Power and Knowledge: Who Decides What Is True?

Foucault: Power/Knowledge

Michel Foucault (1926–1984) argued that power and knowledge are inseparable — so tightly linked that he often wrote them as a single hyphenated term, pouvoir-savoir ('power-knowledge'). In works such as Discipline and Punish (1975), he claimed that those who hold power shape what counts as knowledge: which questions are researched, which methods are considered legitimate, which voices are heard, and which conclusions are accepted. Medical knowledge was shaped by which bodies were studied (historically, white male bodies). Economic knowledge was shaped by which questions economists considered important (GDP growth rather than inequality).

Crucially, Foucault did not mean that all knowledge is simply power in disguise — that would be nihilistic. He meant that knowledge and power produce each other: institutions that gather knowledge (prisons, clinics, censuses) also exercise control, and structures of control generate the knowledge they need. Recognizing this is not relativism; it is a call to broaden the voices and perspectives that contribute to knowledge production.

The Tension Between Expertise and Democracy

Modern democracies face a fundamental epistemic tension. Complex policy issues (climate science, monetary policy, pandemic response) require specialized knowledge that most citizens do not have. But democratic legitimacy requires that citizens, not experts, make the final decisions.

Philosopher Philip Kitcher, in Science, Truth, and Democracy (2001), proposes 'well-ordered science' where public values guide what gets researched, experts conduct the research with methodological integrity, and results are translated for democratic deliberation. This respects both expertise (scientists know more about climate than politicians) and democratic legitimacy (citizens decide what trade-offs are acceptable).

The populist attack on expertise ('elites do not know what is best for ordinary people') and the technocratic dismissal of public opinion ('trust the science') are both epistemologically wrong. The correct relationship is division of labor: experts contribute knowledge, citizens contribute values, and democratic institutions integrate both.

Worked Example: Three Ways to Handle a Carbon Tax

Suppose a country is deciding whether to impose a carbon tax. Climate scientists can estimate emissions reductions; economists can model price effects; but the decision also involves value trade-offs (how much present cost to bear for future benefit). Watch how three models of authority handle it.

The worked point: the technocrat and the populist make the same mistake from opposite directions. Both collapse the fact/value distinction. The technocrat smuggles value judgments in under the banner of 'science'; the populist rejects verified facts because they dislike the values attached to them. Kitcher's model is harder to run — it demands institutions citizens can trust — but it is the only one that keeps facts and values in their proper lanes. This is also where 'manufactured doubt,' documented by historians Naomi Oreskes and Erik Conway in Merchants of Doubt (2010), does its damage: by faking a scientific debate, it corrupts the factual input before the value debate can even begin.