I’m listening to Lutz Prechelt’s keynote at the German Software Engineering research community conference. He is talking about how we should not be undertaking research that has no relevance, and he is demonstrating this by presenting research based on ludicruous assumptions (that will never be real, not in this nor another world).
Nobody could disagree with this, no?
Well, everyone wants high relevance of their work, but as research is inherently risky, you may end up with research that ran into a dead end. Then, we won’t learn much about reality and how to do better engineering. As a professor, I can shrug it off, as a Ph.D. student who just saw three years of work go down the drain maybe not.
The solution is to distinguish between scientific progress and the social validation of research capabilities that Ph.D. students need to demonstrate. In theory, they are two separate things, in practice they all mean one thing: More journal articles.
Nobody wants useless journal articles, but publications are required for a scientific career. I don’t think the Ph.D. title itself or recommendation letters or anything else can compensate for that. Hence, we need to identify places where we can submit work but it really is only to receive peer validation of research competencies, not for the scientific insight.
These are the Elsevier journals. They are write-only journals were colleagues review work for free but which, because of their predatory pricing, are not read by the German scientific community. An involuntary solution, I guess, but one that works.