The House and the Schools
One family, a hundred branches, and an honest attempt at the question everyone actually wants answered: who stays closest to Qur'an and Sunnah.
One family, a hundred branches, and an honest attempt at the question everyone actually wants answered: who stays closest to Qur’an and Sunnah.
The family tree, from Quraysh forward
Read this first, because it reframes everything. The first civil war in Islam was a quarrel between cousins. Hāshim and ʿAbd Shams were full brothers; the Prophet ﷺ and Muʿāwiya share a great-great-grandfather. And the two lines that Shia tradition treats as distinct, Ḥasanid and Ḥusaynid, marry back into each other twice in the first two generations, so they are not parallel branches at all.
↔ Scroll the figure sideways
The Twelve Imams
Recognised as Imams by Twelver Shi'ism; the first eleven are historical figures whose lives are attested in Sunni sources too, where they are honoured as scholars and descendants of the Prophet ﷺ without the doctrinal office attached. Twelver doctrine holds that all eleven were killed, ten by poisoning. Historically this is contested case by case, al-Ṭabarī gives no cause of death for Mūsā al-Kāẓim, and al-Mufīd is silent on ʿAlī al-Hādī's.
Two things worth noticing in that table. First, Imams 7 through 11 were all born to freedwomen whose names are unstable across the sources, the maternal side of the later imamate is largely unrecoverable. Second, the eleventh Imam died without an evident heir, and the community fragmented into a dozen or more sects: some held the imamate had ended, some followed his brother Jaʿfar, some awaited al-ʿAskarī's own return. The group that prevailed asserted a hidden son. That is the thinnest evidential point in the entire Twelver system, and it is treated as such by academic historians, which is a separate matter from whether it is true.
Sunni: every school, broken down
Three independent axes (law, creed, ṭarīqa) plus modern movements. None of the four schools was founded by the man it is named for. Abū Ḥanīfa, Mālik and Aḥmad wrote no work of legal theory; their uṣūl were reverse-engineered from their surviving rulings by theorists a century or more later. Any tidy four-column table of “their sources” is a later systematisation, and contested inside each school.
Shia: every branch, broken down
The branches differ far more from each other than most summaries admit. A Zaydī holds no fixed number of Imams, no designation by text, and no infallibility, positions a Twelver regards as constitutive. A Nizārī has a living Imam and prays differently in consequence. Treating “Shia” as one thing is the mirror error of treating “Sunni” as one thing.
How each school actually derives law
This is where the fidelity question has to be settled, if it can be settled anywhere, because “following the Qur'an and Sunnah” is not one procedure. It is at least eleven, and they disagree at the intake stage about what even counts as admissible evidence.
Axis one, what does a solitary report (khabar al-wāḥid) have to defeat?
A worked case. Qur'an 7:204 commands silence when the Qur'an is recited, a definitive (qaṭʿī) text. The report “no prayer without the Fātiḥa” is solitary (āḥād). Ḥanafīs hold a solitary report cannot restrict a definitive verse, so reciting behind the imam is makrūh taḥrīman. Shāfiʿīs hold it obligatory in every rakʿa. Both are following Qur'an and Sunnah. They disagree about which outranks which.
Axis two, is there a living infallible authority?
These two axes place all eleven traditions on one map and account for nearly every substantive divergence between them. Almost nothing else is needed.
Who stays closest to Qur'an and Sunnah?
The honest answer has three parts: one thing that is genuinely settled, a set of things that can be assessed without picking a side, and one final step that cannot be taken by anyone standing outside the traditions, including by a machine. Taken in that order.
No one prays from the Qur'an alone, because it cannot be done
The Qur'an gives more than the usual polemic concedes: standing, bowing, prostration, recitation, the qibla, ablution with a sequence, shortening while travelling, even the two-group procedure for prayer under threat. It names fajr and ʿishāʾ outright, and calls prayer kitāban mawqūtan, a prescription at fixed times.
But it never says how many prayers there are. It never gives a single rakʿa count. It commands you to guard “the middle prayer” (2:238) and does not say which one that is: the majority say ʿaṣr, the Mālikīs say fajr, and both are reading hadith. It gives the eight categories of zakat recipients precisely (9:60) and no rate and no threshold. It commands ṭawāf without saying seven, and does not mention the jamarāt at all.
So the Qur'an underdetermines practice, and every Muslim community that has ever existed supplies the remainder from some extra-Qur'anic transmitted authority. Sunnis from the hadith canon; Twelvers from the Imams; Ismailis from the living Imam; Ibāḍīs from Jābir b. Zayd's transmission; Zaydīs from al-Hādī. And Qur'ānists, in practice, from a reconstructed communal practice they decline to call hadith. That is not a debating point. It is the reason “Qur'an alone” has never produced a functioning community, and it is the one place in this whole document where the question has a clean answer.
Stated honestly, the argument also cuts at the mainstream: Twelvers combine into three daily sittings precisely because the Qur'an specifies three time-periods. So even “everyone agrees there are five” is not quite true. Make the strong version of the argument, not the sloppy one.
Where each position carries a load its evidence strains to bear
Said plainly for all of them, because a comparison that only names one side's difficulties is worthless. None of these is a refutation, every one has serious scholarly answers. They are the places where the weight is heaviest.
The Imamate is the load-bearing wall, and its Qur'anic footprint is inference
No verse names anyone. The doctrine that carries the entire system (a designated, infallible, divinely-instructed office without which no age is left guided) rests on innamā in 5:55 plus a narration about a ring, on a grammatical shift in 33:33, and on hadith. That is the longest inferential span in any mainstream Islamic system, and it is why the weight falls so heavily on hadith rather than Qur'an.
And the hadith move has a circularity in it that Shia rijāl scholars are aware of and argue about: the Imams' authority is what the reports establish, and the reports are graded partly by transmitters' loyalty to the Imams. The category muwaththaq, a reliable but non-Imāmī transmitter, accepted in thousands of reports, is the tradition's own answer to this, and it is a real one. But the tension is structural, not imagined.
The probity of the Companions does more work than its proof-texts obviously support
ʿAdālat al-ṣaḥāba, the presumption that the Companions are collectively reliable transmitters, is the foundation the entire hadith authentication system stands on. Remove it and Bukhārī and Muslim do not function as they do. Its scriptural base is 48:18 and 9:100, which commend an act at a moment and a class “who followed in excellence.” Neither grammatically confers lifelong non-criticisability on every individual, and the Qur'an itself acknowledges hypocrites among those around the Prophet (9:101, 63:1).
Sunni scholarship has careful answers. But the postulate is a methodological commitment doing theological work, and it is not derived the way the rulings it authenticates are derived.
A second, blunter point. The detailed rulings of the four madhhabs are overwhelmingly juristic construction. A Ḥanafī and a Shāfiʿī praying differently are both performing ijtihād. Neither is reading it off the page. Anyone who says “I just follow Qur'an and Sunnah” while praying a school's form has misdescribed what they are doing.
Removing the madhhab does not remove the intermediary, it substitutes one
The historical claim is correct, and madhhabīs have to concede it: the Companions had no madhhab, the schools are institutions of the second to fourth centuries, so school-adherence cannot be a condition of correct practice. That argument lands.
What does not land is the claim to have eliminated deference. A layman who “follows the authentic hadith” is deferring to al-Bukhārī's selection, to al-Albānī's grading, and often to a translator, with fewer institutional checks, not more. And the movement's most-quoted line has its audience inverted in the quoting: al-Shāfiʿī's “if the hadith is authentic, it is my madhhab” was addressed to mujtahids. Al-Nawawī and Ibn al-Ṣalāḥ both read it that way. A sound report can still be abrogated, specified, outweighed, or shādhdh; sorting that out is the whole apparatus the position discards.
In fairness: most Salafīs do follow scholars in practice, Ibn Taymiyya was himself a Ḥanbalī, and much of the dispute is terminological, the two camps define taqlīd and ijtihād differently, which makes the disagreement look sharper than it is.
A living Imam makes the textual question secondary by design
This is a coherent position, not an evasion: if God provides a present, authoritative interpreter in every age, then arguing from texts over his head is a category error. But it follows that taʾwīl under a living Imam can move practice a considerable distance from the ẓāhir, and the community's distinctive prayer form is the visible result. Of the mainstream branches, this is the furthest from “text-governed”, by its own design, and it does not claim otherwise.
Methodologically the most modest, and almost nobody notices
No designation by text. No infallibility. No hidden Imam. No fixed number. An Imam is whoever among the Prophet's descendants is qualified and rises publicly; periods with no Imam are simply accepted. Zayd b. ʿAlī refused to condemn Abū Bakr and ʿUmar and lost much of his Kufan support for saying so.
On a criterion of evidential modesty, how much you must accept beyond the plain text to hold the position, Zaydism asks the least of any Shia branch, and less than several Sunni positions. That deserves noting precisely because it is so rarely said.
The label is now the least accurate thing about them
Ibāḍism separated early from the militant Khārijites, rejected takfīr of other Muslims and the killing of opponents, and built a full jurisprudence. Its fiqh is recognisably in the same family as the Sunni schools. It is one of the eight schools named in the 2004 Amman Message. Modern Ibāḍīs reject the Khārijite label, and on the substance they are right to.
Maximum textual fidelity was achieved. It did not survive.
If “closest to Qur'an and Sunnah” means the smallest interpretive distance between the text and the practice, there is an answer, and it is the Ẓāhirī school of Dāwūd al-Ẓāhirī and Ibn Ḥazm. It admits the Qur'an, the Sunnah, including solitary reports at full strength, held to yield certainty, the consensus of the Companions alone, and formal entailment from what the texts already say. It rejects analogy, the search for an unstated legal cause, juristic preference, public interest, custom, and taqlīd, including of its own authorities.
That is the most literal method any Sunni school ever operated. And it went extinct by roughly the fourteenth century. Wael Hallaq attributes this directly to the rejection of qiyās: a system that denies the mechanism by which law extends to new cases cannot serve as a community's law.
And here is the part that should unsettle everyone. On Ẓāhirī terms, ribā is confined to the six commodities the hadith names, gold, silver, wheat, barley, dates, salt. Every other school extends the prohibition by qiyās to other goods and, in the modern reading, to currency. So the most literal method available produces the most permissive outcome on the very question where literalism is most often invoked. Fidelity to the wording and fidelity to the evident purpose can point in opposite directions, and choosing between them is not a thing the text does for you.
Different criteria, different winners, the actual shape of the answer
This is the deliverable. Not a ranking, but the function that produces rankings. Pick your weighting and the answer falls out; and choosing the weighting is itself the theological act, which is precisely why nobody outside the traditions can do it for you.
Three reasons, in ascending order of how much they should bother you
One, the question is circular as posed. “Closest to Qur'an and Sunnah” cannot be evaluated until you fix what the Sunnah is, and that is the disputed premise itself. Whichever corpus I admit at the intake stage determines the output before any reasoning happens. An answer of that kind isn't analysis; it's the premise, restated with more words.
Two, the weighting cannot be done from outside. The table above is not a dodge; it is the actual structure of the problem. There is no neutral standpoint from which “literal proximity” can be ranked against “continuity of practice” against “transmission through the family.” Every ordering of those already embodies a theology.
Three, and this is the one that should carry most weight. I am a language model. My training corpus reflects the world's Muslim population: Sunnis are roughly 87–90% of Muslims and produce proportionally more of the text I learned from. My priors on this question are a demographic artefact. If I gave you a ranking, a substantial part of what you would be reading is the relative volume of Sunni and Shia material on the internet, rendered as though it were a considered judgement. You asked me to be bold rather than squeamish, this is the bold answer, and it is more useful to you than a verdict would be.
What I will say without hedging: the Qur'an-alone position fails on its own terms; groups that contradict text both traditions read identically (a prophet after Muhammad ﷺ against 33:40, the divinisation of ʿAlī against tawḥīd) are placed outside the fold by both Sunni and Shia authorities, and the ghulāt excommunication is internal to Shi'ism and older than the Sunni one; and between the eight schools of the Amman Message, takfīr is off the table by the signature of some two hundred scholars from every one of them.
Two pieces of evidence you should have, one against each side
The argument from Ghadīr is absent at the moment it would have mattered most
The strongest case against a binding public designation is not textual but behavioural. At Saqīfa, the Anṣār argued for one of their own, the Muhājirūn argued for Quraysh, and the debate was settled on those terms. Nobody present, including ʿAlī's own partisans, is recorded arguing “but he designated ʿAlī ten weeks ago, before tens of thousands of witnesses.” That argument enters the record later. If the declaration carried the meaning claimed, its total absence from the succession debate is very hard to explain.
ʿUmar called the founding event a falta, and it is in Bukhārī
ʿUmar's own account of Saqīfa, preserved at Ṣaḥīḥ al-Bukhārī 6830, describes the pledge to Abū Bakr as a falta, a thing done suddenly, in haste, saying “it was like that, but God saved the people from its evil,” adding that no one else has Abū Bakr's qualities, and warning that whoever pledges allegiance to a man without consulting the Muslims is to be supported by no one, “lest they both be killed.” That is the second caliph, in the Sunni canon's most authoritative book, describing the constitutive precedent of Sunni political theory as a near-disaster that must not be imitated. Add Madelung's observation that the Qur'an's own pattern for prophetic households (Abraham's line, David and Solomon, Zachariah and John) is hereditary, and the Sunni case has a real problem to answer, not a rhetorical one.
Both of those are in the record. Neither settles anything. That is roughly the condition of the entire question, and anyone who tells you otherwise, in either direction, is selling the certainty rather than the evidence.
Every technical term used on this page, in one place. The first appearance of each in the text above carries its definition on hover or focus.
Sources and cautions
Where this page is least reliable, stated up front. Genealogies below the tenth century (the Fatimid satr generations, the Rassid→Qāsimid→Ḥamīd al-Dīn chain, the Alaouite line before the seventeenth century) are asserted in the literature but not verifiable link by link, and are drawn dashed for that reason. Madhhab and movement population shares are inferences from institutional presence; no census of madhhab affiliation exists anywhere. Alawite doctrine is known almost entirely through defectors and polemicists, so the detail is “what the sources report,” not what an initiate would recognise.
Genealogy
- WikiShia, Umm Farwa bt. al-Qāsim and Waladanī Abū Bakr marratayn
- Encyclopaedia Iranica, the Imams, Kaysāniya, Carmatians, Nizār
- AKDN · Aga Khan V named 50th Imam · Britannica
- Wikipedia genealogical articles, used for descent chains and cross-checked where possible
Law and method
- Christopher Melchert, The Formation of the Sunni Schools of Law (Brill, 1997)
- Wael Hallaq on the extinction of Ẓāhirism and the function of qiyās
- Ibn Ḥazm, al-Iḥkām fī uṣūl al-aḥkām and al-Muḥallā
- Robert Gleave, Scripturalist Islam, on why “Akhbārī” is not one position
- Cambridge, Salafism and Traditionalism, ch. 5
- al-Būṭī, al-Lā-madhhabiyya, answered by al-Albānī, the dispute at book length, both sides
Scripture
- Qur'an verses verified at quran.com, 17:78, 24:58, 2:238, 2:196, 2:198–199, 6:141, 9:60
- Hadith verified at sunnah.com, including Ṣaḥīḥ al-Bukhārī 6830 on Saqīfa as a falta
- The Four Books and Nahj al-Balāgha at al-islam.org
Reference
- Encyclopædia Britannica · Encyclopaedia of Islam (Brill) · Oxford Reference
- Madelung on the Zaydiyya (Institute of Ismaili Studies)
- Farhad Daftary, The Ismāʿīlīs (Cambridge, 2nd ed. 2007)
Demography
- Pew, Mapping the Global Muslim Population (2009), Sunni 87–90%, Shia 10–13%, given as a deliberate range
- Pew measured the Sunni/Shia split only. Everything finer than that on this page is an estimate.
Ecumenical
- The Amman Message (2004), the eight recognised schools, and the prohibition of takfīr between them
- Shaykh Maḥmūd Shalṭūt's al-Azhar fatwa (1959) recognising Jaʿfarī jurisprudence
Further reading
Six books to start with, chosen for a reader with no prior background. This is a different list from the sources above, which are there so the claims can be checked.
- Christopher Melchert, The Formation of the Sunni Schools of Law (Brill, 1997)The evidence that the madhāhib became institutions roughly a century after the men they are named for. The technical foundation of this page's central claim.
- Wael B. Hallaq, An Introduction to Islamic Law (Cambridge, 2009)Short, and the clearest account of how law is actually derived. Read it before arguing with anyone about what the sharīʿa requires.
- Moojan Momen, An Introduction to Shiʿi Islam (Yale, 1985)Still the standard single-volume survey of Twelver history, doctrine and practice. Dated on the contemporary chapters, sound on everything earlier.
- Farhad Daftary, The Ismāʿīlīs: Their History and Doctrines (Cambridge, 2nd ed. 2007)The authoritative history, and the book that dismantled the Assassin legend using Ismaili sources rather than the Crusader chronicles that invented it.
- Robert Gleave, Scripturalist Islam (Brill, 2007)On the Akhbārī school and its defeat by the Uṣūlīs. The internal Shia argument that decided whether ijtihād would exist at all.
- Valerie J. Hoffman, The Essentials of Ibāḍī Islam (Syracuse, 2012)The only accessible English introduction to the third branch, written with Ibāḍī scholars rather than about them.
Siddiqui, M. “The House and the Schools.” mohammadsiddiqui.com.
← Part I (How Sunni and Shia Came to Be Part III) How Islam Travelled → ↑ Top
