Research is often described as the pursuit of knowledge. In practice, it is better described as the navigation of ignorance. One of the most useful frameworks for thinking about this comes from a now-famous classification of knowledge and uncertainty. While simple on the surface, it turns out to be surprisingly powerful when applied to scientific research.
Known Knowns: The Solid Ground
Known knowns are things we know that we know. These include established theories, standard formalisms, and results that have been repeatedly tested and reproduced. In physics, this might be a well-defined Lagrangian, a trusted computational technique, or a symmetry limit that has been validated across many systems. Known knowns form the foundation of research. They are rarely exciting, but without them nothing else stands. Good researchers respect them, use them carefully, and clearly state when they rely on them.
Known Unknowns: The Engine of Research
Known unknowns are things we know we don’t know. These are the clearly articulated open questions:
- unknown parameters,
- uncalculated effects,
- regimes where approximations may break down.
Most research papers live here. A good paper is often nothing more (and nothing less) than a precise statement of a known unknown, followed by a careful attempt to reduce it. A well-defined known unknown is a sign of research maturity.
Unknown Unknowns: Where Discovery Hides
Unknown unknowns are the most dangerous—and the most exciting—category. These are effects, assumptions, or mechanisms that we are unaware of until they surprise us. They appear as unexpected divergences, unexplained discrepancies with data, or results that contradict intuition. Many real discoveries come from unknown unknowns. So do many sleepless nights. The best defense against them is not arrogance, but robustness: checking limits, varying assumptions, and comparing with multiple approaches.
Known Unknown Knowns: What Others Know
Some things are unknown to us, but known to others. These may be buried in old literature, common knowledge in a neighboring subfield, or simply things a more experienced researcher takes for granted. Seminars, workshops, informal discussions, and referee reports are often how we discover these. This category is why research is a community activity, not a solitary one.
Unknown Knowns: Hidden Assumptions
Unknown knowns are things we know implicitly but don’t realize we know. They include:
- assumptions we never write down,
- approximations we apply automatically,
- intuitions formed through experience.
Referees are remarkably good at finding these. Turning unknown knowns into explicit statements almost always improves clarity—and often reveals weaknesses before others find them.
Ignored Knowns: Strategic Neglect
Sometimes we knowingly ignore certain effects. Higher-order corrections, small symmetry-breaking terms, or competing models may be left out deliberately. This is not a flaw—as long as the choice is explicit and justified. Good research is not exhaustive; it is controlled.
Manufactured Unknowns: The Trap
Finally, there are manufactured unknowns—questions that appear open only because the literature was not read carefully enough. Reinventing the wheel is a rite of passage, but it is also a warning sign. A thorough literature review is not just academic courtesy; it is protection against wasted effort.
The Real Goal of Research
The goal of research is not to eliminate ignorance. That is impossible. The real goal is to reclassify ignorance:
- turning unknown unknowns into known unknowns,
- turning known unknowns into known knowns.
Progress is measured not only by answers, but by better questions. A good researcher is not the one who claims to know everything, but the one who can clearly map what is known, what is unknown, and why it matters. That map is where science actually moves forward.
