In Chap. 1 we saw that the basic epistemic cycle includes in an essential way an action on the segment of the world being studied. This action is predicted, on the basis of the previous version of the epistemic model, to produce a certain effect; the fundamental item providing further knowledge is then the correlation of the effect actually found with that predicted. We also noted that it is by no means certain that iteration of the basic epistemic cycle will converge to a satisfactorily functioning model; it is therefore necessary to be able to build up a further epistemic cycle, whose action is now carried on, not directly on the segment being examined, but on the model in the lower- level cycle. This new cycle can also fail to converge satisfactorily, of course: perhaps it does not converge at all, or it converges to something that later proves not to work as had been hoped. Thus there is need for yet further levels of epistemic cycles. Each cycle in such a hierarchy must evidently involve a model, and when the cycle is activated, the model is appropriately revised, as in the basic cycle, whenever needed.
Unable to display preview. Download preview PDF.