Arabic Hollow Verbs (الأَجْوَف): قَالَ، كَانَ، بَاعَ Made Simple

7 min read

Open any Arabic text and you will meet hollow verbs within the first sentence. كَانَ (he was) and قَالَ (he said) are two of the most frequent verbs in the entire language, and both belong to the class Arab grammarians call الأَجْوَف (al-ajwaf), literally the hollow one. The name is apt: the middle root letter is a weak consonant, و or ي, and in many conjugated forms it seems to vanish, leaving the verb hollow in the middle.

This is why قَالَ (he said) becomes قُلْتُ (I said) — the long alif disappears and a short damma takes its place. Learners often experience this as chaos, but it is one of the most rule-governed corners of Arabic morphology. There is essentially one rule to learn, plus three vowel classes to sort verbs into. This guide walks through both, with fully vowelled examples you can check against a conjugation table.

What Makes a Verb Hollow

Arabic verbs are built on three-letter roots. In a hollow verb, the middle radical is و or ي — the two so-called weak letters, which double as long vowels. The root of قَالَ is ق-و-ل, the root of كَانَ is ك-و-ن, and the root of بَاعَ (he sold) is ب-ي-ع. In the basic past-tense form, that weak middle letter surfaces as a long alif: instead of the expected قَوَلَ, Arabic says قَالَ.

Knowing the hidden middle letter matters, because it determines how the verb behaves everywhere else. You can usually recover it from the present tense: يَقُولُ (he says) reveals the و of ق-و-ل, and يَبِيعُ (he sells) reveals the ي of ب-ي-ع. Dictionaries and root-based apps list these verbs under the full root, so a verb that looks two-lettered in the past tense is filed under three letters.

قَالَ - يَقُولُqala - yaquluhe said - he says (root ق-و-ل)
بَاعَ - يَبِيعُba'a - yabi'uhe sold - he sells (root ب-ي-ع)
نَامَ - يَنَامُnama - yanamuhe slept - he sleeps (root ن-و-م)

The One Rule That Explains Everything

Here is the core principle: Arabic does not allow a long vowel to stand in a closed syllable — a syllable that ends in a consonant. When a suffix beginning with a consonant attaches to the verb, the syllable containing the long vowel closes, and the long vowel must shorten.

Watch it happen in the past tense. In قَالَ the syllable qa-la stays open, so the long alif survives. Add the suffix تُ for I, and the stem syllable would become qal-tu, a closed syllable with a long vowel inside — forbidden. So the vowel shortens: قُلْتُ (qultu, I said). The same logic gives كُنْتُ (I was) from كَانَ, بِعْتُ (I sold) from بَاعَ, and نِمْتُ (I slept) from نَامَ.

Suffixes that begin with a vowel do not close the syllable, so the long vowel stays. That is why they said is قَالُوا (qalu) with the alif intact, and she said is قَالَتْ (qalat) — the ت of the feminine ending follows a vowel, leaving the stem syllable open.

  • Consonant-initial suffixes (تُ، تَ، تِ، نَا، تُمْ، تُنَّ، نَ) shorten the vowel: قُلْتُ، قُلْنَا، قُلْتُمْ.
  • Vowel-initial suffixes (َتْ، ُوا، َا، َتَا) keep it long: قَالَتْ، قَالُوا، قَالَا.
  • The same rule runs through the present tense: the feminine plural يَقُلْنَ shortens because the نَ suffix closes the syllable, while يَقُولُونَ stays long.

The Three Vowel Classes

So the long vowel shortens — but shortens to what? قَالَ becomes قُلْتُ with a damma, بَاعَ becomes بِعْتُ with a kasra, and نَامَ becomes نِمْتُ with a kasra too. The choice is not random: the short vowel in the past tense echoes the verb's identity, and the present-tense vowel is your most reliable clue. This gives us three classes.

Class one: و verbs with a damma. These have يَفْعُلُ presents with a long و, like يَقُولُ and يَكُونُ. When the past shortens, it takes a damma — the short counterpart of و. So قَالَ gives قُلْتُ and كَانَ gives كُنْتُ. Other members include زَارَ - يَزُورُ (to visit, زُرْتُ) and عَادَ - يَعُودُ (to return, عُدْتُ).

Class two: ي verbs with a kasra. These have يَفْعِلُ presents with a long ي, like يَبِيعُ. The shortened past takes a kasra, the short counterpart of ي: بَاعَ gives بِعْتُ. Similarly عَاشَ - يَعِيشُ (to live, عِشْتُ), صَارَ - يَصِيرُ (to become, صِرْتُ), and طَارَ - يَطِيرُ (to fly, طِرْتُ).

Class three: verbs with a long alif in the present, like يَنَامُ and يَخَافُ. These descend from an original فَعِلَ pattern (نَوِمَ, خَوِفَ), and that historical kasra resurfaces when the past shortens: نَامَ gives نِمْتُ and خَافَ (to fear, root خ-و-ف) gives خِفْتُ — a kasra even though the middle letter is و.

كَانَ - كُنْتُ - يَكُونُkana - kuntu - yakunuhe was - I was - he is (class 1, damma)
بَاعَ - بِعْتُ - يَبِيعُba'a - bi'tu - yabi'uhe sold - I sold - he sells (class 2, kasra)
خَافَ - خِفْتُ - يَخَافُkhafa - khiftu - yakhafuhe feared - I feared - he fears (class 3, kasra)

Shortening in the Present and Jussive

The closed-syllable rule also explains two present-tense surprises. First, the feminine plural: the suffix نَ begins with a consonant, so the stem vowel shortens — هُنَّ يَقُلْنَ (they say, f.), يَبِعْنَ (they sell, f.), يَنَمْنَ (they sleep, f.).

Second, the jussive mood, used after لَمْ for negated past meaning. The jussive removes the final vowel of the verb, closing the last syllable, so the long vowel inside must shorten: لَمْ يَقُلْ (he did not say), لَمْ يَكُنْ (he was not), لَمْ يَبِعْ (he did not sell), لَمْ يَنَمْ (he did not sleep). If you have wondered why the Quranic phrase لَمْ يَكُنْ has no و, this is the entire reason. The imperative follows the same pattern: قُلْ (say!), كُنْ (be!), بِعْ (sell!), نَمْ (sleep!).

How to Practice Hollow Verbs

Reading rules is the easy half; the hard half is producing قُلْنَا instead of قَالْنَا under time pressure. Drill one verb per class until its pattern feels physical — قَالَ for the damma class, بَاعَ for the kasra class, نَامَ for the alif class — and every other hollow verb becomes a substitution exercise.

Fahm's conjugation trainer includes hollow verbs such as قَالَ and نَامَ with all thirteen persons fully vowelled in past and present, so you can test yourself on exactly the forms where the vowel shortens. Pair that with reading practice: hollow verbs are so common that any graded story or Quranic passage will serve up كَانَ and قَالَ dozens of times.

Frequently asked questions

Why does قَالَ become قُلْتُ instead of قَالْتُ?

Arabic phonology forbids a long vowel in a closed syllable. The suffix تُ closes the stem syllable, so the long alif must shorten, and it shortens to damma because قَالَ belongs to the و class (present يَقُولُ).

How do I know whether the shortened vowel is damma or kasra?

Check the present tense. A long و in the present (يَقُولُ) means damma in the shortened past (قُلْتُ). A long ي (يَبِيعُ) means kasra (بِعْتُ). A long alif (يَنَامُ) also means kasra (نِمْتُ), because these verbs come from an original فَعِلَ pattern.

Is كَانَ conjugated like other hollow verbs?

Yes, completely. كَانَ follows the damma class exactly like قَالَ: past كُنْتُ، كُنَّا، كَانُوا, present يَكُونُ, jussive لَمْ يَكُنْ, imperative كُنْ. Learning one damma-class verb gives you كَانَ for free.

Do hollow verbs behave the same in the derived forms?

Not always. In Forms II and III the middle radical behaves like a strong consonant — قَوَّلَ and حَاوَلَ conjugate regularly. Weak behavior returns in several other forms, such as Form IV أَرَادَ - يُرِيدُ (to want) and Form X اِسْتَطَاعَ - يَسْتَطِيعُ (to be able).

Put it into practice

Fahm covers this with interactive lessons, spaced repetition, and quizzes — start free, no account needed.

Start Learning Free

Keep going