Nothing in what we have mentioned is open to dispute except the self-evident propositions, for these possess generality. One might think that this generality provides knowledge of things outside the mind, and thus suppose that these propositions yield universal knowledge. But the remaining principles contain no universal knowledge. Only purely rational, self-evident propositions should therefore serve as premises of demonstration, since these alone are universal. The remaining propositions are particular, so how can they properly serve as premises of demonstration?

Unless it is argued:

Through these propositions, particulars are known, while through reason, universals are known; together, they complete the demonstration. For example, sensory perception establishes that one man has a thousand dirhams and another two thousand, while reason establishes that two is greater than one. It is thus known that the latter has more wealth.

The reply is:

This is correct, but it yields only a specific, particular proposition: that one man’s wealth exceeds the other’s. Knowledge of specific particulars does not require a syllogism. They may be known without a syllogism, through analogical inference, or through inference from two particular propositions. For you know through sensory perception that this thing is like that thing, and you know that this thing has such-and-such attributes; you therefore know that the other is like it, and that what holds true of a thing holds true of its like. Likewise, it may be known that Zayd is older than Amr and that Amr is older than Khālid, along with similar specific matters that can be known without the subsumptive syllogism for which they have laid down their stipulated conditions.