Until the end of the previous chapter, we had been taking observation and the idea of observational evidence pretty much for granted. Some scientific questions, we have assumed, can be answered just by looking. Even if the logical implications about the truth of theory are unclear in the process of testing, the outcomes of the tests, whether the predictions hold true or not, have been taken as unproblematic. The test results, the observations, are of a purely manifest world of appearance. They are there for anyone who takes the time to look, and they can be used by anyone as impartial evidence. When in doubt about theory, there is always observation.

Observation is the touchstone of external virtues. Empirical adequacy, that is, compatibility with observation, is presumably easy and uncomplicated to evaluate. And observation is the distinction of external virtues, since it is the contact of theory to the outside world. Observation is the source of the world's input and guidance over our describing and modeling. We have separated external from internal virtues under a presumption that the world's input through observation puts the external features beyond our control in the sense that it's not possible to force the observations to come out one way or another. They are forced on us and we must take them as given. Observation is, if anything is, a source of objectivity in science.