The first category consists of sense perceptions. It is well known that the senses do not perceive anything universal or general at all. Mere sense perceptions therefore contain no universal proposition suitable to serve as a premise in a demonstration yielding certainty.
When they illustrate this by saying that fire burns, or something similar, they have no knowledge that this proposition holds universally. All they have is experience and customary regularity, which are of the same kind as analogical reasoning.
As for knowing this through fire’s possession of a power to burn, the knowledge that every fire must possess this power is itself a universal judgment. If it is said, “The form of fire must contain this power, and whatever lacks this power is not fire,” this claim, even if sound, does not establish with certainty that everything possessing this power burns whatever it encounters. Even if this is what usually happens, analogical reasoning, the inclusive syllogism, customary regularity, and incomplete induction all share in this—assuming this is conceded to them.
How, then, can their claim stand when it is known that fire does not burn the salamander, the ruby, or objects coated with specially prepared substances? I know of no universal proposition based on sense perception that admits of no counterexample. Indeed, a universal proposition is not itself sensory. The sensory proposition is simply that this particular fire burns, for the senses perceive only particular things.
As for intellectual judgment, they say that when the soul sees these particular instances, it is prepared to receive, by emanation, a universally applicable proposition. It is well known that this is of the same kind as analogical reasoning. Its universal applicability cannot be trusted unless it is known that the general judgment necessarily follows from the feature shared by those instances. If this is known, it is known in all the particular instances; knowledge of those instances therefore does not depend upon that universal proposition. Moreover, all rational people agree that no proposition based on customary regularity is universal in such a way that it admits of no counterexample.