On valuing research

Last fall, as preparation for the launch of the new Faculty of Technology, we, its future personnel, used the collaborative Viima tool to gather ideas and suggestions for the new faculty. One particularly lively discussion thread was about the kind of research that should be encouraged in the new faculty: basic, applied, or both.  Some saw engineering research as inherently applied with an ultimate goal of commercially successful innovations, while others wished for openness towards basic research, where practical applications of research results are less obvious.  Yet others questioned the point of the whole discussion without clear definitions of basic and applied research.

Reading this commentary led me to ponder the research emphases in our new faculty, and more generally to reflect on how we judge the value of research.  Then, an email from a researcher colleague refined the abstract questions to a more specific one: ”Who gets to decide what research is worthy?”

This email was an alert about a new ”strategic plan” of my colleague’s home university, University of Leicester, a reputable UK research university. The plan talks about ”focusing efforts and building on core strengths”, and ”taking advantage of emerging opportunities in research and education”, but in practice means layoffs of academic personnel.

Sometimes, financial realities force universities to make cuts, but this particular case has caught the attention of the global research community: there are several petitions opposing the plan. A part of the plan is to merge Informatics and Mathematics into a combined school that focuses exclusively on AI, data science, computational modeling, and digitalization. This would mean ending research in the foundations of computer science, which according to the plan is ”highly theoretical and not directly linked with applications”.

Consider digitalization: society’s functioning today depends crucially (as in there are no manual backup systems) on myriad computerized systems.  We know that these systems are vulnerable, many of them are difficult or at least cumbersome to use, and they have defects: we pay a constant price on all of this. How do we then make systems that are less vulnerable, more usable, and (more) correct? There is no silver bullet, but we have, over the past few decades, made steady progress in programming tools, languages, and methods; this has translated into increased programming productivity, scalable distributed systems, and an improved ability (given sufficient resources) to build complex but reliable software. In large part, this is the result of advances in algorithm design and analysis, logic in computer science, foundations of software science, programming languages, concurrent systems, and testing and verification.  These areas (apparently not sufficiently directly linked to applications) are focus areas of the University of Leicester’s internationally recognized and productive Foundations of Computing group, which now finds itself on the chopping block.

So who should decide what research should be done?

In the case of the University of Leicester, this task seems to have landed in the wrong hands, away from the researchers. Ending research that has delivered meaningful results for decades and would continue to do so and replacing it with something else to “meet the rising market demand of artificial intelligence, computational modelling, digitalization and data science” seems ill-advised.  Intentions may be good, but the decision-making is clearly based on a superficial understanding of the research landscape. Yes, AI, computational modelling, digitalization, and data science are all topical and important, but it is then all the more misguided to throw out a well-functioning research operation that works on the foundations of those very topics.

Again, who should decide what research should be done?

I am content with the simple answer: the researcher with their academic freedom of inquiry. Claiming this freedom creates a fair contract. As researchers, our obligation is, in the long term, to justify the relevance of our inquiry. We must convince our peers to accept the papers we write and persuade funding agencies and our employer to provide us resources for the work we do – and eventually we must convince society that our results have value. To be sure, society’s values change over time, so we must adapt our research, too.

From time to time, governments and politicians may not make the best decisions on how to allocate funds to research. Yet, I think we may consider ourselves fortunate at the University of Turku and in Finland in that researchers’ freedom of inquiry is exemplarily supported. Even though research may be chronically underfunded, researchers have significant say in what the emphases are when our (scarce) resources are distributed.

At the Faculty of Technology, most research is and will be applied, and the goal of commercially useful innovations is at the forefront. But there will also be research where direct applications are farther off in the future, or whose applications benefit society in other ways than through commercial activity.

Whatever the emphases are, building up research operations that produce lasting results takes time; hasty swings to follow fast moving trends do not help. We as a community of researchers are well-positioned to recognize when research has value and can have impact, and to make prudent decisions on where to place our effort. And sometimes, after careful consideration, these decisions can even be bold, like establishing completely new research programs – and a new faculty.

Jaakko Järvi
The author is the Dean of Faculty of Technology in the University of Turku