The title of doctor (dr.) stands for the ability to do independent research in a responsible way. Just as a driver's license indicates that the person holding the license can participate in traffic in a responsible way, and just as someone with a swimming diploma can enter the water without drowning.

The process that leads to a doctoral degree can be described as follows: a PhD student carries out research under supervision of a promotor. After some time, the promotor decides that the student's findings, written down in a thesis, should be judged by a committee of scientists. This process culminates in the so-called PhD defense, where the PhD student defends the thesis, after which the doctoral degree is obtained by the PhD student who has now become dr. and can be regarded as a scientist.

The predicate 'cum laude' can be attached to this degree, and is supposed to signify excellence. An explicit criterion is that the thesis is among the five percent best theses. And this is exactly where things go wrong: there is no best five percent. Using such a criterion assumes the theses can be ordered. That, however, is not the case. A thesis is an aptitude test in science - theses need not be ordered and they cannot be ordered.

Indeed, a thesis can be deep and specialized, a thesis can be broad and interdisciplinary, a thesis can be well-written or not, it can be immediately relevant or not - in short, a thesis has many properties. And all these properties cannot be projected on a single dimension without losing the claim that this projection still maintains a reflection of the quality of a thesis.

And there are other arguments that support getting rid of cum laude. The final decision whether cum laude is handed out, is taken during the day of the PhD defense. In the Dutch system, the PhD defense is a ceremony, which is not aligned with taking any merit-based decision; there is now the risk of disappointment on the candidate's side.

And who decides on cum laude? These are the scientists that make up the committee based on advice of some external experts - in other words, human beings. And, hence all properties that are present when humans make decisions apply here as well.

Let's assume you read this, and you are not convinced. Obviously, that's fine. But allow me to make one proposal to decide whether cum laude has any relevant meaning. So let us assume that the predicate cum laude is a proxy for scientific quality. Thus, those PhD students that received the predicate cum laude should make their way faster thru the scientific ranks than their sisters who did not receive the cum laude. 

Now let's select at random a hundred promotions at TU/e that received a cum laude, and let us also select at random hundred promotions at TU/e that did not receive a cum laude. In both sets, we simply count how many rose to the rank of full professor in, say ten years. If the difference between these two numbers is large (statistically significant), we should conclude (with grinding teeth on my part) that cum laude says something about being able to rise through the scientific ranks. However, if the difference is small, then even you, my dear reader, should agree with the title of this piece. I leave it to my statistical friends to (i) express devastating criticism on this test, and (ii) improve the test, and carry it out.

Frits Spieksma | professor Combinatorial Optimization