Contradictions in the Qur’an: The Book’s Own Test

Every sacred book has defenders who insist it contains no contradictions. The Qur’an does something bolder: it stakes a claim on it. “Then do they not reflect upon the Qur’an?” asks Surah 4:82. “If it had been from [any] other than Allah, they would have found within it much contradiction.”¹ This is not a defensive aside. It is offered as a proof—an open invitation to the doubter to go looking, with the book’s divine origin wagered on what he will find.

That is admirable. A text that invites falsification is braver than one that merely demands belief, and Muslims are right to point to this verse with a certain pride. But an invitation only means something if someone may accept it. This essay accepts it—carefully, charitably, and on the book’s own terms. We are not importing a foreign standard or holding the Qur’an to a test its own author never posed. We are applying the test Surah 4:82 itself proposes, with the seriousness the proposal deserves.

The Argument, Stated Carefully

The claim of this essay can be put in four steps:

1. The Qur’an teaches that a book from God would be free of “much ikhtilâf”—divergence, discrepancy, contradiction—and offers this as evidence of its divine origin (Surah 4:82).
2. The canonical printed editions of the Qur’an differ from one another in ways that sometimes affect meaning—in at least one case, what God commands.
3. The text itself contains internal tensions serious enough that Islamic scholarship developed an entire literary genre devoted to resolving them.
4. Therefore, measured by its own stated standard, the Qur’an does not exhibit the flawless self-consistency that popular apologetics claims for it. At minimum, Surah 4:82 cannot serve as the proof it is offered as; and if any of the tensions above is a genuine contradiction, it counts against the very claim it was meant to secure.

Notice what the argument does not say. It does not say the Qur’an is riddled with errors, or that Muslims are foolish to revere it. Nor does it re-argue ground covered elsewhere in this library: how the list of canonical readings was assembled is the subject of the qira’at deep dive; the Uthmanic standardisation and the lost-verse reports belong to the preservation essay; and the most discussed internal tension of all—verses that supersede earlier verses, as with the staged prohibition of wine—is handled in the abrogation (naskh) essay. The question here is narrow: does the Qur’an, examined honestly, pass the test of 4:82 in the strong form in which that test is popularly claimed?

And it is worth being precise about what that strong form says, because it is what Muslims are commonly taught. The claim is not the modest one—that the Qur’an is broadly consistent, its difficulties resolvable. The claim is that the book contains no contradictions at all: not one, anywhere, of any kind—and that this is itself the proof of its divine origin. (How perfectly the text was transmitted is a separate argument, treated in the preservation essay; here the question is the content.) A claim of that shape has a particular logic: it is not weighed on balance, it is falsified by a single genuine counterexample. So nothing below requires the Qur’an to be “riddled” with anything. And if, at the end of the examination, the defence has quietly become “there are no serious contradictions” or “the tensions can mostly be resolved,” the original claim has already been given up—for “mostly resolvable” is precisely the kind of thing one says about an ordinary book.

Part One: Between the Printed Editions

Begin with a fact established in detail in the qira’at essay: there is more than one canonical printed Qur’an. The reading of Hafs dominates the Muslim world; Warsh is printed across North and West Africa; Qalun prevails in Libya; and the reading of Abu ‘Amr of Basra, in the transmission of al-Duri, remains popular in Sudan. All are certified as authentic by Muslim authorities, and the King Fahd Complex in Medina has itself printed multiple transmissions.² How that list came to be is another essay’s business. The present question is simpler: do the editions ever say different things?

First, a Difference That Is Not a Contradiction: King or Owner (Surah 1:4)

Start with the example Muslim scholars themselves reach for, because it shows the phenomenon at its most benign. In al-Fatiha—recited in every unit of every daily prayer—verse 4 is read by ‘Asim (whose transmission by Hafs is the world’s standard text) and by al-Kisa’i among the Seven—joined by Ya‘qub and Khalaf among the Ten—with a long vowel, mâliki, “Owner of the Day of Judgment”; the remaining canonical readers give maliki, “King of the Day of Judgment.”³ Is that a contradiction? Plainly not. God is both king and owner, and the traditional judgment that the two readings enrich one another is entirely reasonable. It should be granted without hesitation—and it is worth granting first, because it marks the standard the harder cases must be measured against. The variants are real, they reach meaning and not just sound, and most of them are complementary in just this way.

Narration or Command? (Surah 21:4)

A step further along the spectrum sits Surah 21:4. In the Hafs edition the verse is a narration: “He said (qâla), ‘My Lord knows whatever is said in the heaven and the earth.’” In other canonical readings, including Warsh and Shu‘ba, it is an imperative addressed to Muhammad: “Say (qul): ‘My Lord knows…’”⁴ No doctrine hangs on this, and one can still call the readings complementary in a loose sense. But notice that they are no longer two facets of one statement; they are two different sentences—a report about the Prophet in one edition, a command to him in the other. The “enrichment” account is beginning to strain.

Whose Amazement? (Surah 37:12)

A step further still, the reader of the editions meets a variant that changes not what is felt but who feels it. In the Hafs edition, Surah 37:12 reads bal ‘ajibta—“Nay, you [Muhammad] marvelled, while they ridicule.” In the canonical readings of Hamza, al-Kisa’i and Khalaf it is bal ‘ajibtu—“Nay, I marvelled”: God speaking of His own amazement.¹⁸ Between certified recitations of the same verse, a human sentiment becomes a divine one.

Unlike the Lot passage below, no command hangs on this. But it brushes something weightier than narration: a divine attribute. Can the all-knowing God marvel—be struck by something as if unforeseen? The tradition itself divided here, and honestly: Ibn Taymiyyah accepted the “I marvelled” reading and argued from it that amazement is a real attribute of Allah, affirmed in a manner befitting His majesty, while others resisted precisely because omniscience seems to leave no room for surprise. The standard harmonization—that God’s “amazement” is to be understood bi-mâ yalîq bi-jalâlih, in the way that befits Him—deserves to be heard, and it may satisfy. Yet even granting it fully, the two editions still ascribe the same marvelling to two different persons: in one printing the Prophet marvels; in the other, God does. That is not two facets of one sentiment; it is two different marvellers—one human, one divine.

The Hard Case: Two Commands to Lot (Surah 11:81)

Before the example, the real Muslim position deserves stating by name: in mainstream Islamic theology the variant readings are not competing human editions but are all divinely revealed—descending from the “seven ahruf” in which, according to hadith, the Qur’an was sent down—so that differences between editions are revelation, not contradiction (the claim is treated in full in the qira’at deep dive). But notice the bridge that position builds: precisely because Muslims affirm every canonical reading as revealed Qur’an, a divergence between readings is a divergence within the Qur’an, on Islam’s own terms.

Now the example on which this part of the argument chiefly rests, and it must be stated with grammatical precision. In Surah 11:81, angels warn Lot to flee the doomed city by night: “So travel with your family during a portion of the night, and let not any among you look back—except your wife; indeed, she will be struck by that which strikes them” (Saheeh International; translators of the same Hafs text themselves split over where to attach the exception—a first sign of the ambiguity at issue). The phrase “except your wife” (illâ mra’ataka / illâ mra’atuka) differs between the canonical readings by a single case vowel, and the classical commentators spell out what each case does.⁵

In the reading of most transmitters—including both Hafs and Warsh—the word is accusative (illâ mra’ataka), and on the standard construal the exception attaches to the command to travel: take your family—except your wife. Leave her behind. But in the canonical readings of Ibn Kathir al-Makki (the early Meccan reader, d. 120 AH / 738 AD—not the later exegete of the same name) and Abu ‘Amr of Basra, the word is nominative (illâ mra’atuka), and the exception can only attach to the clause about looking back—so the command to travel covers his whole household, his wife included: take everyone; let none of you look back—except your wife. She will look back, and be destroyed. The medieval commentator al-Baghawi lays out exactly these two senses side by side: on one reading Lot is commanded not to take his wife; on the other she travels with the family and perishes for turning to look.⁵ And both readings sit today in certified print: the accusative in every Hafs and Warsh mushaf, the nominative in the reading of Abu ‘Amr that Sudanese Muslims recite through al-Duri.²

Here the complementarity defence meets its limit. “King” and “Owner” can both be true at once. But “take your family—except your wife” and “take your whole household, your wife included” are two different divine commands—and they give the same woman two different last nights. The strongest traditional response deserves stating: some grammarians note that in a negated clause an exception may take the accusative even when it attaches to the nearer clause, so the accusative reading could be construed to mean the same as the nominative; and the exegete Ibn Kathir duly records both scenarios in his commentary—that Lot was ordered not to take her, or that she went along and looked back.⁶ That reply is genuinely available, and it should be weighed. But notice what it concedes: the tradition itself transmits two readings whose most natural senses diverge, is not sure which of two incompatible stories the verse tells, and must resolve a divine command by grammatical arbitration after the fact. It is also worth observing that in the parallel telling at Surah 15:65 the same command appears with no exception clause at all.⁷ Whatever one concludes, this is ikhtilâf—the very word of Surah 4:82—located not between the Qur’an and its critics but between the Qur’an’s own certified editions.

Other meaning-affecting variants—“fought” versus “were killed” alongside the prophets (3:146), “wash” versus “wipe” the feet in the ablution law (5:6), the purity ruling of 2:222—are documented in the qira’at deep dive and need not be repeated; they reinforce, from several directions, that the differences between editions are not always cosmetic.

Part Two: Within the Text Itself

Suppose we set the editions aside and take a single text—the Hafs edition on the shelf—on its own. Does it display the seamlessness the popular apologetic claims? A striking fact should frame everything here: the Islamic tradition itself never thought the question impertinent. By the third Islamic century, Ibn Qutayba (d. 276 AH / 889 AD) had written Ta’wil Mushkil al-Qur’an—“The Interpretation of the Qur’an’s Difficulties”—and the genre of mushkil al-Qur’an became a standing department of Muslim scholarship.⁸ Two things follow: apparent contradictions are not a modern polemical invention, and harmonisation is a legitimate scholarly move—Christians use it constantly for the Bible, and it would be a double standard to deny it to Islam. The question in each case is whether the harmonisation succeeds naturally or has to be purchased at a price. Consider four cases across that spectrum.

How Many Days, and in What Order? (Creation)

Seven times the Qur’an says the heavens and the earth were created in six days (e.g., 7:54; 10:3; 57:4). But Surah 41:9–12 gives an itemised account: the earth in two days, its mountains and provisions in four days, then (thumma) the seven heavens in two days—which sums to eight.⁹ The classical harmonisation, found in al-Tabari and Ibn Kathir, is that the “four days” include the first two: two days for the earth, two more for its provisions, two for the heavens—six.¹⁰ That is a possible reading, and it should be stated at full strength: Arabic can use an inclusive count, and the commentators were not fools. But it must also be said that the four days of 41:10 are attached to verse 10’s own activities and joined to verse 9 by a simple “and” (wa); reading them as a running total—“in four days, complete”—is attested in the tafsir, but it is not the sense an unassisted reader supplies. A related tension concerns order: 2:29 and 41:9–11 present the earth (and what is in it) before God “turned to the heaven,” while 79:27–30 says God built the heaven and “after that He spread the earth.” The harmonisation attributed to Ibn ‘Abbas—created before, but spread out after—turns on giving the verb dahâhâ a deliberately narrow sense, and it must reckon with the connective thumma (“then”) of 2:29, which ordinarily marks sequence, not mere conjunction.¹¹ Possible, again. But the pattern is becoming visible: the text is consistent only with assistance.

Who Was the First Muslim?

Muhammad is commanded to be “the first of those who submit” and declares “I am the first of the Muslims” (6:14; 6:163; cf. 39:12). Yet Moses, centuries earlier, says “I am the first of the believers” (7:143), and Abraham and his sons submit (aslamtu—the same root as “Muslim”) earlier still (2:131–132).¹² The standard gloss, reported from the early exegete Qatadah, is “the first Muslim of this community”—each prophet the first of his own people.¹² As harmonisations go this one is comparatively easy, and honesty says so: “first” can carry an idiomatic, foremost-of-my-people sense. But note the method by which the difficulty dissolves: words are supplied that the text does not contain. That method is fine—so long as its availability is remembered when the same tradition presses far smaller tensions in the Bible as fatal.

The Sister of Aaron

The hardest internal case. Surah 19:28 has Mary’s neighbours address the mother of Jesus as “O sister of Aaron”; Surah 66:12 calls her “Mary, the daughter of ‘Imran”; Surah 3:35–36 narrates her birth to the wife of ‘Imran.¹³ But in the Bible the sister of Aaron and daughter of Amram (‘Imran) is Miriam—the prophetess of the Exodus, some fourteen centuries before Mary of Nazareth. The two women share a name (Maryam in both Arabic and Hebrew), and the Qur’an appears to give the mother of Jesus the family of her ancient namesake. Here the steelman is as strong as a steelman can be, because it comes from Muhammad himself. In Sahih Muslim, Mughira ibn Shu‘ba reports that the Christians of Najran pressed him with exactly this question; the Prophet answered that people “used to give names (to their persons) after the names of Apostles and pious persons who had gone before them”—that is, “sister of Aaron” is honorific or tribal, not literal.¹⁴ If “sister of Aaron” stood alone, that answer would settle the matter; such naming conventions are real. What the answer does not address is the convergence: Mary is called Aaron’s sister and ‘Imran’s daughter and the child of ‘Imran’s wife—precisely the family of the biblical Miriam, reproduced in full. In fairness, the hadith’s naming-convention principle extends naturally to a father named after Amram; the improbability lies in the triple convergence taken together—and in the fact that the Qur’an nowhere distinguishes the two Maryams. One may hold that Mary’s actual father coincidentally bore the ancestral name and the honorific coincidentally matched; that cannot be strictly disproven. But in our judgment—flagged as a judgment—the simpler explanation of the whole pattern is a conflation of the two Maryams, and the harmonisation survives only by treating each thread in isolation.

A Case Where Harmonisation Works: Who Takes the Soul?

Fairness requires showing the other kind of case. Surah 39:42 says “Allah takes the souls at the time of their death”; 32:11 says “the angel of death…will take you”; 6:61 says “Our messengers take him.”¹⁵ Polemical lists sometimes parade this as a contradiction. It is not a good example, and we decline to use it as one: the obvious harmonisation—angels act as God’s agents, and the angel of death has assistants—is natural, costs nothing, and mirrors patterns Scripture itself uses. Weighing each case on its merits, rather than counting every tension as a kill, is what the test of 4:82 deserves—and it is what we would ask of a Muslim reader examining the Bible.

What Should Be Conceded

Several concessions belong on the table, gladly. The word ikhtilâf in 4:82 has a semantic range—“discrepancy,” “divergence,” “inconsistency”—and Muslim translators render it variously; the test should not be caricatured as demanding that no two verses ever need reading together.¹ The strongest rejoinder to this whole essay deserves the same candour: 4:82 speaks of much contradiction (ikhtilâfan kathîran), and a critic who searches fourteen centuries and returns with a case vowel and a handful of harmonised difficulties might be said to have confirmed that standard rather than embarrassed it. But Muslim orthodoxy has never claimed that the Qur’an contains only a little contradiction; the doctrine of the Qur’an as God’s own speech, and the entire mushkil al-Qur’an genre, presuppose that the operative standard is zero—every difficulty is resolved, none is tolerated. This essay measures the book against the tradition’s own operative standard, not a lowered one. The great majority of differences between the editions are minor and vanish in translation. Harmonisation is a legitimate tool, and some proposed contradictions genuinely dissolve under it, as the soul-taking case shows. And nothing here impugns the sincerity or intelligence of the Muslim scholars who built the mushkil literature; their honesty about the difficulties compares favourably with the popular apologetic that insists there are none. The claim of this essay is therefore calibrated: not “the Qur’an is riddled with error,” but that, measured by Surah 4:82’s own standard, the text does not exhibit the flawless self-consistency the popular argument claims—its editions diverge in meaning and once, on the standard construals, in command, and its internal tensions are real enough that the tradition itself built a genre to manage them.

A Word About the Double Standard

This point should be made gently, because it is an invitation rather than an accusation. A staple of dawah polemic sets Bible “versions” and variant manuscripts against a supposedly variant-free Qur’an, and catalogues alleged biblical contradictions against a Qur’an said to contain none. A Muslim friend who argues this way is not being dishonest; he is usually relying, in good faith, on what he was taught. But the measure must be one measure. If a case-ending’s divergence between Greek manuscripts counts against the New Testament, the case-ending that divides the printed Qur’ans at 11:81 counts by the same scale; if biblical difficulties may be harmonised, so may Qur’anic ones—and vice versa.

Yet one asymmetry must not be lost, because it is the heart of this essay: the two books make different claims for themselves, and each is answerable to its own. The Bible nowhere stakes its divine origin on containing no difficulties; Christians claim that its text is well preserved—and argue its truth separately, on evidence, since good manuscripts establish preservation, not truth, for either book. The Qur’an, in Surah 4:82, stakes its origin on containing no contradiction. So when the examination reaches 11:81, the Muslim cannot answer as the Christian answers—“the text is well preserved, and the difficulties can be worked through”—because “well preserved, with workable difficulties” is the Bible’s kind of claim, not his book’s. And it is not the Christian who holds the Qur’an to the higher standard. The Qur’an issued the challenge; this essay merely accepts it. The fairness we ask—each book judged by the claims it makes for itself—we owe.

The State of the Scholarship

Where does serious scholarship stand? On the Muslim side, the harmonising tradition is alive and sophisticated, and it has modern academic advocates. M. A. S. Abdel Haleem of SOAS, for instance, has argued influentially that features which strike readers as incoherence—abrupt shifts of person, number, and tense—are in fact a deliberate rhetorical device (iltifât) prized in classical Arabic; his case is a standing caution against finding “contradiction” in what may be style.¹⁶ That caution is well taken for grammatical shifts; it does not reach cases like 11:81 or the creation arithmetic, which concern content rather than register. On the academic side, historical-critical scholars such as Nicolai Sinai read the Qur’an as a corpus of proclamations developing over two decades, in which tensions and revisions are unsurprising marks of a living text—a framework that is not polemical, but which also quietly declines the premise of 4:82 that internal ikhtilâf is unthinkable in the book.¹⁷ It is common ground in scholarship that the canonical readings differ, sometimes in meaning; the live debate is over how much the real variants and tensions matter, and that debate should be reported as live, not settled.

Conclusion: What the Test Does and Does Not Show

What has the book’s own test shown? Not that the Qur’an is worthless, incoherent, or unrevered by honest people—none of that follows, and none of it is claimed. Not even, by itself, that Islam is false: a determined believer can harmonise every case above, and some harmonisations are better than others. What the test shows is narrower and, for the popular apologetic, more uncomfortable: Surah 4:82 invites the world to examine the Qur’an for ikhtilâf, and the examination does not come back empty. Certified editions diverge—once, at 11:81, in their natural senses over what God commanded a prophet to do. The single text carries tensions its own scholars spent centuries smoothing. A book in that condition may still be loved, studied, and defended; what it cannot honestly be is proved divine by its flawless consistency. The showpiece argument claims more than the data will bear. And notice where the honest defender is left standing: not on “no contradictions” but on “no contradictions that cannot be harmonised”—a real position, held about many ancient books, but a different claim from the one the conversation began with. The retreat is quiet, and it is the concession.

And the Christian should say this without a trace of gloating, for the same test awaits us, and we welcome it: our Gospels too have been searched for contradiction by every generation of critics, and we argue their coherence case by case, not by slogan. If neither book can be established simply by pointing to its textual perfection, the question that remains is the one that always mattered: which message is true, and who has God shown Himself to be? We believe that question leads to Jesus—crucified, risen, and worth discussing with every Muslim neighbour with gentleness and respect (1 Peter 3:15). This essay is offered in that spirit: the opening of a fair conversation, not the winning of an unfair one.