I teach at a business school, so many of my students are practically minded. Their research projects have problem statements like, “How can Xompany, Inc. meet ESG requirements while maintaining longterm growth?” When I talk to them, I sometimes challenge them to tell me what their empirical question is. They say things like, “Why is it hard to increase both shareholder value and ESG ratings?” Is that really an empirical question? I ask. It sounds more like a theoretical one. What is it you don’t know about Xompany that you want to learn with this project? What is it you can’t know until you’ve collected some data and analyzed it? That is your empirical question.

The how-question will be answered in their discussion and the why-question will be framed by their theory. The theory will provide the warrant for their practical recommendations; it will tell us why what they are proposing will work. But the real content of the answer will come from what-questions, which will be answered in the analysis.

This simple heuristic may be another way into the thorny issues of the philosophy of science, which I’m working on very deliberately these days. Empirical research tries to settle questions that are “framed” by concepts (theories) and “driven” by norms (values). But our theories and values can’t interact directly. They must meet on the solid ground of objective fact. Why-questions lead us to formulate explanations. How-questions lead us to formulate recommendations. But what-questions demand that we describe reality as it is, even if that reality puzzles us or thwarts our aims.