For any given hypothesis test, that is, for any specification of the null and alternative hypotheses, explain why you could make only a type I error or a type II error, but not both.
When would you make a type I error? When would you make a type II error? Answer as generally as possible.