The claimant’s contention that a definition is no more than a single, meaningful word is of a piece with their further claim that conception—one of the two kinds of knowledge—is a mental image completely stripped of every affirmation and negation. It is well-known among scholars that such a thing cannot be called “knowledge” at all. If a notion simply crosses a person’s mind without his considering in any way whether it exists or does not exist, he has acquired no knowledge whatsoever. Picture someone to whom there occurs the idea of a sea of mercury or a mountain of ruby, with no reflection on whether that image corresponds to something outside the mind, or whether it is possible or impossible; such musings belong to the realm of idle whisperings, not the realm of knowledge.
We have treated this matter at length in several places, such as our commentary on al-Muḥaṣṣal, where we explained that the prerequisite for assent is of the same kind as the knowledge required for speech. Whoever gives assent to what he has not conceived is speaking without knowledge, whereas whoever assents to what he has conceived speaks as one informed. Their remarks about verbal definitions, therefore, are of the same stripe as their remarks about mental conceptions.
Second point: A solitary report, unaccompanied by proof, does not confer certain knowledge.
They also maintain that a definition neither compels assent nor rests upon evidence; it can, they say, be overturned only by invalidation and counterexample—unlike analogical argument, which admits of direct objection and contradiction.
We reply: If the definer has produced no proof for the soundness of his definition, it is impossible for the hearer to know the thing defined, once he allows that the definition may be mistaken. When the correctness of a definition is known solely through the definer’s assertion—and that assertion may be true or false—the hearer cannot attain knowledge through it.
What is astonishing is that these people claim such procedures to be certain rational methods. They make knowledge of the single term the foundation for knowledge of the compound statement, and in all of that they rely upon a definition that is nothing more than the definer’s own assertion, without