A video, unfolded
A narration is only accepted after every narrator, every link and every word in it has been checked. This page walks through the video's crash course on how that checking works.
The video's argument: people who reject hadith wholesale usually don't know how hadith are graded. The grading method, it says, checks sources more strictly than most historical scholarship does.
A hadith is ṣaḥīḥ only if all five checks pass: good character, accurate memory, an unbroken chain, no clash with stronger reports, and no hidden defect.
Meet all five with complete precision and it is ṣaḥīḥ. If a narrator's precision is a little lighter, it is ḥasan: one rank lower, and still accepted as proof. Lose a condition and it is weak, though mild weakness can be lifted by other chains. See the full ladder ↓
Go to the five gates →Next, count how many independent routes a report took: one, two, three or more. If all routes pass through a single narrator at any point, they count as one.
The labels for one, two and three-plus chains are gharīb, ʿazīz and mashhūr.
Try the chain counter →When many independent eyewitness chains from different cities and schools give the same report, the report is mutawātir: a shared mistake or a coordinated lie is ruled out.
The video compares this to a clinical trial: when hundreds of people independently report the same effect, the effect is real.
See the four conditions →The minimum a narration must meet before it is called ṣaḥīḥ. ("Ṣaḥīḥ" literally means healthy.) The stricter standards of al-Bukhārī and Muslim come on top of these. Missing a condition doesn't always mean weak: a lighter version of the second makes a hadith ḥasan, and some weakness can be lifted by other chains.
Each narrator must be known as honest and upright. A report from someone of bad character is not accepted.
Scholars of al-jarḥ wa al-taʿdīl (narrator criticism) visited narrators, asked their neighbours about them, and recorded whether they prayed in the mosque and whether they were honest in business. These records became biographical dictionaries, a field called ʿilm al-rijāl. The speaker says he owns a 55-volume set of them.
A narrator who appears in no record is majhūl, unknown, and his reports can't be graded authentic.
The video's well-known example: a scholar saw a narrator lure his donkey forward with a carrot tied to a stick, and refused his hadith. If he would trick his donkey, he might trick people too.
"If he can trick his donkey, he could trick me."
How a narrator's character is judged · Who did the checking, and when · The biographical works
A good person with a weak memory still fails. A narrator must memorise and write accurately.
Someone may pray, give charity and be entirely sincere, yet misremember through age or illness. That is no fault of theirs, but it disqualifies them. Scholars checked whether each narrator wrote and memorised correctly, and whether they swapped or flipped words.
Classical hadith science adds a middle grade here. When a narrator's precision is good but not complete (khafīf al-ḍabṭ), the hadith is ḥasan rather than weak: accepted, one rank below ṣaḥīḥ.
The speaker explains ḍabṭ and how it compares to the study of the Gospels
Every link, from whoever wrote the hadith down back to the Prophet ﷺ, must be named, known and connected. A single gap makes the hadith weak.
A named but unidentified narrator also fails. If the chain says only "Abdullah", the question is which Abdullah, and if nobody can tell, he is majhūl. The chain explorer below shows how one missing or unknown link changes the grade.
Schools differ here. A mursal hadith, where a Successor quotes the Prophet ﷺ without naming the Companion, is graded weak by the hadith critics, but Abū Ḥanīfa and Mālik accepted it as evidence, and al-Shāfiʿī did with conditions.
The rule · Broken chains in other sources · "Which Abdullah?"
When ten reliable routes agree and one route differs, even by a single word, the odd one out is rejected.
Each route is checked on its own. Changes as small as two words swapping places (qalb), an added word or a dropped word are caught. The majority wording is kept as maḥfūẓ (preserved). The outlier is shādhdh.
Because of this check, the speaker argues, a hadith that truly contradicted the Qur'an could never be graded ṣaḥīḥ in the first place. Try the slip finder ↓
Ten versus one · Word-for-word memory · "This hadith contradicts the Qur'an" · Ibn al-Qayyim on the "kneeling like a camel" wording
Any scribal or textual error, even a grammar slip, weakens that route. Scholars then trace exactly where the error entered.
The Prophet's ﷺ Arabic was eloquent, so a grammatical error in a narration points to a mistake further down the chain. Scholars compared every route carrying the same hadith to find the narrator after whom the wording changed. The speaker calls them "better than Sherlock Holmes".
Long chains of 12 to 14 narrators, which come from the later oral tradition, are always checked against the short early ones, such as the Muwaṭṭaʾ, the Musnad and Ṣaḥīḥ al-Bukhārī. In the speaker's words: "Hadith are never in a silo."
Textual defects · Grammar as evidence · Tracing the error · Long chains checked against short ones
Interactive · test a narration
Added context · not from the video
The video presents two outcomes, ṣaḥīḥ or weak. Classical hadith science, as set out in Ibn Ḥajar's Nukhbat al-Fikar, has accepted grades in between, and weak hadith differ in how weak they are.
Accepted (maqbūl). Ṣaḥīḥ li-dhātihi meets all five conditions with complete precision. Ḥasan li-dhātihi meets them with lighter precision and is still used as proof. A ḥasan hadith with several chains rises to ṣaḥīḥ li-ghayrihi, and a mildly weak hadith supported by other chains of similar strength rises to ḥasan li-ghayrihi.
Rejected (mardūd). Mild weakness, such as a narrator with poor memory, a narrator of unknown status (mastūr) or a mursal chain, can be lifted by support. Severe weakness cannot: a narrator accused of lying makes the hadith matrūk, and a fabricated report (mawḍūʿ) is not counted as hadith at all.
Where scholars differ. Many allow acting on a mildly weak hadith for virtuous deeds, on conditions usually attributed to Ibn Ḥajar: it is not severely weak, it falls under an established principle, and one does not believe the Prophet ﷺ certainly said it. Others reject acting on weak hadith altogether.
Sources: Ibn Ḥajar, Nuzhat al-Naẓar fī Tawḍīḥ Nukhbat al-Fikar; Ibn al-Ṣalāḥ, Muqaddima.
The video's main example is Imam Mālik. Only two narrators stand between him and the Prophet ﷺ, and his book still exists today.
Select a narrator
Each narrator has a biography in the books of rijāl. Pick one, then mark him unknown or remove him and watch the grade change.
Mālik → Nāfiʿ → Ibn ʿUmar · The example in detail
Claim · "Hadith weren't written down for centuries"
The video rejects the idea that hadith were first written two to four hundred years later. It lists written records starting in the Prophet's ﷺ own lifetime.
The video names ʿAmr ibn al-ʿĀṣ and ʿAlī ibn Abī Ṭālib, writing on parchment, leather and bone. Paper existed but was rare.
Abū Hurayrah's student (the speaker says "Munabbih") kept written scrolls. Imam Aḥmad later quoted from those scrolls.
The caliph had hadith written down, working with al-Zuhrī. According to the speaker, this led to Mālik writing a full book.
The speaker places it within about a century of the Hijra. It survives complete and has an English translation.
About 40,000 hadith, far more than al-Bukhārī's collection, and al-Bukhārī took some of his hadith from Aḥmad. It also survives complete.
Written within 100 years · Parchment and scrolls · The official order to write · Al-Bukhārī and earlier collectors
Once a route passes the five gates, the next question is how many independent routes there are. Add routes with the slider, then pinch them through a single narrator.
The rule: "Even if it has 30 chains, if at any stage there's only one, we consider it to be one." · The four labels · Imam Aḥmad reportedly memorised a million chains
Interactive · why the rules work
The video's rules (prefer short chains, count only independent routes) follow from simple probability. Adjust how error-prone a single narrator is and watch both effects.
An illustrative model, not from the video. It assumes routes fail independently, which is exactly what the mutawātir conditions (different cities, schools and teachers) exist to guarantee. A route through a single narrator can't do better than that narrator, however many branches follow. Hover or use the arrow keys on a chart to read values.
Interactive · gates ٤ and ٥ in action
Eleven routes carry the same saying. One route has two words swapped. First find the odd one out, then trace which narrator introduced the error.
Illustration only. The wording is the well-known hadith on sea water, and the swap was invented for this demo. The narrator letters are placeholders, not real chains.
Mutawātir is the highest level. It is not reached by a set number of chains. It is reached when the conditions make a shared mistake or a coordinated lie impossible.
The conditions
A large number of chains, independent at every level, spread across different cities, and coming from different schools and teachers.
Each chain must also start with someone who heard or saw it directly (سمعت samiʿtu, "I heard"; رأيت raʾaytu, "I saw"). Every narrator after that must be known by name.
There is no fixed number, because quality counts more than count: fifty unknown narrators are worth nothing, while thirty strong ones can be enough. The speaker says some hadith have more than 70 independent chains with identical wording.
The four conditions · First-hand hearing · Quality over number · 70+ chains
Analogies from the video
Two people inventing a rumour have to coordinate. A hundred strangers in different places can't, so when they agree, they are right.
The sermon. If a hundred worshippers each describe what the speaker wore at the Friday sermon (white thobe, black overcoat, white shemagh), together they make a strong report.
The clinical trial. The speaker works in clinical trials. If 298 of 300 participants react the same way, the two exceptions are outliers. If all 300 agree, the effect is accepted.
Rumour · Sermon · Clinical trial
Examples given
Raising the hands in prayer, purity of sea water, and wiping over leather socks. The speaker says each comes through 50 or more independent chains.
Raising the hands (rafʿ al-yadayn) when going into and rising from rukūʿ. Some jurists still hold it only at the opening; the speaker notes fiqh is a separate question from hadith grading.
Its water is pure for wudu and its catch is lawful. The speaker reads this, with the Qur'an, as making all seafood halal, shrimp and shark included, with exceptions only for animals that hunt on land.
Wiping over khuff instead of washing the feet, reported through more than fifty chains.
Raising the hands · Sea water · Khuff
Throughout the video, the speaker measures other sources against these five gates.
This section summarises the video's arguments as they were made. It is not an independent assessment, and scholars in these fields would dispute some of the points.
They carry names, but the speaker says their writers are unknown, so their moral character can't be checked (gate ١).
He dates Codex Sinaiticus to the 3rd–4th century and says the author of Hebrews is unknown. That leaves gaps of centuries, which gate ٣ would not allow.
Accounts of Alexander or Julius Caesar are accepted because they appear in history books, with no named chain back to an eyewitness.
Differing numbers (700 or 7,000) and two accounts of Judas's death are copyist errors that, he says, Christian apologists acknowledge. Gate ٥ would reject them.
Paul says 500 people saw the risen Jesus, but none of them are named and Paul did not see it himself. A hadith scholar would ask who the 500 were.
He says the Dead Sea Scrolls, dated 200 BC to 70 AD, are the earliest copies: roughly a thousand years after Moses, by unknown writers, found in a cave.
Sonny's aside: with AI editing, a well-checked chain of narrators may now be more trustworthy than video footage.
Clinical trials don't check participants' character or reporting accuracy. They only count results, which the speaker sees as a shallower check than hadith grading.
The closing point
The speaker calls this only a short introduction. A full analysis also covers whose words a report records, how chains break, and hidden concealment (tadlīs). His conclusion is that rejecting the science without studying it is like dismissing chemistry without having learned any.
Other checks he mentions briefly: whether a report goes back to a Companion, a Successor or the Prophet ﷺ, or is a ḥadīth qudsī; where and how a chain breaks; and who studied under whom, in which generation and in which city.
Further checks · The chemistry analogy · Closing