al-Nisāʾ 4:34-35

Al-rijālu qawwāmūna ʿala'l-nisāʾi bimā faḍḍala'llāhu baʿḍahum ʿalā baʿḍ wa bimā anfaqū min amwālihim, fa'l-ṣāliḥātu qānitātun ḥāfiẓātun li'l-ghaybi bimā ḥafiẓa'llāh, wa'llātī takhāfūna nushūzahunna fa-ʿiẓūhunna wa'hjurūhunna fi'l-maḍājiʿi wa'ḍribūhunna, fa-in aṭaʿnakum fa-lā tabghū ʿalayhinna sabīlā, inna'llāha kāna ʿAliyyan Kabīrā · Wa in khiftum shiqāqa baynihimā fa'bʿathū ḥakaman min ahlihī wa ḥakaman min ahlihā, in yurīdā iṣlāḥan yuwaffiqi'llāhu baynahumā, inna'llāha kāna ʿAlīman Khabīrā

"Men are qawwām over women, because God has favoured some over others and because they spend of their property. Good women are qānit; they guard, where no one sees, what God has guarded. As for those women whose nushūz you fear, admonish them, leave them alone in their beds, and ḍarb them. If they obey you, then seek no way against them. God is exalted, great. — And if you fear a rift between the two of them, send an arbitrator from his family and an arbitrator from hers. If the two of them wish to set things right, God will bring them together. God is Knowing, Aware."

This is the most debated place in the sūra. So I set down at the outset what is done in this section and what is not.

One. The two verses are treated together. The thirty-fourth verse is not a unit separable from the thirty-fifth: the first verse takes as its subject the settling of a dispute from within, the second the settling of the same dispute from outside by arbitration, and the block ends with the word iṣlāḥ. To cut the block in the middle is to damage the text.

Two. No legal ruling is issued. The readings of the classical commentaries and of the schools of law are related as a table of the disagreement, with their grounds. None of them is preferred and none of them is binding.

Three. Current debates are not entered into. How this verse is understood today, how it is answered in this or that legal order, and the social debates around it are not the subject of this text.

Four. No demeaning language is used and the reader is not scolded. No general judgement of worth about either party is drawn from the verse.

Five. What is not in the wording of the verse is not added to it. Below, the words and grounds the verse itself gives are established first; the readings come only after that.

The root ق-و-م: to rise to one's feet, to stand up, to keep a matter standing.

The root has occurred twice already in this sūra: at the fifth verse, allatī jaʿala'llāhu lakum qiyāmā ("the property God has made a means of standing upright for you"), and it was treated there under the heading "the weave of roots in the sūra". I rely on that treatment.

And the pattern must be recorded: qawwām is on the فَعَّال pattern. In Arabic this pattern states two things:

The same word was treated at al-Māʾida 5/8, and this was recorded there:

"That is, a qawwām is not 'one who acts justly on one occasion' but one who makes it his business to keep justice standing. The command asks not for a behaviour but for a character."

I rely on that treatment. And setting the word's three occurrences in the Qurʾān side by side must be recorded as a datum of the text:

The same pattern is present in all three. In two of them the word is a command; in the first it is a statement of a situation. This is a countable fact.

And this must be recorded: the verse does not explain the word qawwām. How much authority and how much obligation the word states is not written in the verse; what is written in the verse is two grounds.

The particle bimā occurs twice in the sentence, and two grounds are set out:

The second ground must be treated in detail, because its wording is plain and not in dispute.

Anfaqū — a perfect verb, third person plural: they spent. Min amwālihim — of their property.

The verse, that is, ties the second ground to an act. And the act is in the same field as the verb iktasabū / iktasabna two verses earlier: property, earning, spending.

I record this as my own reading, and its ground is the tense of the verb: anfaqū is past and states an action. A ground that states an action is bound to that action. The verse has thus tied the ground not to an attribute but to something done.

That this explanation is found in the classical commentaries too must be recorded: the connecting of qawwām-hood with the obligation of maintenance is widely transmitted. But there is disagreement over the relative weight of the two grounds and over the scope of the first, and it is set out in the table below.

First the datum of the text: the sentence does not say "men over women". It says baʿḍahum ʿalā baʿḍ — "some of them over others".

And two verses earlier, in the thirty-second, the same structure occurred (mā faḍḍala'llāhu bihī baʿḍakum ʿalā baʿḍ) — and there the sentence was the subject of a prohibition: do not covet this. This is a datum of the text that can be verified, and it was set out in a table above.

There is disagreement in the classical commentaries over how this expression is to be read. I put it in a table; I impose no preference.

All three readings have been defended in the classical and later literature. I prefer none of them and none of them is binding.

And a limiting record is written plainly, per the method (Method): no conclusion that one sex is superior to the other in respect of being human can be drawn from this sentence. The sūra's first verse (min nafsin wāḥida) and its twenty-fifth (baʿḍukum min baʿḍ) were treated in this commentary, and the records made there hold here too.

قَٰنِتَٰت — root ق-ن-ت: to keep at a task with constancy and reverence; to submit.

And the root's distribution in the Qurʾān must be recorded, because it shows the direction of the word:

Two tendencies are transmitted in the classical commentaries over towards whom the word is directed here: some explain it as obedience to God, some as constancy to the order within the family. I relate both tendencies; I impose no preference.

What can be said from the text is this, and it is an observation about word order: in the Qurʾān the dominant use of the root is directed towards God; and at al-Aḥzāb 33/35 the word is counted for both parties. This establishes no ruling beyond the wording of the verse, but it is a datum that ought to be put in the table.

حَٰفِظَٰتٌ لِّلْغَيْبِ بِمَا حَفِظَ ٱللَّهُ — "those who guard, where it is not seen, what God has guarded."

The root ح-ف-ظ occurs twice in the clause: ḥāfiẓāt and ḥafiẓa'llāh. And the word order is worth recording: the ground of the guarding is tied again to a guarding — bimā ḥafiẓa'llāh.

غَيْب — root غ-ي-ب: to be hidden, to be out of sight. Here the word states the place where no one sees.

The root ن-ش-ز. And its concrete sense must be recorded: nashz — high ground, a hillock, a rise.

The family of senses that follows from this:

The picture in the word is a rising and a leaving of one's place. One party's going out from the position it holds within a relationship.

And now comes the most important datum of the text in this discussion.

The word occurs twice in this sūra, and in its second occurrence it is used of the MAN.

This is a symmetry that can be verified from the text and a countable fact: the same root, in the same sūra, is applied to both parties.

And the detail of the symmetry must be recorded too:

It can be seen that both blocks open with the same word and close in the same field: iṣlāḥ and ṣulḥ.

I record this as my own reading, and its ground is the two tables above: the sūra recognises the same phenomenon for both parties, and ends both blocks in the direction of agreement.

The verse sets out three verbs: fa-ʿiẓūhunna, wa'hjurūhunna fi'l-maḍājiʿi, wa'ḍribūhunna.

The table below gives, with their grounds, the readings transmitted in the classical exegetical and juristic literature about these three steps. None of them is this text's preference, none of them is binding, and no practice can be derived from here.

This table is not a fatwā and cannot be read as a guide to practice. The reason it is placed here is the rule on disagreement in the method (Method): the disagreement is not concealed, no preference is imposed.

And there are three qualifications that follow directly from the wording of the verse; these are not disagreement but data of the text:

The third qualification must be stressed separately: having counted the steps, the verse sets an end to them, and states that end as a prohibition — fa-lā tabghū ʿalayhinna sabīlā: "seek no way against them."

The root ب-غ-ي: to want, to seek; and to exceed the bound. These two sides of the word are worth recording: what is forbidden is the search for a pretext.

The verse closes with two attributes: ʿAliyy (exalted) and Kabīr (great).

The sūra's order of verse-endings was set out in a table in the introductory section, and most of the endings were attributes of watching. Here there are two attributes of greatness.

About the function of this verse-ending the following is widely transmitted in the classical commentaries: because the ruling addresses the party into whose hand authority is given, the closing reminds that party that there is one greater than himself.

I relate this explanation as it is transmitted. And I add the following as my own reading: the verse-ending is directed at the party the verse addresses — because it is he who is addressed by the commands in the verse.

The thirty-fifth verse is the continuation of the thirty-fourth, and the block is completed only here.

And the first thing to be recorded is the change in the word order:

In the second verse, that is, the matter ceases to be one party's conduct and becomes what is between the two.

Shiqāq — root ش-ق-ق: to split, to cleave in two. Shaqq — a fissure. The word draws the picture of a whole being split in two. And the qualification baynihimā (between the two of them) does not tie the responsibility to one side.

فَٱبْعَثُوا۟ حَكَمًا مِّنْ أَهْلِهِۦ وَحَكَمًا مِّنْ أَهْلِهَآ — and the symmetry here is a countable fact:

The clause is built twice and only the pronoun changes. This is the same structure as the naṣīb symmetry in the thirty-second verse — the sūra repeats this construction.

Ḥakam — root ح-ك-م: to make a thing firm, to hold it in check. From the same root come ḥikma and ḥukm; and ḥakama — the bit of a horse, the thing that holds it. The picture in the root is holding in check.

إِن يُرِيدَآ إِصْلَٰحًا يُوَفِّقِ ٱللَّهُ بَيْنَهُمَآ — and here there is a disagreement of word order:

Both readings are transmitted; I impose no preference. What can be verified from the text is this: there are two dual pronouns in the sentence (yurīdā and baynahumā), and the verse does not specify either of them plainly.

The root ص-ل-ح was treated in this sūra at the fifteenth and sixteenth verses (fa-in tābā wa aṣlaḥā fa-aʿriḍū ʿanhumā), and it was recorded there under the heading "the block's real qualification": a door of return is left open after the ruling. I rely on that treatment.

Now the block must be looked at as a whole.

The block ends not with a sanction but with the bringing together of the two.

And the sūra does the same thing once more in the hundred and twenty-eighth verse:

وَإِنِ ٱمْرَأَةٌ خَافَتْ مِنۢ بَعْلِهَا نُشُوزًا أَوْ إِعْرَاضًا فَلَا جُنَاحَ عَلَيْهِمَآ أَن يُصْلِحَا بَيْنَهُمَا صُلْحًا وَٱلصُّلْحُ خَيْرٌ "If a woman fears nushūz or turning away from her husband, there is no fault upon the two of them in setting things right between themselves by a settlement. And settlement is better." (4/128)

I set the two blocks side by side. The table compares the wording of the two verses; it establishes no ruling.

I record this as my own reading, and its ground is the table above: the sūra describes two separate paths in two separate places for the same phenomenon, and closes both paths with the same family of words — ṣulḥ and iṣlāḥ.

And this is the sūra's general line: in the twenty-fifth verse wa an taṣbirū khayrun lakum, in the hundred and twenty-eighth wa'l-ṣulḥu khayr, in the hundred and forty-ninth in tubdū khayran aw tukhfūhu aw taʿfū ʿan sūʾin. These three were set out in a table in the discussion of the twenty-fifth verse; I rely on that.

One. The structure of the block itself provides for a dispute to be handled by stages: first talking, then distance, then an arbitrator from outside. And at every stage there is a qualification of stopping: fa-lā tabghū ʿalayhinna sabīlā.

Two. The arbitration arrangement of the thirty-fifth verse asks that the dispute be handled by an equal number of people from the families of both parties. This is a construction that rules out the circle of one party settling the matter on its own.

Three. That the word nushūz is applied to both parties in two separate verses shows that the sūra does not build this matter in one direction only. This is not an interpretation but the distribution of the word within the sūra.

And a limit: this section makes no decision about how the verse is to be applied. None of the tables of disagreement above is binding; to derive a ruling is the business of the literature of jurisprudence, which has the competence for it.