Hif'il Past: התחלתי, הסברת, הרגיש

The hif'il past uses the same personal endings as every other binyan — ־תִּי, ־תָּ, ־תְּ, ־ָה, ־נוּ, ־תֶּם, ־תֶּן, ־וּ — so there is nothing new to memorise at the back of the word. Everything hif'il-specific happens in front of it: the ה prefix is visible here, and the stem vowel alternates.

The nine cells

Nine cells, eight distinct written strings — התחלת is both hitkhálta and hitkhált, exactly as כתבת is in pa'al — and, again as in pa'al, no separate feminine plural: התחילו covers הם and הן.

Long i or patach — the rule is visible

Read the Pointed column downwards and the alternation jumps out:

• Third person keeps the long i: הִתְחִיל, הִתְחִילָה, הִתְחִילוּ.
• Every person whose ending begins with a consonant shortens it to a patach: הִתְחַלְתִּי, הִתְחַלְתָּ, הִתְחַלְנוּ, הִתְחַלְתֶּם.

Pi'el does the same thing for the same reason — דִּיבֵּר shortens to דִּיבַּרְתִּי — because a consonant-initial ending adds a syllable behind the stem and the long vowel gives way to a patach. The practical consequence is a spelling one — the yod disappears with the vowel. התחיל has one; התחלתי has none. Same for הסביר → הסברתי and הרגיש → הרגשתי.

The retraction in that last example is not a hif'il quirk. Pealim marks the ־תֶּם and ־תֶּן cells as "usually unstressed in spoken language" on לשמור and לקרוא too; the whole language pulls this ending back, and the normative hisbartém survives in careful reading aloud. See stress inside a verb paradigm.

One of only three places the ה is visible

הִתְחַלְתִּי has its ה; מַתְחִיל and יַתְחִיל do not. The past shares that privilege with just two other forms — the infinitive לְהַתְחִיל and the imperative הַתְחֵל — because those are the three forms with nothing else competing for the front of the word. Everything else in the paradigm puts a מ or a person letter there and the ה is squeezed out; the mechanics are on where did the ה go?.

התחיל is not a hitpa'el, and unpointed text will not tell you

This is the one genuine reading problem on the page. הִתְחִיל and הִתְלַבֵּשׁ both begin הת. In hitpa'el that ת belongs to the prefix; in התחיל it is the first letter of the root ת־ח־ל, and the same holds for הִתְאִים "fitted, suited" (ת־א־ם) and הִתְמִיד "kept at it" (ת־מ־ד).

The quickest practical test is to check the present tense: matkhíl with an a is hif'il, mitlabésh with an i is hitpa'el. The historical curiosity that makes ת־ח־ל stick is that it is a young root. Klein files it as a secondary base back-formed from the noun תחילה tkhilá "beginning" — and תחילה is itself built on ח־ל־ל, from the hif'il הֵחֵל "began", with the noun-forming prefix תַּ־. So the ת really was a prefix once: it belonged to the noun, and got read back into the verb system as a root letter. התחיל is a hif'il built on a noun built on a hif'il. That story is on finding a root in the dictionary.

And one hazard in the other direction: a word beginning הש־ or הס־ is often not hif'il. הִשְׁתַּמֵּשׁ hishtamésh "used" and הִסְתַּכֵּל histakél "looked" are hitpa'el verbs whose ת has swapped places with a sibilant first radical — see hitpa'el metathesis.

Spelling: the prefix takes no yod

The prefix vowel of the past is a chirik — hit-, his-, hir- — and it is never written with a yod, because a silent shva follows it. That is a general clause of the ktiv male rules rather than anything special about this binyan: the same reason gives משפחה, not *מישפחה. So התחלתי, הסברתי, הרגשתי, הזמנתי, all bare. The Academy's rules list the hif'il prefix explicitly.

"I have started" is one word

English has three past shapes where Hebrew has one. I started, I have started and I had started are all התחלתי, with no auxiliary anywhere in the sentence and nothing to translate have with.

Those two are worth reading together, because the second is the form learners reach for when they want have started. הייתי מתחיל is real Hebrew — it just means "I used to start", a repeated habit in the past. For a single completed event, plain התחלתי is the whole answer.

Common Mistakes

1. Keeping the long i in front of a consonant ending. Learners build the past from the dictionary form הסביר and hang the endings on it unchanged.

2. Writing a yod in the prefix. Hi- is an i, and ktiv male usually writes those with a yod — but not before a silent shva.

3. Following התחיל with a participle instead of an infinitive. English started working uses an -ing form; Hebrew uses ל plus the infinitive, always.

4. Pushing the beat onto ־ָה and ־וּ. The pa'al rule katvá, katvú is real, and it does not reach hif'il.

Key Takeaways

• The endings are the universal past-tense suffixes; only the stem is hif'il. Nine cells, eight written strings, and no separate feminine plural — התחילו serves הם and הן.
• The stem vowel alternates on a clean rule: third person keeps the long i (הִתְחִיל, הִתְחִילָה, הִתְחִילוּ), consonant-initial endings shorten it to a patach (הִתְחַלְתִּי, הִתְחַלְנוּ) — and the yod disappears with the vowel.
• Hif'il keeps the stress on that i even before ־ָה and ־וּ: hitkhíla, hisbíru are מלעיל where pa'al's katvá, katvú are מלרע.
• The 2nd-person plural retraction (hizmántem for normative hizmantém) is general to the language, not a hif'il fact.
• The past is one of only three forms where the ה is visible — with the infinitive and the imperative.
• הת at the front is not proof of hitpa'el: הִתְחִיל, הִתְאִים, הִתְמִיד are hif'il with ת as first radical (i–i, against hitpa'el's i–a–e), and הש־ / הס־ words like השתמש and הסתכל are hitpa'el with metathesis.
• No yod in the prefix: התחלתי, הסברתי, הרגשתי.
• התחלתי covers "I started", "I have started" and "I had started". הייתי מתחיל exists but means "I used to start".