Pa'al Geminate Roots (סבב, תמם, חגג)

גִּזְרַת הַכְּפוּלִים is the class of roots whose second and third letters are the same — ס־ב־ב, ת־מ־ם, ק־ל־ל, ג־ל־ל, ח־ג־ג. Say at the outset what this page is for: in Modern Hebrew the special geminate paradigm is mainly something you read, not something you use to build new forms. The productive pa'al survivors have largely been rebuilt on the ordinary sound-root pattern. What the older paradigm still gives you is the ability to parse a headline, a legal text, or a literary line without stopping.

Careful which forms you call verbs

The obvious member list is a trap, because it silently mixes verbs with adjectives. Check each form against a dictionary before you conjugate it:

• סָבַב "revolve, turn" is a genuine pa'al past — and Pealim gives it two infinitives, the classical לָסוֹב lasóv and the rebuilt לִסְבּוֹב lisbóv.
• תַּם "come to an end" is a genuine pa'al past, with the future יִיתַּם yitám.
• קַל kal "light, easy" is not a verb at all. It is an adjective on the root ק־ל־ל, and no modern pa'al past stands behind it — the productive verbs are in other binyanim: pi'el קִילֵּל "curse", hif'il הֵקֵל "ease, make lighter".

That last point is worth its own line, because קל is the form everyone reaches for first and it belongs to a different part of speech entirely.

The recognition problem

On the surface a geminate root looks exactly like a hollow root: both show only two consonants in the citation form. תַּם and קָם have the same two-letter shape, even without niqqud. Only the rest of the paradigm separates the classes — and the future does it cleanly.

Notice that the present of a geminate verb is adjective-shaped, exactly like the hollow one: תַּם · תַּמָּה · תַּמִּים · תַּמּוֹת, not *תּוֹמֵם. So even the participle gives nothing away. The signal is the vav in יָקוּם against the doubled letter in יִיתַּם, and the ־וֹתִי of תַּמּוֹתִי against the plain ־תִּי of קַמְתִּי. The hollow class is set out on pa'al ע״ו.

Three stages of one collapse

Line four roots up and you can watch the class dissolve in real time. The evidence is in the dictionary itself: Pealim lists both paradigms under one entry for חגג and for סבב, and for גלל it notes that the root has no special conjugation properties at all.

Read down the last column and the direction is unmistakable: the special geminate paradigm has been giving way to an ordinary one. ח־ג־ג is the clearest single piece of evidence, because both sets sit in one dictionary entry while ordinary contemporary speech overwhelmingly uses the rebuilt forms. ס־ב־ב carries the variation further: Pealim's entry prints two infinitives (לָסוֹב ~ לִסְבּוֹב), two past paradigms (סַבּוֹתִי ~ סָבַבְתִּי) and three futures (יִיסּוֹב, יָסוֹב, יִסְבּוֹב). These pa'al forms are rare or literary; everyday meanings normally use pi'el סוֹבֵב or hitpa'el הִסְתּוֹבֵב instead.

Where the roots went

ס־ב־ב is the model case. The pa'al is essentially unused, and three other binyanim — four verbs between them — carry the meanings:

• לְסוֹבֵב lesovév — pi'el on the polel model: to turn or spin something. תסובב את המפתח.
• לְסַבֵּב lesabév — ordinary pi'el: to cause, to bring about. Not the same verb.
• לְהִסְתּוֹבֵב lehistovév — hitpa'el, also on the polel model: to turn around, to hang around, to wander.
• לְהָסֵב lehasév — hif'il: to turn something, to transfer, to redirect.

Note where the doubled letter went. In the two polel-model forms it did not go anywhere — סוֹבֵב and הִסְתּוֹבֵב carry both ב's, and the vav between them is the polel mould's cholam — and in the hif'il it became a long tsere, הֵסֵב. Only לְסַבֵּב looks like an ordinary sound root, because pi'el doubles its middle radical anyway and the geminate has nowhere to show. That vav is why הִסְתּוֹבֵב is not *הִסְתַּבֵּב. The full picture across the binyanim is on geminate roots across the binyanim.

What is genuinely alive: the vocabulary

This is the useful half of the class. The verbs may have gone quiet, but the nouns and adjectives built on these roots are everyday words, and knowing the root makes them a family rather than a list.

מְסִיבָּה "party" is the nicest of these, because the connection is invisible until you see the root: it comes from the hif'il הֵסֵב "recline at table" — a party is where people sit around. Same root as סביב, and the same root as the reason you came.

The frozen residue

A handful of older geminate forms also survive in fixed expressions. They are formulaic or higher-register: learn them as whole expressions, not as models for freely building new verbs.

• תַּם וְנִשְׁלַם — "finished and complete", the formula copyists wrote at the end of a manuscript. Still used for rhetorical closure.
• קַל וָחוֹמֶר — the Talmudic a fortiori argument: if the light case holds, all the more so the heavy one. Formulaic and higher-register, but also heard in spoken argument.
• סוֹבֵב סֹבֵב הוֹלֵךְ הָרוּחַ — Ecclesiastes 1:6, "round and round goes the wind": the geminate participle doubled for effect.

Common Mistakes

The pa'al of these roots is barely produced at all, so the errors learners actually make are not in pa'al — they are in the binyanim that replaced it. That is itself the lesson.

1. Dropping the vav from the hitpa'el. A geminate root in hitpa'el follows the polel model and keeps a vav in the middle.

2. Building the hitpa'el without metathesis. When a root begins with ס (or שׂ, שׁ, ז, צ), the ת of the hitpa'el prefix swaps places with it.

3. Producing the classical past. Both sets below are ✅ Hebrew — but only one of them is spoken. A learner who has read the paradigm chart and says khagóti will be understood and will sound like a nineteenth-century letter.

4. Assuming לחוג and לחגוג are two different verbs. They are two infinitives of one verb: the second is ordinary in speech, while the first is formal or literary.

Key Takeaways

• כפולים are roots whose 2nd and 3rd letters are identical: ס־ב־ב, ת־מ־ם, ק־ל־ל, ג־ל־ל, ח־ג־ג. In Modern Hebrew the pa'al of this class is a reading skill.
• Check the part of speech first: סָבַב and תַּם are pa'al verbs; קַל is an adjective, with pi'el קילל and hif'il הקל carrying its verbs.
• On the surface a geminate looks like a hollow root — תַּם and קָם are the same shape. The future separates them: יִיתַּם against יָקוּם, and so does the 1 sg. past, תַּמּוֹתִי against קַמְתִּי.
• The class shows three visible stages: ת־מ־ם keeps the special forms (formal); ח־ג־ג and ס־ב־ב show variation between special and rebuilt forms — לָחוֹג ~ לַחְגּוֹג for one, and לָסוֹב ~ לִסְבּוֹב, סַבּוֹתִי ~ סָבַבְתִּי, יָסוֹב ~ יִסְבּוֹב for the other; ג־ל־ל has been fully re-analysed as a sound root.
• The meanings moved to other binyanim, which keep a vav in the middle: סוֹבֵב, הִסְתּוֹבֵב, הֵסֵב.
• What is genuinely alive is the derived vocabulary: סָבִיב, סְבִיבָה, סְבִיבוֹן, סִיבָּה, מְסִיבָּה, תָּמִים, תּוֹם, קַל, קַלּוּת, קְלָלָה, חַג, חֲגִיגָה.
• תַּם וְנִשְׁלַם and קַל וָחוֹמֶר are formulaic, higher-register expressions — learn them whole and do not generalise their shapes into a productive paradigm.