Page History

...

Each antecedent phase of process involves various more or less well defined alternatives for the next phase. And the particular emerging from the next phase is the actualization of one of these alternatives. But "actualization" is not a simple change from "merely possible" to "actual," whatever that could mean, but always connotes "some additional definiteness, or determinateness," not already contained in any of the antecedently obtaining alternatives. This is not to say (with Bergson and others) that there are no antecedent possibilities or that there is something absurd about the concept of such. What we mean by "the antecedent possibility of particular p" is simply that the antecedent phase of process defined itself as destined to be superseded somehow, within certain alternatives, by a next phase of process. This "somehow" is not a wholly undifferentiated question mark, but involves certain more or less well defined alternatives, none of which can coincide in character with the particular that emerges from the next phase of process, but some one of which, or some one region of the continuum of possible quality, will later be recognizable as the nearest alternative or region, the one that with the least further definition is equivalent to the particular once it has emerged and is given as such.

A distinction is necessary, then, between pure, or unrestricted, potentials, on the one hand, and impure is equivalent to the particular once it has emerged and is given as such. (= mixed), or restricted, potentials, on the other. But, far from being either mere selections from pure potentials, or mere (re-)arrangements of them with respect to gradations of relevance, the impure potentials are really creations out of them, further determinations of them, without which the pure potentials would remain determinables but not determinates, or classes thereof. 

...