נפעל vs התפעל: Where They Overlap

The clean statement of the difference is about the semantic role of the subject. In nif'al, something happens to the subject: it is the affected party, and often there is nobody doing it at all. In hitpa'el, the subject does something to or for itself: it is both the doer and the one affected.

That is a real distinction and it is worth carrying around. What follows is the distinction applied to the roots where both binyanim genuinely exist — and then an honest account of how far it will actually take you. The step-by-step decision procedure, for when you have a sentence in front of you, is on choosing between nif'al and hitpa'el.

Five roots that fill both slots

Take the glosses from a dictionary rather than from the shape of the word. These are Milog's.

Two of those entries need their fine print printed. הִתְכַּנֵּס has two senses, and glossing it "convene" alone loses half the verb: Milog gives נאסף; התקבץ and also הסתגר עם עצמו; התבודד. The second is the one you meet about a person, and it is the purer hitpa'el of the two — turning oneself inward. And נִשְׁמַר is not only a passive: Milog's second sense is נשאר באותו מצב, which is exactly השתמר's territory. The two overlap; השתמר is simply the specialised word for it.

Why הִשְׁתַּמֵּר is not הִתְשַׁמֵּר

Before the semantics, one piece of shape that stops learners recognising the hitpa'el half of these pairs at all. When the first root letter is a sibilant — ס, שׂ, שׁ, צ, ז — the ת of the prefix swaps places with it: הִתְשַׁמֵּר is impossible, and the verb is הִשְׁתַּמֵּר. Two of the letters change the ת as well as move it: צ turns it into ט (הִצְטַלֵּם "had a photo taken"), and ז turns it into ד (הִזְדַּקֵּן "grew old").

So a hitpa'el partner may be sitting in the dictionary under a shape you would not have looked for. הִשְׁתַּמֵּר, הִשְׁתַּמֵּעַ, הִסְתָּרֵק, הִצְטַלֵּם, הִזְדַּקֵּן are all ordinary hitpa'el verbs wearing a rearranged prefix.

The honest limit of the diagnostic

Now the part most references leave out. For the overwhelming majority of roots only one of the two binyanim exists at all. There is no hitpa'el of ש־ב־ר beside נשבר, no nif'al of ל־ב־ט beside התלבט. The choice you are being taught to make is not a choice the language usually offers.

And when a root does fill both slots, the second member is often not the one you would have guessed. ל־ב־שׁ has both — התלבש "got dressed" and the nif'al נִלְבַּשׁ "was worn, was put on" — but נלבש is a plain passive of the garment, not a reflexive of the person, so it does nothing to help you choose. Check the dictionary; do not reason from the gap.

And where both do exist, the meanings are lexicalised, not derived. Nothing in "the subject does it to itself" predicts that הִתְמַצֵּא means know one's way around rather than find oneself, or that הִתְרָאָה means see one another rather than appear. You could not have generated those from נמצא and נראה, and neither could a native speaker.

So state what the diagnostic is for. It sorts pairs you already half-know; it does not manufacture verbs. Meeting הִשְׁתַּמֵּעַ in a sentence, the rule tells you it is about something the subject does or shows of itself rather than something done to it, and that is enough to read the sentence. Reasoning in the other direction — I need "be preserved", so I will build a hitpa'el — is how learners produce strings Israelis do not recognise. This is the same warning that runs through what each binyan means, and it applies here with full force.

The productivity asymmetry — and how to use it as a prior

If the diagnostic will not generate verbs, something else will help you guess. The two binyanim are not equally alive.

Bolozky (1978, Afroasiatic Linguistics 5, 111–136) finds no recent inchoatives realised in nif'al at all, and ties that closure to the near-zero productivity of pa'al, the binyan nif'al pairs with. His later measurements make hitpa'el the default pattern for new inchoative, reflexive and reciprocal verbs — הִתְעַצְבֵּן, הִתְמַסְטֵל, הִתְפַיֵּיד, and the whole slang class.

That gives you a genuine, usable prior: an unfamiliar verb that looks reflexive, reciprocal or "become-ish" is far more likely to be hitpa'el than nif'al, and the newer it feels, the stronger the bet. It is a probability, not a rule — נִבְהַל and נִרְדַּם are old nif'al inchoatives and are not going anywhere — but it is the right way to bet on a word you have not met.

Sort these eight

Cover the two right-hand columns and work down the list. The point of the exercise is the last row.

The last row is where the diagnostic breaks, and it breaks in a systematic direction: the non-agentive "become" verbs are exactly the class hitpa'el has taken over. התרגש, התעייף, התאהב, התייאש are all things that happen to you, and all of them are hitpa'el. Which is why the productivity fact above is worth more to a learner than the semantic rule.

Common Mistakes

1. Picking the wrong member of a real pair. Both sentences below are correct Hebrew and they do not mean anything like the same thing.

2. Building the hitpa'el without the metathesis. A sibilant first root letter changes places with the ת, and no amount of regularity will get you the right form.

3. Reaching for a hitpa'el whenever English has a reflexive pronoun. English found myself, hurt myself, enjoyed myself all carry a pronoun; Hebrew uses a bare verb, a nif'al, or את עצמי far more often than it uses a hitpa'el.

4. Confusing נראה with נתראה. One letter apart on the page, a whole syllable apart in the mouth, and a very common slip when arranging to meet.

Key Takeaways

• The diagnostic: in nif'al something happens to the subject; in hitpa'el the subject does something to or for itself.
• Real pairs are few — נכנס/התכנס, נמצא/התמצא, נשמר/השתמר, נשמע/השתמע, נראה/התראה — and their meanings are lexicalised: התמצא is know one's way around, which no rule would have produced.
• התכנס has two senses (gather, and withdraw into oneself) and נשמר overlaps with השתמר in the sense "stayed as it was". Do not tidy either of these up.
• A sibilant first radical metathesises with the ת: הִשְׁתַּמֵּר, הִסְתָּרֵק. With צ the ת becomes ט (הִצְטַלֵּם) and with ז it becomes ד (הִזְדַּקֵּן).
• The rule sorts, it does not generate. For most roots only one of the two binyanim exists at all — and where both do (ל־ב־שׁ: התלבש beside the passive נלבש), the second is often no help. Check the dictionary rather than reasoning from the gap.
• Productivity is the usable prior: Bolozky finds no recent inchoatives in nif'al and makes hitpa'el the default for new inchoative, reflexive and reciprocal verbs. An unfamiliar reflexive-looking verb is far more likely to be hitpa'el.
• The diagnostic fails cleanly on התרגש, התעייף, התאהב, התייאש — things that happen to you, all hitpa'el. That failure is the productivity fact showing through.
• English's reflexive pronoun is not a signal to build a hitpa'el: Hebrew uses מצאתי את עצמי (pa'al plus את עצמי) and נפגעתי (nif'al).