Lost Verses, the Sana'a Palimpsest, and Uthman's Codices

Critics often combine three different issues - reports about abrogated recitation, the Sanaa palimpsest, and Uthman's standardization of the mushaf - and present them as proof that the Quran once existed as competing scriptures. The evidence does not show the Quran being rewritten centuries later. It shows an early period of oral and written transmission, private codices and variant written materials, then a public standardization of the written mushaf while the Companions were still alive.

What Quran 15:9 actually promises

Indeed, it is We who sent down the Reminder, and indeed, We will surely guard it.

Quran 15:9

The "Reminder" here is the Quran. This promise does not require that every private sheet, teaching note, or personal codex written by every Muslim be identical. Nor does it mean revelation was never recited in more than one authorized mode during the Prophet's lifetime. The claim is that the Quran revealed to Muhammad (PBUH), in its final form after the completion of revelation, was preserved by Allah and transmitted to the Muslim community through public recitation and written preservation. It is not that every early personal manuscript had canonical authority.

Lost verses or abrogated recitation?

Some authentic reports mention wording that was once recited and later was no longer recited as Quran. These reports must be separated from the claim that part of the final Quran was accidentally lost. The Quran itself recognizes abrogation during the period of revelation:

We do not abrogate a verse or cause it to be forgotten except that We bring forth one better than it or similar to it.

Quran 2:106

It also states:

We will make you recite, and you will not forget, except what Allah wills.

Quran 87:6-7

So the key distinction is this: wording deliberately removed from Quranic recitation by revelation is not a "lost verse." It is abrogated recitation. Classical Sunni scholars discussed this under naskh al-tilawah, abrogation of recited wording. Abrogation occurred only during the Prophet's lifetime; after his death, no one had authority to remove anything from the Quran.

• The report about stoning: Umar's statement about wording connected to stoning is authentically reported in Sahih al-Bukhari and Sahih Muslim. The Sunni explanation is not that the Quran is missing a verse, but that the recitation of that wording was abrogated while the legal ruling remained established by the Sunnah, the practice of the Prophet (PBUH), and the consensus of the Companions.
• The report about suckling: Aisha's report in Sahih Muslim mentions "ten known sucklings" and then "five known sucklings." It also contains a phrase which scholars discussed carefully: that the Prophet (PBUH) died while this was among what was recited. A standard Sunni reconciliation is that the abrogation of its recitation occurred very late, so knowledge of that abrogation was not immediately equal among every individual. In any case, this wording is not part of the final Quran by the consensus of the Companions and the Muslim community.
• The report of Abu Musa al-Ashari: Sahih Muslim reports Abu Musa speaking of passages that had once been recited and were then forgotten except for remembered phrases, including wording about the son of Adam desiring more wealth. Sunni scholars treat this as material whose recitation did not remain part of the final Quran, not as proof that a surah from the final Quran disappeared after the Prophet's death.
• The report about a goat or sheep eating a page: This common internet citation is not in Sahih al-Bukhari or Sahih Muslim. It is reported in Sunan Ibn Majah, and the specific addition about a domestic animal eating the sheet has been disputed by hadith scholars, with significant criticism of its chain and wording. Even if one assumed the report to be acceptable, it would still not prove loss of Quran: a private sheet being damaged would not erase anything preserved by the public recitation and memorization of the Muslim community.

The Prophet (PBUH) remained the living authority who taught the Companions what counted as Quran until his death. Authentic reports state that Jibril reviewed the Quran with him, and that in the final Ramadan this review occurred twice. This is central to the Muslim understanding that the final recited Quran was known before the Prophet (PBUH) passed away.

What the Sanaa palimpsest is

In 1972, restoration work at the Great Mosque of Sanaa in Yemen uncovered a large cache of ancient Quranic manuscript fragments in a space between the ceiling and the roof. The cache is commonly described as containing roughly 12,000 Quranic parchment fragments, later grouped into about 926 incomplete manuscripts. One of them, known as Sanaa 1, part of it catalogued as DAM 01-27.1, is a palimpsest: parchment that was written on, erased by washing or scraping, and then reused. Traces of the erased writing remain visible, so the folios preserve two layers of Quranic text:

Both layers are Quranic material. Behnam Sadeghi and Mohsen Goudarzi argue that the two layers reflect codices rather than random excerpts, with the upper layer belonging to the Uthmanic textual tradition and the lower layer representing a very early non-Uthmanic written tradition. Some details of interpretation remain debated among manuscript specialists, but the basic point is clear: the lower text is very early and contains Quranic passages with variants from the standard Uthmanic text.

How old is it?

The lower text is exceptionally early. On paleographic grounds it belongs to the seventh century. Radiocarbon testing of one folio gave roughly a 99% probability that the parchment predates 671 CE, 95.5% before 661 CE, and 75% before 646 CE. For perspective, the Prophet (PBUH) died in 632 CE and Uthman became caliph in 644 CE.

Two qualifications matter. First, radiocarbon dating measures the age of the animal parchment, not the ink, so it cannot prove the exact year the words were written. Second, because the lower text includes material from Surah al-Tawbah, which belongs to the final phase of revelation in the standard Islamic chronology, the written text cannot be earlier than that final phase. A fair description is that the lower Sanaa text was produced very early, during or close to the generation of the Companions.

What is different in the lower text?

The differences should not be dismissed as mere spelling. Measured against the standard Uthmanic text, the lower Sanaa text includes:

• substitution of certain words;
• omission or addition of words and short phrases;
• differences in pronouns, prefixes, and suffixes;
• grammatical and verbal variations;
• changes in word order;
• occasional apparent scribal errors;
• and a different ordering of some surahs.

But the evidence must be kept in proportion. The lower codex survives only in fragments, so its complete original contents cannot be inspected. In the surviving portions, however, we find familiar Quranic surahs and passages, not an alternative scripture with a different prophet, theology, law, or narrative world. The variation occurs within the transmission of Quranic material, not between Islam and some other religion.

The strongest objection, stated plainly

The serious objection is not "an old Quran has spelling differences." It is this:

If the Quran was already known and recited in the Prophet's lifetime, why does such an early manuscript contain additions, omissions, substitutions, and a different surah order? Does this mean several different Qurans existed and Uthman simply chose one?

Does Sanaa show the Quran was rewritten? No

The manuscript fits remarkably well with what Islamic sources already say. Muslim historians never claimed that, before Uthman, every Companion's personal written collection was an identical mass-produced book. The Quran was carried first and foremost by a living community of reciters and memorizers, while Companions also kept written materials and personal codices.

Islamic literature openly records readings and codices associated with Companions such as Abdullah ibn Masud, Ubayy ibn Kab, and Abu Musa al-Ashari, though individual reports vary in reliability and interpretation. These reports existed for centuries before the Sanaa manuscripts were discovered.

So Sanaa did not expose a secret history that Muslims had hidden. It provided physical evidence for the existence of early non-Uthmanic written traditions of the general kind already known from Muslim sources. It should not be over-read: Sanaa 1 cannot be identified with certainty as the codex of Ibn Masud, Ubayy, or any other named Companion, and one cannot simply declare every variant to be one of the seven ahruf. Accepted Quranic recitation is established by transmission from the reciting community, not by a solitary manuscript.

What Uthman actually did

As Islam spread, disputes over recitation appeared among Muslims from different regions. Hudhayfah ibn al-Yaman, alarmed by differences between Muslims of Iraq and Syria, urged Uthman to act before the community differed over the Book as Jews and Christians had differed over their scriptures.

Uthman obtained the Quranic sheets preserved with Hafsa, which went back to the earlier collection made under Abu Bakr. He appointed a committee whose names in Sahih al-Bukhari include Zayd ibn Thabit, Abdullah ibn al-Zubayr, Said ibn al-As, and Abd al-Rahman ibn al-Harith ibn Hisham. His instruction was that if they differed with Zayd, they should write it in the dialect of Quraysh, because the Quran was revealed in their tongue.

Uthman sent to every Muslim province one copy of what they had copied, and ordered that every other Quranic material, whether written in fragmentary sheets or whole codices, be burned.

Sahih al-Bukhari 4987

There is no reason to soften that wording. The honest response is to state it directly and then explain why it was done. This was not a secret editing project centuries later. It occurred in the lifetime of the Companions, under a caliph who had known the Prophet (PBUH), using the sheets traced back to Abu Bakr's collection, with leading Companions involved.

Why the other codices were burned

The burning of non-standard codices is sometimes presented as Uthman discovering several contradictory Qurans and destroying the evidence. That goes beyond what the sources establish. Personal codices could contain different arrangements, non-standard written forms, readings not chosen for the public written standard, explanatory notes, abrogated recitations, or scribal mistakes. Once an official mushaf existed, leaving competing private written materials in public circulation would have defeated the purpose of standardization.

Uthman standardized the written reference of the community. He did not invent a new Quran. Nor did the Uthmanic standard eliminate all valid recitational variety, since verified readings compatible with the Uthmanic rasm continued to be transmitted. What it eliminated was the public use of private written codices as rival reference texts.

This is where the manuscript evidence becomes important. The surviving early manuscript tradition after this point is overwhelmingly Uthmanic in text type, and Sadeghi and Goudarzi found that variation within that Uthmanic family is comparatively small. The manuscript record does not show the standard Quran gradually emerging centuries later. It shows an Uthmanic textual tradition already distinct and highly stable around the middle of the seventh century.

What Sanaa actually shows

The Sanaa palimpsest does not support the story of an original Quran, centuries of uncontrolled corruption, and a later reconstruction. The evidence fits a different sequence:

Revelation to Muhammad (PBUH) -> memorization and written transmission among the Companions -> early personal codices and readings -> Uthmanic standardization during the Companions' own generation -> exceptionally stable written transmission afterward.

The lower Sanaa text is not evidence that someone rebuilt Islam's scripture long after the fact. It belongs to the earliest Islamic period and represents the kind of non-Uthmanic written material that Uthman's standardization was meant to remove from public use. Whether a particular folio was copied just before the standardization or from an earlier exemplar soon after it, it belongs to that early horizon, not to a late medieval rewriting of the Quran.

Conclusion

The Sanaa palimpsest matters precisely because Muslims need not fear manuscript evidence. Its lower text contains genuine variants that should not be waved away as spelling. But acknowledging them does not prove that the Quran was corrupted. It gives a rare physical window onto an early stage of transmission that matches what Islamic sources already described: oral preservation, written materials, Companion codices and readings, then an authoritative written standardization under Uthman.

The real question is not whether every private early sheet was identical. It is whether the Quran taught by the Prophet (PBUH) reached the community intact and whether the Uthmanic standard faithfully preserved it. Sanaa does not show otherwise. If anything, when placed beside the extraordinarily stable Uthmanic tradition, it shows how early the standardization occurred and how well it held.

For more on how the Quran was compiled and preserved, see the related topics below.

Sources

• Behnam Sadeghi and Mohsen Goudarzi, "Sanaa 1 and the Origins of the Quran," Der Islam 87 (2012) - discovery background, manuscript description, radiocarbon probabilities, and analysis of the lower and upper texts.
• Sahih al-Bukhari 4987 - Uthman's standardization of the mushaf and the order that other Quranic written materials be burned.
• Sahih al-Bukhari and Sahih Muslim - reports concerning the final review of the Quran and Umar's statement about the stoning wording.
• Sahih Muslim - Aisha's report concerning abrogated suckling wording and Abu Musa al-Ashari's report concerning wording whose recitation did not remain in the final Quran.
• Sunan Ibn Majah - the report attributed to Aisha about a domestic animal eating a page; cited here with the note that it is not in Sahih al-Bukhari or Sahih Muslim and that the specific animal-eating wording is disputed.
• The Institute of Ismaili Studies, overview article on the Sanaa palimpsest and early Quran manuscripts.