1 Leave a comment on block 1 0
Arabic English Commentary

10 — Abrogating and Abrogated Speech

The ordinary meaning of abrogation is to cause to pass away. One says “the sun has abrogated the shade” if it removes it and makes it disappear. Some say its meaning is to copy, because people say “I have abrogated1 what is in this book,” meaning I have copied it. وأما النسخ فمعناه2 الإزالة، يقال نسخت الشمس الظل إذا3 أزالته ورفعته،4 وقيل معناه النقل من قولهم نسخت ما في5 الكتاب أي نقلته.
Abrogation is defined technically as speech that indicates the removal of a legal value established by previous speech, in such a way that without the second speech the legal value would still be in force. There must be a delay between the first speech and the second.6 وحده الخطاب الدال على رفع الحكم الثابت بالخطاب المتقدم على وجه لولاه لكان ثابتا7 مع تراخيه عنه.8
A text can be abrogated while the legal value it indicates remains, and a legal value can be abrogated while the text remains.9 Some abrogation substitutes a new requirement for the old one, and some does not. Some abrogation results in a tougher requirement, and some in a lighter requirement. ويجوز نسخ الرسم وبقاء10 الحكم، ونسخ الحكم وبقاء11 الرسم، والنسخ12 إلى بدل وإلى غير بدل، وإلى ما هو أغلظ وما هو13 أخف.
The Book can be abrogated by the Book, and the Sunna can be abrogated by the Book and by the Sunna. What is collectively transmitted can be abrogated by what is collectively transmitted, and what is individually transmitted can be abrogated by what is individually or collectively transmitted, but what is collectively transmitted cannot be abrogated by what is individually transmitted.14 ويجوز نسخ الكتاب بالكتاب، ونسخ السنة بالكتاب وبالسنة،15 ونسخ المتواتر بالمتواتر، ونسخ الآحاد بالآحاد وبالمتواتر،16 ولا يجوز نسخ المتواتر بالآحاد.17

2 Leave a comment on block 2 0

Another intertextual relationship

Al‑Juwaynī has already presented several types of ambiguity in the language of revelation that allow scholars to establish intertextual relationships in which one piece of evidence modifies the meaning of another. Now he offers one more type of intertextual relationship: abrogation, a relationship in which one text completely supersedes (rather than merely modifies) another text, and thus changes the law, by virtue of its being revealed at a later date.

3 Leave a comment on block 3 0

The Qurʾān speaks mainly (or, arguably, solely) of the abrogation of one prophet’s law by that of a later prophet: the law of Adam was abrogated by that of Moses, and that of Moses by that of the Prophet Muḥammad. (Jewish scholars argued against Muslims that such abrogation was impossible because it would represent a change in God’s mind; Muslim scholars responded that abrogation reflected God’s eternal knowledge that different time periods would be best served by different laws.) Legal theory, however, is interested in changes to the law that occurred within the Prophet Muḥammad’s lifetime. One classic example is the law concerning wine: it was initially permitted, then restricted so that people would not pray while drunk, then forbidden outright. Another example, more controversial today, is the law of warfare: according to one common view, Q 2:109 and numerous other verses commanded forbearance toward opponents, 2:190 commanded fighting in self-defense, and 9:5 commanded seeking out and killing unbelievers. The latter verse, revealed late in the Prophet’s life, is said by some to have abrogated over one hundred earlier verses.

4 Leave a comment on block 4 0

Al‑Juwaynī mentions this type of intertextual relationship after his discussion of the Prophet’s nonverbal actions because actions can abrogate and be abrogated just as well as speech can. He defines abrogation as “interpersonal address,” but this must be understood broadly to include both speech and other means of communication. To my knowledge, no legal theorist ever doubted that actions as well as words could abrogate earlier laws.

5 Leave a comment on block 5 0

Notice that al‑Juwaynī’s technical definition of abrogation once again treats revelation as indicative rather than performative. The first ordinary linguistic meaning of “abrogation” that he quotes is performative: to abrogate is “to cause to pass away.” But technically abrogation is just revelation that “indicates” that a legal value has passed away. Al-Juwaynī is very consistent in filtering out any performative function that revelation might have, and reducing it to indicative statements about the legal values of acts.

6 Leave a comment on block 6 0

Abrogation is quite similar to the other intertextual relationships al‑Juwaynī has already discussed, in that his purpose in defining it is to enable interpreters to use one text to modify the legal impact of another. This is useful first of all for resolving contradictions: if two texts seem to be at odds, and no other way of reconciling them can be found, then they can be said to represent two different stages in the development in the law, and that resolves the contradiction. The ability to reconcile conflicting texts was one of the main reasons al‑Shāfiʿī introduced linguistic intertextual relationships: it enabled him to justify his claim that revelation dictated a coherent law, despite the many apparent conflicts between different revealed texts. Abrogation was just one more way to do the same thing. In fact, according to the historical research of John Burton, the earliest Muslim interpreters of the Qurʾān used the term abrogation in a very general way, to describe any kind of contradiction between two revealed texts. As scholars like al‑Shāfiʿī introduced more sophisticated linguistic devices for resolving contradictions, such as clarification and particularization, the term abrogation came to be used only for those contradictions that could not be resolved linguistically and that therefore had to be chalked up to the passage of time.

7 Leave a comment on block 7 0

Another source of flexibility

The second function of this intertextual relationship (as with linguistic ones) is to give the interpreter as much control as possible over the legal implications of a text. Al‑Juwaynī helps to maximize the reader’s interpretive options by allowing him to define many different forms of abrogation. A text’s legal implications can be replaced by the legal value assigned in a later text, or, if no later text offers a suitable legal value, the first text can simply be discounted. That is what it means for a legal value to be abrogated without a new legal value being imposed in its place. There is no limitation on what kind of legal value can supersede an earlier one: the later one can be either stricter or more lenient.

8 Leave a comment on block 8 0

The only limitation al‑Juwaynī imposes is that a ḥadīth that is not collectively transmitted, and therefore could possibly be inauthentic, cannot abrogate a text that is known with certainty to be authentic, such as the Qurʾān or a collectively transmitted ḥadīth. This is an important limitation, motivated by the general principle that uncertain knowledge cannot trump what is known with certainty.

9 Leave a comment on block 9 0

Beyond this, however, al‑Juwaynī does not care what kind of text abrogates what kind: a Qurʾānic verse can be abrogated by either another verse or a collectively transmitted ḥadīth, while a ḥadīth can be abrogated by any other ḥadīth that has at least the same degree of epistemological certainty. This is a relatively liberal position: some other legal theorists, following al‑Shāfiʿī, argued that the Sunna simply does not have the authority to abrogate the Qurʾān, no matter how authentic the ḥadīth that report it. Some commentators and copyists even changed al‑Juwaynī’s text so that it upheld this less liberal view, but the original states quite clearly that something collectively transmitted (presumably including both Qurʾān and ḥadīth) can abrogate another collectively transmitted text, and in his Burhān he says there is no reason to make a distinction between the Qurʾān and collectively transmitted ḥadīth: Prophetic revelation is all from God, regardless of the form in which it is transmitted, and God can abrogate whatever he wants.

10 Leave a comment on block 10 0

Even the limitation on weaker ḥadīth abrogating the Qurʾān can be bypassed: al‑Shāfiʿī, who held that ḥadīth cannot abrogate the Qurʾān at all, could still say that the penalty of stoning for adultery (which is reported only in individually transmitted ḥadīth) had abrogated the Qurʾānic penalty of imprisonment and/or flogging. He just had to construct a more elaborate argument: he used a different ḥadīth to prove that another Qurʾānic verse had changed the original Qurʾānic penalty of imprisonment to that of flogging, and then argued that this ḥadīth was itself abrogated by the ḥadīth about stoning. Al‑Juwaynī complained this was a needless subterfuge, but his legal theory nevertheless allows for it.

11 Leave a comment on block 11 0

The only major limitation that al‑Juwaynī seems to impose on the interpreter who wishes to use abrogation to modify the legal impact of a text, is that the abrogating text must be revealed after the abrogated one. This is similar to the Muʿtazilī restrictions on delayed clarification and particularization, in that the chronology of revelation limits the intertextual relationships that an interpreter can establish between two texts; the only difference is that here the relationship must involve a delay. This limitation is built into the very definition of abrogation: it would be an oxymoron to say that an earlier text abrogated a later one.

12 Leave a comment on block 12 0

But how much does this really limit the interpreter? Very few Qurʾānic texts and ḥadīth contain indications of when they were revealed, so the chronology of revelation is a subject of much dispute. In fact, the conflicting information on when texts were revealed is often cited as one reason why Muslim jurists disagree on points of law: they rely on different chronological information, and therefore establish different relationships of abrogation. As long as chronology is uncertain, interpreters have room to decide which text should be followed; or they can just claim that the chronology is uncertain and that therefore other evidence is needed, which allows them to make some other kind of argument in favor of one text or the other; or they can argue that abrogation is the wrong device to use, and that a text they admit to be later is nevertheless modified (even if not completely superseded) by an earlier text. That is how some modern interpreters refute the claim that Q 9:5 makes religious warfare against unbelievers an obligation: they can argue, for example, that rather than abrogating 2:190, 9:5 is actually particularized by 2:190, so that 9:5 actually refers only to defensive warfare.

13 Leave a comment on block 13 0

Abrogation can modify the text itself

Finally, al‑Juwaynī allows interpreters not only to modify the legal impact of a text, but also, in a sense, to modify the text itself. That is the upshot of his statement that “a text can be abrogated while the legal value it indicates remains.” This is a surprising statement, since according to his definition abrogation always involves a change of legal value. In the kind of abrogation we have been talking about so far,  “a legal value is abrogated while the text remains”—in other words, the legal value indicated by one text supersedes the legal value indicated by another, even though both texts are still part of the canon of revelation. But now al‑Juwaynī points to another kind of abrogation, in which the abrogated text is no longer part of the canon of revelation.

14 Leave a comment on block 14 0

The classic example of this is the penalty of stoning for adultery: some scholars claimed that at one time the Qurʾān did in fact contain a verse about stoning, and that this verse (rather than the ḥadīth about stoning) was actually responsible for abrogating the earlier penalties of imprisonment and flogging, but that the verse disappeared from the Qurʾān, and was no longer recited. Other such instances are claimed; one scholar has even calculated that if one adds up all the material that was reportedly dropped from the Qurʾān, it amounts to nearly one tenth of the existing text! According to one old story, some of this material was lost when some paper on which Qurʾānic verses had been written down was eaten by a goat; but in general this loss of material is regarded not as an accident but as a deliberate kind of abrogation on God’s part, in which the verse but not the legal value is abrogated (or, as some copies of al‑Juwaynī’s text add, the verse and the legal value can be abrogated together).

15 Leave a comment on block 15 0

On the surface, the fact that some verses were omitted from the Qurʾān sounds like a textual given—a fact that the interpreter must take into account, but which he cannot change. But remember that these verses are not in fact givens; quite the opposite, they are not in any known Qurʾānic text, so the claim that they once were part of the text is a claim made by interpreters, who thus actually change the content of revelation. Of course interpreters cannot change the accepted text of the Qurʾān, but by claiming that the Qurʾān once contained a verse about stoning, for example, they are able to garner support for a law that otherwise would have no Qurʾānic support whatsoever. Missing verses are not often invented today, because today it is hard to fabricate a report out of thin air; but at the time when ḥadīth had not yet all been reduced to writing and collected in well known books, it was still possible for an interpreter to effectively add to the legal (if not the textual) content of the Qurʾān. Even today, it is still possible to dig up such reports about lost verses, which are mostly forgotten but can be found buried in Qurʾān commentaries and other classical texts.

16 Leave a comment on block 16 0

Abrogation supports both the status quo and change

Once again, then, al‑Juwaynī has given interpreters a toolbox of powerful and flexible devices that can be used to make revelation fit just about any legal conclusions one wishes—as long as those conclusions will be appealing enough to one’s audience that they will be disposed to accept the new intertextual relationships one must establish between revealed texts. Like al‑Juwaynī’s linguistic devices, abrogation can be used most easily to defend the legal status quo, because the intertextual relationships required to do so have already been imagined, sorted out, and taken for granted by previous generations of scholars. Defending the legal status quo was probably the main goal for which abrogation was invented, if John Burton is right to say that it was originally designed to maintain both the law and the Sunna in the face of their contradictions with each other and with the Qurʾān. But the rules for using abrogation are not rigid at all, so as the Muslim community’s values evolve, the kinds of abrogative relationships that they will be willing to see in revelation may also evolve.

17 Leave a comment on block 17 0

The flexible potential of abrogation as an interpretive tool is evident from the great disagreement that exists about it today. Throughout Islamic history there have been widely differing opinions about which and how many revealed texts have been abrogated—anywhere from a handful of Qurʾānic verses to several hundred. Medieval scholars tended to keep claiming more and more instances of abrogation, but in modern times the trend has been to reduce the number of claimed instances, or even to reject the concept of abrogation altogether, thus bringing more and more texts back into play, and making them available as evidence that can be used to modify other texts. One modern thinker, the Sudanese reformer Mahmoud M. Taha, even turned abrogation on its head, arguing that the earlier parts of the Qurʾān, revealed while the Prophet was at Mecca, contained a more universal and timeless message that should be applied today, whereas the later portions, revealed at Medina, reflect only situation-bound attempts to implement those principles, and therefore do not have to be implemented in the same form today. Other modern writers have used the doctrine of abrogation not to support individual intertextual maneuvers, but rather to support their general claim that revelation was gradual and progressive, changing with changing circumstances; this in turn supports their view that God intended for the law to continue to evolve even after the death of the Prophet—something that al‑Juwaynī never dreamed of.

18 Leave a comment on block 18 0

Although al-Juwaynī’s theory of abrogation offers a flexible intertextual relationship, just as his discussions of language did, it goes beyond his analysis on language to highlight another crucial interpretive issue: history. We have already mentioned that modern Muslim reformers like Fazlur Rahman have refocused interpretation away from language and onto history, calling for the recovery (or reinvention) of the historical context of revelation as a key to its implementation today. Now we see that modern thinkers are also rethinking the chronology of revelation, and the significance of its chronology, as a justification for reconsidering long-accepted intertextual relationships and interpretations. Analyzing the language of a text is not the only way to modify its meaning; rewriting its history can also change its meaning quite dramatically.

19 Leave a comment on block 19 0

Further reading

John Burton, The Sources of Islamic Law: Islamic Theories of Abrogation (Edinburgh: Edinburgh University Press, 1990).

20 Leave a comment on block 20 0


  1. The Arabic nasakha means to copy as well as to abrogate. ‘Copy’ here translates naqala. ↩
  2. Spr. 601 has والنسخ معناه instead of وأما النسخ فمعناه . ↩
  3. When Ibn al‑Firkāḥ repeats this phrase in his comments, he says أي instead of إذا ; a later marginal note in BM 3093 corrects this to إذا . ↩
  4. In Spr. 601 ورفعته is a marginal addition. It is treated as a gloss in the Maḥallī tradition. ↩
  5. BM 3093 and SH add هذا . When Ibn al‑Firkāḥ repeats this phrase in his comments, he includes هذا according to SH, but not according to Lbg. 256 or BM 3093. ↩
  6. If the second speech followed the first speech immediately, it would constitute something other than abrogation, such as one of the forms of connected particularization described above. ↩
  7. One of SH’s mss, Kuwait 231, has instead تابعا . ↩
  8. From this point until the beginning of chapter 12 – الإجماع there is a gap in Lbg. 256 due to six missing folios, so I will follow BM 3093. ↩
  9. Certain verses which were once included in the Qurʾān, but are not now part of it, are said to have been abrogated as text, whereas other verses were left in the Qurʾān even though their legal implications were superseded by later revelations. ↩
  10. Spr. 601 has instead ويبقى . ↩
  11. Spr. 601 has instead ويبقى . ↩
  12. The Maḥallī tradition has ونسخ الأمرين معا وينقسم النسخ instead of والنسخ . In Spr. 601 ينقسم is added in the margin after والنسخ . ↩
  13. Spr. 601 omits ما هو . ↩
  14. Collective and individual transmission are defined in Section 13 on Reports. ↩
  15. SH has instead والسنة . Spr. 601 has بالسنة instead of بالكتاب وبالسنة ; this reflects al‑Shāfiʿī’s view, abandoned by most of his followers, that the Qurʾān and Sunna cannot abrogate each other. Ibn al‑Firkāḥ’s matn, however, undoubtedly includes نسخ السنة بالكتاب وبالسنة , for he comments on both parts.
    Al‑Maḥallī comments at this point that al‑Juwaynī has not mentioned abrogation of the Book by the Sunna, but then says that one copy of the Waraqāt adds ولا يجوز نسخ الكتاب بالسنة . Ibn al‑Firkāḥ does not include this clause in the matn he reproduces, but at this point in his commentary he says that “he decided to disallow abrogation of the Book by the Sunna” (قطع بمنع نسخ الكتاب بالسنة). (Some scribes felt this could not be al‑Juwaynī’s position, and therefore changed the text: a ḍamma has been added by a later hand above the qāf in BM 3093 to make قطع passive, and one of SH’s mss, Kuwait 231, adds قوم after قطع .) Since Ibn al‑Firkāḥ is generally scrupulous about reproducing al‑Juwaynī’s words even when he disagrees with them, it seems most likely that this is not his paraphrase of something in the matn, but his interpretation of al‑Juwaynī’s conspicuous silence about the Sunna abrogating the Book. This does not necessarily follow, however, from what al‑Juwaynī does say, especially since he allows naskh al‑mutawātir bi‑l‑mutawātir. In his Burhān (ed. ʿAbd al‑ʿAẓīm al‑Dīb, vol. 2, pp. 1307 and 1311) he allows naskh al‑kitāb bi‑l‑sunna al‑mutawātira. ↩
  16. At this point Spr. 601 adds ويجوز نسخ الكتاب بالسنة . This fits poorly at this point in the text, and is difficult to reconcile with Spr. 601’s conspicuous omission of abrogation of the Sunna by the Book, but does agree with al‑Juwaynī’s position in his Burhān (see the previous note). A marginal note, partly unreadable, appears to suggest لعلّه ولايجوز , which would agree with the textual variant mentioned in al‑Maḥallī’s commentary, in SH’s editorial note 2 on p. 223, and in the modern commentary of ʿAbd Allāh ibn Ṣāliḥ al‑Fawzān, pp. 92 and 94. Both the positive addition and the more widespread negative addition were probably early attempts to make explicit what al‑Juwaynī deliberately left unsaid. ↩
  17. Spr. 601 adds لأن الشيء [ينسخ] بمثله أو بما هو أقوى منه . Cf. the matn in the commentary of ʿAbd Allāh al‑Fawzān, p. 92. ↩
Page 16

Source: https://waraqat.vishanoff.com/v/v10/