Why בירך but דיבר: Where the Dagesh Is Compensated

On the page, דיבר, בירך and מיהר are the same shape: three root letters with a yod after the first. In the mouth they are dibér, berékh and mihér — three different vowels in that first syllable, and nothing written tells you which. This page is about why.

Note before anything else that בירך and גירש are not a contrast. Both have ר as their middle radical and both behave identically. The two contrasts that matter are בירך against דיבר on one side, and בירך against מיהר on the other.

What pi'el is made of

The pi'el's structural signature is a dagesh chazak in the middle root letter — the dot in the בּ of דִּיבֵּר, מְדַבֵּר, יְדַבֵּר. It is what identifies the binyan in a pointed text before anything else is parsed, and it is why nothing can be inserted between the person prefix and the stem in the future.

But א, ה, ח, ע and ר never take a dagesh chazak. Try holding a long r or a long 'ayin and you will hear why. When the morphology demands a doubling in one of them, something else has to happen to the syllable in front.

Two outcomes, not one

The traditional names are תשלום דגש — "payment for the dagesh" — for the first case and דגש מדומה, virtual doubling, for the second. The second name describes rather than explains: the dot is imagined to be there, and the syllable behaves as though it were. That is the honest state of the rule, and it agrees with the dagesh page.

The split is clean in pi'el. Every ח verb checked against Pealim — שִׂיחֵק, נִיחֵם, כִּיחֵשׁ, רִיחֵם, נִיחֵשׁ, רִיחֵף, שִׁיחֵד, גִּיחֵךְ — keeps the short chirik, with no lengthened variant printed beside it. Only א and ר compensate.

The half of the rule that is usually left out

The same split runs through the participle and the infinitive, where it shows as kamatz against patach — and this is the more visible half, because those are the forms a learner meets first.

Israeli Hebrew pronounces patach and kamatz alike, so this half of the rule is silent. You meet it when you point a text, or when you have to decide whether a spelling takes a yod.

The dot you can still see is a different dot

בֵּירֵךְ, תֵּיאֵר and כִּיחֵשׁ each have a dot in their first letter, which invites the conclusion that some dagesh survived. It did not. That dot is a dagesh kal in a בגדכפ״ת letter at the start of a word — the same dot that sits in the ד of דִּיבֵּר, alongside the chazak in its בּ. The chazak is simply gone.

The proof is in the verbs whose first letter is not a בגדכפ״ת letter: סֵירֵב, מֵיאֵן, נִיהֵל, מִיהֵר, שִׂיחֵק carry no dot at all, and they belong to exactly the same two classes.

Where the yod in בירך comes from

Ktiv male writes an i vowel with a yod. That covers מיהר, ניהל, שיחק, כיחש — ordinary chirik, ordinary rule, the same one that gives דיבר and מילה.

It does not cover בירך, whose first vowel is an e. The yod there comes from a separate clause of the Academy's ktiv male rules: a tsere standing before א, ה, ח, ע or ר as compensation for a missing dagesh is written with a yod. The rule names תיאר among its examples, alongside חירש kherésh, קירח keréakh, היריון herayón, גירה gerá and ליהנות. מגירה megerá "drawer" is בירך's exact twin.

Two consequences, both load-bearing:

Only the first tsere earns a yod. בֵּירֵךְ has two of them. The first, under the ב, is the compensating one — it stands where a chirik would have stood before a doubled ר. The second, under the ר, is the ordinary pi'el theme vowel, the same tsere that sits in the בּ of דִּיבֵּר. So the spelling is בירך, never בירייך; and *תיאר, never *תיאיר.

The two classes end up looking identical for opposite reasons. שיחק and ניהל get their yod because they genuinely have an i. בירך and תיאר get theirs because they have an e that used to be an i. Same letter on the page, two different vowels in the mouth.

The same clause is behind the closed list of nouns that write a yod for an e — פירות perót, שיער se'ár, בריכה brekhá and עבירה averá among them, with the full list on the ktiv male rules. In every one of them the e is a demoted i, exactly as in בירך.

Two readings for the 3ms, and no ranking

Pealim prints two vocalizations for the third-person masculine past of most of the compensating verbs:

No source ranks them, and this page will not either. What makes the patach variant less arbitrary than it first looks is the fourth column: it is the vowel every suffixed cell of the paradigm already has. שֵׁירֵת and פֵּיאֵר have the same doublet; מֵיאֵן is the one verb of this class checked here for which Pealim prints no second reading at all.

A detail the classical rule does not cover

The two classes also differ in the cells where the vowel reduces — before ־ָה and ־וּ, and in the plural participle. ר takes an ordinary shva; א, which cannot, takes a chataf-patach; and the ה/ח/ע class takes a chataf-patach too, for the same reason.

This one is audible: medabrím has three syllables, mesakhakím has four.

Common Mistakes

1. Reading the yod as a chirik. In this class it looks exactly like the yod of מיהר and is not the same vowel.

2. Writing a yod for the second tsere as well. If a tsere earns a yod, why not both of them?

3. Carrying the yod into the infinitive and participle. The compensating tsere exists only in the past; everywhere else the class has a kamatz.

4. Dropping the yod from the ה/ח/ע class. Reference grammars point מִהֵר, and the consonants get copied straight off the page.

Key Takeaways

• Pi'el's signature is a dagesh chazak in the middle radical. א, ה, ח, ע and ר cannot carry it.
• The five letters split in two. א and ר compensate — chirik becomes tsere (בֵּירֵךְ, תֵּיאֵר), patach becomes kamatz (מְבָרֵךְ, לְתָאֵר). ה, ח and ע do not — the vowel stays short (מִיהֵר, שִׂיחֵק; מְמַהֵר, לְנַהֵל).
• בירך and גירש are the same case, both ר. The contrasts are בירך vs דיבר and בירך vs מיהר.
• The dot in בֵּירֵךְ and תֵּיאֵר is a dagesh kal in a word-initial בגדכפ״ת letter. סֵירֵב, מֵיאֵן, נִיהֵל and מִיהֵר have no dot at all.
• The yod of בירך spells a tsere written לתשלום דגש; the yod of מיהר spells an ordinary chirik. Identical on the page, different vowels.
• Only the first tsere takes a yod — בירך, not בירייך; תיאר, not תיאיר — and the compensating tsere lives in the past only, so the infinitive is לברך.
• Pealim prints two readings for the 3ms of בירך, גירש, פירש, סירב, תיאר and ביאר. No source ranks them; the patach variant matches every suffixed cell. מיאן has only one.
• In the reduced cells, ר takes a shva (בֵּירְכָה) while א and the ה/ח/ע class take a chataf-patach (תֵּיאֲרָה, מִיהֲרָה) — and that difference is audible: medabrím against mesakhakím.