He says:

“Nothing that has an actual reality is without a definition—whether that definition be expressed negatively or affirmatively. The point of defining is to address the reality of the thing by which it is set apart from everything else, for a thing is distinguished from what is other than it by its very essence and reality, not by anyone’s mere utterance.”

Abū al-Maʿālī then said:

“The verifiers maintain that forward-inclusion (al-iṭrād) and reverse-inclusion (al-inʿikās) are among the necessary conditions of a proper definition. Because the aim of a definition is to single out the definiendum from what does not pertain to it by means of a particular attribute, that aim can be realized only when both conditions are met. Forward-inclusion means that the thing defined is found whenever the definition is found; reverse-inclusion means that the thing defined is absent whenever the definition is absent.

So, if someone says, ‘The definition of knowledge is “accident,”’ forward-inclusion fails, for not every accident is knowledge—hence the definition is broken. And if one says, ‘The definition of knowledge is “every created cognition,”’ reverse-inclusion fails, because there is knowledge that is not created. The person asking for the definition of knowledge does not seek the definition of one particular kind to the exclusion of others; rather, he wishes to encompass the meaning of all forms of knowledge.

If we say, ‘Knowledge is cognition,’ then every cognition is knowledge and every knowledge is cognition; whatever is not knowledge is not cognition, and whatever is not cognition is not knowledge. These are four statements—two affirmative and two negative—and without this two-sidedness no definition is sound.

Abū al-Maʿālī continued:

‘If someone were to ask, “Is it permissible to construct a definition from two descriptive elements, or not?” we reply: the speculative theologians differ on the matter. Many of them hold that a composite is not a proper definition, and our teacher Abū al-Ḥasan inclines to that view and criticizes composition. By rejecting composition we do not mean to oblige the respondent to produce a single word as the definition of whatever he is asked about; what is sought is unity of meaning, not of wording. Expressions are not desired for their own sake, and they are not themselves definitions; rather, they merely point to definitions.

Our teacher Abū al-Ḥasan, while denying that a definition may be composite, nevertheless defined knowledge by saying: “It is that which renders its locus a knower.” Although this consists of several words, he did not regard it as a composite definition, for the purpose of a definition is to address a single attribute—namely, that knowledge confers its ruling. Likewise, when one says in defining substance, “That which receives accidents,” it is not