Rationalists claim that real knowledge about self, God, the world, is possible le only if it is certain or based on something that is certain. Since sense experience cannot guarantee certainty, it cannot be the basis for knowledge. Only reason can reveal propositions that are indubitable or that cannot be doubted, so only reason can be trusted to provide knowledge as opposed to mere beliefs.
The problem with relying on reason alone, though, is that a priori propositions or propositions whose truth or falsity is known prior to and impendent of any sense experience, do not provide any useful information about the world. To know, for example, that bachelors are unmarried, that unicorns have horns, or that triangles have three sides, does not tell us whether there are such things in the world as bachelors, unicorns, or triangles. For that information, we have to rely on experience. But if experience is ruled out by rationalists as unjustified bases for knowledge, then it seems that we will never be able to know anything about the world.
Descartes attempted to get around this problem by saying that we know that we exist and that there is some information about the world that is not based on sensory experience. We also know according to Descartes that God exists and that God does not deceive us when we limit our beliefs about the world to clear and distinct ideas. So we can know things about the world insofar as it is clearly and distinctly organizable. That last feature means that, if we think about things in the world not in terms of what we learn from relying on our senses, for example that grass is green, or that it is cold outside, but only in terms of how things have to be, that in order to be grass, a plant has to have certain characteristics, we will then be able to say that we know something about them. Otherwise, we have to admit that we have beliefs about things not knowledge.
Empiricists disagree with this conclusion however. For empiricists, fact about the world are known a posteriori or they depend on experience for their truth of falsity. For empiricists also facts are publicly verifiable and exhibit enough order that they can be the gasis for generalizations and predictions. Admittedly experience does not provide the absolute certainty that rationalists require for saying that we know something. But even if our knowledge is only probable, at lest it is of some use rather than being simply about how we under definitions.
