ע״ו and ע״י (Hollow Roots) Across the Binyanim

A hollow root carries a ו or a י in its middle slot. In its characteristic short stem that radical is not realised as a consonant: its history is visible in the vowel between the other two radicals. ק־ו־ם gives kam, שׁ־י־ר gives shar. Most of this page follows from one structural fact — the stem has two consonants where a binyan expects three — and from the different ways the seven moulds cope with that.

Start with the fact that decides how much of this you have to learn. ע״ו and ע״י are different classes on different roots — ק־ו־ם and שׂ־י־ם are as distinct as any two roots in the language — and pa'al keeps them apart most transparently: לָקוּם with a u, לָשִׂים with an i. In the productive short-stem patterns, the letter on the page belongs to the mould rather than identifying the root. Polel writes a cholam ו for both classes (קוֹמֵם, שׁוֹרֵר from שׁ־י־ר), huf'al a shuruk ו (הוּקַם, הוּשַׂם from שׂ־י־ם), and hif'il neither radical at all, giving both classes a long i (הֵקִים, הֵבִין). The ordinary pi'el route described below is the important exception: כִּיוֵּון and קִיֵּים do preserve the root's own middle letter.

The grid, on ק־ו־ם

בנייןpast (he)present m.sg.infinitiveMeaning
pa'alקָם kamקָם kamלָקוּם lakúmgot up, stood up
nif'algap — there is no לְהִיקּוֹם; see below
pi'el (polel)קוֹמֵם komémמְקוֹמֵם mekomémלְקוֹמֵם lekomémoutraged, angered
pu'al (polal)קוֹמַם komámמְקוֹמָם mekomám— none existswas outraged
hif'ilהֵקִים hekímמֵקִים mekímלְהָקִים lehakímset up, founded
huf'alהוּקַם hukámמוּקָם mukám— none existswas set up
hitpa'el (hitpolel)הִתְקוֹמֵם hitkomémמִתְקוֹמֵם mitkomémלְהִתְקוֹמֵם lehitkomémrebelled, rose up

Read the ו down that column and it is a different creature in every block. One letter, four fates.

קמנו בארבע בשביל טיסה שבסוף התעכבה בשלוש שעות.

kámnu be'árba bishvíl tisá shebasóf hit'akvá beshalósh sha'ót

We were up at four for a flight that ended up three hours late.

הקימו שם מרכז מסחרי ענק והשכונה השתנתה לגמרי.

hekímu sham merkáz miskharí anák vehashkhuná hishtantá legamréi

They put up a huge shopping centre there and the neighbourhood changed completely.

המחסום הוקם ביום שישי ופורק אתמול בבוקר.

hamakhsóm hukám beyóm shishí uforák etmól babóker

The roadblock went up on Friday and was dismantled yesterday morning. (journalistic)

הכיתה כולה התקוממה נגד מועד הבחינה.

hakitá kulá hitkomemá néged mo'éd habkhiná

The whole class rose up against the exam date.

The pi'el slot, and the thing the textbooks get wrong

A hollow root cannot form an ordinary pi'el the usual way, because pi'el doubles the middle radical and the short stem has no middle consonant to double. One Hebrew solution is the פּוֹלֵל pattern: repeat the third radical and let the pattern supply a cholam between the two stem consonants. That is why קוֹמֵם has two מs and הִתְקוֹמֵם keeps its vav.

RootpolelhitpolelDerived word you already know
ק־ו־םקוֹמֵם komémהִתְקוֹמֵם hitkomémהִתְקוֹמְמוּת hitkomemút uprising
שׁ־י־רשׁוֹרֵר shorér(literary)מְשׁוֹרֵרmeshorér poet
ס־ב־ב (geminate)סוֹבֵב sovévהִסְתּוֹבֵב histovévסְבִיבָה svivá environment

Now the qualifier, and it is a real one: the polel is not the only route. Some hollow roots build a perfectly ordinary pi'el instead, by treating the ו as a full consonant with a vowel of its own. כ־ו־ן is the clearest case: pi'el כִּיוֵּון kivén "aimed, directed", hitpa'el הִתְכַּוֵּון hitkavén "meant, intended" — neither of them a polel, and both among the commonest verbs you will use. So the rule to carry away is "the pi'el column of a hollow root will not look like מְדַבֵּר", not "it is always a polel".

This route is also the one place outside pa'al where the root's own middle letter still shows, for a plain reason: a radical with a vowel of its own is written as itself. כִּיוֵּון shows the ו of כ־ו־ן; a ע״י root sent the same way shows its י — Pealim roots לקיים lekayém "to uphold; to hold (an event)" as ק־י־ם.

הוא משורר מוכר, אבל בחיים לא הרוויח מזה שקל.

hu meshorér mukár, avál bakhayím lo hirvíakh mizé shékel

He's a well-known poet, but he's never made a shekel from it.

לא התכוונתי לזה, יצא לי לא נכון.

lo hitkavánti lazé, yatsá li lo nakhón

That's not what I meant — it came out wrong.

nif'al: thin, and not on ק־ו־ם

Pealim lists no nif'al for ק־ו־ם — there is no לְהִיקּוֹם, and inventing one is the commonest way learners try to make a passive here. Hollow nif'als do exist on other roots, but the few useful items are lexical and do not share one prefix vowel:

  • נָכוֹן nakhón (כ־ו־ן) — the rare verbal paradigm means "was prepared, was established"; in everyday Hebrew you meet the separately listed adjective "correct, true".
  • נִידּוֹן nidón (ד־ו־ן) — "was discussed, was tried". This one is genuinely current in legal and journalistic writing.
  • נָבוֹן navón (ב־י־ן) — an adjective, "wise, discerning", rather than a productive verb. It still makes the spelling point in one word: the middle radical is a י and the form is written with a ו.

That ו belongs to the mould, so a root with no use for it does not get one: Pealim heads the geminate ס־ב־ב as לְהִיסּוֹב, but its finite forms drop the vav — past נָסַב, present נָסָב, both nasáv.

Where the passive of a hollow verb is wanted in ordinary Hebrew, it is usually the huf'al that supplies it: הוּקַם, הוּכַן, הוּבָא.

התיק נידון בבית המשפט המחוזי כבר שלוש שנים.

hatík nidón bevéit hamishpát hamekhozí kvar shalósh shaním

The case has been before the district court for three years now. (formal / legal)

The hif'il merge — and the third class inside it

Hif'il is the most conspicuous of the merges, because it is the one binyan that writes neither letter. Every hollow root, whichever it hides, lands on הֵ־ִי־ with a long i that belongs to no class:

VerbRootMiddle letter
הֵקִים hekím foundedק־ו־םו
הֵכִין hekhín preparedכ־ו־ןו
הֵבִיא heví broughtב־ו־או
הֵבִין hevín understoodב־י־ןי
הֵסִיר hesír removedס־י־רי

Nothing in the hif'il paradigm identifies the class, which is why הֵבִין is ב־י־ן and not ב־ו־ן however often you see the second in a word list. And there is a third group in the same room: geminate roots take the identical prefix vowels, because they too have a two-consonant stem. ס־ב־ב gives הֵסֵב, with future יָסֵב, and ק־ל־ל gives הֵקֵל, with future יָקֵל. Those futures share the kamatz of יָקִים and יָבִין rather than the patach of יַסְבִּיר. So "a kamatz prefix means a hollow root" is one step too confident: it means a short stem, which may be hollow or geminate.

הכינו לנו חדר בקומה השנייה, ליד המעלית.

hekhínu lánu khéder bakomá hashniyá, leyád hama'alít

They put a room aside for us on the second floor, next to the lift.

מי הביא את העוגה? היא מדהימה.

mi heví et ha'ugá? hi madhimá

Who brought the cake? It's incredible.

Stress does the work the letters cannot

In pa'al the masculine singular past and present are one form — הוא קם is both "he got up" and "he gets up" — and in the feminine singular the two tenses are spelled alike and separated by stress alone: past is מלעיל, present מלרע. The plurals do not collide, because the present has endings of its own: past קָמוּ against present קָמִים and קָמוֹת.

היא גרה בלונדון שנתיים ואז חזרה לארץ.

hi gára belóndon shnatáyim ve'áz khazrá la'árets

She lived in London for two years and then came back to Israel.

היא גרה בלונדון, אבל מגיעה לארץ כל קיץ.

hi gará belóndon, avál magi'á la'árets kol káyits

She lives in London, but she comes to Israel every summer.

💡
Same two words on the page, two different tenses in the air — gára against gará, and likewise bá'a / ba'á and káma / kamá. Niqqud will not help: it marks vowels and not stress, so גָּרָה is pointed identically in both sentences above and the romanization line is the only place the contrast is written down. Say them aloud and put the beat in yourself. This is drilled on when stress is the only difference. Hif'il behaves differently and it is worth knowing why: there the stem's long i refuses to hand the beat to a ־י or ־וּ ending, so it is hevínu, tavíni, hekímu — stress on the stem, not the ending.

Common Mistakes

1. Keeping the yod in the huf'al. The long i belongs to the active hif'il stem, and huf'al has an a there.

❌ הבניין הוקים בשנות החמישים

Incorrect — there is no such form. The huf'al of הקים is הוקם: הבניין הוקם בשנות החמישים, hukám. Likewise הוכן from הכין and הובא from הביא.

✅ הבניין הוקם בשנות החמישים ומאז שיפצו אותו פעמיים.

habinyán hukám bishnót hakhamishím ume'áz shiptsú otó pa'amáyim

The building went up in the fifties and has been renovated twice since.

2. Giving the pa'al an object. קם is intransitive in every register; putting something up is a different cell of the grid.

❌ הם קמו אנדרטה בכיכר העיר

Incorrect — קם takes no object at all: you get up, you do not get something up. Erecting something is the hif'il: הם הקימו אנדרטה בכיכר העיר.

✅ הם הקימו אנדרטה בכיכר העיר בלי לשאול אף אחד.

hem hekímu andárta bekhikár ha'ír bli lish'ól af ekhád

They put up a monument in the town square without asking anybody.

3. Dropping the polel's vav. With only two visible radicals it is tempting to build the hitpa'el straight onto them.

❌ התושבים התקמו נגד השלטון

Incorrect — a hollow root takes the hitpolel shape, with a cholam and a repeated third radical: התושבים התקוממו נגד השלטון, hitkomemú. Compare הסתובב from ס־ב־ב.

✅ התושבים התקוממו נגד השלטון עוד לפני שהחוק עבר.

hatoshavím hitkomemú néged hashiltón od lifnéi shehakhók avár

The residents rose up against the regime before the law had even passed.

4. Confusing the pa'al future with the hif'il one. אָקוּם and אָקִים are one letter apart on the page and mean unrelated things.

❌ אני אקים מוקדם מחר בבוקר

Incorrect — אָקִים is the hif'il ‘I will set up’ and needs an object. ‘I'll get up early’ is the pa'al אָקוּם: אני אקום מוקדם מחר בבוקר.

✅ אני אקום מוקדם מחר בבוקר וניסע לפני הפקקים.

aní akúm mukdám makhár babóker venisá lifnéi hapkakím

I'll get up early tomorrow and we'll set off before the traffic.

Key Takeaways

  • In a hollow root's characteristic short stem, the middle ו or י is not realised as a consonant, so the stem has only two consonants — and each binyan compensates differently.
  • ק־ו־ם fills six of the seven: קָם · קוֹמֵם · קוֹמַם · הֵקִים · הוּקַם · הִתְקוֹמֵם. The gap is nif'al — there is no לְהִיקּוֹם.
  • The pi'el and hitpa'el slots are usually taken by the polel / hitpolel shape, which repeats the third radical around a cholam: קוֹמֵם, הִתְקוֹמֵם, שׁוֹרֵר (whence מְשׁוֹרֵר "poet"). But not always — כ־ו־ן builds an ordinary pi'el, כִּיוֵּון, and hitpa'el הִתְכַּוֵּון. That route is also the only place outside pa'al where the root's own letter still shows.
  • Hollow nif'als are thin and lexical, not one uniform נָ־וֹ־ paradigm: compare verbal נָכוֹן and current legal נִידּוֹן; נָבוֹן is now an adjective. The everyday passive of a hollow verb is usually the huf'al — הוּקַם, הוּכַן, הוּבָא.
  • Pa'al keeps ע״ו and ע״י apart most transparently — לָקוּם against לָשִׂים. Polel and huf'al print the mould's ו, and hif'il prints neither radical (הֵקִים, הֵבִין). Ordinary pi'el is the exception: כִּיוֵּון and קִיֵּים preserve the root letter. When the form neutralises the contrast, look the root up.
  • Geminate roots share hif'il's hollow prefix vowels: הֵסֵב, future יָסֵב; הֵקֵל, future יָקֵל. A kamatz prefix means a short stem, not specifically a hollow one.
  • In pa'al the m.sg. past and present are identical (הוא בא), and in the feminine singular stress carries the tense: gára she lived / gará she lives. The plurals stay apart (קָמוּ / קָמִים).
  • Two consonants with a long vowel between them ⇒ hollow, and ten of the commonest verbs in the language are built that way: קם, בא, שם, גר, רץ, שר, טס, מת, שב, דן. Running the other way, הֵבִין, הֵכִין, הֵבִיא, הֵקִים are hif'il, not pi'el — מֵבִין looks like מְדַבֵּר, but the infinitive settles it: לְהָבִין has ל plus ה where pi'el has a bare ל.

Now practice Hebrew

Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.

Start learning Hebrew

Related Topics

  • Pa'al ע״ו Hollow Verbs: קם, גר, בא, טסB1Roots whose middle letter is ו lose it entirely — you hear two consonants, never three. The past is stressed on the stem and the future is built on a long u, which together produce the most useful stress contrast a Hebrew learner will meet: bá'a 'she came' against ba'á 'she comes'.
  • Pa'al ע״י Hollow Verbs: שם, שר, רבB1ע״י verbs are ע״ו verbs with one substitution: a long i where the other class has a long u. Past and present are identical in shape, which means nothing on the page tells you whether a hollow verb gives אָקוּם or אָשִׂים — that has to be learned with the word.
  • Hif'il Hollow Roots (ע״ו / ע״י): הבין, הכין, הביא, הקיםB1ע״ו and ע״י are two different classes everywhere except here: in hif'il both collapse onto one template, הֵ־ִי־, so הֵבִין (ב־י־ן), הֵכִין (כ־ו־ן) and הֵקִים (ק־ו־ם) are indistinguishable. The forms are uniform; the spelling of the past is where learners come unstuck.
  • Pa'al Geminate Roots (סבב, תמם, חגג)C1Roots whose second and third letters are identical. In Modern Hebrew the special geminate paradigm is mainly a reading skill: many meanings moved to pi'el and hitpa'el, productive survivors were rebuilt on the sound-root pattern, and the roots remain alive throughout the vocabulary.
  • שלמים: The Sound Root in Every BinyanB1A root with no weak letter behaves identically in all seven binyanim — which makes it the control case. Here is one grid, two roots, and the honest gaps in both, so that every weak class on the pages that follow can be read as a documented deviation from this.