Undecidable questions

The cybernetician Heinz von Foerster proposed a distinction that is surprisingly practical for leadership teams. Some questions can be decided because, in principle, they are already decided: a calculation, a rule or expertise settles them. Answering them means working something out. Only questions that cannot be settled this way call for a real decision. Someone has to choose and own the choice.

Strategic relevance

The questions a transformation depends on are almost always undecidable in this sense: which business gets its own logic of success, which initiatives end, who decides what from now on. No analysis takes this choice away. Organisations still avoid it by treating an undecidable question as a decidable one: another analysis, a business case, an expert opinion. The typical signal is a decision backlog in which every meeting ends with the remark that more figures are needed.

Common misconceptions

The most common misconception is that every question becomes decidable with enough information. Under uncertainty this does not hold, because the missing information only emerges through action. A second misconception is that working something out is worth less than deciding. On the contrary: whatever can be worked out should be handed to someone with a mandate and a deadline, so that the leadership team keeps its time for what is undecidable.

Decision architecture perspective

For work with the Compass this yields a first test for every candidate question: work it out or decide it? Whatever can be worked out leaves the room. What remains is decided as a hypothesis with a review date, supported by heuristics rather than false precision, and where possible as a decision about future decisions (decision premises).