al-Ṭalāq 65:1

Yā ayyuha'l-nabiyyu idhā ṭallaqtumu'l-nisāʾa fa-ṭalliqūhunna li-ʿiddatihinna wa aḥṣu'l-ʿidda, wa'ttaqu'llāha rabbakum, lā tukhrijūhunna min buyūtihinna wa lā yakhrujna illā an yaʾtīna bi-fāḥishatin mubayyina, wa tilka ḥudūdu'llāh, wa man yataʿadda ḥudūda'llāhi fa-qad ẓalama nafsah, lā tadrī laʿalla'llāha yuḥdithu baʿda dhālika amrā

"O Prophet! When you divorce women, divorce them towards their waiting period, and count the period. Guard yourselves against God, your Lord. Do not put them out of their houses, and let them not go out — unless they commit a clear indecency. These are the limits of God; and whoever transgresses the limits of God has wronged himself. You do not know; it may be that God will bring about something new after this."

The longest verse of the sūra, and its whole programme is here. It contains six separate rulings or qualifications. I shall take them in order.

The verse opens with a grammatical oddity, and it is something that ought to be noticed at once.

يَا أَيُّهَا ٱلنَّبِىُّ — "O Prophet". Singular, definite, to one person.

إِذَا طَلَّقْتُمُ — "when you divorce". Plural.

That is, the sentence calls out to one person and then immediately issues a ruling to a community. In the classical works of grammar several explanations are given for this construction:

All three are defensible and they do not exclude one another. I want to add this — as my own reading: the whole sūra advances by this double structure. The first verse begins with a singular address, continues with plural imperatives, and at the end returns to the singular: lā tadrī — "you do not know". That is, the verse draws a circle. The rule is for everyone, but what is said at the end is said to each person one by one. I shall take up the reason for this in the final section.

It is also transmitted in the sources that there is a singular variant reading of the verse, "O Prophet! When you divorce women…". The existence of this reading is on record; since I am not sure to which imām it belongs, I give no name. And since it does not substantially change the meaning, I do not dwell on it here either.

The word that carries the name of the sūra. The root is ط ل ق (ṭ-l-q), and the concrete sense of the root sits strikingly ill with what the English word "divorce" calls to mind.

The centre of the root: to untie a bond, to let loose, to set free.

• طَلِيق (ṭalīq) — one who has been set free; one released from captivity or from a bond.
• إِطْلَاق (iṭlāq) — releasing, setting loose. The derivative muṭlaq (unrestricted, unbound) belongs here as well.
• نَاقَةٌ طَالِق (nāqa ṭāliq) — a she-camel whose tether has been untied and which has been let out to pasture. This is the most concrete use of the word: the animal's rope is untied and the animal is left free to go.
• طَلْقُ الْوَجْه (ṭalqu'l-wajh) — open-faced, cheerful. The face's untying from its tension.
• طَلْق — the pangs of childbirth. The connection the lexicographers give: birth is the child's being untied and parted from the mother's body.

And here there is something worth pausing over. In Arabic the name of divorce is not breaking but untying. The word describes not a demolition but a setting free.

This has two consequences, and both need to be recorded:

First — what the word assumes. For a thing to be "let loose", it must first have been bound. The word ṭalāq assumes a bond behind it, and the name of that bond is set down explicitly in the Qurʾān: مِيثَاقًا غَلِيظًا — "a solemn covenant" (al-Nisāʾ 4/21). The same expression is used in the Qurʾān for the covenant taken from the prophets as well (al-Aḥzāb 33/7). That is, marriage is counted at the level of the word as a weighty contract; and ṭalāq is not the annulment of that contract but its untying.

Second — the manner the word imposes. The one who performs ṭalāq is the one who sets the other free. He is not disposing of a piece of property, he is not driving off an enemy; he is untying a bond. This choice of word agrees directly with the two qualifications bi-maʿrūf in the rest of the sūra (verse 2): both the keeping and the letting go are to be done "with the good that is known".

I offer this as a moral inference, not as a ruling that follows necessarily from the dictionary sense of the word. Languages do not always remember the origins of their words; but the Qurʾān had alternatives before it when it chose this word — faskh, qaṭʿ (to cut), farq — and in the second verse it will use one of them (fāriqūhunna).

Pattern: تَفْعِيل (tafʿīl). Ṭallaqa is Form II. This form generally conveys causation or intensity: ṭalaqa (he became free) → ṭallaqa (he set free, he let loose). That is, the subject of the verb is the party that sets the other free.

إِذَا — in Arabic this is used for conditions whose occurrence is certain or strongly probable; in conveys possibility (this distinction was recorded in al-Muzzammil and in al-Inshirāḥ).

The use of idhā here is an acknowledgement: divorce will occur. The text does not position it as a deviation, an exception, a thing that ought not to be. It takes it as a situation that will occur and regulates it.

This is the basic posture of making law, and it is not always to be seen in religious texts. A text either forbids an act it does not approve of or ignores it. The Qurʾān here takes a third way: it does not forbid the act, but neither does it leave it unattended.

The first ruling of the verse and one of its most debated expressions. The difficulty lies in the letter لِ (lām).

Why is it difficult? Because in Arabic the particle li- covers a wide field: belonging, purpose, time, cause, direction. And which of these it is here is not said. The sentence does not say "divorce them during their waiting period" (it does not say fī ʿiddatihinna); it says "for/towards their waiting period".

The readings proposed in the classical commentaries:

All three readings come out at the same result in practice: a divorce cannot be pronounced at a random moment; it is pronounced at a time at which the period can begin. The difference between them is a technical point of grammar; it does not change the ruling.

But which time is this time? Here we are in the field of jurisprudence, and it will not be settled here. It is transmitted in the sources that there is wide acceptance of the view that the divorce should be pronounced at a time when the woman is not menstruating and when she has not been approached during that period of purity; a divorce pronounced in observance of this is called ṭalāq sunnī, one pronounced without it ṭalāq bidʿī. The details of this distinction, whether a bidʿī divorce is to be counted valid, and the consequences attached to that, have been disputed. This commentary adopts no preference here.

But something can be said about the logic of the regulation, and it needs to be said:

The condition of time separates the decision from the moment of anger. To set a condition of time is to insert a compulsory interval between the moment the decision is taken and the moment it is carried out. The man cannot divorce at the moment he is angry; he has to wait for the suitable time. And the waiting tests the decision itself.

This is the functional counterpart of the mechanism modern law calls a cooling-off period. Today the divorce law of many countries has similar intervals, and the ground given is the same: to prevent an irreversible decision from being taken in a passing state of feeling. I record this parallel not as a claim but as a comparison that makes the function of the regulation visible.

A second function: the determinability of the period. If the divorce could be pronounced at any time at all, when the waiting period began and when it would end would be open to dispute. To fix the time is to fix the period as well. And the next command comes precisely on top of this.

One of the most important words of the sūra is here.

عِدَّة — root ع د د (ʿ-d-d). To count; number, tally. From the same root: ʿadad (number), taʿdād (a counting), ʿidda (a counted quantity), maʿdūd (numbered).

The word is rendered "waiting period"; but what it says in Arabic is narrower and sharper: the counted period. There is no idea of waiting in it; there is the idea of counting.

This root was treated in al-Humaza, and a finding was made there. I do not repeat that analysis; a short reminder is enough: in treating the verb ʿaddadahu (he kept counting it) in Sūrat al-Humaza, this was recorded — "number makes uncertainty measurable; it draws out of an uncontrollable field a controllable figure." There this was the pathology of the man counting his wealth over and over: the counting did not remove the real uncertainty, it merely stood in its place.

Here the same mechanism works in the opposite direction, and this is one of the most striking contrasts between the two sections:

That is, counting is in itself neither a good nor a bad operation. What is counted, and in whose favour it is counted, determines it.

And what does the counting prevent here? Concretely, this: one party's lengthening or shortening the period at his own pleasure.

In an order in which the period is not counted, this happens: the man says "the waiting period is over", the woman says "it is not over"; the man says "I have not divorced you yet", the woman says "you have". Where there is no number, every word carries as much weight as the power of the one saying it. And in that order the balance of power is known.

Number puts measure in the place of power. This is one of the most basic functions of law, and here it is done with a single word.

أَحْصُوا۟ — root ح ص ي (ḥ-ṣ-y). And the choice of this word is further striking, because in Arabic ʿuddū could also have been said for "count" (from the same root, directly).

Iḥṣāʾ is akin to the word ح ص ى (ḥaṣā): pebble. The lexicographers record that the concrete sense of the root is "to count with pebbles" — before the number system and writing were widespread, counting was done with small stones; for each unit a stone was set aside.

The sense that comes out of this is counting completely, one by one, skipping none. Iḥṣāʾ is not "to know approximately"; it is to take an inventory.

The Qurʾān uses this word in the same way elsewhere too — in the sense of a thing's being recorded exactly, with no part of it left outside.

That is, the command is not "wait roughly a period"; it is "count them one by one". And the verb is a plural imperative: aḥṣū — count. Who is to count is not specified. This indeterminacy is meaningful: the task of counting has not been consigned to one party.

Root و ق ي (w-q-y). This root was analysed in full in al-Baqara in connection with 2/2; I do not repeat that analysis. A short reminder: at the centre of the root there is not fear but guarding — wiqāya (protection, a shield), muttaqī (one who guards himself). The muttaqī is not the timid man but the careful one.

What is new here is where the word has been placed.

The command comes immediately after two technical rulings (time + counting). That is, this order is set up:

1. Divorce at the suitable time. 2. Count the period. 3. Guard yourselves against God.

The third item is the overseer of the first two. For the first two rulings had — in this period — no institutional oversight. Who will count the period? Who will check that he has counted rightly? There is no judge, no register, no record of persons.

Taqwā is put in the place of oversight. And this is the first instance of a structure that will be repeated to the end of the sūra: beside the ruling an inner mechanism is placed that holds the ruling up.

I record this as an observation from the point of view of the sociology of law: in orders with a low capacity for sanction, the working of rules depends in large measure on internalised norms. That almost all the legal passages of the Qurʾān close with a divine name or a qualification of taqwā has to do with this structure. This is not a "proof"; it is a finding about how the text itself works.

رَبَّكُمْ — and note: it does not say "guard yourselves against God" and stop; it adds "your Lord". The word rabb means the one who nurtures, who raises, who watches over (the analysis was made in al-Fātiḥa). That is, the authority to be guarded against is at the same time the authority that looks after. The function of the addition is to take the command out of being a threat and to tie it to a relationship.

The most concrete ruling of the sūra, and the one that says the most.

And the first thing to be noticed is a pronoun: بُيُوتِهِنَّ — "THEIR houses".

The house is ascribed to the woman. And that in the very middle of the period in which the marriage is coming to an end. Had the man been speaking of his own house, buyūtikum would have been said; had it been left neutral, al-buyūt would have been said. It is not said.

What this ascription means has been debated, and there are two main readings:

The two readings do not exclude one another; the second is the legal consequence of the first. The result in any case is the same: throughout the waiting period the woman's right to be there takes precedence over the man's power to put her out.

What does the regulation prevent? Something very concrete: the woman's finding herself homeless from one moment to the next.

And to see the weight of this the conditions of that period must be recalled:

• A woman normally had no independent dwelling or income.
• A divorced woman would return to her father's house or to a relative. That door was not always open; her family might be dead, far away, or poor.
• The options before a woman with no shelter were extremely narrow in that social order.

That is, the ruling prevents a second blow arriving at the most defenceless moment. The divorce is already a loss; the ruling forbids the loss of shelter from being added on top of it.

A second function: the process's being lived in fact. If the woman remains in the house, the parties remain inside the same life for the whole period. For the possibility spoken of in the next sentence (lā tadrī laʿalla'llāha yuḥdithu…) to be able to come about, the parties must not have been physically severed from one another. The reversibility of the separation depends on the preservation of the place. Two people who have gone apart cannot make peace — they have no place in which to do it.

I offer this as an inference, but the verse's own word order supports the inference: the ruling about the dwelling and the sentence "it may be that God will bring about something new" stand inside the same verse, and side by side at that.

وَلَا يَخْرُجْنَ — "and let them not go out". The second half of the prohibition is directed at the woman.

This symmetry must not be passed over. The ruling is not a one-sided protection; it lays an obligation on both parties:

• The man cannot put her out — his power is limited.
• The woman does not go out — her freedom is limited.

The ground of both is the same: that the process should run to its end. If the woman could go out whenever she wished, the period would in fact be emptied of meaning; if the man could put her out whenever he wished, the period would not turn into a protection.

Views are transmitted in the classical sources to the effect that the woman may go out during this period for necessary needs, and the extent of this has been debated. I do not go into the detail of that debate.

There is a disagreement about the scope of the ruling, and it would not be right to conceal it:

This disagreement is real and has been treated at length in the literature of jurisprudence. I adopt no preference here. What I can say is this: the wording of the verse draws no distinction; the distinction has been built with other evidence and other reports. This is a finding, not a ruling.

The verse sets an exception to the ruling about the dwelling: "unless they commit a clear indecency."

فَاحِشَة — root ف ح ش (f-ḥ-sh). The sense of the root: an ugliness that exceeds measure, an evil that passes the bound. Fuḥsh is a thing's exceeding the limit in ugliness; in Arabic an excessively high price is also called fāḥish. That is, the word announces not the kind of the evil but its degree: not an ordinary fault, but what is openly and excessively ugly.

مُبَيِّنَة — root ب ي ن (b-y-n). This root was analysed in full in al-Bayyina; I do not repeat the analysis given there. A reminder: at the centre of the root there is the space between two things, and from this both the sense "to separate" and the sense "to be evident" arise. Mubayyina: what is plain, what stands there in a way that can be told apart.

The function of the adjective here is narrow and important: the exception is not open to suspicion. Had it said "indecency" and stopped, every allegation would have opened the door of the exception. The qualification mubayyina makes an allegation insufficient: there must be something evident, beyond dispute.

But what is this fāḥisha? Several explanations are transmitted in the classical commentaries:

These explanations are present in the sources and I adopt no preference among them. The dictionary sense of the word is not proper to the sexual field; but a part of the uses in the Qurʾān are in that field. The text does not specify here.

One thing, however, must be said plainly: how this exception is to be operated — who is to establish it, by what procedure — is not in the verse. And this gap is a place open to abuse. An exception clause, when no procedure of establishment is laid down, can be widened so far as in effect to abolish the ruling. It is for exactly this reason that the literature of jurisprudence has tried to narrow the scope of the exception. I record this as an observation; not as a ruling.

حَدّ — root ح د د (ḥ-d-d). The concrete sense: the dividing line between two things, the edge of a piece of land, the place where a thing ends.

The derivatives of the root confirm this picture:

• حَدِيد (ḥadīd) — iron. According to the lexicographers the connection is the sharpness of iron and its cutting apart.
• حَدَّاد (ḥaddād) — blacksmith; and at the same time gatekeeper, warder — the one who holds the boundary.
• مَحْدُود — bounded.
• حِدَّة — sharpness.
• تَحْدِيد — delimiting, defining. (The derivative taḥdīd, "delimitation", belongs here.)

The choice of the word states the character of the ruling: the ḥudūd are not things one is asked to do; they are lines one is asked not to cross.

This distinction is fine but real. A "command" tells you what you are to do; a "limit" tells you how far you may go and leaves you free within. What the sūra sets is the second: divorce is not forbidden, the edges of divorce are drawn.

Accordingly the same word occurs again and again in the middle of the divorce rulings in Sūrat al-Baqara as well (al-Baqara 2/229-230). Two sūras do the same work with the same word.

وَمَن يَتَعَدَّ حُدُودَ ٱللَّهِ فَقَدْ ظَلَمَ نَفْسَهُ — "whoever transgresses the limits of God has wronged himself."

يَتَعَدَّ — root ع د و (ʿ-d-w): to pass the bound, to go beyond, to trespass. From the same root ʿaduww (enemy — the one who violates the boundary), ʿudwān (aggression), iʿtidāʾ.

and فَقَدْ ظَلَمَ نَفْسَهُ — this is the crux of the sentence.

The expected sentence would have been: "whoever transgresses the limits has wronged his wife." After all, what is being regulated is a ruling that protects the woman's right.

The verse does not say this. It says "has wronged himself".

If one looks at the use of this expression in the Qurʾān, the pattern is widespread and it does this work: it turns the real address of the harm from the victim to the doer. When the rule is broken the first party harmed is of course the other side; but the verse brings forward the loss on the account of the one who breaks it.

The practical function of this is as follows — and I write it as my own reading: the person who breaks the rule has for the most part already accepted the harm to the other party. His relationship is broken as it is; the other's loss will not stop him. The sentence therefore carries the reckoning elsewhere — to the loss of the one who breaks it. And this is perhaps the only effective thing that can be said to a person on bad terms with the other side.

ظُلْم is from the root ظ ل م (ẓ-l-m): darkness; and from this "to put a thing somewhere other than the place it belongs to". Ẓulm is the disturbing of a balance. A person's wronging himself is his disturbing his own balance.

After six rulings the verse suddenly shifts into another language. And this last sentence is the ground of everything that came before it.

First the word order.

The whole verse was plural: ṭallaqtum, ṭalliqū, aḥṣū, ittaqū, lā tukhrijū. Five verbs, all second person plural.

And in the last sentence, suddenly: لَا تَدْرِى — "you do not know". Second person singular.

In Arabic rhetoric this shift is called iltifāt (a shift in the direction of address), and it is a technique often used in the Qurʾān. But not every iltifāt does the same work; two different functions were recorded earlier in this body of work (drawing the reader onto the stage in al-Humaza, handing the criticism to its owner in ʿAbasa). The function here is a third, and it is this:

The rule is set upon the community; the ignorance is said to each single person.

A rule of law addresses everyone — "when you divorce, do thus". But the sentence "you do not know" cannot be said to a community. Who does not know what is the affair of each single person. The sentence takes the ground of the rule out of being a general principle and carries it into the head of the man making the decision.

And sharper still: the sentence is said at the moment when the man is most certain. A person who has resolved on divorce is at that moment certain of his decision; had he not been certain he would not have made it. The verse enters exactly there: lā tadrī.

Root ح د ث (ḥ-d-th). The centre of the root: a thing's coming to be that was not there before, being new.

• حَدِيث (ḥadīth) — new; and speech, news. Both are from the same root: a piece of news is what is new. The term ḥadīth (a saying of the Prophet) comes from here as well.
• حَادِث / حَادِثَة (ḥāditha) — what happens, what comes about; an event.
• مُحْدَث — what has come about afterwards, newly arisen.
• إِحْدَاث (iḥdāth) — to bring forth something new.
• حَدَاثَة (ḥadātha) — newness, youth. (The word now used to render "modernity" is from this root as well.)

And the choice of this word changes what the sentence says.

The verse does not say "it may be that God will bring back the former state". It does not say yurjiʿu (he returns), it does not say yuʿīdu (he restores). It says يُحْدِثُ: he brings about something new.

The difference is not trivial. The expected possibility is the parties' "making peace and returning to what was". The word does not say this. What it says is wider and more realistic: a situation may arise that was not there before.

Why is this more realistic? Because in a broken relationship "returning to what was" is not possible in any case. A couple cannot, after a fracture has come between them, return to what preceded the fracture. The only place they can go is a new situation — either a different agreement, or a changed attitude, or (not out of the question) a cleaner form of the separation itself.

The verse does not say which it will be. أَمْرًا — indefinite, without the article: "a matter". What it is is not specified.

And the function of the indefiniteness here is not magnification but indeterminacy. The sentence gives no promise. It does not say "you will be reconciled". It says only this: you do not know what will happen.

And this word sets the tone of the sentence.

لَعَلَّ in Arabic conveys tarajjī (hoping). There is no certainty; there is possibility.

Had the verse said "God will bring about a new situation", the period would have been a waiting room: the outcome is settled already, only let the time come. Laʿalla prevents this. The outcome is not settled. The period does not guarantee the outcome.

Then what is the ground of the period? Not the outcome, but the possibility. The door is held open not because it is known that someone will come through it; because the probability of coming is not zero.

This sets the logic of the regulation on a very clear footing:

The law here does not command an outcome. It preserves a possibility.

And the manner in which the possibility is preserved is wholly material: the period will be counted, the woman will remain in the house. That is, the text does not give an exhortation such as "try to make peace"; it establishes an order in which making peace remains physically possible. An exhortation is recommended, an order is imposed. The verse chooses the second.

I want to gather up the structure that comes out of this verse as a whole, because the rest of the sūra is built upon it.

The verse does this:

1. It does not forbid divorce. 2. But it takes it out of being an operation completed in a moment. 3. It makes the period countable — so that neither party can define the period in his own favour. 4. It keeps the woman in place — so that the period is in fact lived and a return remains physically possible. 5. And it grounds all of this in a human being's not knowing his own future.

The fifth item carries the first four. And what is striking is this: the ground is not rested on the goodness or the morality of the parties. The verse does not say "treat one another well, and perhaps you will be reconciled". It says: you do not know.

That is, the support of the regulation is not a virtue but an epistemic limit: a human being's knowledge about the rightness of his own decision is incomplete.

And this is a fairly solid ground on which to make law. Rules that rest on virtue do not work on those who are not virtuous. A rule that rests on ignorance works on everyone — because no one knows the future.

One. Irreversible decisions and time.

The mechanism the verse sets up — putting a compulsory interval between an irreversible decision and its being carried out — is today a deliberate design principle in many fields. Withdrawal periods in consumer law, notice periods in labour law, the divorce waiting periods of many countries, even the "undo" button in software interfaces all carry the same idea: a decision's remaining reversible raises the quality of the decision.

The ground is the same too: the state of mind at the moment of decision is not a good indicator of the long-term rightness of the decision. The verse says this as lā tadrī.

I record this not as a "miracle" but as a parallel that makes the function of the regulation visible.

Two. The protective side of "counting".

Today "measuring" is generally taken to be a cold and technical business; in human matters the intrusion of measure looks like a defect. The verse shows the opposite direction: where there is no measure, power fills the gap.

The counting of a period, the keeping of a record, the finding of a witness — these are not formalities that chill a relationship; they are the only support in the hand of the weaker party. In every order in which a word is not a document and a period is not a number, who is in the right is determined by who is strong.

Three. The tendency of exception clauses to expand.

The qualification noted above on illā an yaʾtīna bi-fāḥishatin mubayyina, though it looks like a narrow technical point, points to a general problem: an exception whose procedure of establishment is not laid down can eat the rule.

This is a known weakness in legal texts, and it meets us in the reading of religious texts as well. To read an exception placed upon a protection broadly and the protection narrowly reverses the work the text does. That the verse has added the adjective mubayyina is the first dam set against that expansion.