Niqqud and Harakat

Summary

Notes

In Sephardic pronunciation, tsere and segol are both pronounced as [e:] and [e] respectively (same sound but different lengths) and kamatz katan is pronounced as [o], which is a shorter version of the holam [o:]. Katan means "small" and the vowel is "smaller" than the long vowels by being shorter. Show (1)

1. When transcribing with intent to preserve the Tiberian pronunciation (rather than the Sephardic) the long versions of these are transcribed as ẹ and ọ respectively, since e and o are already used for the tsere and holam.
2. A colon written on this row indicates that the tsere and the holam are always long, although the other vowels may also sometimes be long (and their proper IPA representations would include a colon also in this case).

3. Two vertical dots are used to indicate two things in the Tiberian system.

   1. Silence, the absence of a vowel between two consonants, called a silent sh'vā.

      English also has cases where two consonants run together, for example s and t as st; t and r as tr, etc. but we don't have a special "no vowel" character, there just isn't a vowel letter written after the first consonant.

   2. An extra short vowel. This also comes in two forms.
      1. Where the two dots appear on their own, without any explicit indication of which particular vowel sound is to be spoken for an extra short amount of time, called a vocal sh'vā (which looks visually identical to the silent sh'vā).
      2. Where the two dots appear alongside another vowel symbol, indicating an extra short vowel where the specific vowel sound is marked, known as a hataf vowel.

   I will defer the discussion until concerning the rules of how to identify a silent sh'vā and how to identify a vocal sh'vā.

   The Palestinian and Babylonian notations didn't have any special symbols to distinguish the hataf vowels. In the surviving vocal traditions they've each been promoted to the corresponding medium length vowel. Show (2)

   The pronunciation of vocal sh'vā in Tiberian Hebrew follows the rules in . These rules are seldom followed anymore Show (3), though the Yemenite Jews and Sephardim from Amsterdam are known for them.

   Rule No.	Case	Next Letter	Vowel Sound
   1	Gutturals	א ,ה ,ח ,ע	Extra short version of the vowel marked on the guttural. ă, ĭ, ŭ, ĕ, ẹ̆, ŏ, ọ̆
   2	Yod	י	ĭ
   3	Default	Anything else	ă	(Sephardim when a meteg ◌ְֽ is present, Tiberian & Yemeni)
   			ĕ	(Other cases)

4. These are described as "short" in some contexts and grouped together with the "long" vowels in others.

5. In the Tiberian system the patterns on this row can represent both medium and long length vowels. With the exception of tsere and holam (which were always long) these patterns historically represented long vowels in open syllables (CVV) and medium length vowels in closed syllables (CVC), the lengthened vowel "filling in" for the "missing" consonant. The distinction was largely allophonic, meaning there were rarely any words which differed only in respect of one of these symbols being pronounced as medium length in one word and long in the other. Outside of the Tiberian tradition the vowel length for the patterns on this row is fixed as being either always medium or always long, with one exception. Going one step further, the realization of "medium" and "long" vowels as two different lengths of sound is now considered completely irrelevant in Modern Israeli Hebrew. Nonetheless, the Tiberian tradition is thought to have once implemented this feature systematically depending on whether or not there followed a ה ,ו ,י or א, in a similar way to how Arabic uses وْ ,يْ and اْ. The possible syllable structures and terminology (such as CVV versus CVC) will be explored fully in .

   Instead of the open versus closed syllable distinction, medieval Sephardic pronunciation (from which Modern Hebrew is ultimately descended) largely differentiated between medium and long length vowels based on the vowel symbol itself rather than the presence or absence of a mater lectionis letter afterwards. Tsere and holam are always long in this tradition (as in Tiberian). Segol is medium length irrespective of whether it's followed by a י. Patach and kubutz are also always medium length because the open/closed syllable distinction now only applies to kamatz. Kamatz is medium or long, depending on whether the syllable is open or closed (as in Tiberian, but for Sephardic Jews it also changes from an o sound to an a sound. It's also pronounced as long-a in stressed syllables, which are usually the last syllables of each word although other syllables can be noted as stressed by adding a meteg or an accent mark (ole), ◌֫, but it's never an a sound if the next letter has a hataf kamatz). Hiriq is either medium or long, but dependent only on whether there is י after it and the open vs. closed syllable distinction is irrelevant. Shuruk is long (as always). The difference in length between segol and tsere exists despite the fact that early examples of Palestinian notation used only one symbol in the places where Tiberian had either segol or tsere, suggesting that perhaps the difference in length is a later development. Later examples of Palestinian notation did use two separate symbols.

6. א can also occur instead of ה.
7. These are usually described as "short", since Arabic doesn't have anything shorter.
8. آ represents the glottal stop–a vowel combination that could theoretically be represented by the combination أَاْ but isn't in practice.
9. Tanwin is pronounced as a medium length vowel sound followed by an n sound.
10. When patach appears underneath the last letter of a word and the last letter is ח ,הּ or ע (i.e. the gutterals except for א) then the vowel is pronounced before the consonant. This phenomenon is called Stolen Patach.

11. In transcriptions the long vowels can be differentiated from medium length vowels in various different ways.

    1. Long vowels may be represented using a line above: a, e, i, o, u.
    2. Long vowels can also be represented using an acute accent: á, é, í, ó, ú.
    3. Two-letter combinations can be used to represent long vowels (and sometimes there is a choice between two different two-letter combinations, see ). However, all two-letter combinations apart from ah, ei / ey and éi / éy are for informal educational use and internet chat only.
    4. Long vowels are sometimes left undifferentiated from medium and extra short vowels: a, e, i, o, u. However this loses information about the both correct pronunciation and the spelling of the original Arabic or pre-modern Hebrew.
    5. Sometimes tsere and segol are differentiated using é and e but the other vowel pairs are left undifferentiated.

    יְ◌ֵ (long) and יְ◌ֶ (medium) are often transcribed by explicitly recording the yod (י) as either i or y (thus ei or ey, or if being precise then éi or éy for tsere). Although é and éi / éy are pronounced identically (and similarly for e and ei / ey), the presence or absence of a yod can change the meaning of a word.

    Transcription
    1 Letter	2 Letters
    a	aa ah (word ending in ה)
    i	ee
    u	oo
    e	eh éy (when with י)

"Diphthongs"

What is a diphthong? A simple definition is that it's when two vowel sounds occur next to each other. That's a good start, but it's insufficient. Consider the word tenuous. There's two vowel sounds next to each other there, but the published analytical descriptions of English don't list this pair as a diphthong. Whereas the sound in beer is considered a diphthong. The u and the ou can clearly be heard as two separate vowels whereas the eer is more of a smooth "swinging" motion. So, English has a lot of diphthongs but not every pair of English vowels is a diphthong. There are two further complications to consider.

• The sounds represented by the English letters w and y sometimes sound like vowels and sometimes sound like consonants. Consider the pairs bow and wet, and, bay and yet. It turns out this isn't a freakish English spelling thing but a general phenomenon that's common to many languages. These sounds are called semivowels.
• According to some definitions of the word diphthong, a pair of sounds is only referred to as a diphthong in cases where two vowels (or a vowel and a semivowel) can take the place where normally (according to the rules for constructing syllables in the given language) a single vowel would be expected. However, this situation never occurs in Hebrew or Arabic. We only have cases where a vowel and semivowel (i.e. a special kind of consonant) are written where a vowel and a consonant are expected. This is why many Arab scholars state that Arabic doesn't have any diphthongs, because according to this narrow definition it doesn't! Nonetheless, the sounds involved are distinctive and they do sound similar to genuine diphthongs that we're familiar with in English.

There are three levels of formality when it comes to transcribing the diphthongs into English letters.

1. The most formal method (primarily used in education and academia) is to transcribe the first vowel "as normal" (including the macron above it if it's a long vowel), and to transcribe the Yod / Ya as y, also as normal.
2. A more informal approach is transcribe the vowel without attempting to distinguish between long and medium length vowels, and to transcribe the Yod /Ya as y, as normal.
3. The difference between the first two methods also can be applied to transcribing Arabic and Hebrew vowels more generally. Additionally, for Hebrew diphthongs in particular, a third scheme is to transcribe the Yod as i when it's part of a diphthong.

1. Formed by a tendency to soften the y sound when it occurs between an a and an i. May sound more like a-yi in other speakers or accents.
2. The ي can also be with shaddah (يّ) instead of with sukoon (يْ), in which case the transcription would be ayy. Similarly, in each case in this table the Yod (י) could have a doubling dagesh inside it (and any vowel mark underneath) instead of a silent sh'vā (see the ayy case for an example fragment).

Diphthongs created using the w semivowel are rarer in Hebrew than they are in Arabic. An aw diphthong in many Arabic words tends to morph into an o vowel in the equivalent Hebrew word, although orthographically (meaning in written form) this difference is small because the holam is often written above a waw, the holam-waw, וֹ. Deriving the correct pronunciations for w diphthongs (where they do exist) requires two considerations.

1. In the most common contemporary Hebrew accents (both in liturgy and common speech), ו overwhelmingly represents a v sound rather than w sound. Since v is not a semivowel then combinations like av, iv and ov are in no way diphthongs according to any sensible definition of the term, and these combinations end up being listed in this section solely because of accidents of history.
2. The pronunciation of kamatz gadol varies between different Hebrew accents. It's either an a-type vowel or an o-type vowel.

3. This is an example of a rare phenomenon called Qere and Ketiv, a difference between what is written in the Bible and what has always traditionally been read aloud in synagogue (literally Aramaic for "read" and "written"). In this way the text has a dual meaning. ו◌ִ is an illegal combination in Hebrew so we can deduce that two words are present. Take, for example, the form הִוא. One word is expressed by the written consonants הוא (the 3rd person masculine singular pronoun), while the vowel marks imply that the word to be read aloud is הִיא, which is the 3rd person feminine singular pronoun. Another example is the consonants יהוה, which are annotated using vowels taken from the word אֲדֹנָי, because Jews are forbidden from reciting God's name aloud so the vowel marks remind them to pronounce Adonai instead. There is no such word in Hebrew as Yehovah or Jehovah. The form יְהוָה is an example of qere and ketiv being used to encode the two words יהוה and אֲדֹנָי simultaneously. The J in Jehovah arises because the letter j in the German alphabet represents the same sound as an English y and Germany was an important place for early Protestant Biblical scholarship.
4. Note this special case where the Yod is silent.

Syllable Structures

Simple Rules

• There are two kinds of syllables.
  • Open syllables with a Consonant–Vowel pattern.
  • Closed syllables with a Consonant–Vowel–Consonant pattern.
• Thus, when reading, one has to distinguish three things.
  1. The first consonant.
  2. The vowel
     (which might be indicated by a symbol beside the first consonant only. This is usually but not always underneath in Hebrew, above or below in Arabic. Or the vowel might be indicated using a combination of a symbol and a mater lectionis letter following onward towards the left hand side of the page).
  3. Whether or not the next letter forms part of the current syllable or belongs as the first consonant of the next syllable.
• If the vowel is a sh'vā or a hataf vowel then the syllable is an open one and the next letter doesn't need to be inspected yet.
• If the second letter is marked with one these two annotations then the second letter forms part of the current syllable:
  • A sukun (◌ْ), or, equivalently a silent sh'vā.
  • A shaddah (◌ّ), or, equivalently a dagesh (which will be a doubling dagesh).

  In most of these cases in Arabic the syllable is a closed syllable. In the cases of يْ preceded by ◌ِ and وْ preceded by ◌ُ the syllable is an open syllable with a long vowel sound but the second letter is nonetheless still the last letter of the current syllable. See the discussion about super heavy syllables in for a few tricky cases.

  In Hebrew the syllable is definitely a closed syllable when there's a dagesh and is usually a closed syllable when there's a sh'vā, although sometimes the sh'vā can be vocal sh'vā and belong to the next syllable instead.

• If the second letter has a vowel sign of its own and doesn't have a shaddah or a dagesh then the second letter plays no part in the current syllable.
• In Hebrew, if a letter doesn't have its own vowel marking and it doesn't have a silent sh'vā or a mappiq then it's a mater lectionis letter. That is, it will be one of ו , ה, א or י and it will be occurring in it's capacity as a vowel rather than these letters' other capacities as consonants. Only ה and א can have a mappiq, which looks like a dagesh (these two letters cannot have a dagesh).

Full Rules

• The possible ways of forming syllables in a language can be described using patterns of Cs and Vs, where C stands for any consonant and V stands for any vowel.
• English syllables are summarized by the structure (CCC)V(CCCC), where one or more of the letters in brackets do not always occur in every syllable and the vowel in this case can be short, long or a diphthong. Thus, the word strengths (/strɛŋkθs/ in IPA) fits this pattern with all of the optional consonant slots filled, and therefore (because all of the sounds have been matched) the whole word is one syllable. Fortunately, Hebrew and Arabic have simpler syllable structures.
• I'll use the convention (not used unanimously by all authors) that VV stands for a long vowel.
• A shaddah or a doubling dagesh signals two identical consonants (CC) (which sometimes belong to separate syllables).
• Syllables can be classified into 3 categories based on how many Vs and Cs they have. Each syllable is either light (1–2), heavy (3) or super heavy (4 or more).
• Some fudges are commonly used to keep things simple.
  • A diphthong counts as a vowel and consonant (VC or VVC) for the purpose of identifying syllables.
  • Stolen Patach isn't counted at all.
  • Analytically, it is simplest to consider vocal sh'vā (ĕ) as a vowel that forms part of an open syllable (CV). In terms of pronunciation however, a sh'vā is a brief burst of sound, shorter than a normal vowel, and it makes sense to think of the sh'vā and the preceeding consonant as attached to the syllable that follows it so as to not accidentally lengthen the sh'vā. This difficulty causes some authors to use a different classification system that includes additional terms such as "doubly closed syllable" or "virtually closed syllable".
• An English speaker won't be able to identify the glottal stop sound produced at the beginning of words beginning with alif (ا / א). It is actually a consonant sound. However, all English words which are written with an initial vowel are actually pronounced by English speakers as a glottal stop followed by the vowel, we just don't notice it. An English speaker may be tempted to describe a word like app as having a VC syllable structure even though the structure is really CVC.
• According to the Tiberian pronunciation rules all CV syllables have either a vocal sh'vā or a hataf vowel.
• All hataf vowels belong to open syllables (CV).
• Super heavy syllables only occur in Arabic in two specific circumstances.
  1. In uses of a word outside of the context of a sentence (such as a dictionary-style definition), where it is written without a short vowel or a tanwin present on the last letter. Final vowels and tanwin perform grammatical functions, so when seen within the context of a word in a sentence of Classical Arabic writing these words would have one of these endings attached (-a, -i, -u, -an, -in or -un). However, when speaking and pausing for breath after the word, coming to the end of sentence, or finishing recitation these word endings are dropped (called pausal pronunciation), and this will cause a super heavy syllable to form in spoken word.
  2. When the long vowel a occurs before a letter with a shaddah (◌ّاْ◌َ).
     1. If the letter with shaddah is the last letter of the word then a CVVCC syllable can occur (if there's no final short vowel).
     2. If it's not the last letter then a CVVC syllable occurs and the third consonant (implied by the shaddah) forms the beginning of the next syllable.
• Words are most reliably broken into syllables by starting from the end of each word and working backwards, matching letters and vowel signs to the patterns in as you go. However, starting from the beginning of a word and reading forwards usually works too.

Sh'vā: Silent or Vocal?

The same symbol is used to indicate both the silent sh'vā and the vocal sh'vā. How do we know when a sh'vā is vocal (a short vowel sound) or silent (no sound)? The purpose of a silent sh'vā is to indicate closed syllables. Thus, where we're expecting the possibility of a closed syllable the sh'vā is silent and if we're not expecting to read the end of a closed syllable then the sh'vā is vocal.