…is merely a report about what exists in the mind, not about what exists outside it. Thus, whatever is evident to a person—known to him and present in his intellect—needs no proof, whereas whatever is not so present does call for proof. The very fact that something is evident or not is a relation between the object known and the knower, and in this people differ. What is self-evident to Zayd may remain obscure to ʿAmr, for the causes of knowledge, the strength of perception, and the quality of minds all vary. Hence, once the existence of certain attributes, their necessity, or the subject’s being characterized by them becomes clear to one individual, that does not oblige its being clear to everyone; nor does its remaining hidden from some entail that it must remain hidden from all.

Accordingly, if one separates first principles from other propositions on the ground that a first principle needs no middle term, the distinction between a first principle and a non-first principle becomes relative, dependent on people’s states. It is like saying that a proposition is either known or unknown, manifest or unmanifest, clear or unclear, obvious or not, or that a person can conceive its truth or cannot. This is a sound distinction in the sense that, with respect to any individual, every proposition is either evident to him or not. Yet the distinction does not partition propositions into genera such that one entire genus consists of non-evident propositions and another entire genus of evident ones. The criterion does not refer back to an intrinsic property of the proposition itself, as is alleged by those who claim that an evident necessary attribute is one that belongs to its subject without a middle term in reality, whereas a non-evident necessary attribute belongs only through a middle term in reality.

In fact, every necessary attribute of a subject attaches to that subject without any intermediary whatever in external reality. That a person needs a proof in order to know of this attachment does not mean that the attachment itself depends on a middle term, as has already been explained. Not everything whose necessity for its subject is known only through evidence is therefore attached to it by means of a middle term. The effect is necessary to its determinate cause; it cannot fail to follow from its complete cause—one that comprises the presence of all conditions and the absence of all impediments. Nevertheless, a person may know of the effect’s existence only through a middle term, for instance, by inferring the cause from another effect.