The notion of "blindness" reflected in the name of Campbell's model (which has been the focus of much of the criticism with regard to scientific advance) enters via the mechanisms of variation. For Campbell, any variation that leads to truly new knowledge--knowledge that has not been attained before--must be blind in the sense of going "beyond the limits of foresight or prescience." It is important to be clear here, since the point tends to be misunderstood, that "blind" (in this sense) does not mean "random" or "unpremeditated" or "unconstrained." It simply denotes, in accord with Campbell's characterization, that the outcome of the variation cannot be foreseen or predicted insofar as the matter in question in concerned--if it could, the knowledge obtained would not be new. . . . Consistent with Campbell's characterization, knowledge grows (that is, blindness is reduced) through extension of the limits of what can be foreseen or predicted. Whatever the disagreement in other fields, such a statement seems to me valid in engineering design." (Vincenti, 1990, pp. 242-243)
From outside or in retrospect, the entire process tends to seem more ordered and intentional--less blind--than it usually is. It is difficult to learn what goes on in even the conscious minds of others, and we all prefer to remember our rational achievements and forget the fumblings and ideas that didn't work out. Luck can also play a role. (Vincenti, 1990, p. 246)
[HOME] [NEXT]In the end, decreasing uncertainty in the growth of knowledge in a technology comes, I suggest, mainly from the increase in scope and precision (that is, the decrease in unsureness) in the vicarious means of selection. Just as expanding scope tends, as we saw, to widen the field that can be overtly searched, so also the increase in both scope and precision sharpens the ability to weed out variations that won't work in the real environment. Blindness in the variations may by the same token even increase--engineers have freedom to be increasingly blind in their trial variations as their means of vicarious selection become more reliable. One sees engineers today, for example, using computer models to explore a much wider field of possibilities than they were able to select from just a decade ago. (Vincenti, 1990, p. 250)