Apologetics · FaithLogic
The Quran's Preservation Problem — Variants, Manuscripts, and Uthman's Fire
Muslims claim the Quran is the perfectly preserved word of Allah — unchanged from the moment of revelation to Muhammad. This is one of Islam's strongest apologetic arguments. This section examines the historical and manuscript evidence and shows that the Quran's preservation is considerably more complex — and more contested — than Islamic tradition acknowledges.
The Historical Record from Islamic Sources
Uthman Burned Quranic Manuscripts — Because Variants Already Existed
Sahih Bukhari 6:61:510 (complete text):
Hudhayfa ibn al-Yaman came to Uthman while people from Syria and Iraq were fighting together in Armenia. Hudhayfa saw their differences in Quranic recitation and came to Uthman saying: "O chief of the Believers! Save this nation before they differ about the Book (Quran) as Jews and the Christians did."
Uthman then instructed Zaid ibn Thabit and a committee to standardize the text. "When they had written many copies, Uthman returned the original manuscripts to Hafsa and sent to every Muslim province one copy of what they had copied, and ordered that all the other Quranic materials, whether written in fragmentary manuscripts or whole copies, be burnt."
What this establishes:
(1) Within 20 years of Muhammad's death, regional communities already had different versions of the Quran
(2) The differences were serious enough that Hudhayfa feared they would cause the same disputes among Muslims that Bible differences had caused among Jews and Christians
(3) Uthman's solution was to standardize ONE version and destroy all others — permanently eliminating our ability to compare what the variants contained
(4) The burning of Muslim scripture by a Caliph is itself extraordinary — if the Quran was perfectly preserved, why burn anything?
Hudhayfa ibn al-Yaman came to Uthman while people from Syria and Iraq were fighting together in Armenia. Hudhayfa saw their differences in Quranic recitation and came to Uthman saying: "O chief of the Believers! Save this nation before they differ about the Book (Quran) as Jews and the Christians did."
Uthman then instructed Zaid ibn Thabit and a committee to standardize the text. "When they had written many copies, Uthman returned the original manuscripts to Hafsa and sent to every Muslim province one copy of what they had copied, and ordered that all the other Quranic materials, whether written in fragmentary manuscripts or whole copies, be burnt."
What this establishes:
(1) Within 20 years of Muhammad's death, regional communities already had different versions of the Quran
(2) The differences were serious enough that Hudhayfa feared they would cause the same disputes among Muslims that Bible differences had caused among Jews and Christians
(3) Uthman's solution was to standardize ONE version and destroy all others — permanently eliminating our ability to compare what the variants contained
(4) The burning of Muslim scripture by a Caliph is itself extraordinary — if the Quran was perfectly preserved, why burn anything?
Ibn Masud's Quran — The Companion Who Refused to Hand Over His Copy
The historical record (Tabarani, Musnad Ahmad, Ibn Abi Dawud's Kitab al-Masahif):
Abdullah ibn Masud was one of Muhammad's closest companions and one of the four people Muhammad himself directed Muslims to learn the Quran from (Sahih Bukhari 6:61:521: "Learn the Quran from four persons: Abdullah ibn Masud..."). Ibn Masud had his own personal copy of the Quran.
When Uthman's committee came to collect copies for burning, Ibn Masud refused to hand over his copy. He said: "I will not give up a copy I received from the mouth of the Messenger of Allah for something that was written by Zaid ibn Thabit." He called the Uthman recension "a poor copy" and said "Ibn Masud's recension was better."
What Ibn Masud's copy contained differently:
According to Ibn Abi Dawud's Kitab al-Masahif, Ibn Masud's version:
— Did not include Surah 1 (Al-Fatiha) as part of the Quran
— Did not include Surah 113 (Al-Falaq)
— Did not include Surah 114 (An-Nas)
— These three chapters are present in every Quran used today
The theological problem:
One of the four people Muhammad himself designated as authorities on the Quran had a version missing three chapters — and refused to accept the standardization. Either Ibn Masud was wrong (meaning Muhammad's own designated Quran authority was unreliable) or Uthman was wrong (meaning the Quran we have today is missing material that was in earlier versions). Neither option is comfortable for Islamic theology.
Abdullah ibn Masud was one of Muhammad's closest companions and one of the four people Muhammad himself directed Muslims to learn the Quran from (Sahih Bukhari 6:61:521: "Learn the Quran from four persons: Abdullah ibn Masud..."). Ibn Masud had his own personal copy of the Quran.
When Uthman's committee came to collect copies for burning, Ibn Masud refused to hand over his copy. He said: "I will not give up a copy I received from the mouth of the Messenger of Allah for something that was written by Zaid ibn Thabit." He called the Uthman recension "a poor copy" and said "Ibn Masud's recension was better."
What Ibn Masud's copy contained differently:
According to Ibn Abi Dawud's Kitab al-Masahif, Ibn Masud's version:
— Did not include Surah 1 (Al-Fatiha) as part of the Quran
— Did not include Surah 113 (Al-Falaq)
— Did not include Surah 114 (An-Nas)
— These three chapters are present in every Quran used today
The theological problem:
One of the four people Muhammad himself designated as authorities on the Quran had a version missing three chapters — and refused to accept the standardization. Either Ibn Masud was wrong (meaning Muhammad's own designated Quran authority was unreliable) or Uthman was wrong (meaning the Quran we have today is missing material that was in earlier versions). Neither option is comfortable for Islamic theology.
The Manuscript Evidence
The Sana'a Manuscripts — The Oldest Quranic Parchments Show Variants
Discovery:
In 1972, workers restoring the Great Mosque of Sana'a in Yemen discovered a cache of ancient Quranic manuscripts in the loft — approximately 40,000 parchment fragments. Carbon dating places the oldest fragments to the late 7th to early 8th century AD — making them among the oldest physical Quranic manuscripts in existence.
What the Sana'a manuscripts show:
German scholar Gerd-R. Puin was given access to study the manuscripts and reported in 1999 (Atlantic Monthly interview):
"Some of the parchment pages in the Yemeni hoard seemed to date back to the seventh and eighth centuries AD or even earlier — they were palimpsests... The margins contain textual variants — some quite significant — suggesting that the text of the Quran was a fluid, developing text."
The manuscripts show:
— Variant readings in the text itself (different words from the standard Uthman text)
— Verse ordering that differs from the standard text
— Some pages appear to be palimpsests — earlier text washed off and rewritten — suggesting the Quran was not fixed but edited
The Yemeni government's response:
The Yemeni authorities restricted further scholarly access to the manuscripts after Puin's report — which prevented full academic investigation. The very restriction of access to the oldest Quranic manuscripts is itself theologically significant.
In 1972, workers restoring the Great Mosque of Sana'a in Yemen discovered a cache of ancient Quranic manuscripts in the loft — approximately 40,000 parchment fragments. Carbon dating places the oldest fragments to the late 7th to early 8th century AD — making them among the oldest physical Quranic manuscripts in existence.
What the Sana'a manuscripts show:
German scholar Gerd-R. Puin was given access to study the manuscripts and reported in 1999 (Atlantic Monthly interview):
"Some of the parchment pages in the Yemeni hoard seemed to date back to the seventh and eighth centuries AD or even earlier — they were palimpsests... The margins contain textual variants — some quite significant — suggesting that the text of the Quran was a fluid, developing text."
The manuscripts show:
— Variant readings in the text itself (different words from the standard Uthman text)
— Verse ordering that differs from the standard text
— Some pages appear to be palimpsests — earlier text washed off and rewritten — suggesting the Quran was not fixed but edited
The Yemeni government's response:
The Yemeni authorities restricted further scholarly access to the manuscripts after Puin's report — which prevented full academic investigation. The very restriction of access to the oldest Quranic manuscripts is itself theologically significant.
The Seven Ahruf and Hafs vs Warsh — The "Perfectly Preserved" Quran Has Multiple Versions
The seven ahruf (modes of recitation):
Islamic tradition records that Muhammad received permission from Allah to recite the Quran in seven different ways (ahruf — dialects or modes). Sahih Bukhari 6:61:514: "This Quran has been revealed to be recited in seven different ways, so recite of it whichever is easier for you."
The qira'at (formal recitation traditions):
From the seven ahruf, fourteen major recitation traditions (qira'at) developed, of which seven are considered "mutawatir" (mass-transmitted) and therefore fully authoritative. The two most widely used today are:
— Hafs 'an 'Asim: used in most of the Muslim world, including Saudi Arabia, Egypt, most of Asia
— Warsh 'an Nafi': used in North and West Africa, including Morocco, Algeria, Libya, Senegal
The differences between Hafs and Warsh:
These are not merely accent differences — they involve thousands of variant readings across the text, including:
— Different words in specific verses
— Different vowelizations changing meanings
— Different verse numberings (the two traditions have different verse counts for many chapters)
— Specific theological differences in certain passages
The implication:
If the Quran is perfectly preserved, then either Hafs or Warsh is perfectly preserved — but not both, since they differ. Or both represent valid recitation traditions, in which case there is no single perfectly preserved Quran — there are at least two (and historically fourteen) authoritative versions. The claim of perfect preservation requires specifying which version is perfect.
Islamic tradition records that Muhammad received permission from Allah to recite the Quran in seven different ways (ahruf — dialects or modes). Sahih Bukhari 6:61:514: "This Quran has been revealed to be recited in seven different ways, so recite of it whichever is easier for you."
The qira'at (formal recitation traditions):
From the seven ahruf, fourteen major recitation traditions (qira'at) developed, of which seven are considered "mutawatir" (mass-transmitted) and therefore fully authoritative. The two most widely used today are:
— Hafs 'an 'Asim: used in most of the Muslim world, including Saudi Arabia, Egypt, most of Asia
— Warsh 'an Nafi': used in North and West Africa, including Morocco, Algeria, Libya, Senegal
The differences between Hafs and Warsh:
These are not merely accent differences — they involve thousands of variant readings across the text, including:
— Different words in specific verses
— Different vowelizations changing meanings
— Different verse numberings (the two traditions have different verse counts for many chapters)
— Specific theological differences in certain passages
The implication:
If the Quran is perfectly preserved, then either Hafs or Warsh is perfectly preserved — but not both, since they differ. Or both represent valid recitation traditions, in which case there is no single perfectly preserved Quran — there are at least two (and historically fourteen) authoritative versions. The claim of perfect preservation requires specifying which version is perfect.
The Birmingham Quran Fragment — And Why Its Dating Creates Questions
The discovery (2015):
The University of Birmingham announced the discovery of Quranic parchment fragments in its Mingana Collection, carbon dated by the Oxford Radiocarbon Accelerator Unit to 568-645 AD (95.4% probability).
The theological problem with the early date:
Muhammad is traditionally dated to 570-632 AD. Carbon dating of the Birmingham fragment gives a range of 568-645 AD. The lower end of this range (568 AD) predates Muhammad's birth. The upper end (645 AD) is after his death but well within the period before Uthman's standardization (~650 AD).
Three possible explanations — all problematic for Islamic theology:
(1) The parchment was old when written on — standard practice (scribes reused parchment). This is the most likely explanation, but it means carbon dating cannot definitively date the writing itself — only the material.
(2) The fragment dates from Muhammad's own lifetime and represents an early version of the text — but then it should match the Uthman standardization perfectly. If it does not, it shows variants existed from the beginning.
(3) The fragment predates Muhammad — in which case questions arise about the Quran's historical origins.
What the fragment shows about the text:
The Birmingham fragment contains text from Surah 18 and 19 and is remarkably similar to the standard Uthman text — which suggests the core text was stable relatively early. However, even small variants are significant for a document claimed to be perfectly preserved.
The University of Birmingham announced the discovery of Quranic parchment fragments in its Mingana Collection, carbon dated by the Oxford Radiocarbon Accelerator Unit to 568-645 AD (95.4% probability).
The theological problem with the early date:
Muhammad is traditionally dated to 570-632 AD. Carbon dating of the Birmingham fragment gives a range of 568-645 AD. The lower end of this range (568 AD) predates Muhammad's birth. The upper end (645 AD) is after his death but well within the period before Uthman's standardization (~650 AD).
Three possible explanations — all problematic for Islamic theology:
(1) The parchment was old when written on — standard practice (scribes reused parchment). This is the most likely explanation, but it means carbon dating cannot definitively date the writing itself — only the material.
(2) The fragment dates from Muhammad's own lifetime and represents an early version of the text — but then it should match the Uthman standardization perfectly. If it does not, it shows variants existed from the beginning.
(3) The fragment predates Muhammad — in which case questions arise about the Quran's historical origins.
What the fragment shows about the text:
The Birmingham fragment contains text from Surah 18 and 19 and is remarkably similar to the standard Uthman text — which suggests the core text was stable relatively early. However, even small variants are significant for a document claimed to be perfectly preserved.
💡 Quran Preservation — The Evidence Summary"The Islamic claim of perfect Quranic preservation faces serious historical challenges from Islamic sources themselves: (1) Sahih Bukhari 6:61:510 records Uthman burning all variant Quranic manuscripts in ~650 AD because regional communities already had different versions — if the Quran was perfectly preserved, why was there anything to burn? (2) Abdullah ibn Masud — one of the four people Muhammad himself designated as Quran authorities (Bukhari 6:61:521) — refused to hand over his copy and called Uthman's standardization 'inferior.' His version was missing Surahs 1, 113, and 114. (3) The Sana'a manuscripts (Yemen, 7th-8th century AD, oldest physical Quran fragments) show textual variants and possible palimpsest editing — scholar Gerd-R. Puin reported 'significant variant readings.' The Yemeni government then restricted access. (4) Two major recitation traditions (Hafs and Warsh) used by hundreds of millions of Muslims today differ in thousands of variant readings — both cannot be the 'perfectly preserved' Quran. Compare with the NT: 5,700+ Greek manuscripts from multiple countries, centuries, and language traditions, all showing remarkable consistency — and full scholarly access has always been available. The NT's textual tradition is better documented and more consistently preserved than the Quran's."
This is part of the FaithLogic study library — the biblical, historical and manuscript case that Jesus is God, with every claim linked to its source. Open the interactive version for search, cross-references, read-aloud, and the full library.