Dorothee Haroske

Dorothee Haroske: The Geometry of Smoothness

There is a chair at the University of Jena devoted not to analysis in general but to function spaces in particular — a single, deep, technical subject. It was founded by Hans Triebel, whose name is attached to one of the two great classification scales of such spaces. Dorothee Haroske took her doctorate under Triebel in 1995 and now holds that chair, leading a research group with more than fifty years of unbroken work on one question: how do you sort the infinitely many possible functions by how smooth they are, and what follows once you have?

Jena, Almost Without Interruption

Haroske was born in 1968 in Germany. She completed her studies at Friedrich Schiller University Jena in 1992, her doctorate there in 1995, and her habilitation there in 2002. From 2003 to 2006 she led a junior research group at Jena on fractal analysis. Her first professorship came at Jena in 2010, a second at the University of Rostock in 2017, and by 2018 she was back at Jena in a third — the chair for function spaces. Apart from that one year on the Baltic coast, her entire career has been spent in a single university department, working on a single circle of problems. In an era of relentless academic mobility this is either a limitation or a form of concentration, and her output suggests the latter.

What a Function Space Is For

The subject sounds forbiddingly abstract and is in fact deeply practical. Suppose you want to solve a differential equation — the heat equation, say, or an elliptic boundary value problem describing a membrane. You need to know what kind of object the solution is: how many times can it be differentiated? Does it stay bounded? Is it continuous?

Function spaces answer this by grouping functions according to their smoothness and their size. Sobolev spaces were the first great success. The Besov and Triebel–Lizorkin scales, developed later, are finer instruments: families of spaces indexed by parameters that let you specify smoothness in fractional amounts and measure it in different ways. Haroske's group works on the general theory of these spaces — decomposing them via atoms, quarks and wavelets, handling spaces with generalised smoothness, anisotropic spaces where smoothness differs by direction, weighted spaces, and spaces defined on domains rather than on all of Euclidean space.

Entropy Numbers

Her doctoral thesis, supervised by Triebel, concerned entropy numbers and approximation numbers in weighted function spaces, and their connection to eigenvalue distributions. That chain of ideas is the heart of her work and worth unpacking.

An embedding is a statement that every function in one space automatically lies in another — that a certain degree of smoothness forces a certain degree of good behaviour. The interesting embeddings are compact, which means, roughly, that the first space is not merely contained in the second but squeezed into it. Entropy numbers measure exactly how tight the squeeze is: how many small balls you need to cover the image of the unit ball. Approximation numbers measure a related thing — how well the embedding can be imitated by an operator of finite rank.

The payoff is spectral. If you can bound the entropy numbers of an embedding sharply, you can bound the eigenvalues of differential operators built on those spaces — and eigenvalues are the physically meaningful quantities: the resonant frequencies of a drum, the energy levels of a system, the decay rates of a diffusion. A purely geometric measurement of how one space sits inside another translates into a statement about how a physical object vibrates. Haroske has spent thirty years making that translation sharper.

Fractals

The fractal-analysis group she led from 2003 pursued the same machinery on rougher terrain. Classical analysis lives on smooth domains — intervals, discs, well-behaved regions with tidy boundaries. Fractals have none of that: infinitely crinkled boundaries, non-integer dimension, no tangent planes. Building function spaces on fractals, and doing spectral theory for operators defined on them, requires rebuilding the theory from foundations that no longer assume smoothness anywhere. Her group continues to work on spaces on fractals and the spectral theory of operators on them, alongside local smoothness theory, Besov regularity, Fourier analysis, approximation theory, and non-linear boundary value problems.

Books

She wrote *Envelopes and Sharp Embeddings of Function Spaces* in 2007. The subject is the borderline: the cases where a standard embedding theorem just barely holds, or just barely fails, and where crude estimates say nothing useful. Envelope functions are the tool for describing precisely how badly a function can behave right at that limiting case. In 2008 she published *Distributions, Sobolev Spaces, Elliptic Equations* with Triebel — a graduate text taking the reader from the general theory to the differential equations it was built to serve. She also edited the Hans Triebel Anniversary Volume in 2003.

The Copernicus Award

In 2026 Haroske and the Polish mathematician Leszek Skrzypczak received the Copernicus Award, given for exceptional achievement in Polish–German scientific cooperation, for their joint work in mathematical analysis. It is a prize for a collaboration rather than for an individual, which suits a career built substantially on long-running partnerships.

Why Dorothee Is Called a Genius

She is not, in any documented sense. No source examined here uses the word, and the Copernicus Award citation honours cooperation and joint achievement rather than singular brilliance. There is no Haroske theorem in general circulation, and her most cited work is co-authored with her doctoral supervisor.

The intellectual quality her record does show is precision at the boundary. The interesting mathematics in function space theory is almost never in the comfortable interior of a parameter range, where results are easy and estimates are generous; it is in the limiting cases, where an embedding is about to fail and every existing tool goes blunt. Choosing to work there, repeatedly, and building instruments — envelopes, sharp entropy estimates — specifically for those cases is a distinctive kind of mathematical taste. It is the analyst's equivalent of insisting on the last decimal place, and it is where the actual difficulty lives.

The counter-case is emphatic and she would likely make it herself. Her achievement is cumulative across three decades, thoroughly collaborative — with Triebel, with Skrzypczak, with a research group she inherited rather than founded — and highly specialised: function space theory is a subfield whose active practitioners could fill a lecture hall, not a stadium. She is a leading figure in a deep and narrow area, and the honest description of that is excellence, sustained, in a discipline that advances by exactly this kind of patient work rather than by individual flashes.

Legacy

Haroske's most consequential act may be custodial. The Jena function spaces group is one of the longest-running continuous research programmes in mathematical analysis anywhere, and she took it over from its founder and kept it at the front of the field, extending it into fractals, weighted spaces and generalised smoothness rather than merely maintaining it. Her books are standard references for anyone entering the subject. And the machinery she has sharpened — entropy numbers governing eigenvalues — continues to do quiet, invisible work every time someone needs to know how a differential operator behaves on a domain that refuses to be smooth.