The Quran, side by side

Strictly speaking, the Quran has never been translated. The book is the Arabic. A Muslim holds that these are the literal words of God, given to Muhammad in Arabic over twenty-three years, and that no other language can carry them. So the English Quran does not exist, exactly. What exists is a translator telling you, in English, what he thinks the Arabic means.

The translators say so themselves, right on the cover. Marmaduke Pickthall called his version The Meaning of the Glorious Koran, because, in his own words, the Koran "cannot be translated" and the result "is not the Glorious Koran." The Cambridge Arabist A. J. Arberry called his The Koran Interpreted, for the same reason. The title is an apology. The real book stayed in Arabic.

That sounds like a technicality. It is the whole subject of this page. If the original can never be reproduced, then every translator is guessing at the same locked door, and you can watch them choose different keys. Here is the line a Muslim says before nearly every act of the day, the one that opens the book and opens almost every chapter in it.

Seven translators, one short line, and they already disagree about the name of God and about the two words for his mercy that come after it. Three keep Allah; four write God. The two mercy words, ٱلرَّحْمَـٰن and ٱلرَّحِيم, both grow from one root that also gives the word for womb, and English has no clean pair for them, so you get Beneficent and Merciful, Gracious and Merciful, Entirely Merciful and Especially Merciful. Sale, in 1734, gave up and folded both into "the most merciful." That small gap, multiplied across the whole book, is what you are about to read.

Below are fourteen passages, chosen the way the tradition itself weights them: the prayer recited in every cycle of the daily prayers, the four lines a hadith calls equal to a third of the book, the verse most often hung on a wall for protection. For each one you get the Arabic, a transliteration to sound it out, a plain breakdown of what it says and where the seven part ways, and then all of them stacked. Use the arrows or the list to move between passages.

The hard words

Five words that will not cross over

Some of the trouble is just poetry, the kind any translator faces. But a handful of words sit at the center of the religion and have no English equal, so the translators either reach for an approximation or leave the Arabic standing. Watch these five, and most of the disagreements in the stack above start to make sense.

The stack

Who the seven are

They span three centuries and almost every angle you can come at the book from: an Enlightenment lawyer, an English novelist who converted, a British-Indian barrister, a Cambridge orientalist, a Jewish-born convert, a team of American women, and an Egyptian who knew the whole Quran by heart. Naming them matters, because the differences below are not random noise. Each translator wanted something different from the Arabic.

• George Sale1734A London lawyer, largely self-taught in Arabic. The first English Quran rendered straight from the Arabic rather than through Latin or French, and the lens through which the Enlightenment, Voltaire and Gibbon and Jefferson, first met the book. Archaic, lawyerly, oddly beautiful.
• Marmaduke Pickthall1930An English novelist and son of a clergyman who converted to Islam in 1917. The first English Quran by an Englishman who was himself a Muslim. Stately, King-James in cadence, and careful to call itself only "the meaning."
• Abdullah Yusuf Ali1934A British-Indian barrister. For most of the twentieth century the most widely distributed English Quran on earth, handed out by the millions, loved and argued over for its dense, poetic footnotes.
• A. J. Arberry1955Sir Thomas Adams's Professor of Arabic at Cambridge. The version scholars most often call the finest in English, the one that tries hardest to carry the Arabic's rhythm instead of flattening it to prose. A non-Muslim's act of reverence.
• Muhammad Asad1980Born Leopold Weiss, an Austrian Jew, journalist and diplomat, who converted in 1926. A rationalist reading with heavy interpretive notes; admired abroad and banned in Saudi Arabia for straying from the official creed.
• Saheeh International1997A team of three American women converts working in Jeddah. The plain, literal, lightly bracketed version that has become the default English Quran online and in mosque racks today.
• M. A. S. Abdel Haleem2004An Egyptian who learned the whole Quran by heart as a boy, later a professor at SOAS in London. The Oxford World's Classics version, prized for reading like clear, natural, modern English.

Here is the part that surprises people. With seven translators this far apart, you would expect them to be working from different Arabic texts, the way the King James and a modern Bible descend from competing manuscripts. They are not. The Arabic is essentially one text the world over.

That uniformity is real, and it helps to know how it was made. Around the year 650, a generation after Muhammad died, the caliph Uthman had one written version declared standard and the rival copies burned. Almost every Quran printed since follows a single recitation of that text, called Hafs, fixed in an edition published in Cairo in 1924. That is not the same as no variation. A few canonical recitations, like Warsh across North Africa, differ in vowels and a handful of words. The fourth line of the opening is read both as "Master of the Day of Judgment" and "King of the Day of Judgment," the whole difference hanging on one unwritten vowel. The oldest manuscripts, like the burned-and-reused Sana'a parchment, preserve an early version that is plainly the same book with real differences in it. But the differences are small and catalogued, and everything descends from that one archetype, which makes the text far more uniform than, say, the New Testament, whose thousands of Greek copies carry hundreds of thousands of variants. The comparison is fair on the result, though the Quran's uniformity was partly engineered, by one man's edit and a fire.

So the real disagreements in Islam are almost never about the words. They are about what the words mean, which is the work of tafsir, the vast science of interpretation, and about who has the standing to say. The original split between Sunni and Shia was a fight over who should lead the community after Muhammad, not over the text; the two read the same Quran. An old accusation that the text was tampered with surfaces in early polemic, and mainstream Shia scholarship rejects it and affirms the same book. The contest, in other words, is exactly the one this page is about: not the source, but the reading. The translators are just the version of it that happens in English.

One more thing worth knowing before you read. The 114 suras are not arranged in the order they were revealed, and not by topic. After the short opening, they run roughly from longest to shortest, and only roughly. The book you can read front to back is not the book in the order it came down, which is part of why a guided path through the keystones, rather than page one to the end, is the honest way in.

Notes

Where the text comes from

Every translation here is reproduced verbatim, not paraphrased. The Arabic, the transliteration, and the modern translations were pulled with scripts and parsed, so nothing in them was retyped or reworded by hand or by a model. The only change to any quotation is curly quotation marks turned straight, to match the rest of the site.

• The Arabic is the Uthmani text in the Hafs reading, and the transliteration, from Quran.com. The breakdowns add the standard academic transliteration with its long vowels.
• Pickthall (1930), Yusuf Ali (1934), Saheeh International (1997), and Abdel Haleem (2004) are from the Quran.com text API. Arberry (1955) and Asad (1980) are from Al Quran Cloud.
• George Sale (1734) is in the public domain. The short suras here come from the Wikisource transcription of his first edition; the rest from a later reprint scanned at the Internet Archive. Obvious scanning errors are corrected; his archaic spelling, and his "Lokman" and "Tagut," are left as he wrote them.
• The hadith cited in the breakdowns are checked against sunnah.com: al-Fatiha as the greatest sura (Bukhari 4647), al-Ikhlas as a third of the Quran (Bukhari 5015), Ayat al-Kursi as the greatest verse (Muslim 810).

The breakdowns are mine. They lean on standard scholarship and on the classical commentaries (tafsir) for the contested lines. Four of the seven translations, Arberry, Asad, Saheeh, and Abdel Haleem, are still in copyright; short passages are quoted here for comparison and study, with every translator named, which is the whole point of a page like this. The four in-copyright versions are not reproduced in full anywhere on the site.

Where I have had to characterize what Muslims believe, I have tried to give the mainstream view and to flag where the tradition itself disagrees, rather than smooth it over or pick a side. Corrections from people who know the text better than I do are welcome.

The image on the homepage card is a folio from the Blue Qur'an, gold Kufic script on indigo-dyed parchment, made in Tunisia (probably Kairouan) sometime between the late ninth and mid tenth century. The Metropolitan Museum of Art, New York (Open Access, public domain), accession 2004.88.