Methodology

The First and Second Models are two strictly independent datasets: they remain separate everywhere on the site and are never merged by default.

To make searching easier, the program may ignore certain marks when comparing text. The displayed Quranic text remains faithful to the source supplied by the project’s author, with its original letters and marks.

Repetition and lexical resemblance: what is the difference?

Al-Zarkashi, in his work al-Burhân, defines “lexical resemblance” (al-mutashâbih al-lafẕî) as recounting the same story in different forms and sequences, a device frequently found in Qur'anic narratives and announcements.

He thus refers to the same Qur'anic statement recurring in similar forms, where the resemblance lies in differences of addition, omission, substitution or word-order inversion — which requires the reader who memorises the Qur'an to double-check and pay increased attention; reading specialists call this type “al-mushkil” (the ambiguous).

• The “repeated” (al-mukarrar): the same word recurring without any variation at several places in the Qur'an, such as the verse ⁧﴿فَبِأَيِّ آلاءِ رَبِّكُمَا تُكَذِّبَانِ﴾⁩ [Ar-Rahman, 13] — identical, literal repetition.
• The case where the meaning is repeated with slight, very close variations in wording — this is precisely lexical resemblance (al-mutashâbih al-lafẕî).
• The case where only the meaning is repeated without the words resembling each other — such as the repetition of the prophets' stories in very different styles and wordings — which falls outside the scope of lexical resemblance.

Methodological note on words of similar form but different meaning

This work is based on cataloguing the words of the Qur'an according to their written form and their locations, keeping every location where a word resembles another in its shape or pronunciation, even if its meaning varies according to context, derivation, morphological structure or its position in the verse.

The reader may thus find in this project words of similar form that do not carry the same meaning at all their locations. This is neither a cataloguing error, nor an unintended repetition, nor a confusion between words: it is an intentional, methodical choice, because a Qur'anic word is never understood in isolation from its context, its construction, or the verse in which it appears.

Graphic resemblance therefore does not always imply a resemblance in meaning, just as a difference in vowelling, derivation or context can shift a word from one meaning to another. All these locations have been kept so that the research material remains complete, and so that both the reader and the researcher can reflect, compare, and return to the verses in their original context.

Thus, “⁧أَلَّفَ⁩” (allafa, “to unite”) and “⁧أَلْفَ⁩” (alfa, “thousand”) are close in written form, but their meaning differs: ⁧﴿وَأَلَّفَ بَيۡنَ قُلُوبِهِمۡۚ لَوۡ أَنفَقۡتَ مَا فِي ٱلۡأَرۡضِ جَمِيعٗا مَّآ أَلَّفۡتَ بَيۡنَ قُلُوبِهِمۡ وَلَٰكِنَّ ٱللَّهَ أَلَّفَ بَيۡنَهُمۡۚ إِنَّهُۥ عَزِيزٌ حَكِيمٞ63﴾⁩ [Al-Anfal, 63] — here, “allafa” means uniting hearts, creating harmony and affection between them; the meaning of uniting hearts is explicit in the verse.

In another verse, however: ⁧﴿وَلَتَجِدَنَّهُمۡ أَحۡرَصَ ٱلنَّاسِ عَلَىٰ حَيَوٰةٖ وَمِنَ ٱلَّذِينَ أَشۡرَكُواْۚ يَوَدُّ أَحَدُهُمۡ لَوۡ يُعَمَّرُ أَلۡفَ سَنَةٖ وَمَا هُوَ بِمُزَحۡزِحِهِۦ مِنَ ٱلۡعَذَابِ أَن يُعَمَّرَۗ وَٱللَّهُ بَصِيرُۢ بِمَا يَعۡمَلُونَ96﴾⁩ [Al-Baqara, 96] — here, “alfa” is a number (a thousand years), with no connection to the meaning of uniting.

Likewise, “⁧سُنَّة⁩” (sunna, “way, immutable law”) and “⁧سِنَة⁩” (sina, “drowsiness”) resemble each other graphically, but their meanings are clearly different: ⁧﴿سُنَّةَ ٱللَّهِ فِي ٱلَّذِينَ خَلَوۡاْ مِن قَبۡلُۖ وَلَن تَجِدَ لِسُنَّةِ ٱللَّهِ تَبۡدِيلٗا 62﴾⁩ [Al-Ahzab, 62] — here, “sunnat Allah” denotes Allah's immutable way, the constant divine law in His creation.

Whereas: ⁧﴿ٱللَّهُ لَآ إِلَٰهَ إِلَّا هُوَ ٱلۡحَيُّ ٱلۡقَيُّومُۚ لَا تَأۡخُذُهُۥ سِنَةٞ وَلَا نَوۡمٞۚ⁩...⁧255﴾⁩ [Al-Baqara, 255] — here, “sina” denotes drowsiness, in a verse that denies any sleep for Allah.

Similarly, “⁧أُمَّة⁩” (umma) changes meaning according to context: ⁧﴿إِنَّ إِبۡرَٰهِيمَ كَانَ أُمَّةٗ قَانِتٗا لِّلَّهِ حَنِيفٗا وَلَمۡ يَكُ مِنَ ٱلۡمُشۡرِكِينَ 120﴾⁩ [An-Nahl, 120] — here, “umma” denotes a man embodying by himself the qualities of goodness, standing in the place of an entire community in piety and obedience.

Elsewhere, it denotes a group of people: ⁧﴿وَلَمَّا وَرَدَ مَآءَ مَدۡيَنَ وَجَدَ عَلَيۡهِ أُمَّةٗ مِّنَ ٱلنَّاسِ يَسۡقُونَ⁩...⁧23﴾⁩ [Al-Qasas, 23] — here, “umma” denotes a group of people.

The same word can also mean a period of time: ⁧﴿وَقَالَ ٱلَّذِي نَجَا مِنۡهُمَا وَٱدَّكَرَ بَعۡدَ أُمَّةٍ أَنَا۠ أُنَبِّئُكُم بِتَأۡوِيلِهِۦ فَأَرۡسِلُونِ45﴾⁩ [Yusuf, 45] — here, “umma” denotes a period of time. Likewise, “⁧آية⁩” (aya) can mean a sign, a clear proof, or a verse of the Qur'an; here too, only the context can settle the matter.

The same applies to “⁧الرُّوح⁩” (ar-rûh): depending on context, this word can refer to revelation, to the angel Gabriel, or to the vital spirit of a human being — none of these occurrences should be reduced to a single meaning without examining the context of the verse.

This is why these words of similar form have been kept, despite their different meanings, for the following reasons:

• Removing certain locations would reduce the catalogued Qur'anic material.
• The reader needs to see all similar locations in order to compare and reflect.
• The precise meaning never appears in the isolated word but in the verse, its context and its construction.
• The difference in meaning despite graphic resemblance opens a door to linguistic, rhetorical and contemplative reflection.
• What does not reveal its wisdom to the researcher today may become clear tomorrow, as linguistic and Qur'anic research expands, without ever claiming to settle what is not established by evidence.

The presence of words that look similar but differ in meaning is therefore an integral part of this project's method, not a classification flaw. The reader should always go back to the full verse, the surah name and the verse number, before basing any linguistic or semantic judgement on a word in the table.

Note: the search only covers words or text matching exactly and in isolation (“isolated” meaning not attached to another word or even a single extra letter). For example, a search for “⁧يعلمون⁩” (they know) does not include “⁧فيعلمون⁩” (and they know). A second clarification: some forms, such as “⁧فيعلمون⁩” (and they know), are not repeated in the Quran. However, the absence of a word as a separate entry does not necessarily mean that it is not repeated: this program retains the longest repeated segment. For example, “⁧فسيعلمون⁩” (then they will know) occurs twice, followed by “⁧من⁩” (who) in both cases. The program therefore lists “⁧فسيعلمون من⁩” (then they will know who), rather than “⁧فسيعلمون⁩” alone.

Why the conjunction waw is separated for indexing and analysis

Separating the conjunction waw (“⁧و⁩”) from the word that follows it, in this work, is in no way an alteration of the Qur'anic script (rasm), nor a call to write the Qur'an other than according to the established script of the Mushaf or the rules of Arabic spelling. In Arabic, the conjunction waw is normally written attached to the following word — such as “⁧والله⁩”, “⁧ويهدي⁩”, “⁧ومن⁩” — because it is a letter that is graphically linked to the following word.

But this project has an analytical and indexing nature, not the nature of a Mushaf transcription. This is why the separation of the conjunction waw has only been applied within the database and the search tables, so that it appears as an autonomous linguistic element, one that can be counted, whose locations can be tracked and whose relationship with what follows it can be studied. The displayed Qur'anic text remains preserved in its original form; this separation remains a technical tool for research and analysis.

Hiding the conjunction waw inside the word it is attached to would prevent distinguishing between two locations that differ in structure and scope: a word appearing alone, and that same word preceded by the conjunction waw. For example, searching for the phrase “⁧يهدي من يشاء⁩” may give ten results, while restricting the search to the form “⁧و يهدي من يشاء⁩” reveals that only five of these locations are actually preceded by the conjunction waw. This result is not a pointless repetition, but distinct analytical information, useful for studying context, linking verses and understanding the transition structure between meanings.

Similarly, the name “⁧الله⁩” appears in 2,393 results, while the form “⁧و الله⁩” appears in only 240 results. The first figure indicates the word's overall presence across the whole text; the second indicates its presence at the precise locations where it is linked by a conjunctive function. Neither piece of information should be dismissed on the grounds that the other has more results: each answers a different research question and carries its own significance.

The frequency of the conjunction waw, which reaches 9,546 locations in these tables, is not a reason to neglect it, but rather proof of its importance in the structure of the text. The conjunction waw is indeed one of the tools most closely tied to the construction of the sentence and the verse; it links words, structures and meanings, and helps mark coordination, contextual order or transitions between the elements of discourse depending on context. Counting it and separating it analytically therefore opens the way to a precise study, inaccessible if it remained merged into the following word.

This method has also revealed particularities that ordinary search does not bring out: notably, Surah At-Takathur (102) is the only surah where the conjunction waw does not appear according to the criteria of this analytical index. This observation is not in itself a definitive rhetorical judgement, but it offers a clear starting point for the researcher wishing to return to the surah and study its structure, context and linking devices.

This principle is not limited to the conjunction waw alone: very frequent words and particles do not lose their value because of their frequency. The name “⁧الله⁩”, pronouns such as “⁧هم⁩” and “⁧إلى⁩”, or words such as “⁧الذين⁩”, “⁧في⁩” and “⁧أن⁩” recur very often; yet cataloguing them and linking them to their contexts reveals precise nuances between their uses. The large number of results therefore does not make the search pointless: on the contrary, it makes classification and analysis more necessary and more rigorous.

This is why this project has adopted the analytical separation of the conjunction waw: to allow the reader to search for the word alone, the word preceded by the conjunction waw, or the conjunction waw itself, and to compare results clearly and verifiably. The goal is not to artificially inflate the figures, but to allow a methodical reading of the text, distinguishing the continuous orthographic script from the grammatical and semantic elements useful to scientific research.

Practical tome/page correspondence rules

The volume/page correspondence for the First Model follows this rule: in volume 1, the PDF page number equals the table page number plus 24 (an unnumbered 24-page introduction); in volumes 2 to 13, the two page numbers match exactly. The Second Model's rules will be documented separately once the exports are received.

The writing used in this project

The project follows the standard (imlāʾī) orthography of the Quran (in Arabic « الرسم الإملائي »), not the Uthmani script (« الرسم العثماني ») used in printed copies of the Quran. In the Quran, the same word can be written in more than one way: the verb “to see” is written « رَءَا » in ﴿فَلَمَّا جَنَّ عَلَيۡهِ ٱلَّيۡلُ رَءَا كَوۡكَبٗاۖ قَالَ هَٰذَا رَبِّيۖ فَلَمَّآ أَفَلَ قَالَ لَآ أُحِبُّ ٱلۡأٓفِلِينَ76﴾ [Al-Anʿām: 76] and « رَأَىٰ » in ﴿لَقَدۡ رَأَىٰ مِنۡ ءَايَٰتِ رَبِّهِ ٱلۡكُبۡرَىٰٓ18﴾ [An-Najm: 18].

In this project it is always written the same way, « رَأَى », so that all its repetitions are counted together. These differences in spelling surely carry wisdom and meaning, even if we do not always know them; this too is part of the greatness of the Quran.

Search figures: one verse, no repetition

When you search for a letter, a word or a text, the table shows each verse found once by default, keeping each model’s order: Model No. 1 follows the order of the verses in the Quran, Model No. 2 the order of its chains.

The figures are then counted for the searched text only, with the project’s rule: “isolated” means the text standing alone, not attached to a longer word; “non-isolated” means the text inside a longer word. For each verse, isolated occurrences are counted, then isolated and non-isolated ones; these numbers are then added up over the whole Quran, together with the number of verses and surahs containing an isolated occurrence. The text is green if it is repeated in isolation at least twice in the surah, black otherwise.

This one-off count only concerns the current search and changes no figure in the database. When the “One verse, no repetition” button is off, every repetition returns, with its original colours and figures.

Normalized search and spellings: why “الله” gives 2395

The groups of the database count each spelling separately: a form written with hamza or madda is not the same spelling as the form without them. So the group “الله” counts 2393 repetitions of the spelling “الله”, and counts separately the 2 repetitions of the spelling “آلله” (Surah 10, verse 59 and Surah 27, verse 59).

The “Normalized search” of the Search page ignores vowel marks and brings together the forms of alif (ا أ إ آ): it therefore finds “الله” and “آلله” together, and its “Total isolated” column shows 2395. Both figures are right; they answer two different questions. To remove any ambiguity, the Search page now shows the “Breakdown by spelling” when several spellings are grouped, for example “الله 2393 · آلله 2”, or “إلا 666 · ألا 99” for the search “الا”.

“Recitation & tafsir” page: sources and checks

Recitations come from mp3quran.net (whole surahs and, for some reciters, the timing of each verse) and EveryAyah.com (one file per verse, for memorization). The 42 verse-by-verse recitations were checked on 24 October 2026; incomplete recitations and those of the Warsh reading, whose verse numbering differs from the text shown, were left out of memorization.

The Uthmani text, translations, tafsirs and the English meaning of each word come from Quran.com. Arabic word meanings come from 4 books of غريب القرآن published by Quranpedia (quranpedia.net): السراج في بيان غريب القرآن (al-Khudayri), التبيان في تفسير غريب القرآن (Ibn al-Ha’im), كلمات القرآن (Hasanayn Makhluf) and تذكرة الأريب في تفسير الغريب (Ibn al-Jawzi); 19427 of their 20244 explanations are linked to the exact word of the verse, the others stay in the verse’s list. Roots, base forms and grammatical analyses come from the Quranic Arabic Corpus (corpus.quran.com). Before publication, the two word divisions were compared: 6236 verses and 77429 words, with the same number of words in every verse; the only spelling differences are pause marks and the way hamza is written.

This data changes nothing in the duplicates database: the page is independent of the table, its figures and its buttons. Received texts are shown as plain text. Bookmarks and the last listening position stay in the visitor’s browser.

Readings after each verse: translations recorded by human voices from QuranEnc (French, Rachid Maach; English, Rowwad) and EveryAyah (English, Saheeh International read by Ibrahim Walk), one file per verse; التفسير الميسر: official recording of the King Fahd Complex for the Printing of the Holy Quran, one file per verse — the official archive (743 MB) was checked on 25 October 2026: it holds all 6236 verses; the files are played from the public copy cdn.quran.ws. “Offline listening”: the chosen files are copied into the browser storage of the device (never to the site) and played by a standalone page that works without internet; the site’s general service worker is not changed. “Tafsir videos”: no public list of videos per verse exists; the links are YouTube searches, with no video chosen or checked by the site.

“Duplicate study”: the same definitions as the table

The study tools read the same database as the table, without changing it. A duplicate there is one spelling of a group, and only its isolated occurrences count, as in the “Total isolated” column. Check made on 24 October 2026: for all 31896 spellings, the recomputed numbers of verses, surahs and “Total isolated” are identical to the table’s figures; the statistics per surah are identical to the “Duplicates of the surah” summary for all 114 surahs, in both models.

The “Duplicate of the day” and the quiz pick duplicates of at least 3 words (4 for the duplicate of the day; 2 possible in a small learned part) found in 2 to 8 verses whose texts are not all identical: these are the most easily confused. The comparison of 2 verses aligns words without taking vowels or the forms of alif and hamza into account, like the “Normalized search”; the text shown remains the verse’s text.

With “The part I have learned”, the quiz keeps only the duplicates found at least 2 times in the range (as isolated occurrences), whatever their number elsewhere, and proposes only verses of the range. Verses are numbered from 1 (1:1) to 6236 (114:6) in the order of the Quran, which allows a range from one surah to another. The statistics keep the figures of the whole surah, to stay identical to the table.

“Without repetition” quiz: the list of possible questions is that of the verses of the chosen part which contain at least one duplicate that can be asked (same rules as above). It is shuffled once with a random number kept only in the page; question No. k is the k-th verse of this shuffle, which guarantees that no verse is asked twice. Check of 25 October 2026: 4 parts tried in full (including Al-Baqara 1 to 286: 248 questions), each verse asked only once, all verses covered.

Locations quiz, verse videos, sources and licences

Locations quiz: a duplicate is used if it has at least one isolated occurrence in the chosen part and occurs, as an isolated occurrence, in 2 to 12 verses of the search scope (the same part or the whole Quran). The number to find is the sum of isolated occurrences in these verses: the same calculation as “Total isolated” (checked on the 31896 spellings of the database). The “trap” verses are taken from the most to the least similar: verses of another spelling of the same group, verses containing the first and last word of the duplicate, verses containing its longest word, then, for a short surah, its other verses. The series has no repetition.

Verse videos: the index (10495 videos, 6195 of the 6236 verses) is built from the titles published by 6 tafsir series on YouTube (surah and verse number or numbers). It is checked by the presence of the surah name and the selected numbers in the title, and, for the Zaid Al-Bahri series, by the presence of the beginning of the quoted verse in the verse text. A video is only loaded on click, by the YouTube player without advertising cookies.

Sources and licences: text and translations, tafsirs and English word meanings: Quran.com; recitations: mp3quran.net and EveryAyah.com; Arabic word meanings (غريب القرآن): quranpedia.net; roots and grammatical analysis: Quranic Arabic Corpus (corpus.quran.com, GNU licence); videos: YouTube (channels of the series listed). The contents remain the property of their authors.

Words of a root and quiz answers

Words of a root in the text of the verses: the position of each word comes from the Quranic Arabic Corpus (the same files as the grammatical analysis). The site’s text sometimes separates a small particle (و, ف, ب, ل, ك, يا, ها) from the following word: it is attached to that word before counting. For 6195 verses the number of words matches and the words of the root are coloured; for 9 verses written differently (for example “بَعْدَ مَا” / “بَعْدَمَا”), the text is shown without colour rather than with a wrong colour.

Extra quiz answers (at least 5): they are always taken from the chosen part and never from the verses of the duplicate. For “Where is this verse?”, first verses that contain words of the duplicate, then nearby verses in the same surahs. For “What comes next?”, what follows the last 2 words of the duplicate (or at least the last one) in other verses; in a small part there may be fewer than 5.

Following the recited verse

While a whole surah is playing, the recited verse is found from the start times published for each recitation. For علي بن عبد الرحمن الحذيفي (حفص عن عاصم), these times lagged behind the sound (often about 3 seconds at the end of a surah, sometimes more) and sometimes skipped a verse, so the frame stayed on the verse already recited. The 6236 start times of this recitation were therefore recalculated from the sound itself, using the silences between verses, then checked by comparing each verse with the same reciter’s verse-by-verse recording. The same check was made for خالد المهنا: his published times drifted little by little (up to about 38 seconds at the end of Surah Al-Baqara); they were recalculated for 109 surahs, and the 5 surahs where the check was not reliable enough keep the published times. For other reciters, the published times are used as they are.

Recited word and surah classification

The recited word is highlighted using the word timings published by Quran.com for 12 verse-by-verse recitations: عبد الباسط عبد الصمد (مجوّد and مرتّل), عبد الرحمن السديس, أبو بكر الشاطري, هاني الرفاعي, محمود خليل الحصري (مرتّل and معلّم), مشاري راشد العفاسي, محمد صديق المنشاوي (مجوّد and مرتّل), سعود الشريم and محمد الطبلاوي. With these reciters, the sound file played is the one that matches these timings. The other reciters, including علي بن عبد الرحمن الحذيفي offered at the start, have no published word timings: only the recited verse is highlighted, and a verse longer than the screen scrolls along with the recitation.

The classification of surahs as Meccan (86) and Medinan (28) is that of Quran.com, taken from the Tanzil project; other classifications exist for a few surahs.

Mushaf pages and new tafsirs

The page layout comes from Quran.com (Medina mushaf, Hafs reading, 604 pages of 15 lines): for each word, its page and line. It is copied into the site; the page of each verse was checked against the duplicates database (6204 of 6204 verses identical). The surah title and basmala lines are deduced from the empty lines of the page (114 titles, 112 basmalas). The text shown is the ʿuthmānī text of Quran.com, in the “Amiri Quran” font.

New tafsirs: “المختصر في التفسير” (Tafsir Center, published by QuranEnc) in Arabic, French, English and Spanish; “تفسير الجلالين” (alquran.cloud); “Ma'arif al-Qur'an” in English (Quran.com). No source of a fully vowelled tafsir, with every vowel sign, was found: the texts are shown as published. Spanish has no translation read by a human voice: the device voice reads it.

Notes, bookmarks and personal documents stay in the device's browser (local storage and IndexedDB database): they are never sent to the site. To find them on another device, use the backup then “Restore a backup”.

Hizb quarters, pause signs and word meanings

The starts of hizb quarters (sign ۞) come from Quran.com: for each, the number of the quarter (1 to 240), the hizb (1 to 60) and the juz (1 to 30). The label says “ربع الحزب”, “نصف الحزب” or “ثلاثة أرباع الحزب” followed by the hizb number, or “الحزب” / “الجزء” at the start of a hizb or a juz. Pause signs (ۚ ۗ ۖ ۘ ۛ ۙ) are only coloured: the Quran text is not changed.

Word meanings: the main source remains “السراج في بيان غريب القرآن”. When a word is not explained there, the word panel shows the word-by-word translation of Quran.com (in English, the only language available), with the source named. In the “What comes next?” and “What comes before?” quizzes, when other verses have exactly the same continuation (or beginning) as the verse asked, they are removed from the choices, so there are never two right answers in the list.

Surah card and mushaf index

The short card shown when choosing a verse gives the number and name of the surah, Meccan or Medinan, the number of verses, the order of revelation (usual Egyptian order, Tanzil data), its mushaf pages and the juz it covers. The mushaf index gives, for each of the 604 pages, the juz and the surah of the first verse of the page. Page parity follows the printed mushaf: odd pages are on the right, even pages on the left.

The Basmala in this project

All scholars recognise “بِسْمِ اللَّهِ الرَّحْمَنِ الرَّحِيمِ” as a verse of the Quran within Surah An-Naml (surah 27, verse 30). Its status at the beginning of surahs, however, is a classic point of scholarly difference.

For the school of Imam Al-Shafi‘i and a group of scholars, the Basmala is the first verse of Surah Al-Fatiha; this is why it is numbered 1 in copies of the Quran based on the Medina or Kufa count, such as the widespread Hafs reading. For the schools of Imam Malik and Imam Abu Hanifa, it is not part of Al-Fatiha: the first verse is then “الْحَمْدُ لِلَّهِ رَبِّ الْعَالَمِينَ”, and the last verse is divided in two to keep seven verses.

At the beginning of the other surahs (it opens 113 of the 114, all except Surah At-Tawba), the Shafi‘i and Hanbali schools regard it as a verse, while the Maliki and Hanafi schools see it as a blessed formula that separates the surahs without being part of them.

In this project, “بِسْمِ اللَّهِ الرَّحْمَنِ الرَّحِيمِ” is counted as verse 1 of Surah Al-Fatiha, following the numbering of the Hafs reading. This numbering does not take sides in the difference between the schools.