that the ḥadith has been abrogated by the verse? Such a notion flatly contradicts what is settled in the discipline of uṣūl al-fiqh (the principles of jurisprudence). Their case resembles that of a person who would claim that the sound ḥadith, “The thief’s hand is not to be cut off except for a quarter of a dinar or more” (agreed upon), has been abrogated by the Almighty’s words, “As for the male thief and the female thief, cut off their hands” (Al-Māʾidah 5:38); or that the Prophet’s statement regarding the sea—“Its water is purifying and its dead creatures are lawful” (sound)—and every report of similar purport has been abrogated by His declaration, “Carrion has been made unlawful for you” (Al-Māʾidah 5:3). Examples of this sort are exceedingly numerous, and for that reason al-Shawkānī explicitly ruled that the face and the hands are excepted, adducing this very ḥadith as proof, as noted a moment ago.

A third point then occurred to me: the interpretation in question would only be valid on the assumption that the decisive proof in our issue were merely the maintenance of the original ruling, namely permissibility; in that case, the proposed construction might stand. But that is not the situation here, for the discussion revolves around this particular ḥadith, whose soundness we have already conceded. Hence the proposed construction is invalid. How can the ḥadith be confined to the period before the legislation of ḥijāb or jilbāb when it actually concurs with the verse in forbidding a woman to appear without her jilbāb—that is, it too departs from the original presumption—and even adds to the jilbāb verse by excluding the face and the two hands, as explained earlier? This is perfectly clear.

This also exposes the weakness of what our brother al-Iskandarānī has written in his book (3/344), where he attempted to shore up the aforesaid interpretation by quoting some of my words about fashioned gold, imagining that they served as evidence against me, while overlooking the distinction between a case in which one leaves the default ruling by means of a general text and a case in which the text that departs from the default is coupled with an explicit exception indicating… ⦗125⦘