And none of this can be conceived while they maintain:
“Allah’s names are shared only in wording, not in meaning.”
For things that share a term but not a meaning are exemplified by mushtarī, which denotes both the planet Jupiter and a buyer, and Suhayl, which denotes both the star Canopus and Ibn ‘Amr. If someone hears another say, “Suhayl ibn ‘Amr came to me, and he is the buyer of this commodity,” he derives no notion of a celestial body from these words at all unless he knows that the term is also used for one. Thus, if Allah’s names did not convey a common meaning across their different applications, His servants would understand nothing whatsoever from His names unless they knew what is peculiar to His essence. Yet they do not know what is peculiar to His essence, so they would know nothing.
Furthermore, the division of existence into the beginningless and the originated, and similar distinctions, is known self-evidently; anyone who challenges this is a sophist. Likewise, it is self-evident that the two names share a common element of meaning.
To say that a term applies literally to both does not require that speakers of the language should have used it without qualification to express the unrestricted, common meaning. For meanings such as life, knowledge, power, and rising over—and indeed a hand and other such things—exist only in relation to something else: either as an attribute subsisting in another or as a body subsisting in another, with no external existence apart from its locus.
However, when rational investigators sought to abstract unrestricted universal meanings, they expressed them through unrestricted universal terms. Speakers of the language, by contrast, begin their discourse by saying, for example, “Zayd came,” and “This is Zayd’s face.” They point to his coming and his face, and the person addressed understands this.