Similarly, scientifically collected evidence can aid us in developing the kinds of normative
theories we need to answer questions about ethics and the meaning of life. There is no simple leap
possible from “this is how things are” to “this is how things should be,” but evidence is nevertheless
highly relevant to questions of value. Such relevance is most easily seen in instrumental reasoning,
where something is assigned value because it is a way of achieving something else already identified
as valuable. For example, if we value truth, and scientific method is a good road to truth, then
scientific method can also be valued. The main problem is how we manage rationally to assign value
to our top-level goals, such as truth and explanation. I have ruled out any transcendent, a priori
arguments for such goals, so it might seem that one must be either arbitrary or circular in defending
them.