Volume 17 Belief System 610–661

Islam

The machinethe standardised codex, and the memoriser

Attested from: 610 CE
Narrates: 610–661

The argument in one sentence

Inside fifty years a recitation carried in trained memories acquired a state, a treasury and a standardised written skeleton — and the one thing the community never standardised, the succession, split it while the memorisers were still alive — over a question the two traditions that came out of it do not categorise the same way: a procedure, on one account, and on the other a designation already made.

Qur’an folios in Birmingham radiocarbon-date to 568–645 CE at 95.4 per cent confidence, a window overlapping the Prophet’s own lifetime — though the date is the parchment’s, not the ink’s.The commissioning line

What it is

At Hammat Gader, in a bath house below the Golan, a Greek inscription dates itself by the year 42 of the Arabs. It was cut while the thing it records was still happening. Wind back fifty-two years and there is none of it: no state to count those years, no treasury, no era, no book.

This is the history of what happened in between, and of how we know.

Islam’s first half-century is usually told as a life. This volume tells it as a transmission problem. A recitation carried in trained memories acquired a state, a payroll and a written skeleton so tightly controlled that every surviving early manuscript descends from a single archetype — and the one thing the community never standardised was the succession, which split it while the memorisers were still alive.

Five movements: the world it began in, the machine that carried it, the canon and who had the authority to close it, the money, and the split.

The book never adjudicates. Where the tradition’s account and the earliest evidence disagree, both are given, attributed, and neither is given the last word. Every translation is named in the text. Every figure carries its source in the same sentence. Where something cannot be established, the book says so and says why.

Belief System, Volume 17.

In full

The oldest object in this book is a paving stone. The most recent is a parchment in a Birmingham library that nobody read for eighty years.

Between 610 and 661 a movement with no state, no treasury and no written scripture became an empire with all three. Most accounts of that half-century are biographies. This one is not. It asks the questions the series asks of every tradition in it: what carried this, who paid for it, what got written down, who had the authority to decide, and what finally divided it.

The answers are unusually well documented on one side and unusually thin on the other, and the difference is the book. The money left paperwork — a bilingual Greek and Arabic receipt from Egypt dated year twenty-two of the new era, a stipend register whose grades survive in al-Baladhuri, a tax schedule written in camels. The manuscripts left a fingerprint: idiosyncratic spellings copied letter for letter across early copies made in different places, which is the mark of one written archetype rather than of a shared memory. The life of the prophet, by contrast, reaches us through a biography composed more than a century afterwards, surviving only in a recension whose editor states in his preface that he cut material.

This book prints both states of the evidence and keeps them apart. It uses five fixed phrases — the record shows, the tradition holds, conventionally dated, is contested, cannot be established — and no sixth, so that a reader always knows which kind of claim is being made. It quotes scripture and names the translation in the visible text every time. It reports no miracle as fact and denies none. It never tells you what anybody was thinking.

The succession dispute that ends the volume is the most consequential argument in the history of the tradition, and it is written here the way the rest of the book is written: both accounts, in their own vocabulary, attributed, with neither given the last word. The reports from the portico and the reports from the pool at Khumm come down through the same collectors, in the same collections, transmitted by the same methods. That is why neither can be used to destroy the other — and it is the volume’s closing argument.

Belief System is a history of world religion, one tradition at a time: where it started, what carried it, who paid for it, and how it divided. Thirty-one volumes, ordered by earliest secure attestation.

The contents

26 chapters in five movements. Every volume in the series has the same five, in the same order, and each chapter below is printed with the idea it carries.

Movement OneThe World

Mecca and Medina as they actually were, and the question the movement answered. The evidence is thinnest here, and the confidence falls with it, in the same sentence.

1The SanctuaryMecca’s economy was a truce with a market inside it — a haram where the sacred months suspended the feud so pilgrimage and trade could happen in the same place at the same time, and the family that administered the shrine ran the town.
2The Caravan QuestionEveryone agrees Quraysh ran caravans; what they carried, how far, and how rich it made them is contested — the incense-route picture and the leather-and-coarse-cloth picture are both on the table, and the chapter reports the fight rather than settling it.
3An Oasis Without a JudgeTribal law worked — blood-money, retaliation and sworn alliance instead of a state — but its unit was the kin group, and Medina in the 610s, an agricultural oasis whose clans had fought each other to a standstill (the tradition dates the last battle, Bu’ath, to about 617), shows what the system could not produce: a verdict both sides had to accept.
4Rahmanan’s CountryArabia in 610 was not a pagan vacuum waiting for monotheism — Himyar’s kings had been invoking “Rahmanan, lord of heaven” in official inscriptions for two centuries, Najran had Christian martyrs, and the question a new prophecy answered was live, local and already argued in Arabic.
5The Exhausted EmpiresFrom 602 to 628 Rome and Persia fought the last and worst of their wars — Jerusalem sacked in 614, Egypt taken, Constantinople besieged in 626, Khosrow II deposed and killed in 628 — and both had already dismantled the Arab client kingdoms that managed the frontier for them, which is why the peninsula’s northern edge stood open at precisely the moment a state formed behind it.
6The Story, and When It Was Written DownThe narrative every reader knows — the cave in about 610, the emigration in 622, the return to Mecca in 630, the death in 632 — is the tradition’s own connected account, first composed as a book more than a century later; what near-contemporary outsiders confirm is smaller and sharper: by 634 there was a movement, it had a prophet, and foreign writers knew his name within a decade.

Movement TwoThe Machine

The transmission technology, named as a concrete object or institution — not an abstraction. Here: the standardised codex and the memoriser, as one working system.

7The MemoriserThe first storage medium was the trained human memory, organised — the huffaz carried the text as a licensed skill, the tradition’s own trigger-story for writing it down is the death of reciters in one battle (Yamama, 633), and the profession is no relic: the memorisation economy runs today, on chains of certification, and it is the reason the written page could afford to be incomplete.
8A Skeleton of ConsonantsThe earliest copies record consonants only — no vowels, few or no distinguishing dots, long vowels often unwritten — so the page could not tell a stranger how to read it and was never meant to: it was a support for a memorised recitation, not a replacement for one, and the script’s deficiencies are the system’s design.
9The Parchment in BirminghamTwo leaves in the University of Birmingham’s Mingana Collection were radiocarbon-dated in 2015 to between 568 and 645 CE at 95.4 per cent confidence — a window overlapping Muhammad’s lifetime — and the same sentence carries the limit: the date is the animal’s death, not the ink’s application, and several specialists said so at the time.
10The Palimpsest in the RoofIn 1972 workmen restoring the Great Mosque of Sana’a found a mass of parchment between roof and ceiling, and one manuscript in it had been erased and rewritten — the washed-off lower layer is a Qur’anic text that differs from the standard one, the clearest surviving physical evidence that variant written copies existed to be withdrawn.
11Two Channels, One BookSeen whole, the machine is a two-channel error-correcting system — a copied consonantal skeleton that fixes what memory drifts on, and a licensed recitation that fixes what the skeleton cannot record — and the proof it worked is philological: the regional codices’ small divergences form a family tree pointing at a single written archetype.

Movement ThreeThe Canon

What is in, what is out, who decided, when, and what the decision cost. The real subject is who had the authority to close it. The movement opens with that argument, never with a manuscript.

12What There Was to CloseAt Muhammad’s death (conventionally dated 632) there was no book — the tradition’s own accounts inventory the text on palm-stalks, shoulder-blades and in the memories of men, and companions kept private codices that differed in order and, the reports say, in contents: Ibn Mas’ud’s lacked three suras of the now-standard text, Ubayy’s carried two extra.
13The Caliph’s File CopyThe tradition holds that the first collection was made under Abu Bakr as sheets kept privately — passing to Umar and then to Umar’s daughter Hafsa, not to the treasury or the mosque — which is to say the community’s first complete text was an archival object, and no authority had yet been asserted over what people recited.
14Uthman Closes the TextConventionally dated around 650, the third caliph had a committee produce master copies, sent them to the garrison cities, and ordered other copies destroyed — the verb varies between reports — an administrative act of state power over scripture which the tradition remembers openly rather than hides, and which is testable: every surviving early manuscript descends from one archetype.
15What the Closing CostThe closing had a price the tradition itself records — Ibn Mas’ud in Kufa told his students to hide their codices rather than surrender them, and withdrawn readings survived for centuries in the commentaries — and it had a remainder: fixing the skeleton did not fix the voice, and readings within the standard skeleton stayed plural until long after this volume ends.
16The Canon That Stayed OpenThe community that closed its scripture inside a decade had a second corpus — reports of what Muhammad said and did — with no Uthman: in this period it was not systematically written, the tradition transmits instructions both to write it down and not to, and everything the succession dispute would need — precedents, appointments, spoken designations — lived in that unclosed canon.

Movement FourThe Money

Who paid, who was paid, and how we know. Numbers get their provenance in the same sentence.

17The Alms That Became a TaxZakat is paired with prayer throughout the Qur’an as the double mark of membership, but the text sets no rates — the familiar percentages are later jurists’ work, and the chapter says so — and under Muhammad, the tradition holds, it became something no Arabian shrine had had before: a levy collected by named agents, on behalf of a centre, from tribes that had submitted.
18The RefusalWithin months of Muhammad’s death, tribes across the peninsula announced they would keep the prayer and stop the payment; the tradition calls what followed the apostasy wars, and its own operational detail says taxation — Abu Bakr is reported saying he would fight any who withheld so much as a hobbling-cord of what they had paid — so the community’s first war was, on the tradition’s own telling, a tax enforcement action.
19The Fifth, and the FarmThe spoils verse of the eighth sura assigns a fifth of captured wealth to God, the messenger and his kin — which made the founder’s estate a fiscal institution — and within months of his death the community’s first constitutional argument was about exactly that: Fatima claimed her father’s lands at Fadak, Abu Bakr refused on the strength of a reported saying that prophets leave no heirs, and money and succession were the same question from the start.
20The RegisterConquest pay was systematised, the tradition holds, under Umar in about 641 as the diwan — a register of stipends ranked not by tribe or wealth but by precedence in the faith, with Muhammad’s widows at the head of the scale — so that within a decade of the founder’s death, piety had a payroll and seniority in belief was a class system with a cash value.
21The Land That Was Not DividedThe richest prize of the conquests, the irrigated Sawad of Iraq, was not carved up among the conquerors — the caliphate kept the land in place, taxed it, and paid the army from the revenue — a decision that created a fiscal state and a standing question, who controls the surplus, to which the transmitted grievance lists against the third caliph are substantially answers.

Movement FiveThe Split

What divided it, what it became, what it handed on. Names the next volume by NAME, never by number.

22The Meeting in the PorticoMuhammad left no surviving written designation of a successor; the choice was made at an unminuted gathering in a Medinan portico on, the tradition holds, the day of his death — and the contradictory reports of it, plus the Ghadir Khumm addresses months earlier, which all parties transmit and each side reads as settling the matter its own way, became the texts over which the community would divide.
23Blood on the CodexIn June 656 the caliph who had standardised the scripture was besieged in his own house by Muslims and killed — the tradition says while reading his codex — and the grievance lists against him (kinsmen appointed, revenue steered, codices burned) survive only in words written down under regimes and sects that were parties to the quarrel that followed.
24Pages on LancesAt Siffin in the summer of 657 two Muslim armies fought under the same scripture, and the battle ended, the accounts say, when one side raised codex leaves on lance-points and called for judgment by what was written on them — the standardised book was at most seven years old, and it was already the one authority both sides had to acknowledge.
25The Arbitration, and the Men Who LeftThe written agreement to arbitrate survives in two transmitted versions — in one of which Ali’s title “Commander of the Believers” is struck out at the other side’s insistence, an erasure the tradition pairs with the treaty of Hudaybiyya a generation earlier — and part of Ali’s own army seceded over the procedure itself, arguing that a dispute under God’s rule may not be submitted to human arbitrators; the first permanent sect in Islam separated over process, not theology, and was cut down at Nahrawan in 658.
26The Year of the CommunityAli was struck down in the mosque at Kufa in Ramadan 40 (January 661) by a Kharijite; his son Hasan settled with Mu’awiya for terms, the tradition calls 661 “the year of the community” — and for the first time the ruler of the believers is visible in datable stone and coin. But nothing settled in 661 answered the question of 632, and the unfinished business — the succession’s legitimacy, the unclosed hadith canon, the fifth owed to the kin — detonates at Karbala in 680, which is where the Sunni and Shia volume begins.

What this volume cannot establish

Printed in the book before the argument is, and printed here for the same reason. A history of a tradition that will not say where its evidence runs out is not a history.

What was recited in Mecca in 610, or what stood in anyone’s memory in 632. The manuscripts show that a stable consonantal text existed extremely early; they cannot reach behind their own parchment. (This is the volume’s signature limit and it is stated inside Movement Two, not saved for a coda.)
Pre-Islamic Mecca from contemporary evidence. No datable pre-Islamic text names the city — the equation with Ptolemy’s “Macoraba” is contested — and the earliest surviving description of its sanctuary economy is al-Azraqi’s, from the mid-ninth century.
The life of Muhammad beyond a hard minimum. Near-contemporary outsiders establish that by 634 an Arab movement with a prophet existed, and Greek and Syriac writers knew his name within a decade; the connected biography is Ibn Ishaq’s, more than a century after the events, surviving only in a recension whose editor states he cut material.
Whether any speech in the succession dispute was said as transmitted. The Saqifa accounts, the Ghadir Khumm addresses and the grievance lists against Uthman were fixed in writing under dynasties and sects that were parties to the outcome. The Ghadir reports exist early and on all sides; what their operative word meant cannot be established from the reports alone.
The date and mechanics of Uthman’s recension. beyond “conventionally dated around 650” — how many master copies, to which cities (the lists vary), and what the withdrawn codices contained beyond the recorded variants and one surviving palimpsest.
Any number. Army sizes, casualties, stipend scales and revenue totals are transmitted in ninth- and tenth-century compilations. They are the tradition’s figures, not audits, and they are printed here only as such.
Anyone’s interior state.

Islam, three ignitions

This tradition ignites more than once in the chronology. The other slots are Sunni and Shia, Sufism. See the thread.

Sources

63 works, dated, and specific enough that a reader can go and look at the thing. Nothing is on this list that the desk did not consult.

The bibliography, as the volume prints it
#	Document or work	Date	Evidence for
1	al-Azraqi, Akhbar Makka	mid-ninth century	earliest surviving description of Mecca’s sanctuary economy
2	Ian D. Morris, “Mecca and Macoraba,” al-'Usur al-Wusta	2018	the contested equation of Mecca with Ptolemy’s Macoraba
3	Qur’an, Sura 2:198, Marmaduke Pickthall, The Meaning of the Glorious Koran	1930	permission to trade during the pilgrimage
4	W. Montgomery Watt, Muhammad at Mecca	1953	the standard trade-republic picture of Quraysh
5	Patricia Crone, Meccan Trade and the Rise of Islam	1987	demolition of the incense-route picture
6	Qur’an, sura of Quraysh, Pickthall	1930	the two named winter and summer caravan journeys
7	Constitution of Medina, in Ibn Ishaq via Ibn Hisham and in Abu Ubayd, Kitab al-Amwal	8th–9th c. transmission	blood-money and alliance clauses of an oasis without a judge
8	Michael Lecker, The “Constitution of Medina”: Muhammad’s First Legal Document	2004	why the document’s archaic terminology is treated as early
9	Robert Hoyland, Arabia and the Arabs	2001	the law of the feud
10	Himyarite royal inscriptions, catalogued by Christian Julien Robin	late fourth century onward	monotheist formulae naming Rahmanan
11	Letter of Simeon of Beth Arsham	c. 524	the Najran massacre
12	Abraha’s Marib dam inscription, CIH 541	548 CE	Christian invocation; dated campaigns that do not name Mecca
13	Ahmad Al-Jallad, The Religion and Rituals of the Nomads of Pre-Islamic Arabia	2022	pre-Islamic desert graffiti
14	Chronicon Paschale	ends 628	Heraclius’s victory dispatch
15	George of Pisidia, court poems	620s	the last Roman–Persian war as a court witness
16	Antiochus Strategos on the sack of Jerusalem	after 614	the 614 sack; casualty figures as genre, not a count
17	James Howard-Johnston, The Last Great War of Antiquity	2021	the 602–628 war and the dismantled Arab client kingdoms
18	Ibn Ishaq, Sirat Rasul Allah, in Ibn Hisham’s recension	Ibn Ishaq d. 767; Ibn Hisham d. 833	the connected biography; the editor’s preface that he omitted material
19	Doctrina Jacobi nuper baptizati	Greek, set in 634	a prophet appearing among the Saracens
20	Thomas the Presbyter, Syriac chronicle	c. 640	earliest dated text to name Muhammad (Seleucid 945 / 634)
21	Robert Hoyland, Seeing Islam as Others Saw It	1997	the file of near-contemporary outside notices
22	al-Bukhari, chapters on the collection of the Qur’an	compiled 9th c.	Yamama deaths and Zayd ibn Thabit’s commission
23	Codex Parisino-petropolitanus, BnF Arabe 328	Hijazi, 7th–8th c.	the defective consonantal skeleton
24	François Déroche, La transmission écrite du Coran dans les débuts de l’islam	2009	analysis of the early Hijazi copies
25	François Déroche, Qur’ans of the Umayyads	2014	Umayyad-period written transmission
26	Marijn van Putten, orthographic studies of defective spellings	2010s	what the skeleton cannot record
27	Mingana Islamic Arabic 1572, University of Birmingham; Oxford radiocarbon result	2015	parchment dated 568–645 CE at 95.4 per cent confidence
28	Behnam Sadeghi and Mohsen Goudarzi, “San’a’ 1 and the Origins of the Qur’an,” Der Islam	2012	the lower text of the Sana’a palimpsest, DAM 01-27.1
29	Asma Hilali, The Sanaa Palimpsest	2017	a more cautious reading of the same object
30	Behnam Sadeghi and Uwe Bergmann, radiocarbon work, Arabica	2010	parchment date of the palimpsest
31	Michael Cook, “The Stemma of the Regional Codices of the Koran”	2004	family tree of regional copies pointing at one archetype
32	Marijn van Putten, “'The Grace of God’ as evidence for a written Uthmanic archetype,” BSOAS	2019	idiosyncratic spellings copied letter-for-letter
33	Ibn Abi Dawud, Kitab al-Masahif	early tenth century	companion codices (Ibn Mas’ud, Ubayy)
34	Arthur Jeffery, Materials for the History of the Text of the Qur’an	1937	catalogue of the masahif literature
35	Harald Motzki, “The Collection of the Qur’an,” Der Islam	2001	isnad-cum-matn tracing the collection account to al-Zuhri (d. 742)
36	Nicolai Sinai, “When did the consonantal skeleton of the Quran reach closure?,” BSOAS	2014	late-closure thesis set against the manuscript file
37	Estelle Whelan, “Forgotten Witness,” JAOS	1998	early epigraphic witness to a closed text
38	Ibn Mujahid, Kitab al-Sab’a	d. 936	the fixing of seven readings after the skeleton closed
39	al-Khatib al-Baghdadi, Taqyid al-'Ilm	eleventh century (d. 1071)	contradictory reports on writing hadith
40	Michael Cook, “The Opponents of the Writing of Tradition in Early Islam,” Arabica	1997	the unclosed second canon
41	Gregor Schoeler, The Oral and the Written in Early Islam	2006 English	oral versus written hadith in this period
42	Abu Ubayd al-Qasim ibn Sallam, Kitab al-Amwal	author d. 838	earliest surviving treatise of Islamic public finance
43	al-Tabari, Ridda narratives (much via Sayf ibn Umar)	turn of the tenth century	the refusal of zakat after 632
44	Elias Shoufani, Al-Riddah and the Muslim Conquest of Arabia	1973	the apostasy wars as tax enforcement
45	Qur’an 8, spoils verse, Pickthall	1930	assignment of the fifth
46	al-Baladhuri, Futuh al-Buldan, Philip Hitti trans.	9th c.; Hitti 1916	Fadak and the no-heirs saying
47	Wilferd Madelung, The Succession to Muhammad	1997	money and succession as one question
48	Hugh Kennedy, The Armies of the Caliphs	2001	reconstruction of Umar’s stipend register
49	PERF 558, bilingual Greek–Arabic papyrus, Herakleopolis	Jumada I, year 22 / April 643	requisition receipt of the new era
50	Zuhayr rock inscription, year 24 (Ali Ghabban; Robert Hoyland trans.)	644; pub. 2008	Umar’s death dated in stone
51	Abu Yusuf, Kitab al-Kharaj	written for Harun al-Rashid, c. 150 years after the decisions	the Sawad kept and taxed
52	Daniel Dennett, Conversion and the Poll Tax in Early Islam	1950	treaty and revenue texts assessed case by case
53	Ghadir Khumm reports in Ibn Hanbal, Musnad	9th c. compilation	the contested word mawla
54	al-Baladhuri, Ansab al-Ashraf	9th century	the killing of Uthman, 656
55	Sebeos	Armenian, 660s	near-contemporary notice of civil war among the conquerors
56	Nasr ibn Muzahim al-Minqari, Waq’at Siffin	author d. 827	earliest surviving monograph on the battle
57	al-Tabari via Abu Mikhnaf	Abu Mikhnaf d. 774	Siffin and the arbitration
58	Martin Hinds, “The Siffin Arbitration Agreement,” Journal of Semitic Studies	1972	the two transmitted versions of the agreement
59	Maronite Chronicle	Syriac, 660s	Mu’awiya’s accession at Jerusalem; coinage that dropped the cross
60	Hammat Gader bathhouse inscription	year 42 of the Arabs (662/3)	“the servant of God Mu’awiya, commander of the faithful”
61	Dam inscription near Ta’if	58 AH (George Miles, 1948)	Mu’awiya named as builder
62	Arab-Sasanian drachms in Mu’awiya’s name	660s	the first ruler of the believers on coin
63	Pew Research Center, global Muslim population estimate	published 2025	the living count in the opening paragraph — not evidence for the seventh century

Common questions

When is Islam first securely attested?

610 CE. That is the earliest date the surviving evidence supports, and it is the key this volume is filed under. It is not the same claim as when the tradition began: the tradition’s own account of its beginning is printed beside it in the book’s front matter, never instead of it.

What period does the book cover?

610–661. The coverage span and the attestation date are two different facts, and the gap between them is often the most interesting thing about a volume.

What carried Islam?

The standardised codex, and the memoriser. Movement Two of every volume is about the transmission technology — the concrete object or institution that made copying possible, and would not have existed a century earlier.

How long is it?

About 32,000 words, roughly 130 minutes’ reading, 134 pages in the paperback. The audiobook runs 3h 12m, measured from the finished masters.

What does this volume not claim to establish?

What was recited in Mecca in 610, or what stood in anyone’s memory in 632. The manuscripts show that a stable consonantal text existed extremely early; they cannot reach behind their own parchment. (This is the volume’s signature limit and it is stated inside Movement Two, not saved for a coda.) Every volume prints its own limits before it prints its argument; there are 7 on this one’s list.

Does the book say which tradition is true?

No. Belief System does not adjudicate. Where the tradition’s account and the earliest evidence disagree, both are given, in their own vocabulary, attributed, and neither is given the last word. No miracle is reported as fact and none is denied.

Can I buy Islam in print?

Yes. It is a 5.5 × 8.5 in paperback, 134 pages, $12.99 a copy, bought direct rather than through a retailer. All thirty-one together are $199.00, with the ebook and the audiobook of every volume included.

The edition

The paperback
Trim	5.5 × 8.5 in / 140 × 216 mm
Binding	Perfect-bound paperback, matte laminate
Paper	60# uncoated cream, black ink
Extent	134 pages

Editions and ISBNs
Edition	Extent	ISBN
Paperback	134 pages, 5.5 × 8.5 in	ISBN not yet issued
Ebook	EPUB	979-8-90623-845-0
Audiobook	3h 12m	979-8-90623-846-7

Filed under
BISAC 1	`REL017000` RELIGION / Comparative Religion
BISAC 2	`REL037010` RELIGION / Islam / History
BISAC 3	`REL033000` RELIGION / History

early Islamic history
Quran manuscripts
Uthmanic recension
Sunni Shia split
Rashidun caliphate
Islamic origins
history of religion
Birmingham Quran
first fitna
comparative religion