The Qur’an and the hadith, audited

Oneness, in the Qur’an’s own words

Select a verse to read the translation and the note. Every verse here states God’s oneness directly; none needs an inference to reach it, which is the difference the companion dashboard measures verse by verse.

• 112:1–4 Al-Ikhlas Explicit +

  قُلْ هُوَ ٱللَّهُ أَحَدٌ · ٱللَّهُ ٱلصَّمَدُ · لَمْ يَلِدْ وَلَمْ يُولَدْ · وَلَمْ يَكُن لَّهُۥ كُفُوًا أَحَدٌ

• 2:163 Al-Baqarah Explicit +

  وَإِلَٰهُكُمْ إِلَٰهٌ وَٰحِدٌ لَّآ إِلَٰهَ إِلَّا هُوَ

• 2:255 Ayat al-Kursi Explicit +

  ٱللَّهُ لَآ إِلَٰهَ إِلَّا هُوَ ٱلْحَىُّ ٱلْقَيُّومُ

• 3:18 Al-'Imran Explicit +

  شَهِدَ ٱللَّهُ أَنَّهُۥ لَآ إِلَٰهَ إِلَّا هُوَ

• 4:171 An-Nisa Explicit +

  وَلَا تَقُولُوا۟ ثَلَٰثَةٌ ٱنتَهُوا۟ خَيْرًا لَّكُمْ إِنَّمَا ٱللَّهُ إِلَٰهٌ وَٰحِدٌ

• 5:73 Al-Ma'idah Explicit +

  لَّقَدْ كَفَرَ ٱلَّذِينَ قَالُوٓا۟ إِنَّ ٱللَّهَ ثَالِثُ ثَلَٰثَةٍ

• 6:102 Al-An'am Explicit +

  ذَٰلِكُمُ ٱللَّهُ رَبُّكُمْ لَآ إِلَٰهَ إِلَّا هُوَ خَٰلِقُ كُلِّ شَىْءٍ

• 16:51 An-Nahl Explicit +

  لَا تَتَّخِذُوٓا۟ إِلَٰهَيْنِ ٱثْنَيْنِ إِنَّمَا هُوَ إِلَٰهٌ وَٰحِدٌ

• 17:111 Al-Isra Explicit +

  ٱلْحَمْدُ لِلَّهِ ٱلَّذِى لَمْ يَتَّخِذْ وَلَدًا وَلَمْ يَكُن لَّهُۥ شَرِيكٌ فِى ٱلْمُلْكِ

• 21:22 Al-Anbiya Explicit +

  لَوْ كَانَ فِيهِمَآ ءَالِهَةٌ إِلَّا ٱللَّهُ لَفَسَدَتَا

• 23:91 Al-Mu’minun Explicit +

  مَا ٱتَّخَذَ ٱللَّهُ مِن وَلَدٍ وَمَا كَانَ مَعَهُۥ مِنْ إِلَٰهٍ

• 37:4 As-Saffat Explicit +

  إِنَّ إِلَٰهَكُمْ لَوَٰحِدٌ

• 42:11 Ash-Shura Explicit +

  لَيْسَ كَمِثْلِهِۦ شَىْءٌ وَهُوَ ٱلسَّمِيعُ ٱلْبَصِيرُ

• 47:19 Muhammad Explicit +

  فَٱعْلَمْ أَنَّهُۥ لَآ إِلَٰهَ إِلَّا ٱللَّهُ

From recitation to the printed page

The route the text travelled, and the physical evidence at each stage. Dates are CE.

• Revelation
• Compilation
• Manuscript
• Reading

1. Revelation and memorisation 610–632 CE

   Revealed in portions across twenty-three years, memorised by the companions as it came and written down by appointed scribes — Zayd ibn Thabit among them — on whatever was to hand. Recitation and writing ran in parallel from the start.

2. The first collection c. 633 CE

   After the battle of Yamama killed a number of memorisers, Abu Bakr ordered the written material gathered into a single copy under Zayd ibn Thabit, kept afterwards with Hafsa.

3. The ʿUthmānic codex c. 650 CE

   ʿUthmān had standard copies made from Hafsa’s collection in the Qurayshi orthography and sent to the main garrison cities, ordering other written material destroyed. Every printed Qur’an today descends from this recension.

4. The Birmingham folios radiocarbon 568–645 CE

   Two parchment leaves at the University of Birmingham, carrying parts of surahs 18–20 in Hijazi script, radiocarbon-dated with 95.4% confidence to 568–645 CE. Their text matches the modern one.

   Check it yourself: University of Birmingham

5. The Sanaa palimpsest c. 650–700 CE

   A manuscript with two layers. The upper text is standard; the lower, erased text is a non-ʿUthmānic companion codex, differing in word order, in some pronouns and synonyms, and in surah order — the most significant variant witness known.

   Check it yourself: Corpus Coranicum (BBAW)

6. The Topkapi and Samarkand codices 8th–9th century CE

   Two of the oldest near-complete copies. Both are traditionally ascribed to ʿUthmān himself; palaeography places them later, and both agree with the received text in wording.

   Check it yourself: UNESCO Memory of the World

7. The readings codified c. 936 CE

   Ibn Mujahid limited the accepted recitations to seven, later extended to ten. The readings differ in vowelling, dialect and occasionally a word; Hafs ʿan ʿAsim is the reading in most print today, Warsh ʿan Nafiʿ across much of North and West Africa.

The hard questions about the Qur’an

Put as an opponent would put them, then answered.

• If the Qur’an is preserved, why are there ten readings?

  The canonical qiraʾat differ: 1:4 is read mālik ("master") in some and malik ("king") in others; other places differ in vowelling, in a pronoun, or in a verb form.

  The answer given: The readings are themselves transmitted, named and mass-transmitted rather than accidental; the tradition treats them as revealed modes of recitation. The differences are of vowelling and dialect, and where a word differs the two readings complement rather than contradict. No reading produces a different doctrine.

• Why did ʿUthmān destroy the other copies?

  Companions held personal codices — Ibn Masʿūd’s and Ubayy’s among them — which differed in surah order and in some annotations. ʿUthmān had them destroyed.

  The answer given: The reported motive is disputes over recitation between garrisons. Standardising on one written form settled the orthography while leaving the ten readings in place. The differences reported for those codices are of arrangement and marginal notes rather than of doctrine, though the destruction means we cannot now examine them.

• What about the erased text under the Sanaa manuscript?

  The lower layer of the Sanaa palimpsest is a companion codex differing from the standard text in word order, pronouns, synonyms and surah order.

  The answer given: It is the strongest physical evidence of the non-ʿUthmānic codices the sources describe, and it matches what they describe: variation of the kind the tradition itself reports for Ibn Masʿūd, not a different message. Its existence is an argument for the sources’ honesty as much as a problem for the text.

How a hadith is graded

Hadith are reports about the Prophet, not revelation, and they are not accepted as a block. Each is graded on its chain of narrators and on its content; the bar shows the weight each grade carries in law and creed.

• Sahih (sound) ×1

  An unbroken chain of upright, precise narrators, free of irregularity and hidden defect. Admissible in law and creed.

• Hasan (good) ×0.75

  As above, but with a narrator of lesser precision. Admissible, and strengthened when corroborated by another chain.

• Daʿif (weak) ×0.25

  A break in the chain, or a narrator criticised for memory or reliability. Not a basis for law or creed; scholars differ on its use for encouragement in virtue.

• Mawduʿ (fabricated) ×0

  A report demonstrated to be invented. Not a hadith at all; catalogues of them exist precisely so they can be named and rejected.

The major collections

Counts follow each collection’s own numbering and include reports repeated with different chains, which is why the distinct totals are much lower.

The hard questions about hadith

Put as an opponent would put them, then answered.

• The collections were compiled two centuries later. Why trust them?

  Bukhari died in 870, some 238 years after the Prophet. Any report in his book passed through several intermediaries first.

  The answer given: Every report carries its chain, and the chain is the object of study: an entire discipline (ʿilm al-rijal) catalogues the narrators by date, teachers, students, memory and character, so a chain can be tested rather than trusted. Earlier written material exists — the Sahifa of Hammam ibn Munabbih, the Muwatta a century before Bukhari.

• Weak and fabricated hadith exist. Does that undermine the corpus?

  The tradition itself records fabrication on a large scale, by sectarians, storytellers and partisans.

  The answer given: It records it because it hunted for it. Weakness is published rather than concealed: the grades are stated, the fabrications catalogued by name, and no weak report may establish law or creed. A corpus that names its own forgeries is in a different position from one that cannot identify them.

• Do Muslims hold hadith to the same standard as the Qur’an?

  The two are often quoted side by side, which can suggest equal authority.

  The answer given: No, and the distinction is explicit. The Qur’an is held to be mass-transmitted in word; hadith are reports about the Prophet, graded individually, and the great majority are transmitted by single chains. A hadith contradicting the Qur’an is rejected on that ground alone.