The following was written in partial fulfillment of the requirements of Dr. Greg Welty's Philosophy: Science and Religion class at Southeastern Baptist Theological Seminary.
The following was prepared as a one-page, single-spaced short response to a question from the readings for this class.
Aristotle’s impetus theory provided an effective and predictively accurate (and indeed, scientific) model of the behavior of objects in terrestrial contexts. Though it was later superseded by Newton’s model, which had no impetuses at all, Aristotle’s method was a perfectly defensible model because it accounted for the known phenomena well. The method worked because a number of constraints in the systems observed were not yet themselves understood. In a limited sense, Aristotle’s view could be construed as a for-certain-conditions subset of Newton’s laws (though, strictly speaking, this is inaccurate). That is, in the conditions in which Aristotle made his observations, and failing to account for some of the phenomena for which Newton’s approach did account, the impetus model for motion was valid.
In systems where there is resistance—e.g., all terrestrial systems—objects do not remain in motion without some other object providing an impetus to them, and objects of different masses but the same basic size do not fall at the same rate. Newton’s revolution was not in denying this but in recognizing that the grounding assumptions were incorrect: terrestrial systems are a subset of a broader possible systems. Because this is so, there exists a more general set of laws which govern the behavior of objects in both terrestrial and non-terrestrial systems. This in turn led to the recognition that Aristotle had made a false generalization from the observed (terrestrial) data: the idea that objects require impetuses to remain in motion was empirically sound, but incomplete. Thus, a feather and a similarly-weighted needle fall at different rates because their different densities and surface areas relative to their masses result in different degrees of resistance. A bowling ball eventually comes to rest because of the resistive forces it experiences due to friction. These are not the general case, but subsets of the broader laws Newton derived; but Aristotle’s formulation can readily be seen just as that subset, under certain conditions.
This is not an unusual pattern in the history of science, and in fact is similarly applicable to Newton’s own system of mechanics, since superseded by both quantum mechanics and general relativity. As the constraints under which a given law operates become better-understood, it is sometimes possible (at least in the case of physical laws) to derive deeper generalizations about the systems in question, generalizations of which the currently-held set of laws are recognized to be a subset. Granted that at times, the new laws do not supersede but wholly replace the old model, and that both the Aristotle-Newton and Newton-Einstein/QM transitions contain some degree of this latter, it is still the case that newer theories ofttimes succeed because they demonstrate themselves capable of solving the same set of problems as the preceding theory, while also addressing many of the unresolved issues in it. So: Newton was able to derive more, and more accurate, predictions than was the Aristotelian model, and to provide more thorough explanations of the systems in question; and likewise Einstein than Newton. In both cases, however, the original model had the utility it did (and the empirical success it did) precisely because it was in some way a for-certain-conditions subset of the more general theory.
A rather more interesting contrast, then, is between theories where the earlier theory is not even construable as a for-certain-conditions subset of the later, as in the phlogiston theory of combustion, since replaced by an oxidizing view. The idea had substantial explanatory power and harmonized quite well with the data, but was ultimately overthrown by the alternative (and now current, albeit revised) view of oxidation. Such cases, unlike the Aristotelian-Newtonian or Newtonian-Einsteinian/quantum mechanical transitions, are much harder to reconcile with a progressive view of the development of scientific theory, or with realist accounts of scientific explanation.