BK: How did you get involved in dynamical systems?

FD: After my Master's research in 1969 on questions of differential topology I was looking for a suitable topic for PhD research. In 1970 I went to a conference in Amsterdam on manifolds to meet and talk to Nico

Kuipers, whose work I knew through his publications. He advised me to switch my interest to the new and emerging field of dynamical systems and to work with his former student Floris Takens. So I started PhD work on singularities of planar vector fields as Floris's first PhD student.

After my promotion in 1973 I spent a whole year at IMPA. In total I must have spent at least three full years of my life there. The atmosphere at IMPA was very open and vibrant with many visitors. Jacob Palis was a main driving force and I took part in his seminar and other seminars. IMPA is also where I met my long-term close collaborator Robert Roussarie for the first time in 1976. He had made use in his work of results in my thesis and we started working together straight away on a question suggested by Jacob Palis. Today Robert and I have more than twenty joint papers and are continuing to work together.

With Robert Roussarie and Jorge Sotomayor I worked on the codimension-three unfoldings that arise at degenerate cases of the Bogdanov-Takens bifurcation. In fact, already in 1975 I wrote a letter to Vladimir Arnold asking him to send me the publications of Professor Bogdanov and inquiring whether the generalization to higher codimension had been done. Arnold informed me that, first, Bogdanov is not a professor and, second, the codimension-three problem had been solved. Therefore, I stopped working on this problem. However, it turned out only much later that the codimension-three problem was actually still unresolved, and I picked it up again with Robert and Soto.

After they appeared as Springer Lecture Notes in Mathematics, our results have been applied quite quickly by other colleagues in fields as diverse as laser physics and ecology. I am confident that my more recent work on singular perturbations will also lead to insights and results that are applicable to models arising in different scientific domains.

To me both aspects seem essential to mathematics and they should continue to influence each other closely. There really is a need for good research ranging from direct applications, as done by engineers, for example, all the way to work that deals with the foundations of mathematical thinking itself. Policy makers should better be convinced of this point of view as well. However, present funding, certainly in Europe, is primarily directed towards subjects

that have a certain market value, such as cancer or telecommunications research. This is not a bad thing as such, but it creates a need to justify fundamental research, which is no longer seen as useful in its own right. A lot of mathematics is of course quite fundamental. While mathematics is perceived as inexpensive --- and the more so the more theoretical it is --- funders nevertheless expect some sort of return after a few years. This easily creates the message that pure mathematics is not essential. Yet I believe that most countries ought to invest more in fundamental research.

What about funding opportunities for the next generation of mathematicians?

In the last years a lot of new positions for PhD research have been created, but generally the number of permanent faculty to supervise them has stayed the same or has even decreased. Not only does this increase the work load of permanent staff, but it also means that there are insufficient opportunities for young researchers after their PhD. Certainly in Belgium there is a shortage of postdoctoral positions with a perspective of a permanent academic career.