ALTHOUGH WE HAVE EXAMINED several forms of Enlightenment rationality in some detail, we have not yet seriously considered rationality per se. For Locke, it has to do with the interpretation of sensory data. According to the Essay, our sensations convey information about various external objects. But how do we know whether that information really amounts to genuine knowledge? Here the role of rationality becomes evident within Locke's scheme. Obviously, we do not need rationality just to assimilate sensory data. To determine how to interpret such data, though, we need to apply rational analysis. Through rational analysis, we can ascertain that the sensory data we receive are not produced by the mind itself. But if not produced by the mind, they must originate from external sources. Thus we arrive at our belief in sensory data as a source of knowledge by means of rational inferences.
In contrast to Locke, Malebranche is less interested in the interpretation of sensory data than in that of the ideas to which such data give rise. From his standpoint, the one fact we cannot deny is that our ideas fail to describe external objects as they actually are. It is at this point that he begins to make use of rational analysis. Ideas that do not properly describe external objects cannot originate from those objects. Instead, there must be some other source. What Malebranche ultimately hopes to prove is that our ideas of external objects could only come from God. To do this, he needs to demonstrate that there is no other way to reconcile their material quality (essentially, their capacity to "modify" the mind via sensation) with their intellectual quality (i.e., their existence in the mind). Thus his analysis attempts to eliminate all other possibilities on the basis of inconsistency. He comes to the notion of God as the source of all our ideas of external objects by rational induction, as the only possible solution to all the requirements in question.
With Leibniz, finally, the definition of rationality changes somewhat. A discussion of necessity versus contingency in the Nouveaux Essais makes it clear that the only propositions endowed with strict necessity are those of geometry and meta-