intend the former. They generally use “interpretation” in the third sense, though they may use it in the second. Sometimes a text is explained in a way consistent with its apparent meaning, and it becomes clear that this is not interpretation in the third sense. Yet they reject this and dislike reflecting on the texts and examining their meanings—I mean the texts whose interpretation, they say, is known only to Allah.

Their approach to these texts then depends on their beliefs. If they are Qadariyyah, they say: “The texts affirming that the servant acts are unambiguous, while the texts affirming that Allah Most High creates the acts of His servants, or wills everything that occurs, are ambiguous texts whose interpretation none knows except Allah”—if they are among those who do not reinterpret them. For most groups include those who reinterpret whatever conflicts with their position and those who do not.

If they are among those who affirm the divine attributes they claim to know through reason, but not the attributes known through revealed reports—as was the case with many later Kullābiyyah, such as Abū al-Maʿālī towards the end of his life and Ibn ʿAqīl in much of his discourse—they say of the texts containing attributes that, according to them, cannot be known through reason: “These are ambiguous texts whose interpretation none knows except Allah.” Many of them have two positions and adopt two different stances: sometimes they reinterpret the texts and deem such interpretation obligatory or permissible; at other times they prohibit it. Such differences of opinion are found in the statements of Abū al-Maʿālī, Ibn ʿAqīl and others like them.

Those who affirm Allah’s being above on rational grounds and regard it as an attribute known through reason—such as Abū Muhammad ibn Kullāb, Abū al-Hasan ibn al-Zāghūnī and those who agree with him, and Judge Abū