Gutturals (א ה ח ע ר) and What They Do to a Pattern

Textbooks present א, ה, ח, ע and ר as a list of exceptions. They are not. They are five letters that physically cannot do certain things an ordinary consonant does, and every "irregular" form built on a root containing one of them is simply the pattern rearranging itself around the refusal.

There are exactly three refusals, and knowing which letter refuses what is worth more than any paradigm table:

Notice immediately that the five letters do not form one club. ר joins for the dagesh and nothing else — it takes a plain shva quite happily (רְחוֹב rekhóv, רְשִׁימָה reshimá, רְפוּאָה refu'á). And the third refusal is positional: whether a guttural colours a vowel depends on where in the root it stands, not on which letter it is. Get that grid right and most of the guttural chapters in a textbook collapse into three facts.

Refusal 1: no doubling dot

A dagesh chazak marks a consonant the grammar wants doubled. Throat consonants cannot be held long, and ר has too much continuous movement, so all five refuse the dot. When a pattern demands doubling and the letter will not take it, one of two things happens — and which one depends on the letter. This split is flattened by almost every learner resource, and it is the single most useful thing on this page.

Before א and ר, the vowel lengthens in compensation. The chirik of the pi'el past stretches into a tsere. Before ה, ח and ע, the vowel simply stays and the doubling is "virtual" — traditional grammar imagines the dot is there without writing it.

So it is בֵּרֵךְ, never בִּרֵּךְ, and מִהֵר, never מֵהֵר. The same split runs through the present tense, where the pi'el's patach lengthens to a kamatz before א and ר only: מְבָרֵךְ mevarékh and מְתָאֵר meta'ér against מְמַהֵר memahér and מְנַחֵם menakhém.

The verb paradigms that fall out of this are set out on dagesh and gutturals in pi'el.

The same split outside the verb — and a spelling rule it explains

This is not a verb quirk. It happens in any pattern that wants a doubled consonant. The adjective mishkal קִטֵּל, which supplies Hebrew's words for physical disabilities, doubles its middle letter — until it cannot:

Refusal 2: no mobile shva

Where any other letter would carry a mobile shva and be read as a short e, א ה ח ע take a chataf instead: a shortened real vowel written as shva-plus-vowel-sign.

אֲנִי aní · חֲלוֹם khalóm · עֲבוֹדָה avodá · חֲצִי khatsí · אֳנִיָּה oniyá

A chataf is what a guttural does instead of a shva — not an extra sign, and not a different vowel quality. Say the vowel quickly and unstressed and you have it. The signs themselves, the prefix rule they trigger (לַעֲבוֹד against לִכְתּוֹב) and the important detail that ר is not in this group belong to chataf vowels, which owns that story in full.

Refusal 3: gutturals pull vowels towards a — but it depends where they stand

The third consequence has nothing to do with dots. A guttural likes the vowel a beside it, and a pattern will swap its own vowel out to supply one. What almost every letter-by-letter list gets wrong is that this refusal is positional: the same letter colours a vowel in one slot of the root and not in another. Two places where the effect is unmissable:

In the pa'al future. Most roots take a cholam; a guttural in the second or third position replaces it with a patach. In the middle slot all four of א ה ח ע do it. In the final slot only ח and ע do — a final א gives a kamatz instead. (A handful of ordinary roots take a patach for an unrelated historical reason — יִלְמַד, יִלְבַּשׁ, יִשְׁכַּב — so a patach does not prove there is a guttural, but a guttural in those slots does force one.)

Be precise about the limits here, because they are easy to over-generalise. A final א gives a kamatz, not a patach — יִמְצָא yimtsá, יִקְרָא yikrá — which is a separate story, even though the same א in the middle of a root gives the patach of יִשְׁאַל and יִדְאַג. ר does it in no position at all: יִשְׁמוֹר and יִסְגּוֹר end in ר, יִדְרֹשׁ has one in the middle, and all three keep their cholam. And the colouring belongs to the future, not the infinitive: לִשְׁלוֹחַ is still lishlóakh with its cholam intact, and לִבְחוֹר is livkhór. A guttural as the first root letter produces an a too — יַעֲבוֹד ya'avód, יַחְשֹׁב yakhshóv, לַהֲרֹג laharóg, where a regular root has יִ־ and לִ־ — but that is not this refusal at work. There the a is the prefix's own original vowel, which the guttural stopped Hebrew from reducing; ה is as good at blocking a reduction as ח and ע are, which is why it appears in that list and not in the table above. The story is on pa'al with a guttural first radical.

In the furtive patach. When a word ends in ח, ע or a ה with mappiq, and the vowel before it is anything other than a, a patach slides in between — written under the final letter, but pronounced before it.

רוּחַ rúakh · יוֹדֵעַ yodéa · מַפְתֵּחַ maftéakh · גָּבוֹהַּ gavóa · לִשְׁמוֹעַ lishmóa

The grid, in one table

Why this matters if you cannot hear א and ע

English speakers routinely skip this whole topic, and the reasoning is understandable: Israeli Hebrew has lost the pharyngeal ח and ע for most speakers, ח is now the same sound as כ, and א and ע are silent. If you cannot hear the consonant, why learn what it does?

Because the consonant is inaudible and its effects are not. Every vowel change on this page is fully audible in ordinary Israeli speech. יוֹדֵעַ really does end yodéa and not yodé; מַפְתֵּחַ really is maftéakh with two syllables and a final akh; יִשְׁמַע really is yishmá and never yishmór. The silent letter has left a loud footprint, and a learner who ignores the footprint mispronounces a large slice of the everyday vocabulary.

Common Mistakes

1. Keeping the chirik before א or ר. These two compensate; the vowel has to lengthen.

2. Lengthening before ה, ח or ע. The mirror-image error, and it appears as soon as the first rule is learned.

3. Giving a ח/ע root the regular future vowel. The cholam of יִשְׁמוֹר is not available here.

4. Swallowing the furtive patach. The extra vowel is not decoration; it is a syllable Israelis pronounce.

Key Takeaways

• The gutturals are three refusals, not one list. Knowing which letter refuses what replaces most guttural paradigms.
• No dagesh chazak in א ה ח ע ר. Before א and ר the vowel lengthens in compensation (בֵּרֵךְ, תֵּאֵר, חֵרֵשׁ); before ה, ח and ע it stays and the doubling is virtual (מִהֵר, נִחֵם, תִּעֵד).
• That compensation is also a spelling rule: a tsere standing in for a missing dagesh is written with a yod — בירך, תיאר, חירש, פירות.
• No mobile shva in א ה ח ע — they take a chataf instead. ר is not in this group and takes a plain shva.
• Vowel colouring is positional, not per-letter. In the middle of the root all four of א ה ח ע force a patach in the pa'al future — יִשְׁאַל, יִנְהַג, יִדְאַג, יִבְחַר, יִכְעַס. In final position only ח and ע do (יִשְׁמַע, יִפְתַּח, and the furtive patach in רוּחַ, יוֹדֵעַ, מַפְתֵּחַ); a final א gives a kamatz (יִמְצָא), and ר does nothing anywhere.
• The consonants may be inaudible in Israeli speech, but their vowel effects are not — which is exactly why the topic cannot be skipped.