What major manuscript families and codices of the Quran exist and how do they differ?
Executive summary
Early Quranic manuscripts cluster into a handful of traditions: early Hijazi fragments (including the Birmingham folios and the Ṣanʿāʾ palimpsest), the companion codices (e.g., Ibn Masʿūd, Ubayy ibn Kaʿb), and the later ʿUthmānic codices that became the standard. Surviving physical evidence includes dozens of 7th–8th century fragments (over 60 fragments and ~2,000 folios reported by Corpus Coranicum) and distinctive codicological families that scholars map as regional or companion traditions [1] [2] [3].
1. The early Hijazi fossils: fragmented but foundational
Collections of Hijazi-style manuscripts — sloping, informal script found in early folios such as the Birmingham leaves and many Sanaʿa fragments — are the oldest physical witnesses to the text; radiocarbon tests on the Birmingham folios place the parchment between roughly 568–645 CE, and the Corpus Coranicum counts more than 60 fragments and some 2,000 folios from before 800 CE [1] [2]. These fragments are important because they preserve consonantal rasm and orthographic features that allow scholars to group them into small regional families rather than a single uniform early text [2].
2. The Ṣanʿāʾ palimpsest: an outlier that provoked debate
The Ṣanʿāʾ palimpsest contains an undertext that several scholars argue does not fit the Uthmānic textual tradition; Behnam Sadeghi and Uwe Bergmann call it “unique” and say it appears not to belong to the Uthmānic tradition, a conclusion that has driven debates about early textual plurality and the limits of standardisation [2] [3]. Later codicological and palaeographic studies have contested some early claims about the undertext, and this debate illustrates how a single high-profile manuscript can be read in competing ways by specialists [4].
3. Companion codices (Ibn Masʿūd, Ubayy, Zaid, others): multiple authored witnesses
Islamic tradition reports that several companions kept their own codices with measurable differences; historians record that Ibn Masʿūd’s and Ubayy ibn Kaʿb’s codices were among the most influential companion collections and that surviving narratives and manuscript evidence show real divergences between them [5]. Modern researchers document hundreds of variants reported among companion codices (Jeffery), and some argue those differences include omissions and additions that go beyond mere dialectal pronunciation [6] [7].
4. The ʿUthmānic codices and the move toward a standard text
Traditional accounts hold that Caliph ʿUthmān commissioned a recension and distributed standard codices to major provincial centers (Madīnah, Basra, Kūfah, Shām), intended to unify recitation; Islamic sources list about forty minor differences among those regional Uthmānic codices, and Muslim scholarly responses characterise these as small (often orthographic or single-letter) variants amounting to a tiny fraction of the text [8] [9]. Modern textual work finds that the Uthmānic codices were highly uniform compared with earlier companion codices, but regional Uthmānic variants still form recognisable patterns that help map textual transmission [10] [9].
5. How manuscripts differ: rasm, dotting, vocalisation and content
Differences between witnesses fall into discrete technical types: differences in the rasm (consonantal skeletal text), divergent dotting or diacritics that change consonants or attached pronouns, and variant vocalisation or orthography that can affect grammar and nuance. Estimates in the literature put canonical qirāʾāt variants at roughly 1,400 word-level differences across transmitted readings (about 2% by one count), with around 40 variants traceable to the regional Uthmānic codices and about 300 involving consonantal dotting differences [10] [11].
6. Competing scholarly narratives and implicit agendas
Two broad interpretive camps appear in these sources: (A) traditionalist and many Muslim scholars emphasise the rapid and conservative standardisation under Abū Bakr and ʿUthmān, portraying surviving variants as minor and compatible with oral transmission; (B) revisionist or critical scholars emphasise plural early codices and textual variation (citing the Ṣanʿāʾ undertext and companion codices) to argue for more complex transmission dynamics. Each camp stresses different evidentiary priorities — manuscript palaeography vs. hadith/narrative texts — and some polemical sources frame variant lists to challenge or defend doctrinal claims about textual perfection [1] [2] [6] [12].
7. What the physical evidence can and cannot show
Physical manuscripts establish that multiple textual witnesses existed early and that the Uthmānic recension produced a widely-disseminated, relatively uniform written archetype; radiocarbon and palaeography date folios to the 7th century and place many surviving fragments within the first two centuries of Islam [1] [2]. Available sources do not mention definitive proof that any single early codex predates the Prophet’s lifetime as a complete book, and they show ongoing scholarly contestation over how to interpret palimpsests and fragmentary evidence [2] [13].
8. Takeaway for readers
The manuscript record shows early plurality followed by rapid standardisation: early Hijazi fragments and companion codices preserve meaningful variants, while the Uthmānic codices — despite minor regional differences — became the basis for the text in use today. Scholarly disagreements persist about the significance of specific manuscripts (notably Ṣanʿāʾ), and readers should weigh codicological evidence alongside the recorded historical traditions that describe compilation and recension [1] [2] [9].