In our introductory grammar books in the seminary, we are taught a simple rule: the word aw (او), which translates to “or” in English, can sometimes carry the meaning of “and” (و), and sometimes the meaning of “rather” (بل). But the moment you move beyond the introductory texts into more advanced grammar and works of research, a different picture will emerge. You will realize that this is not a settled matter at all. It is, in fact, a disputed matter stretching back to the earliest days of Arabic grammar, and the conditions under which aw might function as bal are far narrower and more contested than the seminary shorthand suggests. I will summarize and present an overview of the discussion and differing opinions on this grammatical rule in this post.

First and foremost, the view we learn in the introductory grammar books is the view of the Kufan school of grammar, which held that aw could indeed carry the meaning of bal in certain contexts. Their proof-texts include verses such as:

وَأَرْسَلْنَاهُ إِلَى مِائَةِ أَلْفٍ أَوْ يَزِيدُونَ

“And We sent him to a hundred thousand people, or more.” (37:147)

The Kufan argument runs as follows: Allah cannot be in a state of doubt. Since He knows precisely how many people were present, the aw here cannot be functioning as a genuine disjunction (“or”); it must instead mean “rather” (bal). On this reading, the verse effectively says: “We sent him to a hundred thousand people, rather [it was] more.” The Kufans cite several other Qur’anic verses and lines of poetry in support of this reading.

On the other hand, the Basran school rejected this view outright. For the Basrans, aw is never used in the sense of and (و) or bal (بل). Rather than concede that the particle’s meaning shifts, they reinterpret the verses the Kufans cite so that the default meaning of “or” is preserved.

Applied to 37:147, the Basran solution is that the doubt expressed by aw is not attributed to Allah at all, but to the onlookers witnessing the crowd, people who might reasonably estimate “a hundred thousand, or perhaps more.” Read this way, no exception to the ordinary meaning of aw is required.

A third, more measured position is attributed to Sibawayh, the foundational figure of Arabic grammar. In his view, aw can carry the sense of bal, but only under specific conditions: the sentence must be negative (or a negative imperative), and the governing noun, verb, or preposition must be repeated. For example:

ما قام زيد، او ما قام عمرو

“Zayd did not stand, rather, Amr did not stand.”

Here the sentence is negative, and the governing verb (qāma) is repeated in both clauses, satisfying Sibawayh’s condition.1

Ibn al-Anbari, in his book al-Insaf, takes up this dispute directly and sides with the Basrans, defending the view that aw is never used to mean bal.

A Contemporary Refinement: Sayyid Hashem Tehrani

Among contemporary grammarians, the Shi’i scholar Sayyid Hashem Tehrani addresses this principle at length in his al-‘Ulum al-‘Arabiyyah (vol. 2, p. 698), and he ultimately sharpens Sibawayh’s condition rather than simply repeating it.

Tehrani argues that aw can function as bal, which is used for iḍrāb (a corrective “rather”), but only between two prohibitions or two negations that are genuinely different from one another. He gives the example of telling one’s son, “Do not interact with the ignorant, or [rather] do not argue with them” – two distinct prohibitions, where the second corrects or sharpens the first.

By contrast, when aw occurs between two prohibitions or negations that are of the same kind, as in “Do not strike me with the stick, or with the whip,” or “Let not Zayd stand, or Amr” – the particle retains its original disjunctive meaning. Each clause is directed at one of two parties; there is no correction taking place.

Applying this to Qur’an 76:24, “Do not obey among them a sinner, or a disbeliever” (لَا تُطِعْ مِنْهُمْ آثِمًا أَوْ كَفُورًا), Tehrani concludes that the prohibition is against obeying one of the two, not an instance of iḍrāb, and not evidence that aw means bal.

He also revisits the verses the Kufans relied on, 37:147, 2:74 (“So they are like stones, or harder”), and 16:77 (“The matter of the Hour is not but as a glance of the eye, or nearer”), and notably rejects the Kufan reading of these too. In his assessment, aw in these verses is being used in its original disjunctive sense, and those who read it as bal have erred, since what is intended is simply that the actual reality corresponds to one of the two stated possibilities.

Finally, on 2:99–100:

وَلَقَدْ أَنزَلْنَآ إِلَيْكَ ءَايَـٰتٍۭ بَيِّنَـٰتٍۢ ۖ وَمَا يَكْفُرُ بِهَآ إِلَّا ٱلْفَـٰسِقُونَ

أَوَكُلَّمَا عَـٰهَدُوا۟ عَهْدًۭا نَّبَذَهُۥ فَرِيقٌۭ مِّنْهُم ۚ بَلْ أَكْثَرُهُمْ لَا يُؤْمِنُونَ

“Indeed, We have sent down to you ˹O Prophet˺ clear revelations. ˹But˺ none will deny them except the rebellious. Rather [if read as aw, instead of awa], every time they took a covenant, a party of them threw it away? Rather (aw), most of them do not believe”

Tehrani notes that on the reading with sukūn on the waw (i.e. aw, “or,” rather than the conjunction awa, “Is it ever”), this verse does belong to the category of iḍrāb, because it occurs between two genuinely different negations: casting away the covenant in verse 100 refers to a failure of action, while disbelief and rebellion in verse 99 refers to an absence of inward faith.

In short, Tehrani agrees with Sibawayh that aw can only take the meaning of bal within a negative sentence, but unlike Sibawayh, he adds the further condition that the two clauses must be topically different, not merely grammatically parallel.

Taken together, these positions show that the claim “aw can sometimes be used in the meaning of bal” is not a universally accepted rule by all grammarians or a rule without conditions. The Basrans deny it altogether. Sibawayh restricts it to negative sentences with a repeated governing element. Tehrani restricts it further still, requiring that the two clauses be genuinely different in meaning. And even among those who accept the possibility in principle, the specific Qur’anic examples usually cited in its defence remain disputed; Tehrani himself rejects several of the classic Kufan proof-texts.

This matters when the same grammatical claim is invoked elsewhere, for instance, in relation to this verse which has recently begun being cited by some polemicists:

أَفَإِن مَّاتَ أَوْ قُتِلَ انقَلَبْتُمْ عَلَىٰ أَعْقَابِكُمْ

“So if he dies or is slain, will you turn back on your heels?” (3:144)

Anyone proposing that aw here means bal first needs to establish why such a reading is even linguistically necessary. Does the default, disjunctive meaning of aw create any theological or grammatical difficulty? It does not. The verse sits comfortably alongside other, very similarly constructed verses:

وَلَئِن قُتِلْتُمْ فِى سَبِيلِ ٱللَّهِ أَوْ مُتُّمْ لَمَغْفِرَةٌ مِّنَ ٱللَّهِ وَرَحْمَةٌ خَيْرٌ مِّمَّا يَجْمَعُونَ

“And if you are killed in the cause of Allah or die — then forgiveness from Allah and mercy are better than whatever they accumulate.” (3:157)

وَلَئِن مُّتُّمْ أَوْ قُتِلْتُمْ لَإِلَى ٱللَّهِ تُحْشَرُونَ

“And whether you die or are killed, unto Allah you will be gathered.” (3:158)

وَٱلَّذِينَ هَاجَرُوا۟ فِى سَبِيلِ ٱللَّهِ ثُمَّ قُتِلُوٓا۟ أَوْ مَاتُوا۟ لَيَرْزُقَنَّهُمُ ٱللَّهُ رِزْقًا حَسَنًا وَإِنَّ ٱللَّهَ لَهُوَ خَيْرُ ٱلرَّزِقِينَ

“And those who emigrated for the cause of Allah and then were killed or died — Allah will surely provide for them a good provision. And indeed, it is Allah who is the best of providers.” (22:58)

It is also worth noting that most of these verses above which use a particular verb tense of mawt and qatl — except 22:58 — are conditional statements. Conditional sentences are not even among the sentence types under discussion in the classical debate over when aw might carry the sense of bal, since that debate is confined to informative (al-jumlah al-khabariyyah) and imperative sentences, negative or positive, and not conditional sentences (al-jumlah al-shartiyyah). Reading aw as bal in a conditional clause therefore has no real basis in the grammatical tradition at all.

A grammatical claim repeated confidently in introductory textbooks can turn out, on closer inspection, to rest on a genuine and long-running scholarly dispute, one where the majority position, and the most careful contemporary treatment of it, actually cuts against the popular assumption.

Footnotes

  1. Source: al-Mughni, vol. 1, pg. 64.