בניין הפעיל: Causative and Much More

After pa'al and pi'el, hif'il is the binyan you will meet most. It supplies התחיל, הסביר, הרגיש, החליט, הגיע, הבין, הזמין, הכניס — verbs you cannot get through a day without. It is also the binyan with a moving part: a ה prefix that is visible in three forms and invisible in the rest, which is why learners recognise התחיל and then fail to recognise מתחיל as the same word.

The five forms of one verb

Here is ת־ח־ל "begin", the verb this whole group is built on:

The ה survives exactly where nothing else is competing for the front of the word. The present tense puts מ there and the future puts a person letter there (א, ת, י, נ), and in both the ה is squeezed out. The past, the infinitive and the imperative have no such prefix, so the ה stays. That single observation explains the whole alternation, and it is worked through in detail on where did the ה go?.

The second signature: a long i in the stem

Look down the readings above: hitkhíl, matkhíl, yatkhíl, lehatkhíl. Whatever happens at the front of the word, the stem holds a long i written with a yod, and it is stressed. That vowel is the most reliable hif'il marker there is, because unlike the ה it never disappears:

Now the qualifier the textbooks drop, and it matters because two of the commonest verbs in the language fall under it. Hollow roots keep the long i but change the prefix vowel. ב־י־ן, כ־ו־ן, ב־ו־א and ק־ו־ם give a tsere where the strong roots have a patach, and a kamatz in the future and the infinitive:

Where the causative story works

The textbook line is that hif'il makes someone else do the thing, and on a well-chosen set of roots that is exactly right:

Notice what the causative does to the grammar: the intransitive verb gains an object. You go in; you put something in.

And where it does not

Now the part to install before the habit forms, because a learner who trusts "hif'il = causative" will mistranslate daily. These are all hif'il, and not one of them is causative:

הגיע is the clearest case: a causative is by definition transitive, and הגיע cannot take an object in any register. The honest description is that hif'il often adds an argument and often does not, that the reading is a property of the individual verb rather than of the pattern, and that the way to use the semantics is backwards — from a form you have met to a guess at its meaning — never forwards. That argument is made in full on does the binyan predict the meaning? and the counter-examples are collected on hif'il verbs that are not causative.

The מת trap

מתחיל opens with the two letters מת, and every English speaker's eye reads that as the hitpa'el prefix — the reflexive pattern of מתלבש "gets dressed". It is not. The ת of מתחיל is the first letter of the root ת־ח־ל, and the same is true of מתאים "fits, suits" (ת־א־ם) and מתמיד "persists" (ת־מ־ד).

The vowels separate them instantly. Hif'il runs a–i, hitpa'el runs i–a–e:

mitlabésh against matkhíl — on the page they share two letters, in the mouth they share almost nothing. The full set of look-alikes is on binyan confusion.

Is hif'il still alive?

You will read that hif'il is closed. It is not. It still takes inchoatives — הִשְׁחִיר "turned black", הִבְרִיא "got better", הִרְזָה "lost weight" — and loans whose consonant cluster survives in its frame, הִשְׁפְּרִיץ "sprayed", הִסְנִיף "sniffed". What Bolozky (2007) documents is a register-conditioned retreat: hif'il remains preferred in literary Hebrew while colloquial Hebrew increasingly reaches for pi'el for the same job. That is a much narrower claim than "dead", and the details are on borrowed roots.

Its passive is huf'al, automatically and with no choice in the matter: הִזְמִין → הוּזְמַן huzmán "was invited", הִסְבִּיר → הוּסְבַּר husbár "was explained". Huf'al has a present, a past and a future and nothing else — no infinitive, no imperative.

Common Mistakes

1. Giving an intransitive verb an object instead of reaching for its hif'il partner. This is the error the causative pairs exist to prevent.

2. Counting the ה as a root letter. A Hebrew root is three letters, and the hif'il ה is never one of them — which matters the moment you look a verb up, or try to connect it to its noun.

3. Giving הסביר a direct object for the person. English explains someone something; Hebrew explains to someone.

4. Expecting every hif'il to be transitive. Several of the most frequent ones are not, and English hands you the wrong frame: I managed it has an object where Hebrew has none.

Key Takeaways

• Hif'il's ה prefix appears in the past, the infinitive and the imperative (הִתְחִיל, לְהַתְחִיל, הַתְחֵל) and is displaced by מ in the present and by the person letter in the future (מַתְחִיל, יַתְחִיל).
• Its second and more reliable marker is a stressed long i in the stem, written with a yod: hisbír, masbír, yargísh, lehazmín.
• Hollow roots keep the i but change the prefix vowel: הֵבִין, מֵבִין, יָבִין, לְהָבִין — tsere and kamatz where the strong roots have patach. Hif'il is the only binyan where ע״ו and ע״י merge completely.
• The causative reading is real but partial: נכנס→הכניס, יצא→הוציא, אכל→האכיל, לבש→הלביש behave as advertised; הסביר, הרגיש, החליט, הבטיח, הצליח, הכיר, הבין and הגיע do not, and הגיע takes no object at all.
• מת at the front is not hitpa'el. הִתְחִיל, הִתְאִים and הִתְמִיד are hif'il verbs whose first radical is ת; the a–i vowels give them away (matkhíl, not mitlabésh).
• Hif'il is not closed: inchoatives (השחיר, הבריא, הרזה) and cluster-preserving loans (השפריץ, הסניף) still enter it, and its retreat is register-conditioned (Bolozky 2007).
• Its passive is huf'al — הוזמן, הוסבר — which has a present, past and future and no infinitive or imperative.