About the Qur’an

Abstract: This paper evaluates the textual composition, linguistic origins, and historical emergence of the Qur'an through a critical, philological lens. It argues that the Qur'an is fundamentally a "text without context," characterized by a discontinuous, non-linear structure and an ambiguous underlying consonantal script (rasm) that relies heavily on external Biblical knowledge to be understood. Drawing on contemporary linguistic and historical scholarship, the paper demonstrates that the text's vocabulary and themes are overwhelmingly rooted in a late-antique, heterodox Judeo-Christian (specifically Ebionite) milieu rather than a native Arabian one. This background is evidenced by the stark statistical dominance of Biblical figures over the Prophet of Islam, as well as distinct narrative conflations and anachronisms resulting from the oral transmission and reworking of Aramaic and Hebrew sources into the Old Hijāzī dialect. Furthermore, the paper examines traditional Sunni and Shia accounts to highlight historical uncertainties regarding the integrity, preservation, and standardization of the Uthmanic recension. Challenging traditional Islamic geography, the paper concludes that Islam developed from a northern messianic movement that was only later localized in Arabia and formalized into a distinct religion under the Marwanid caliphs, leaving a text that remains inherently obscure and open to subjective interpretation and that it operates as an obscure document stripped of geographical and narrative context, and its inherent ambiguity allows it to function primarily as a mirror for diverse reader interpretations.

Keywords: Qur'an, philology, rasm, Judeo-Christian substrate, Ebionite, iltifāt, Uthmanic recension, Old Hijāzī, Sami Aldeeb, Patricia Crone, textual criticism, anachronism, Marwanid caliphate

Language and composition

The Qur’an is a difficult, non-linear composition. Its chapters (suras) are arranged roughly from the longest to the shortest rather than by chronology or theme, so the book does not read as a continuous narrative, and it frequently shifts subject, addressee, person and tense within a single passage. Muslim rhetoricians studied many of these shifts as a deliberate ornament of eloquence — the device of iltifāt [the rhetorical shift of person, number or tense]. A reader looking for narrative flow experiences the same shifts as discontinuity.

Critics sometimes draw up catalogues of linguistic errors in the text — Sami Aldeeb counts more than 2,500 that include genuine lexical ambiguity; inconsistency of spelling; the differences among the canonical reading traditions (qirāʾāt); the rhetorical person- and tense-shifts just mentioned; and grammatical features that depart from later Classical norms. Several of these may be traces of an earlier stage of the language.

The vocabulary carries a considerable layer of loanwords from Aramaic, Syriac and Hebrew, especially in its religious terminology. This is not a recent or controversial discovery: it was established by mainstream scholarship long ago — by Abraham Geiger in 1833 and catalogued by Arthur Jeffery in 1938. The contents draw mainly on a Jewish-Christian Ebionite milieu — carried through material that was originally Aramaic and Hebrew.

Some of these traces sit in plain view once one reads with the donor languages in hand. The term ḥanīf, the Qur’an’s word of approval for the pure Abrahamic monotheist, derives from a root that in Christian Aramaic (ḥanpā) meant the opposite — “pagan, heathen” — a semantic inversion that betrays the Aramaic-Christian substrate. And at Q83:19 the highest celestial register is named ʿilliyyūn — rendered “the Highest / the heavens” by Irving and “highest places” by Maulana Muhammad Ali, as Aldeeb notes — a word from the Semitic root of height, ʿ-L-W, and the cognate of Hebrew ʿelyon, “the Most High.” The Hebrew name of God’s exaltation sits embedded in the Arabic text.

It is worth saying why such readings deserve weight. Sami Aldeeb works directly, in his translation notes, with Biblical and Mishnaic (Talmudic) Hebrew, with Aramaic, Syriac and Koine Greek — a philological range a reader can verify simply by opening his apparatus. Much of the debate over the Qur’an’s origins, on both the traditional and the revisionist sides, is conducted without that competence; the Semitic substrate is therefore often missed by those arguing over it. 

The surviving consonantal text appears, on recent linguistic work, to reflect the Arabic dialect of the Hijaz, not a northern Arabic of Syria or Iraq. This does not weaken the case for a northern, Jewish-Christian origin of the contents; it strengthens it. If the movement that produced these texts moved south into Arabia and needed an Arabian audience to receive them as a revelation in their own tongue, then casting northern, Aramaic-rooted material in the Hijazi dialect is what one would expect. The origin can be northern while the language is, deliberately, Hijazi.

Judeo-Christian content

For the greater part, the Qur’an is composed of material drawn from a Judeo-Christian background, carried through texts originally written in Aramaic and Hebrew. Its vocabulary and its themes are overwhelmingly those of the Jewish and Christian scriptures, not of an Arabian milieu. A simple count of names makes the point at a glance.

The contrast between the two extremes is striking: Moses is named 171 times, Muhammad only four. The figures of the Hebrew Bible — Moses, Abraham, the patriarchs, Joseph — together with Jesus and Mary, fill the text, while the Prophet of Islam barely appears in it by name.

And if the Qur’an was “sent down” and “revealed” in Arabic — the word Arabic itself occurs eleven times — why does it dwell so much more on the ancestors of the Jews, Isaac (17) and Jacob (20, with one further mention as Israel), than on Ishmael (12), the ancestor of the Arabs? A book addressed to Arabs in their own tongue gives the lion’s share of its attention to the lineage of Israel.

This is reinforced by a feature that Mohammad Ali Amir-Moezzi and Guillaume Dye have emphasised: the Qur’an is, in a phrase that has become a historiographical commonplace, un texte sans contexte — a text without context. It alludes to the narratives of the Bible rather than recounting them, and presupposes a hearer who already knows them; Amir-Moezzi observes that it is filled on every page with Jewish and Christian references and continually casts itself as the continuation of the messages of Moses and Jesus. To be followed at all, it requires a reader who brings that biblical knowledge from outside.

The Christianity behind it, however, is not the canonical orthodoxy of the imperial Church but the heterodox Christianity of the late-antique Near East: anti-Trinitarian currents of an Arian cast, Docetism — whose claim that Christ only seemed to die is echoed almost exactly in the Qur’an’s denial of the crucifixion (Q4:157, “they did not kill him … but it was made to appear so to them”) — Ebionites, the Church of the East, Nestorianism,  and the Ethiopian church, together with the apocryphal gospels. Aldeeb’s translation often points to this apocryphal material: the infancy gospels lie behind episodes such as the infant Jesus speaking from the cradle or fashioning living birds from clay. A reader can follow the Qur’an without having read the apocrypha, but its Jesus is unmistakably the Jesus of these traditions — a human Messiah and prophet, not God incarnate. The Marian material shows the same imprint: the Qur’an keeps Mary’s virginity and the miraculous birth, but never calls her the Mother of God — the very title (Theotokos) that the Nestorian current refused — because a Jesus who is not God can have no God-bearing mother. The same milieu shows in the figure of Dhūl-Qarnayn (Q18:83-86-94,), the bicorn. It indicates either Alexander the Great, or Cyrus the Great to whom Daniel 8:20 may  refer, which mentions two horns and whose tale closely follows the Syriac Christian Alexander Legend composed around 629–636, a source virtually contemporary with the Qur’an’s emergence.

Two passages show what “without context” means in practice — places where material received at second hand has been merged or misjoined, leaving a visible seam.

Mary and Miriam. The Qur’an calls Mary “sister of Aaron” (Q19:28) and “daughter of ʿImrān” — ʿImrān being Amram, the father of Moses, Aaron and their sister Miriam in the Hebrew Bible. The mother of Jesus has thus been given the family of Miriam, the sister of Moses, a woman separated from her by something like a thousand years. The two namesakes — Maryam and Miriam are the same name — have been conflated: exactly the slip that occurs when biblical narrative is carried orally and reworked rather than read from the text.

The Cow. The sura that gives the Qur’an’s longest chapter its name (al-Baqara, “The Cow,” Q2:67–71) fuses two distinct Pentateuchal rites into one. The red heifer of Numbers 19, whose ashes purify, is merged with the broken-necked heifer of Deuteronomy 21, the rite prescribed for an unsolved murder — and indeed the verses that follow (Q2:72–73) turn to a slain man and a cow used to resolve his killing. The result matches neither biblical rite exactly. Tellingly, the Muslim commentator Maududi, explaining the passage, reaches for the biblical source himself and cites Numbers 19:1–10. John of Damascus (c. 675–749), among the earliest non-Muslim writers to discuss the Qur’an, singled this sura out by name as an independent book: this Muhammad wrote many ridiculous books, to each of which he gave a title. There is, for instance, the book On Woman, then the book The Table, and The Cow, and other things worthy only of laughter, which I think I will pass over because of their number.
He legislated that they be circumcised, women as well as men. He also commanded them not to keep the Sabbath and not to be baptized. And while he permitted them to eat some things forbidden by the Law, he ordered them to abstain from others which the Law allowed. He also absolutely prohibited the drinking of wine.

Haman. The Qur’an makes a man named Haman the minister of Pharaoh and the builder of a tower meant to reach the God of Moses (Q28:38, Q40:36–37). But the Haman of the Bible is the minister of the Persian king Ahasuerus in the book of Esther — a thousand years after Moses and in another empire — and the tower belongs to a third story, the Tower of Babel of Genesis 11, older than Moses still. Three separate narratives have been fused into one. This is not a hostile reading: the Encyclopaedia of Islam itself records Haman as the name the Qur’an associates with Pharaoh “because of a still unexplained confusion” with the minister of Ahasuerus, and notes that the account modifies Exodus “by misplaced recollection of both the book of Esther and the story of the tower of Babel.”

Saul and Gideon. In Q2:249 the army of Saul (Ṭālūt), marching against Goliath, is tested at a river: those who drink from it are rejected, those who refrain are kept. In the Hebrew Bible that water-test belongs to Gideon (Judges 7), a different leader in a different episode generations earlier. Saul’s story has been merged with Gideon’s, and the whole attached to the march against Goliath.

The Trinity as Father, Jesus and Mary. The Qur’an’s rebuttal of the Trinity has God ask Jesus whether he told people to take “me and my mother as two gods besides God” (Q5:116) — as though the three were Father, Jesus and Mary, not the Father, Son and Holy Spirit of orthodox doctrine. This is most likely not a simple misunderstanding but an engagement with a real heterodox sect: the Collyridians, described by Epiphanius in the fourth century as worshipping Mary as divine and offering her cakes, and reported precisely in Arabia (a connection Gabriel Said Reynolds draws in his commentary). The Qur’an is not garbling orthodox Christianity so much as answering the Mary-venerating Christianity actually present in its world — itself a mark of the milieu from which it comes.

Ezra. The Qur’an says “the Jews say ʿUzayr is the son of God” (Q9:30). ʿUzayr is the Hebrew ʿEzra, the scribe of the return from exile, who has a book of the Bible in his name. Mainstream Judaism never deified him, and so the verse is often dismissed as an invention. But Samaritans and Jews share substantially the same Torah and the same Hebrew, and were readily taken for one another; and Aldeeb — himself of Samaritan origin — reports a Samaritan tradition of a sect that did make a god of Ezra. On that reading the verse is not a fabrication but a real sectarian belief, recorded accurately and misattributed to “the Jews” through the blur between Samaritan and Jew. It is of a piece with the Qur’an’s loose use of “the Jews” for the whole Israelite spectrum — the same spectrum on which the Samaritan of the golden calf (above) sits.

Alongside these conflations sit anachronisms — figures, peoples and practices set in centuries to which they cannot belong, the residue of material remembered and retold rather than read from a fixed text. A Samaritan (al-Sāmirī) makes the golden calf in the time of Moses (Q20:85–95), though the Samaritans did not exist as a people until many centuries later; some Muslim commentators, aware of the difficulty, read the word as something other than “the Samaritan.” Pharaoh threatens crucifixion in the days of Moses and Joseph (Q7:124, Q20:71, Q12:41), though crucifixion as a mode of execution is not attested until roughly the sixth century BCE, far later. Abraham’s father is named Azar (Q6:74), whereas the Bible calls him Terah. Each is the kind of seam left when a tradition is carried by memory across a linguistic and cultural gap.

The script and the ambiguity of the text

The oldest manuscripts are written in a defective script — the rasm [the bare consonantal skeleton], without the dots that distinguish many consonants and without vowel marks. Twenty-three of the twenty-eight Arabic letters are therefore ambiguous in isolation, and one written skeleton can support more than one reading. This is a real source of uncertainty about the original sense of particular words and sentences. The entire science of the qirāʾāt [the canonical reading traditions] grew up around the different ways the skeleton was vocalised. A reader who has only the bare written text, cut off from the recitation that accompanied it, will find much of it underdetermined.

Doubts about the integrity of the text

The text Muslims use today is the recension associated with the caliph Uthmān. The traditional sources report that Uthmān had a standard codex (mushaf) prepared and ordered the variant copies burned, so that disputes over recitation would cease. One Companion, Ibn Masʿūd, is reported to have refused to surrender his copy. Uthmān was assassinated — and the tradition's own sources preserve the hostility against him: he was reviled as the burner of the Qur’an and as a falsifier of it, and that charge was among the grievances for which he was killed. Sami Aldeeb states it plainly: Uthmān falsified the Qur’an, and Ḥafṣ was a liar.

The recension did not fix a single recitation. Several reading traditions remain canonical. The reading transmitted by Ḥafṣ (from Āṣim) is the one most Muslims follow, printed in Saudi Arabia and Egypt; Morocco follows the reading of Warsh, and Sudan that of al-Dūrī. Their verse-totals differ — about 6,236 for Ḥafṣ, 6,214 for Warsh, 6,204 for al-Dūrī. These totals differ chiefly because the readings divide the verses at different points, not because one contains substantially more text than another.

Aldeeb presses a further point about Ḥafṣ. The same Ḥafṣ whose recitation the majority of Muslims follow was graded a weak and unreliable transmitter of ḥadīth [the reports of Muhammad’s words and deeds] by the ḥadīth critics, Ibn Ḥajar among them. A man set aside as a narrator in one field is trusted as the principal transmitter of the text in another. The reply offered within the tradition is that reliability was assessed field by field: the transmission of the recitation rested on broad parallel chains (tawātur [mass transmission through many independent lines]) rather than on one man’s character.

Beyond the question of transmission, the tradition’s own literature preserves reports that the text was once larger than it now is.

On the Sunni side:

• The sura al-Aḥzāb (33), now 73 verses, is reported to have once been longer than al-Baqara (2), which has 286 — implying that more than two hundred verses are no longer present. The report is recorded by al-Suyūṭī in al-Itqān.

• The sura al-Tawba (9), now 129 verses, is reported to have once been comparable in length to al-Baqara; by some accounts only a fraction remains.

• Two short suras used in early prayer, al-Ḥafd and al-Khalʿ, are not in the present text, though they appear in the codex attributed to Ubayy ibn Kaʿb.

• Umar is reported in al-Bukhārī to have held that a “verse of stoning” was once recited and is no longer in the text. The tradition accounts for such cases through the doctrine of naskh al-tilāwa [abrogation of the wording — a passage once recited but later withdrawn].

On the Shia side:

• Al-Kulaynī (d. 941), in al-Kāfī, transmits a report that the revelation Gabriel brought Muhammad contained seventeen thousand verses, against the roughly 6,236 of the present text.

• The same source speaks of a codex of Fāṭima, said to contain material not in the current text.

• Al-Fayḍ al-Kāshānī (d. 1680) held that the present text is incomplete and disarranged, omitting the name of ʿAlī.

• Certain Shia accuse Uthmān of removing passages that referred to Alī, including two alleged suras, al-Wilāya and al-Nūrayn.

Mainstream Shia scholarship rejects any corruption of the text (taḥrīf [alteration of the wording]) and regards the alleged missing suras as fabrications; these reports sit in its canonical collections without being held as doctrine. What stands at the end is narrow and hard to dismiss: the tradition’s own sources preserve reports of a text once in motion, alongside the doctrine that the text was preserved without change.

Where the text comes from

The Qur’an effectively names neither Mecca nor Jerusalem. The name Makka occurs just once (Q48:24) — a token Aldeeb and others take to be a later insertion — and Bakka, a variant name usually identified with Mecca, once more (Q3:96); Jerusalem and its later Arabic name al-Quds do not appear at all. For a book traditionally held to have been composed in Mecca and about it, the silence is striking: the text carries almost none of the local geography — the streets, the hills, the circumstantial detail — that a work rooted in a place displays. Patricia Crone, in Meccan Trade and the Rise of Islam, dismantled the economic premise that might have anchored the narrative there: the great pre-Islamic Meccan trading city of tradition does not survive examination of the evidence. The earliest mosques, for their part, face a scatter of directions rather than one — most simply because the early community lacked the geographical and mathematical means to determine the direction of a distant place with precision.

The lone variant name repays a closer look. Bakka (Q3:96), which tradition identifies with Mecca, recalls the Valley of Baca of Psalm 84: 

Happy are those who find refuge in You,
whose mind is on the [pilgrim] highways.
They pass through the Valley of Baca,
regarding it as a place of springs,
as if the early rain had covered it with blessing.

— a station on the Jewish pilgrimage up to the Temple in Jerusalem (Arabic wādī al-bukāʾ, which Aldeeb connects to the ḥayt al-mabkā, the Wailing Wall) — and Bonnet-Eymard reads the “House” of the adjacent verse (Q3:97) as the Temple of Jerusalem rather than the Kaʿba. On this reading the Qur’an’s sanctuary language points north, to the Temple, before the tradition relocated it to Mecca. Baca may also refer to a valley near Jerusalem where trees called bakha grew.

What the text carries instead is an Ebionite Judeo-Christian world, and its language offers a clue to how that came about. Recent linguistic work — Marijn van Putten and Ahmad al-Jallad, who calls the underlying dialect “Old Ḥijāzī” — has shown, from the Qur’an’s own rhyme and orthography, that the consonantal text reflects a Hijazi Arabic distinct from the Classical Arabic of the later reading traditions. The loss of the glottal stop (hamzah), for example, is a Hijazi feature visible beneath the text, onto which the reading traditions later imposed a glottal stop the dialect never had. This points to a layered composition. In the first layer, Ebionite Judeo-Christian material — biblical and lectionary texts originally in Aramaic and Hebrew — was freely rendered into Hijazi Arabic: not invented but translated and reworked, which is exactly what produces the conflations and seams above. In a later layer, further material was composed into the same Hijazi register, by then an established and prestigious scriptural voice, so that additions would be received as of a piece with what came before. On this reading the few explicit tokens of Mecca and of the name Muhammad — four occurrences of the name in the whole text — belong to that later, localizing layer; so Aldeeb, who takes them to be insertions.

This is broadly the picture developed by Patricia Crone, and followed by me, that Islam emerged from an Ebionite Judeo-Christian, messianic movement of the seventh-century Near East, the Qur’anic material taking shape in that milieu before being localized, under the Marwanid caliphs, in an Arabian setting. Lafontaine, among others, points also to the early Islamic coinage. For decades after the conquests the caliphate’s coins still carried the cross and other Christian imagery; under ʿAbd al-Malik, at the end of the seventh century, the cross was progressively effaced — its arms cut down to a bare staff — and the explicitly Islamic epigraphy, including the formula Muḥammad rasūl Allāh, introduced in its place. The earliest coins bearing that formula were struck at Jerusalem. The distinct identity of the new religion, on this evidence, crystallized late and by deliberate program rather than being present from the first.

Conclusion

As the Qur’an is supposed to be God’s perfect and unaltered word — created, or even uncreated (!), before the creation of the world — most believers seek deeply hidden mystical values in the typos, mistakes and plain gibberish it contains. An exception is Ibn Khaldūn (1332–1406), who cautiously wrote that the men around Muhammad, and the generations that imitated them, wrote the Qur’an in their own script, which was not of a firmly established good quality.

The Qur’an is a text you can’t make head or tail of. Stripped of the context it never supplies, drawn from sources it does not name, broken in its arrangement and uncertain in its very letters, it resists the reader who comes to it expecting a book that means one thing. It is the Emperor’s new clothes of Andersen’s tale: a splendour everyone is expected to admire and no one will confess he cannot see, for to admit the text is obscure is to be thought unfit to read it. So those who profess to have read and understood it are fooling themselves and others — repeating the praise that is required of them rather than the sense they have actually found. And because the text carries no settled meaning of its own, it becomes a mirror: anyone can make it say whatever he wishes, and call his own reflection revelation.

References

Sami Aldeeb (2016), The Koran: Arabic text with the English translation, Amazon.

Sami Aldeeb (2021), Introduction aux erreurs linguistiques dans le Coran.

Sami Aldeeb, “Uthman a falsifié le Coran… et Hafs est un menteur.” 

Mohammad Ali Amir-Moezzi & Guillaume Dye, eds. (2019), Le Coran des historiens, 3 vol., Paris, Cerf.

Patricia Crone (1987), Meccan Trade and the Rise of Islam, Princeton.

Patricia Crone (2015) Jewish Christianity and the Qurʾān (Part One) in Journal of Near

Eastern Studies Vol. 74, No. 2

Patricia Crone (2016) Jewish Christianity and the Qurʾān (Part Two) in Journal of Near

Eastern Studies Vol. 75, No. 1

Édouard-Marie Gallez (2005), Le messie et son prophète: aux origines de l’Islam.

Encyclopaedia of Islam (s.v. “Hāmān” and “Firʿawn”).

Abraham Geiger (1833), Was hat Mohammed aus dem Judenthume aufgenommen? Bonn.

Ahmad al-Jallad (2020), The Damascus Psalm Fragment: Middle Arabic and the Legacy of Old Ḥigāzī, Chicago.

Arthur Jeffery (1938), The Foreign Vocabulary of the Qurʾān, Baroda.

John of Damascus (8th c.), On Heresies (ch. 100/101, on the “heresy of the Ishmaelites”).

Ibn Khaldūn (14th c.), Muqaddima.

Odon Lafontaine (2024), The Great Secret of Islam.

Abu al-Aʿlā Maududi, Tafhīm al-Qurʾān (commentary on Q2:67–71).

Francis E. Peters (1991), “The Quest of the Historical Muhammad,” International Journal of Middle East Studies 23.

Richard Preschel (2021), The Zionist Origins of Islam.

Richard Preschel (2025), Everything You Always Wanted to Know About the Red Heifer and The Cow* (*But Were Afraid to Ask).

Marijn van Putten (2022), Quranic Arabic: From its Hijazi Origins to its Classical Reading Traditions, Brill.

Gabriel Said Reynolds (2018), The Qurʾān and the Bible: Text and Commentary, Yale.