Did Jesus die on the cross?

Summary

• The crucifixion of Jesus is one of the best-attested events in ancient history. It is reported in a creed that Paul says he received, which critical scholars date to within a few years of the event; by a hostile Roman senator; by a Jewish aristocrat writing for Romans; and by a professional debunker of religious frauds who thought Christianity was nonsense.
• Scholars who reject Christianity entirely still affirm the crucifixion. John Dominic Crossan, who denies the resurrection, calls it "as sure as anything historical can ever be."¹ Gerd Lüdemann, an atheist, writes that "Jesus' death as a consequence of crucifixion is indisputable."² Paula Fredriksen calls it "the single most solid fact about Jesus' life."³
• The Quran denies it in one verse, Quran 4:157, recited about six hundred years later. The key phrase is three Arabic words: wa-lakin shubbiha lahum. The verb shubbiha is an impersonal passive with no stated subject. It appears nowhere else in the Quran in that form.⁴
• The Quran never says a substitute died in Jesus' place. That story is in the commentaries, not the text. This matters, because most Christian apologetics attacks the substitution theory as though the Quran taught it. It does not. Todd Lawson, whose monograph is the standard treatment, puts it bluntly: "The point is that both John of Damascus and tafsīr, not the Qur'an, deny the crucifixion."⁵
• The commentators could not agree on what happened instead. Al-Tabari (d. 310/923) reports eleven traditions on this one verse, split between two irreconcilable accounts of what God is supposed to have done.⁶ Across the classical tradition the candidate who died in Jesus' place is a nameless volunteer disciple, or a disciple called Sergius, or the traitor Judas, or a Jewish guard also called Yahudha, or all seventeen disciples at once, or an assassin called Tityanus. These are not variants of one memory.
• A first-rank Muslim theologian said the substitution theory would destroy the basis of all prophecy. Fakhr al-Din al-Razi (d. 606/1209) argued that if God can put one man's face on another, no chain of historical testimony is safe — and that includes the chain that carries Muhammad's own prophethood.⁷
• A serious minority of scholars argues the Quran does not deny the crucifixion at all. Gabriel Said Reynolds argues in BSOAS that the verse is anti-Jewish polemic, not Christology: it denies a Jewish boast, not a Roman execution.⁶ This is a minority position, and Reynolds says so himself.
• The Study Quran, a major mainstream Muslim commentary in English, concedes most of the structure of that reading. It says the passage does not come "in the context of a critique of Christian belief," that the substitution reading is what the verse "is widely understood in the Islamic tradition" to mean rather than what it says, and that Jesus, not being divine, "must share in human mortality."⁸

Verdict:  Jesus of Nazareth was crucified under Pontius Pilate.  The evidence for that is early, multiple, and includes hostile witnesses;  no serious historian of any religion or none disputes it.  Quran 4:157 denies it six centuries later, without naming an alternative event, and its own commentators then produced a dozen contradictory reconstructions of what happened instead.  On ordinary historical method the first-century sources win.  But the popular Christian argument here is aimed at the wrong target:  the Quran does not teach the substitution theory, the commentators do, and the strongest reading of the Arabic — defended by Muslim and non-Muslim scholars alike — is that the verse denies a Jewish claim to have defeated God's messenger, not the fact of a Roman execution.

• Where this article's own case is weak. Five things must be said plainly. (1) The historical evidence establishes a death, not a resurrection; Crossan, Lüdemann, Fredriksen and Bart Ehrman all grant the crucifixion and none grants the resurrection, and quoting them proves less than apologetics usually claims. (2) If Reynolds and Lawson are right that the Quran does not deny the crucifixion, then the Quran is not refuted by the historical evidence — only the commentators are, and a Muslim can drop them. (3) The Quran nowhere denies that Jesus died; it uses the ordinary Arabic verb for God taking a life about Jesus in two places, so "the Quran says Jesus never died" is false as stated. (4) A Muslim who holds that the Quran is God's speech is not being irrational in preferring it to human testimony; he is being consistent, and the disagreement is about whether the Quran is God's speech, which this article does not settle. (5) Some non-Christian evidence commonly wheeled out in these debates — Mara bar Serapion, the Talmud's Sanhedrin 43a — proves nothing, and padding the list with it damages the case.

The claim being tested

The claim is not one proposition but four, and they are usually blurred.  Separating them is the whole job.

1. The textual claim: Quran 4:157 denies that Jesus was crucified.
2. The exegetical claim: someone else was substituted for Jesus on the cross and died in his place.
3. The Christological claim: Jesus did not die at all; he was taken up alive and will return.
4. The historical claim: the Christian and Roman records of a crucifixion under Pontius Pilate are mistaken.

Almost every popular treatment of this subject treats these as one claim.  They are not.  Claim 2 is not in the Quran — it is in the commentaries, and different commentators tell it differently.  Claim 3 is contradicted by the Quran's own vocabulary, as Muslim exegetes have noticed.  Claim 1 is disputed inside Quranic studies, by scholars who are not Christian apologists.  Only claim 4 is a straightforward historical question, and it is the one this article can settle.

Here is what the Quran actually contains on the subject.  Quran 4:157, in Saheeh International's rendering — the translation a reader meets first at quran.com, produced by Muslims and freely licensed:

And [for] their saying, "Indeed, we have killed the Messiah, Jesus the son of Mary, the messenger of Allāh."  And they did not kill him, nor did they crucify him;  but [another] was made to resemble him to them.  And indeed, those who differ over it are in doubt about it.  They have no knowledge of it except the following of assumption.  And they did not kill him, for certain.⁹

The words in square brackets are the translator's, not the Arabic.  The next verse:  "Rather, Allāh raised him to Himself.  And ever is Allāh Exalted in Might and Wise."¹⁰

That is the entire Quranic basis for the claim.  Two verses.  No named substitute, no description of the event, no alternative account of what the witnesses saw.

The strongest case for the Muslim claim

Stated at full strength, without evaluation.

First, the Quran is not making a historical guess.  A Muslim does not hold that Muhammad, in seventh-century Arabia, reasoned his way to a better account of a first-century execution.  He holds that God, who was present at the event, corrected a report that human beings had got wrong.  Testimony is testimony;  revelation is knowledge.  The verse itself makes exactly that distinction:  those who dispute the matter "have no knowledge of it except the following of assumption" — the Arabic word is zann, conjecture as opposed to certainty.  On this view the length of the gap is irrelevant.  An eyewitness who is mistaken is still mistaken a thousand years later.

Second, the Christian record is not as independent as it looks.  Al-Razi made this argument around the year 1200 and it has not aged badly.  Jesus, he says, was "a man little given to social intercourse, and thus known only to a few chosen companions."  Therefore "the Christian agreement in the transmission (of the crucifixion event) goes back to a few people whose agreement on a false report is not improbable."⁷  Four Gospels that share sources are not four witnesses.  Tacitus, writing around AD 116, most likely repeats what Christians said about themselves;  he did not consult a Roman docket, and the standard critical treatment notes that his calling Pilate a procurator rather than a praefectus "shows that he is not directly citing any official record."¹¹

Third, the Quran's denial fits a real problem in the Christian sources.  The Gospels themselves say the disciples fled (Mark 14:50).  The women watched "from a distance" (Mark 15:40).  There is no named Christian who claims to have stood at the foot of the cross and watched Jesus stop breathing, except in the Gospel of John, the latest of the four.  A skeptic who doubts John's eyewitness claims and a Muslim who doubts the Christian record are, on this narrow point, making the same move.

Fourth, God does not lose.  This is the theological core and it should not be caricatured.  The Quran's point in Quran 4:157 is that the enemies of a prophet do not get to write the ending.  The passage runs through a list of Israelite offences — the golden calf, breaking the covenant, disbelieving God's signs, murdering the prophets, slandering Mary — and the claim to have killed Jesus is the last item on that list.⁶  Read that way, the verse is not a forensic report about a Friday in Jerusalem.  It is a refusal to let the killers claim a victory.  Reynolds, who is not a Muslim, states the Quran's own reason for the denial as "the insistence that God is the one who gives life and death."¹²

Fifth, the substitution stories, whatever their problems, are not stupid.  Second-century Christians told a version of the same story.  Irenaeus reports that the Gnostic teacher Basilides taught that "Simon, a certain man of Cyrene, being compelled, bore the cross in his stead;  so that this latter being transfigured by him, that he might be thought to be Jesus, was crucified … while Jesus himself received the form of Simon, and, standing by, laughed at them."¹³  Whatever else it is, a substitution at the cross is not an idea invented in Arabia in the 620s.

What Quran 4:157 actually says

The Arabic of the crux, in the Uthmani text served by quran.com:

وَمَا قَتَلُوهُ وَمَا صَلَبُوهُ وَلَـٰكِن شُبِّهَ لَهُمْ

Transliterated:  wa-ma qataluhu wa-ma salabuhu wa-lakin shubbiha lahum.

Word by word, following the morphological analysis at corpus.quran.com:⁴

Two features of that table carry the whole argument.

The subject of "they killed" and "they crucified" is the Jewish opponents named in the previous clause, the ones who said "we have killed the Messiah."  The verb is third person plural, and its antecedent is the boasters.  Nothing in the sentence mentions Romans, a cross-detail, or a substitute.

The verb shubbiha is a passive with no stated subject at all.  It is what grammarians call an impersonal passive.  In English you can only render it with a dummy subject:  "it was made to appear."  The verse does not say who was made to resemble whom.  It does not say a man was substituted.  It says something was made ambiguous, to them.  The form is unique in the Quran:  this is the only place the root sh-b-h appears in form II anywhere in the book.⁶

Because the Arabic is genuinely underdetermined, the English translations split — and the split does not run between Muslims and Christians.  It runs straight through the Muslim translators.

Pickthall, Yusuf Ali, Maududi and Abdel Haleem — all Muslims, three of them among the most widely used translators in the English-speaking Muslim world — render the phrase as being about the event being made unclear, not about a person being swapped.  Saheeh International and Hilali-Khan insert the substitute.  Hilali-Khan puts the entire substitution theory inside square brackets that correspond to nothing in the Arabic.

Reynolds sorts the non-English translators the same way, and names them:  Blachère's French has "his double was substituted before their eyes";  Paret's German supplies "(another) appeared to them similar (so that they confused him with Jesus and killed him)";  Arberry has "only a likeness of that was shown to them."⁶  His comment on the difference between the two jobs is worth keeping:  "A translator has recourse to ambiguity.  A commentator, on the other hand, does not."⁶

That is the first result, and it is the one most Christian apologetics gets wrong.  The substitution theory is not a translation of Quran 4:157.  It is an interpretation of it, and about half the major Muslim translators decline to put it in the text.

What the grammar allows

The best statement of the grammatical problem is not modern, and not Christian.  It is by al-Zamakhshari (d. 538/1144), one of the great Arabic grammarians and the author of al-Kashshaf, still the reference commentary for the linguistic side of the Quran.

Zamakhshari asked the obvious question:  in the passive shubbiha, who or what is the understood subject?  Mahmoud Ayoub, a Muslim scholar of Islamic thought, summarises his reasoning from the Arabic:

He begins by asking to what the verb shubbiha refers.  If it is made to refer to Christ, then Christ is the one to whom something else is likened, not the one likened to something else.  The verb, however, cannot have as its subject the one killed, since he is never mentioned in the Qur'an.  Thus it must refer to the preposition "to" (them), that is, "they were made to imagine it."⁷

Read that slowly, because it is the hinge of the whole subject.  Jesus is the only person named in the sentence.  If Jesus is the subject of shubbiha, the verse says Jesus was made to look like someone else — which is the opposite of the substitution story, and useless to it.  The substitute cannot be the subject, because the substitute does not exist in the sentence;  he has to be imported from outside.  What is left is the pronoun in lahum, "to them" — and then the verse means the situation was made confusing to the people who claimed the kill.

Reynolds gives the same analysis of Zamakhshari's argument independently, from the Arabic edition.⁶  So does a modern mainstream commentary:  The Study Quran notes that "concerning him can also be read as 'concerning it,' meaning the entire affair of the crucifixion, a reading preferred by al-Zamakhshari."⁸

Al-Baydawi (d. 685/1286), Zamakhshari's most influential successor, repeats the same objections and then adds a possibility that neither Muslim nor Christian polemicists usually quote:  "or it may be that no one was killed;  rather his being killed was falsely claimed and spread among men."⁷

So the range of readings the Arabic permits, laid out honestly, is:

1. Someone else was made to look like Jesus and was killed in his place. This is the majority reading in the later tradition. It requires supplying a subject that is not in the sentence — Zamakhshari's objection.
2. Jesus was made to look like someone else. Grammatically the most natural reading if Jesus is the subject. It destroys the substitution story.
3. The event was made ambiguous to them. "They were made to imagine it"; "the matter was made dubious to them." This is what most Muslim translators actually print.
4. Nobody was killed at all and the killing was a false rumour. Al-Baydawi's aside.

Todd Lawson, whose 2009 monograph The Crucifixion and the Qur'an is the standard scholarly treatment of the verse's whole reception history, sorts every reading he found — Muslim, Christian and academic — into three families:

1  No one was crucified.  2  Jesus was crucified, but this happened only because God decided so;  it was not a result of the plotting of the Jews.  3  A person other than Jesus was crucified, this is the view most widely held in the contemporary Muslim world.⁵

Lawson's own judgement on the text is the sentence a Christian apologist should read before writing anything else on this subject:

The Qur'an only asserts that the Jews did not crucify Jesus;  which is obviously different from saying that Jesus was not crucified.  The point is that both John of Damascus and tafsīr, not the Qur'an, deny the crucifixion.⁵

Lawson adds a detail that is easy to miss and hard to explain away.  The earliest surviving reading of Quran 4:157 as a denial of the crucifixion is not by a Muslim.  It is by John of Damascus (d. 749), a Greek Christian civil servant in the Umayyad administration writing a polemic against Islam — "the earliest textual evidence for such an interpretation is not Muslim at all, rather it is from the pen of none other than the last great Church Father, John of Damascus."⁵  The first person on record to read the verse the way modern debates read it was a Christian arguing against Islam.

The commentaries: how many stories are there?

If a text is silent about what happened, and a community needs to know, the community will fill the silence.  That is what happened here, and the classical commentators did not hide it.

Al-Tabari (d. 310/923) wrote the foundational Sunni commentary, a compilation of earlier traditions with their chains of transmission.  Reynolds, working from the Beirut edition, summarises what Tabari found on this verse:

According to Abū Jaʿfar al-Ṭabarī the mufassirūn are in agreement that Jesus did not die … Their principal disagreement is only whether God cast the image of Jesus on a number of people, from whom the Jews chose one to crucify, or whether God cast the image of Jesus only on one specific person.  Ṭabarī reports two traditions which reflect the first opinion. … Ṭabarī then reports nine different traditions which reflect the second opinion.⁶

Eleven traditions, in one commentary, on one verse.  In one of them all seventeen disciples in the room receive Jesus' face at once and the Jews pick one.⁶

Muqatil ibn Sulayman (d. 150/767), the earliest complete commentary to survive, says the image of Jesus was cast on "a man who had struck him named Judas (yahūdhā)," who "was therefore justly punished."⁶  Lawson adds a correction that matters:  in Muqatil this Yahudha is a Jewish guard, "not the disciple Yāhūdhā (Judas), who other exegetes reported was substituted for Jesus."⁵  Two different men with the same name, in two different traditions, doing two different jobs.

Al-Zamakhshari (d. 538/1144) reports both the volunteering-disciple version and the punished-betrayer version, in that order, without reconciling them.  His account of what happened next is remarkable, because it describes the disciples themselves as having no idea what they had seen:

Some of [the disciples] said, "He is a god and it was not right to kill him".  Some of them said, "He was killed and crucified".  Some of them said, "If that was Jesus then where is our companion?" and "If that was our companion then where is Jesus?"  Some of them said, "He was raised to heaven".  Some of them said, "The face was the face of Jesus but the body the body of our companion".⁶

Ibn Kathir (d. 774/1373), whose commentary is the most widely read today, records five traditions.  In one, a young disciple volunteers, an opening appears in the ceiling, and Jesus is lifted through it while he sleeps.  In another, the same happens and then "they killed him and crucified him."  A third, from Wahb ibn Munabbih, follows the Gospel Passion closely — a last supper, Simon's threefold denial, thirty dirhams, the scourging, the spitting, the crown of thorns — until God raises Jesus at the last moment;  Ibn Kathir's own verdict on this one is that it is "a very unusual (gharīb) narrative."  A fourth, from Ibn Ishaq, has a cruel king named David, thirteen disciples, and one named Sergius who receives the image.  A fifth, again from Ibn Ishaq and attributed to "a Christian who became a Muslim," has Judas receive the image and be dragged off shouting "I am not the one you seek.  I am the one who pointed him out to you!"⁶

Ayoub records a fifth kind of story altogether, which he calls punishment substitutionism:  the man sent in to kill Jesus is himself transformed, and "the man's name is variously given as Tityanus, Titabus, or Titanus."  In that version "God, however, cast the likeness of Jesus only on the man's face and not on his body" — an explanation offered to account for why the verse says people "differed" about the matter.⁷

Put the candidates side by side:

The number of disciples in the room is given as "twelve, thirteen, or seventeen," depending on which report a commentator is following, and even the traitor's name comes through the tradition garbled — Ayoub lists "Yudas Zechariah, Yutah, or Butah" as forms of it in a single chain.⁷

These are not variant details in a single memory, the way four Gospels differ over the number of women at the tomb.  They are incompatible answers to a question the text does not answer.  The natural conclusion is not that Muslims are dishonest, which they were not — Tabari and Ibn Kathir printed the contradictions in full and let the reader see them.  The natural conclusion is Reynolds':  these narratives "appear to be a product of exegesis, and not its source."⁶

Reynolds turns this into an argument that actually works, unlike most arguments from silence:

If the Prophet Muḥammad announced to his companions that Jesus never died, but rather someone who was made miraculously to look like him died in his place, i.e. if he gave a historical account of the crucifixion which fundamentally contradicts that which Jews and Christians had been reporting for hundreds of years, then certainly such a revolutionary account — if any — would be well remembered and well preserved.  But, quite to the contrary, the reports of the mufassirūn are inconsistent and often contradictory.  They have all of the tell-tale signs of speculative exegesis.⁶

One more thing should be said about al-Tabari, because it cuts against the modern polemical use of this verse in both directions.  Because most of the accounts he collected say the illusion deceived not only the Jews but Jesus' own followers, al-Tabari concluded that no blame or charge of dishonesty can be laid on Christians who believe Jesus was crucified.⁸  The most influential of the Sunni commentators did not think Christians were lying about the cross.  He thought they had been fooled.

Al-Razi's objection: the substitution theory costs too much

The sharpest attack on the substitution theory in a thousand years of writing on this verse was made by a Muslim.  Fakhr al-Din al-Razi (d. 606/1209) was one of the most formidable theologians in Islamic history, and his al-Tafsir al-kabir is a work of philosophy as much as commentary.  He did not repeat the received view;  he tested it.

Ayoub summarises and then quotes his argument:

The second and more important question concerns what would happen if it is supposed that the likeness of one man could be cast on another.  Two problems would result.  First, it would open the gate of sophistry so that no social norm such as marriage or ownership rights could be ascertained.  Further, this would lead to doubt in historical testimony, that is, the ongoing transmission of historical reports (tawātur). … "In sum, the opening of such a gate necessitates doubt in tawātur, and this in turn necessitates doubt in fundamentals (uṣūl), and this in turn necessitates doubt in the prophethood of all prophets.  This is a branch (farʿ) necessitating doubt in fundamentals and must therefore be rejected."⁷

Follow the reasoning.  Tawatur is the technical term for a report transmitted by so many independent people that collusion in a lie is impossible.  It is the foundation of Islamic epistemology:  it is why a Muslim can be certain the Quran is the text Muhammad recited, and why he can be certain Muhammad existed and did what the sources say.  If God can make one man look exactly like another, then any mass of eyewitnesses can be mass-deceived, and tawatur stops guaranteeing anything.  Al-Razi saw that the price of explaining away the crucifixion was the certainty of Islam's own historical foundations, and he refused to pay it.

The Study Quran states the same objection independently, without citing Ayoub:

A notable exception to this is al-Rāzī, who finds the idea that God would perpetrate such a deception in the physical realm, particularly as it relates to individual identity, unacceptable.  If we cannot rely on our senses to identify individuals, then the proper application of Islamic Law, which is dependent upon physical witnessing and upon the certainty of people's identities in matters of marriage and so forth, would be called into doubt.⁸

Al-Razi's own alternative was not the crucifixion.  He proposed that when Jesus was taken up, the Jews seized some man and killed him and claimed it was Jesus, because Jesus was "a man little given to social intercourse, and thus known only to a few chosen companions."  That fallback is a real historical argument, and it is answered later in this article.⁷

The point to hold onto is narrower and stronger.  The best-known criticism of the substitution theory was written by a Sunni theologian, in Arabic, around the year 1200, and it says the theory would undermine the evidential basis of prophecy itself.  Christians who treat the substitution theory as "what Islam teaches" are ignoring a critic Muslims themselves rank among their greatest.

Reynolds: the verse is polemic, not Christology

Gabriel Said Reynolds is a historian of the Quran at Notre Dame.  In 2009 he published "The Muslim Jesus:  Dead or Alive?" in the Bulletin of the School of Oriental and African Studies, arguing that the Quran does not deny that Jesus died — and in fact assumes he did.  His abstract states the position and its minority status in the same breath:

According to most classical Muslim commentators the Quran teaches that Jesus did not die. … Most critical scholars accept that this is indeed the Quran's teaching, even if the Quran states explicitly only that the Jews did not kill Jesus.  In the present paper I contend that the Quran rather accepts that Jesus died, and indeed alludes to his role as a witness against his murderers in the apocalypse.⁶

His case rests on the Quran's own vocabulary, not on Christian sources.

The Quran has the infant Jesus predict his own death.  In Quran 19:33 the newborn Jesus says, in Saheeh International, "And peace is on me the day I was born and the day I will die and the day I am raised alive."¹⁴  Eighteen verses earlier the identical formula is used of John the Baptist:  "And peace be upon him the day he was born and the day he dies and the day he is raised alive."¹⁵  Nobody argues that John's death is still in the future.  Reynolds notes Geoffrey Parrinder's point that "there is no futurity in the grammar of the Qur'ān to suggest a post-millennial death."⁶

The Quran uses the ordinary verb for dying about Jesus.  The verb is tawaffa, which in Arabic means God taking a soul.  Reynolds counted:

Tawaffā appears in twenty-five passages in the Quran, and twice in relation to Jesus (here and Q 3.55).  For twenty-three of those passages the Muslim commentators generally follow the standard definition of this term, namely God's act of separating the soul from the body, or making someone die.⁶

His conclusion about the other two:

there can be hardly any other reason to argue that while tawaffā refers to death on twenty-three occasions in the Quran, on the two occasions on which it is applied to Jesus it refers either to falling asleep or being taken to heaven.⁶

The two occasions are Quran 3:55, where God says inni mutawaffika wa-rafiʿuka ilayya, and Quran 5:117, where Jesus says falamma tawaffaytani — "when You took me up," or, using the same verb's ordinary sense, "when You caused me to die."¹⁶  Abdel Haleem renders Quran 5:117 "Ever since You took my soul."¹⁷  And Quran 3:55 gives the order death-then-ascent, not ascent-instead-of-death:  "According to this sequence, God indeed raised Jesus to heaven, but only after He first caused him to die."⁶

The passage is about the Jews, not about Christians.  Reynolds reads Quran 4:157 in the six-item list of covenant offences that runs from Quran 4:153:  the golden calf, breaking the covenant, disbelieving God's signs, murdering the prophets, slandering Mary, and claiming to have killed Jesus.⁶  On that reading the denial is not a correction of Christian history.  It is a refusal of a boast:

in this passage the Quran is in no way denying Jesus' death.  Instead, the Quran is using the Crucifixion as one example of Israelite infidelity.⁶

the Quran not only leaves open the possibility that Jesus died on the Cross, it uses his death on the Cross as a paradigmatic example of Jewish infidelity.⁶

In his 2018 commentary on the whole Quran he says it in one sentence:  "In fact the verse does not deny that Jesus was crucified or that he died — it only denies that the Jews killed him."¹²

Do not overstate this.  Reynolds is not a Christian apologist and he is not conceding the Christian case.  He is a historian arguing that the Quran is not what the commentators made of it.  He never says the Quran affirms a redemptive crucifixion, and he says the Quran's reason for the denial is "quite particular to the Qurʾān:  the insistence that God is the one who gives life and death."¹²  He also states in his own abstract that most critical scholars read the verse the traditional way.  His position is a serious minority argument, not the consensus.

Two Western scholars of Islam reached a related conclusion earlier.  William Montgomery Watt argued that a Christian could accept the Quran's statement "since the crucifixion was the work of Roman soldiers;  and it is also true in a deeper sense, since the crucifixion was not a victory for the Jews in view of his resurrection."⁶  Kenneth Cragg argued that the emphasis of the passage falls on the human instinct to think one can outwit God, so that shubbiha lahum means the event, not the figure, was made to appear otherwise.⁶

Muslims who read the verse differently

The substitution theory is the majority view.  It is not the only one, and it never has been.

The Isma'ilis affirmed the crucifixion.  Lawson notes that "the Ismaʿili scholars of the 10th and 11th century saw perfect harmony between this Qurʾanic verse and the Gospels," reading the denial in terms of Jesus' words about those who can kill the body but not the soul.⁵  Martin Whittingham's 2021 study names the individuals — Abu Hatim al-Razi (d. c. 322/933–934), Abu Ya'qub al-Sijistani, Ja'far b. Mansur al-Yaman, and the Brethren of Purity — and, on Whittingham's account, they affirmed the crucifixion because the Gospels stated it so clearly, and read the Quranic denial as saying that Jesus did not die in spirit.  This is at second hand and should be treated as such.  Whittingham's book could not be obtained, and the summary of his page 106 comes from Saqib Hussain's review of it in the Review of Qurʾanic Research — in Hussain's words, not Whittingham's.¹⁸  What Hussain does quote from Whittingham verbatim is the reason:  "Ismāʿīlī thought stresses the underlying unity of all revelations, Qurʾanic and pre-Qurʾanic."¹⁸

Al-Hallaj recited the verse from his own gibbet.  Lawson reports, on al-Ghazali's authority, that the mystic Mansur ibn al-Hallaj (d. 923), while being crucified in Baghdad, quoted this very verse — "They did not kill him and they did not crucify him, it only appeared so to them."⁵  He was manifestly being crucified as he said it.  He cannot have meant that no crucifixion was occurring.  He meant that what the executioners thought they were destroying was not the real thing.

Mahmoud Ayoub rejected the substitution theory outright.  Ayoub (1935–2021) was a Lebanese Muslim scholar of Islamic thought who taught for decades in American universities.  His essay "Towards an Islamic Christology II:  The Death of Jesus, Reality or Delusion?" appeared in The Muslim World in 1980 and was reprinted verbatim in 2007;  page numbers here are the 2007 reprint.  His objections are moral and theological, not historical.

First, he denies that Islam has any room for the docetic instinct that produced substitution stories in the first place:

Islam, however, does not admit of docetism in any form. … neither the Qur'an nor later Islamic tradition suggests a phantomlike appearance of Christ.  He was, rather, a man, born in the usual way (Q. 19:22, 23), lived like other men, and like them must die and be resurrected for the final reckoning (Q. 19:33).⁷

Second, he narrows what the verse denies:  "It must therefore be argued that the Qur'an only denies the death of Jesus on the cross and leaves open the question of his actual death."⁷

Third, and most sharply:

The substitutionist theory will not do, regardless of its form or purpose.  First, it makes a mockery of divine justice and the primordial covenant of God with humanity … Would it be in consonance with God's covenant, his mercy and justice, to deceive humanity for so many centuries?⁷

And his verdict on the commentators:

Why then, it must be asked, does the Qur'an deny the crucifixion of Christ in the face of apparently overwhelming evidence?  Muslim commentators have not been able convincingly to disprove the crucifixion.  Rather, they have compounded the problem by adding the conclusion of their substitutionist theories.⁷

His own reading is that the verse is not a historical statement at all:  "The words wa lakin shubbiha lahum do not disclose, therefore, a long-hidden secret of divine deception;  rather, they constitute an accusation or judgment against the human sin of pride and ignorance."⁷

Be honest about what Ayoub is doing.  He is not conceding the cross.  He says the statement "belongs not to history but to theology in the broadest sense," and he declines to read it as a factual claim about a Friday in Jerusalem.⁷  A Christian who quotes Ayoub as an admission has misread him.  What Ayoub does establish is that a respected Muslim scholar found the substitution theory unworthy of God, and said so in print, in a mainstream academic journal.

The Ahmadiyya swoon theory.  Mirza Ghulam Ahmad (1835–1908), the founder of the Ahmadiyya movement, took the one position on this list that is testable by ordinary historical and medical evidence, because it concedes that Jesus was fastened to a cross.  In Jesus in India, written in Urdu in 1899 and published in 1908, he states the thesis in the first pages:

I shall prove in this book that Jesus, the Messiah, may peace be on him, did not die on the cross, nor did he go up to the heavens;  nor should the hope be entertained that he would ever again come down to the earth from the heavens.  On the contrary, the fact of the matter is that he died at the age of 120 years¹⁹

— in Kashmir, having survived the cross and travelled east.  He reads the sign of Jonah as a promise of exactly this:  Jesus "would not die on the cross, nor would life depart from him on the accursed wood;  on the contrary, like the prophet Jonah, he would only fall into a state of swoon."¹⁹  On the verse itself he writes that "though it is true that Jesus, the Messiah, was apparently placed on the cross, and that the intention was to kill him;  yet it was merely a delusion for the Jews and Christians to suppose that Jesus, the Messiah, may peace be on him, had actually died on the cross."¹⁹

Two things must be said about this reading, and they pull in opposite directions.  It is the only Muslim reading of Quran 4:157 that can be argued about with evidence rather than grammar, because it agrees with the Gospels, Tacitus and Josephus that a crucifixion took place.  And it is a minority position within a minority:  mainstream Sunni and Shi'a Islam reject Ahmadiyya teaching, and this teaching in particular is among the reasons.⁶  Treating it as "the Muslim view" would be a misrepresentation, and the Ahmadiyya community is persecuted in several Muslim-majority countries.  It is included here because it is a real reading with a real primary text, not because it is representative.

Note also, for anyone checking references:  Ahmadiyya editions count the opening formula of each surah as verse 1, so what everyone else calls Quran 4:157 they number 4:158.  A mismatch of one in an Ahmadi citation is not an error.

What The Study Quran says

The Study Quran (HarperOne, 2015), edited by Seyyed Hossein Nasr with a team of Muslim scholars, is the standard one-volume English commentary for educated Muslim and academic readers.  It is not a fringe work and not an apologetic one.  Its note on Quran 4:157 concedes more of the structure of the argument above than most Christian readers expect.

On what the verse is aimed at:

Although the Quran addresses and criticizes the Christian belief in the Trinity (v. 171;  5:73) and the divinity of Jesus (5:17, 72, 116;  9:31), this is the only passage that addresses the belief in Christ's crucifixion and death.  It does not come in the context of a critique of Christian belief, however, but rather as part of a lengthy passage criticizing historical incidents of Jewish unfaithfulness to their covenant.  It is important to note that here the critique is not aimed directly at the belief in Christ's crucifixion and death, but rather at the Jews' claim to have killed him.⁸

On the distance between what the verse says and what the tradition made of it:

Although the verse only directly criticizes a Jewish claim to have killed and crucified Jesus, it is widely understood in the Islamic tradition as meaning that Jesus was not crucified or killed at all;  it only appeared so unto them, that is, to the Jews as well as to most of Jesus' followers.⁸

On the grammar, it reports the alternative readings without dismissing them, including the possibility that the last clause means "they did not kill [their doubt] about the matter" — that is, that the doubters remained in doubt.⁸

And on Quran 4:158 it concedes the mortality point in plain words:  "The Quran is clear elsewhere, however, that Jesus is not Divine (5:116).  Logically, then, he must share in human mortality.  Islamic tradition holds Jesus will return near the end times to fight the Antichrist (al-Dajjāl) and will thereafter eventually die and be resurrected with the rest of humanity."⁸

Three concessions, from an establishment Muslim commentary, in print:  the verse's target is a Jewish boast rather than Christian belief;  the substitution reading is what the tradition understands rather than what the verse states;  and Jesus dies eventually in any case.  That is the single most useful "from their own side" citation on this subject, and it should be quoted rather than argued around.

The historical evidence that Jesus was crucified

Everything above is about what the Quran says.  This section is about what happened.  It is deliberately built out of non-Christian and anti-Christian material wherever possible, because a reader who does not accept the Gospels needs a case that does not depend on them.  The same sources, examined at greater length, carry the separate question of whether Jesus existed at all, which is treated in the article on Jesus mythicism.

The earliest source is a creed, not a Gospel

The oldest surviving statement that Jesus died is not in any Gospel.  It is a formula Paul quotes in 1 Corinthians, written about AD 54, and introduces with the technical language of handing on received tradition:  "For I delivered to you as of first importance what I also received:  that Christ died for our sins in accordance with the scriptures, that he was buried, that he was raised on the third day."²⁰

Critical scholars date the formula itself far earlier than the letter, on internal grounds:  its vocabulary is not Paul's ("for our sins" in the plural, "according to the scriptures," "the twelve" where Paul elsewhere says "the apostles"), and its structure is Semitic, so Joachim Jeremias concluded that "the core of the kerygma was not formulated by Paul, but comes from the Aramaic-speaking earliest community."²¹  How early?  Gerd Lüdemann, an atheist, put it inside two years:  "We can assume that all the elements in the tradition are to be dated to the first two years after the crucifixion of Jesus."²²  Michael Goulder, who also rejected the resurrection, dated Paul's receipt of it to "a couple of years after the crucifixion."²³  Richard Carrier, the best-known living defender of the view that Jesus never existed at all, agrees on the dating:  the creed "definitely existed and was the central 'gospel' Christians were preaching in the early 30s A.D."²⁴

Paul is also a hostile witness turned insider.  He persecuted the movement before joining it, and he says he spent fifteen days in Jerusalem with Peter and met "James the Lord's brother."²⁵  Whatever else is uncertain, the claim that Jesus was executed was being made by people who had known him, within a few years, in the city where it happened.

The non-Christian sources

Josephus, a Jewish aristocrat writing for a Roman audience about AD 93, mentions Jesus twice.  The longer passage, the Testimonium Flavianum, says Pilate condemned him to the cross;  it has been touched up by a Christian hand, and the debate is over how much.²⁶  Alice Whealey's study of the Syriac and Arabic witnesses concludes that "the only major alteration that has been made to Josephus' original passage about Jesus is the alteration of the phrase 'he was thought to be the Messiah' to the textus receptus phrase 'he was the Messiah'."²⁷  The shorter passage, about the execution of "the brother of Jesus, who was called Christ, whose name was James," is accepted as authentic by nearly everyone, because its tone is offhand and its wording is not Christian usage.²⁶

Tacitus, a Roman senator and consul who despised Christianity, wrote about AD 116 that "Christus, from whom the name had its origin, suffered the extreme penalty during the reign of Tiberius at the hands of one of our procurators, Pontius Pilatus, and a most mischievous superstition, thus checked for the moment, again broke out."²⁸  John Meier's assessment is that "despite some feeble attempts to show that this text is a Christian interpolation in Tacitus, the passage is obviously genuine. … such a short and dismissive description of Jesus hardly comes from a Christian hand."¹¹  Even Brent Shaw, the most aggressive modern skeptic of the Neronian persecution story that surrounds this sentence, grants that "the passage is probably genuine Tacitus."²⁹

Lucian of Samosata, a satirist who made a career out of exposing religious frauds, wrote about AD 165 that Christians "still worship, the man who was crucified in Palestine because he introduced this new cult into the world."³⁰  Lucian thinks Christians are gullible idiots and says so at length.  He does not think the execution is in doubt.  That is what a hostile witness looks like:  he concedes the fact he has no motive to concede, and mocks everything built on it.

What should not be used.  Two items appear in nearly every apologetics list and should be dropped from this one.  Mara bar Serapion's letter about the Jews killing their "wise king" never names Jesus and is dated anywhere from AD 73 to the third century.  The Babylonian Talmud's line at Sanhedrin 43a about hanging "Yeshu" on the eve of Passover comes from a compilation finished centuries later and is probably reacting to Christian claims rather than preserving independent memory.  Neither adds anything, and padding a list with weak items invites a fair reader to discount the strong ones.

Meier's summary of what is left is the honest one:  "With Josephus and Tacitus we exhaust the early independent witnesses to Jesus' existence, ministry, death, and ongoing influence."¹¹  Two independent non-Christian witnesses is not a hundred.  It is also more than survives for most events of the period, and both of them report the execution.

What scholars hostile to Christianity say

The strongest form of this argument for a skeptical reader is not a Christian scholar's verdict.  It is the verdict of scholars who reject the resurrection, reject the Gospels' reliability, and in some cases reject God.

John Dominic Crossan co-founded the Jesus Seminar, thinks most of the Passion narrative is prophecy read backwards into history, and does not believe Jesus rose.  In the same paragraph in which he argues that the Passion story was largely constructed, he writes:

That he was crucified is as sure as anything historical can ever be, since both Josephus and Tacitus, in texts to be seen in the final chapter, agree with the Christian accounts on at least that basic fact.¹

Gerd Lüdemann was an atheist New Testament scholar who wrote a book arguing the resurrection did not happen.  Under the heading "Historical Facts behind 1 Cor. 15:3–8," his first item reads:

Jesus' death as a consequence of crucifixion is indisputable.²

Paula Fredriksen, a historian of ancient Christianity who converted to Judaism, puts this first among the facts she builds her account of Jesus on:

The single most solid fact about Jesus' life is his death:  he was executed by the Roman prefect Pilate, on or around Passover, in the manner Rome reserved particularly for political insurrectionists, namely, crucifixion.³

E. P. Sanders, whose work reshaped the field's understanding of Jesus and Judaism, opens his outline of Jesus' life with a list of statements about Jesus that he says "are almost beyond dispute;  and they belong to the framework of his life, and especially of his public career."  One item on that list is "he was executed on the orders of the Roman prefect, Pontius Pilate."³¹

Bart Ehrman, who describes himself as an agnostic with atheist leanings and has spent a career arguing that the New Testament is unreliable, writes that "historians have almost universally agreed that whatever else one might say about him, Jesus of Nazareth lived in first-century Palestine and was crucified by the prefect of Judea."³²

Read these carefully and do not oversell them.  Not one of these five believes Jesus rose from the dead.  Crossan thinks the Passion narrative is largely invented;  Lüdemann thinks the appearances were visions;  Ehrman thinks the Gospels contradict each other on many points.  What they agree on is the narrow fact of the execution — and that is exactly the point.  The crucifixion is not a Christian doctrine that scholars accept out of deference.  It is a datum that survives even after the doctrine has been stripped away.

There is one more argument, and it is Ehrman's, not a Christian's.  A crucified messiah was not a story anyone in first-century Judaism wanted to tell.  Crucifixion was the punishment Rome used to humiliate rebels;  a messiah was supposed to defeat Rome.  Ehrman's version, in the book:  "Who would make up the idea of a crucified messiah?  No Jew that we know of."³²  In the newspaper piece he put it at more length:  "prior to Christianity, there were no Jews at all, of any kind whatsoever, who thought that there would be a future crucified messiah. … The Christians did not invent Jesus.  They invented the idea that the messiah had to be crucified."³³  Groups do not fabricate the single fact that most damages their own case.

The chronology argument, stated plainly

Strip away the technical detail and the shape of the disagreement is simple.

A first-century event is reported by first-century sources.  One of them is a creed being recited within a few years, in the city where it happened, by people who had known the man.  Two of them are non-Christians with no reason to help — a Roman senator and a Jewish historian — writing within about eighty years.  A fourth is a professional satirist who thought the whole religion was a fraud and still took the execution for granted.  None of them disputes the death.  The disputes in the ancient world were about who was to blame and what it meant, never about whether it happened.

Against that stands one sentence in a text recited in Arabia in the 610s to 630s.  It is about six hundred years later.  It names no alternative event.  It does not say Jesus was rescued, or hidden, or that a substitute was used.  It says the boasters did not kill him and that something was made to appear to them.  When Muslims later asked what did happen, the answer had to be constructed, and the construction produced at least eleven incompatible stories in al-Tabari alone.

Ordinary historical method gives the earlier sources the presumption.  Not because early sources are always right — they are often wrong — but because a late source that contradicts early ones has to explain how it knows better, and this one does not attempt to.  The Quran does not say it is correcting a record on the basis of a preserved older report, or a document, or a surviving eyewitness line.  It simply asserts.

A Muslim answer to this is available and it is not stupid.  It runs:  the Quran is God's own speech.  God was there.  Human testimony, however early and however multiple, is zann — conjecture — and the verse says so in as many words.  If the earliest sources for the Battle of Hastings all said the arrow missed Harold's eye, and God said otherwise, God would be right.  The gap of six hundred years is irrelevant to an omniscient speaker.

The reply is a question, and it should be asked without heat:  what evidence could then ever count?

If early, multiple, independent, hostile testimony cannot establish a public execution in a Roman province, then no historical claim can be established by evidence at all.  That includes the historical claims Islam itself rests on.  A Muslim believes on evidence that Muhammad lived, that he recited the Quran, that the Companions transmitted it, and that the text in his hand is the one Uthman had copied.  He believes all of that on exactly the kind of testimony he is being asked to discount here — earlier, more numerous, and better-attested testimony, in the crucifixion's case, than exists for most events in Muhammad's biography.

This is al-Razi's point arriving from the other direction, and he saw it eight hundred years ago.  If God will overturn the united testimony of witnesses about a public event, then tawatur — the mass-transmission that guarantees the Quran's own text — "necessitates doubt in the prophethood of all prophets."  A method that discards the crucifixion discards too much.

The honest statement of the disagreement is therefore this.  The Muslim position on Quran 4:157 is not irrational;  it is derivative.  It follows from the prior claim that the Quran is God's speech.  Anyone who grants that prior claim will accept the verse whatever the historians say, and no argument in this article will move him.  Anyone who does not grant it is left with a seventh-century assertion against sixth-decade evidence, and the evidence wins easily.  So the crucifixion is not really the point at issue.  The point at issue is whether the Quran is what it claims to be — which the other articles in this series examine.

The strongest Muslim responses

"The Quran is revelation. Historical testimony is conjecture."

This is the serious answer, and it is what the verse itself says:  those who dispute the matter "have no knowledge of it except the following of assumption."  It is not a dodge;  it is a claim about sources of knowledge.

The reply is not that revelation cannot outrank testimony.  If God speaks, God is right.  The reply is that the argument is circular in this context:  it presupposes that the Quran is God's speech, which is the proposition under examination.  Used as a response to historical evidence, it converts the Quran into a text that cannot be checked against anything — and a text that cannot be checked cannot be confirmed either.  The Muslim tradition itself has never accepted that position generally;  it argues for Muhammad's prophethood from evidence, from tawatur, from the Quran's literary quality, from fulfilled prediction.  It cannot both appeal to evidence and exempt itself from it.

Al-Razi's argument that Christian testimony traces back to very few people

This is a genuine historical argument and deserves a genuine reply.  Al-Razi held that Jesus was known to few, so the whole Christian tradition could go back to a small group whose agreement on a false report "is not improbable."⁷

Three things are wrong with it as history.  First, the tradition does not trace back to one group.  It is attested by Paul, who was an enemy of the movement before he joined it and who names the people he checked with;  by four Gospels with at least partly independent Passion material;  by the letter to the Hebrews, by 1 Peter, by Revelation;  and by Josephus and Tacitus, who were not Christians and had no interest in the group's internal story.  Second, al-Razi's premise is false:  the execution of Jesus was a public Roman punishment carried out in Jerusalem at Passover, when the city was full, which is the opposite of a private event known to a few.  Third, the Jewish opponents of Christianity in antiquity never used this argument.  They disputed the charges, the claims, and the resurrection — never the execution.  A rumour invented by a handful of people would have been easy to deny in the city where it was said to have happened, and nobody denied it.

"The verse only denies that the Jews killed him, and Christians can agree with that"

This is Watt's and Cragg's reading, and Reynolds' too.  Watt argued a Christian could accept the Quran's statement "since the crucifixion was the work of Roman soldiers;  and it is also true in a deeper sense, since the crucifixion was not a victory for the Jews in view of his resurrection."⁶

A Christian can accept this response without giving anything up, and should say so.  The New Testament agrees that Jesus was not defeated by his enemies:  it says he laid down his life, that no one took it from him, and that the powers that crucified him did so in ignorance.  If Quran 4:157 means that the killers did not win, there is no conflict.  The conflict only appears when the verse is read as the commentators read it.  That is a large concession from a Christian side, and it is one the evidence supports.

"Jesus was crucified but did not die"

This is the Ahmadiyya position, and it is listed here for completeness rather than because Muslims generally offer it — mainstream Sunni and Shi'a Islam reject Ahmadiyya teaching, so a Muslim apologist would not use it.  It is the only reading here that makes a testable historical claim, and it fails on the same two grounds as the substitution theory plus one more.

It is late:  the case was first argued at length in 1899, more than eighteen centuries after the event, using no source that predates the Gospels.  It has no ancient support:  no first-, second-, or third-century writer, Christian, Jewish, pagan or Gnostic, reports that Jesus survived the cross — the Gnostics who denied the death denied that Jesus was on the cross at all, which is the opposite claim.  And it must explain away the specific detail in John 19:34 that a soldier pierced Jesus' side with a spear to confirm death, and the detail in Mark 15:44–45 that Pilate refused to release the body until the centurion certified the death.  A defender can call both details Christian invention, but then the Ahmadiyya reading is resting on the Gospels for the crucifixion and rejecting them for the death, which is not a stable position.

"You are attacking a doctrine the Quran does not teach"

This objection is correct, and it should be conceded before a Muslim has to make it.

Most Christian material on this subject attacks "the Islamic substitution doctrine" as though Quran 4:157 stated it.  It does not.  The verse contains no substitute.  Roughly half of the standard English translations decline to insert one.  Zamakhshari, the greatest grammarian among the commentators, argued the Arabic will not bear it.  Al-Razi rejected it on theological grounds.  Ayoub rejected it outright.  The Isma'ilis affirmed the crucifixion.  The Study Quran describes the substitution reading as what the tradition understands the verse to mean, not as what it says.

An argument aimed at a doctrine the text does not contain is aimed at the wrong target, and a Muslim reader who spots that will discount everything else the arguer says.  The defensible claim is narrower and harder to answer:  the majority tradition of Quranic interpretation denies the crucifixion, and that tradition contradicts the historical evidence, and the verse it depends on is too underdetermined to support any of the dozen accounts built on it.

Common objections and responses

"The Gospel of Barnabas says Judas was crucified in Jesus' place"

This text is still circulated as independent confirmation of the Quran.  It is not, and the reason has nothing to do with Christian bias.

The Gospel of Barnabas is known from one Italian manuscript in Vienna, first published with an English translation by Lonsdale and Laura Ragg at the Clarendon Press in 1907.³⁴  The Raggs examined the physical object.  The paper is not oriental;  its watermark, an anchor within a circle, was identified by the paper historian C.-M. Briquet as "distinctively Italian," in a form belonging "to the second half of the sixteenth century."³⁴  Their conclusion about the date was blunt:  "The true date lies, we imagine, nearer to the sixteenth century than to the first," and the alternatives they weighed were an earlier Italian archetype or "a deliberate forgery of the latter half of the sixteenth century."³⁴

Its use in Muslim exegesis is recent and traceable.  Todd Lawson records that Rashid Rida (d. 1935) "was the first exegete to rely upon the Gospel of Barnabas," and points out the problem this creates:  "Riḍā condemns the Christian canon as unreliable but is able to accept the apocryphal Barnabas at face value."⁵  A text cannot be used to prove the New Testament unreliable if the only reason for trusting it is that it agrees with a conclusion already held.

"The disciples fled, so no follower actually watched Jesus die"

This is the best of the popular objections, and it is half right.  Mark says the disciples "all forsook him and fled" (Mark 14:50), and that the women watched "from a distance" (Mark 15:40).

But it does not do the work required.  A crucifixion was a public Roman spectacle conducted precisely so that people would see it.  The question is not whether a particular follower had a clear view.  It is whether Jesus was executed at all — and that is attested by the enemies of the movement, not only by its members.  Josephus and Tacitus were not standing at Golgotha either, and neither had any reason to invent a Roman execution of a Galilean.  A movement can be wrong about what it saw;  it cannot easily be wrong about whether its founder was killed, in a city full of Passover pilgrims, by an occupying army that kept records and posted charges.

"The Gospels contradict each other about the crucifixion, so they establish nothing"

They do differ:  on the wording of the charge, the hours, the last words, the number of women, the reaction of the thieves.  Any honest treatment says so.

Differences of that kind are what independent reports of a chaotic public event look like, and they are the reason historians treat the shared core as strong rather than weak.  All four report the same core:  arrest, a hearing before the Jewish authorities, delivery to Pilate, crucifixion, death, burial.  And the argument in this article does not rest on the Gospels anyway.  Strike all four out and the case still stands on Paul's creed, Josephus, Tacitus and Lucian.  Crossan, who thinks large parts of the Passion narrative were composed out of Old Testament texts rather than memory, still calls the crucifixion "as sure as anything historical can ever be."

"The Bible is corrupted, so its account of the crucifixion proves nothing"

This is the tahrif argument, and it is a different subject with its own evidence — the manuscripts, the Quran's own statements about the Torah and Gospel, and the history of the charge in Muslim writing.  It is examined separately in the article on whether the Bible is corrupted.

Two points belong here, though.  First, the corruption charge cannot rescue the substitution theory, because the substitution theory is not in the Quran either;  showing that the Gospels are unreliable would leave the Quran still silent about what happened.  Second, corruption arguments do not touch the crucifixion specifically.  The death of Jesus is in the creed Paul quotes, in every Gospel, and in two non-Christian authors whose texts nobody alleges were tampered with for this purpose.  Textual variants are variations in words and phrases, and the state of that evidence is examined in the article on New Testament manuscripts.  A corruption argument that could remove the crucifixion would have to name a manuscript, a reading, and a date, and no Muslim writer has produced one.

"Islam simply borrowed the substitution story from Gnostic Christians"

Something like the substitution story really does exist in second-century Christian heresy.  Irenaeus reports that Basilides taught that Simon of Cyrene was crucified in Jesus' place, "so that this latter being transfigured by him, that he might be thought to be Jesus, was crucified … while Jesus himself received the form of Simon, and, standing by, laughed at them."¹³

But this should not be turned into a genealogy, and the scholar who supplies the quotation says so himself.  Samir Khalil Samir's own caution:  "This does not mean that the Gnostic theory was current among the Christians of Arabia, but simply that the idea of a substitute for Jesus on the Cross was present."¹³  Ayoub's objection is also fair:  Islam rejects docetism in every form, so if Muslims took up a substitution story they took it stripped of the theology that generated it.  Basilides substituted a man because he thought divinity could not suffer.  Islam has no such problem, because it does not think Jesus is divine.  The most that can be said is that the idea was in circulation.  That is worth saying, and no more.

"The sign of Jonah proves Jesus was not meant to die — Jonah did not die in the fish"

This is Mirza Ghulam Ahmad's argument:  the parallel to Jonah shows Jesus "would only fall into a state of swoon."¹⁹

The parallel Matthew draws is stated in the text and it is about location, not survival:  "as Jonah was three days and three nights in the belly of the great fish, so will the Son of Man be three days and three nights in the heart of the earth" (Matthew 12:40).  Being in the heart of the earth is a description of burial.  In the very next chapter of the same Gospel Jesus tells the disciples plainly that he "must go to Jerusalem and suffer many things … and be killed, and on the third day be raised" (Matthew 16:21).  An argument that reads a simile in one verse against an explicit statement three verses later has to explain why the simile wins.

"Christian sources are biased because Christianity needs the crucifixion"

True, and it is why this article leans on the sources that do not need it.  Tacitus needed nothing from Christianity but a sneer.  Lucian was in the business of exposing religious con men.  Josephus was writing for the Flavians.  Crossan, Lüdemann, Fredriksen, Sanders and Ehrman have careers built on questioning Christian claims, and all five affirm the execution.  Bias arguments cut in the direction of whoever has something to gain.  On this particular fact, nobody with something to gain is on the affirming side.

What evidence would change the conclusion?

What would strengthen the Muslim claim.

• A pre-Islamic source, of any religion, reporting that Jesus was not crucified or that another man was killed in his place, with a date and a provenance. The Gnostic substitution stories are the closest thing that exists, and they are second-century Christian heresy, not independent testimony about the event.
• Evidence that the substitution story circulated among Muhammad's contemporaries as a received account rather than appearing later in the commentaries. Reynolds' argument is precisely that the opposite is true — that the tafsir accounts have "all of the tell-tale signs of speculative exegesis." A demonstrably early, single, consistent Muslim account of what happened instead would undercut that.
• A first- or second-century document, Christian or not, in which Jesus is known to have survived. This would rescue the Ahmadiyya reading, which is otherwise unsupported by any ancient source.
• Serious evidence that the Testimonium Flavianum is a wholesale Christian forgery and that the Tacitus passage is interpolated. Both have been argued; neither is the mainstream view, and the strongest recent studies argue the other way.

What would weaken it further.

• Documentary evidence from Roman Judea naming the execution. None is expected; provincial records of this kind do not survive.
• Earlier manuscript evidence for the Pauline creed material. This is incremental; the dating already rests on internal linguistic grounds that manuscripts would not change.

What would change this article's framing rather than its verdict.

• If the Reynolds–Lawson reading became the consensus in Quranic studies, the honest conclusion would shift from "the Quran conflicts with the historical evidence" to "the Muslim exegetical tradition conflicts with the historical evidence, and the Quran does not." That is a real possibility, it is argued by non-Christian scholars, and this article treats it as a live minority position rather than a settled one.
• Conversely, if a strong grammatical case were made that shubbiha lahum can only mean a person was substituted, the exegetical tradition would be vindicated and the conflict with history would become direct. Zamakhshari and al-Razi both concluded otherwise, and no modern grammarian has made that case.

What would not change anything.  Further arguments about whether the Gospels are historically reliable in general.  The crucifixion does not depend on them.

Sources