideas originate, how data art comes into existence!
I do want to be honest today, and add that being close to such a brilliant mind is not always easy. In a discussion, or even just a conversation, I’d say that Martin would often be “poignant”. I was lucky to find this word in a dictionary when working on this speech; it captures perfectly what I wanted to express, meaning “painfully affecting the feelings, piercing”, but also, “pleasurably stimulating”.
The closer you would be, the harder it could be - in that sense, my own interests were distinct enough to never really feel the heat, but I too had to find how to best collaborate. What made Martin special, however, was that he himself was always looking for the best way to cooperate as well. Once you’re on his team, he’d create opportunity; I do not think I would have been full professor if it were not for his coaching, and if the science community would not be the best path for someone dear to Martin, he’d find them a different way to stay on board; he’d invent or identify a different role in the institute, or found a new company; his creativity as a manager could certainly rival the originality of his scientific work - that data art.
To me, Martin personally invented the “Erkennen & Waarderen” policy that is making the rounds in the Dutch science community today (Recognition & Rewards, for English speakers). He realised very well that for him to shine, he should not seek ten almost identical players, but rather build a team with diverse competences. He would ask you to decide for yourself what is your position, where you’d play the best. He also had a sharp eye for talent, and would not hesitate to help you see where you should play (or he would simply tell you).
I want to close with personal words of thanks.
Martin, you were a tremendous source of inspiration. You made us see our talents, and go after developing them further, making the most of those. You were very important for my career, both inside academia and outside, as aspiring entrepreneur. You have coached me for more than twenty years, and I’m so grateful for this influence on my life!
Our relationship was predominantly a professional one, but I also remember how you were very supportive when I needed a little help to restart my life, after leaving a long relationship; and as an advisor to Spinque, you have always encouraged us, myself and co-founders Wouter Alink and Roberto Cornacchia, in choosing our own path in entrepreneurship; advising us, paired with the right degree of being critical and pushy - to ensure we would have success too, in spite of our non-standard views on how to run a high tech software business.
Martin, you taught me to be independent and I thank you for that more than anything else.
Would there ever be a Museum of Data Art, “The works and life of Martin Kersten” would be the founding exhibition, that’s a sure thing.
May you rest in peace, while living on as inspiration in us all.
Thank you for everything.