Observation and Interpretation

There is no such thing as pure, presuppositionless observation. Following Popper, CF holds that perception is always guided by prior ideas: we select what to look at according to our goals, and we interpret what we see using background theories. Popper dramatized this by opening a lecture with “Observe,” then waiting silently until someone asked what to observe — the point being that observation needs a question or expectation to direct it. There is far too much data to absorb everything, so the senses are used to search for what some idea has already marked as relevant.

This reverses the empiricist picture. We do not first gather neutral facts and then generalize from them; we make guesses first and then use targeted observation to test those guesses. Evidence never “speaks for itself” — raw data has no single inherent meaning, so we only learn from evidence via arguments and interpretations. This connects to the searchlight theory of mind: the learner actively shines a light rather than passively soaking up impressions.

The consequence is that observations are fallible in two ways: the senses can err, and the interpreting ideas can be wrong. So observation cannot serve as the certain foundation that induction and naive empiricism need; this is part of why CF rejects justificationist appeals to “what the data shows.” Instead, observation does its real work as criticism — noticing when results clash with expectations. CF extends this by building discussion structures that force the prior assumptions guiding observation out into the open, where they can themselves be scrutinized rather than smuggled in as neutral fact.