User:Sartma/Hebrew transliteration

Old and new transliteration systems compared edit

Transliteration/Transcription Table edit

Note that:

  • Transliteration = sign-to-sign transcription. This is in line with what is currently done by traditional Biblical Hebrew scholars, since it's important to represent in Roman letters the exact signs that are used to write a word. It allows to correctly transcribe the different spellings a word can have in the Bible, like: כָּמוֹךָ (kāmôḵā) vs כָּמֹכָה (kāmōḵâ) (Exodus 15:11).
  • Transcription = transcription of how a word is pronounced. Biblical Hebrew being a dead language, all considerations about its pronunciation are obviously based on reconstructions. This makes it necessary to choose one of the existing pronunciation proposals. It does not show how a word is written. Both כָּמוֹךָ and כָּמֹכָה would be spelt the same (kāmōḵā).
Comparison between transliteration systems
Transliterations (sign-based) Transcription (pronunciation-based)
Letter traditional
transcription
superscript Sartma 1 Sartma 2 Traditional
(5 vowels)
Modern
(7 vowels)
Simple
vowels
סָ ā/o ɔ ā/o ɔ
סַ a a a a
סֵ ē e ē e
סֶ e ɛ e ɛ
סִ i i i i
סֹ ō o ō o
סֻ u u u u
Matres
lectionis
סִי î ī ī i
סֵי ê¹ ē ē e
סֶי ê¹ (ệ, é⁴) ɛʸ ɛ̄ e ɛ
סוּ û ū ū u
סוֹ ô ō ō o
סָה â ɔʰ ɔ̂ ɔ ā ɔ
סַה ah³ â a a a
סֵה ê¹ ê e ē e
סֶה eh³ (ê²) ɛʰ ɛ̂ ɛ e ɛ
סֹה ōh³ ô o ō o
Shwas סְ ə/e ə ə ə̆
סֲ ă/ă ă ă ă
סֱ ĕ/ĕ ɛ̆ ĕ ɛ̆
סֳ ŏ/ŏ ɔ̆ ŏ ɔ̆
Syllable-final
(quiescent)
aleph
־א ʾ ˀ ʔ ʾ ʔ
Consonants aleph ʾ ʔ ʔ ʾ ʔ
ʿayin ʿ ʕ ʕ ʿ ʕ
other
consonants
b/ḇ, g/ḡ, d/ḏ, h, w, z, ḥ, ṭ, y, k/ḵ, l, m, n, s, p/p̄, ṣ, q, r, š/ś, t/ṯ
Notes
¹ Ambiguous transcription. It doesn't allow to distinguish סֵי, סֶי, and סֵה.
² Ambiguous transcription. It doesn't allow to distinguish סֵי, סֶי, סֵה and סֶה.
³ Ambiguous transcription. It doesn't allow to distinguish סַה, סֶה, סֹה from סַהּ, סֶהּ, סֹהּ.
⁴ Individual notation in some textbooks to disambiguate סֶי from סֵי, סֵה, סֶה.

Examples edit

See below for a comparison of words transliterated/transcribed in the different systems.

Traditional transcription edit

Description edit

  • Five-vowel system. It distinguishes qamats qatan (o) and qamats gadol (ā).
  • It uses macrons on putative long vowels and circumflex on vowels spelt with matres lectionis.
  • Transliterates quiescent aleph, as traditionally done for Biblical Hebrew.
  • Depending on the publication, it can always indicate the accent, show it only on the penultimate syllable (notably in segolates), or not show it at all.

Pros edit

  1. It's widely used on textbooks to learn Biblical Hebrew, so it's likely to be familiar to people new to the language.
  2. The distinction between qamats qatan (o) and qamats gadol (ā) agrees with some modern traditions used to read Biblical Hebrew (and with Israeli Hebrew pronunciation).

Cons edit

  1. Some transliterations are ambiguous, since this system doesn't allow to distinguish certain letter combinations (see notes in the table above).
  2. It's confusing in mixing diacritics with different values: some are supposed to indicate vowel length (macron, breve), others try to describe spelling (circumflex).
  3. It gives the false impression that vowel length was phonemic in Tiberian Hebrew showing vowels as long by nature (with a macron) when it is now widely accepted that vowel distinction was purely qualitative, while length was determined by position (vowels in open syllables or stressed syllables were long).
  4. By distinguishing the pronunciation of qamats qatan and qamats gadol, it arbitrarily imposes to the transliteration one particular reading tradition (Sephardi) that is not otherwise represented in the script (which represents the Tiberian tradition).

Public comments edit


Superscript transcription edit

Description edit

  • Seven-vowel system (no distinction between qamats qatan and qamats gadol).
  • It writes aleph and ayin with IPA symbols (א (ʾā́lep̄) = ʔ, ע (ʿáyin) = ʕ). This allows aleph to be written with a superscript when quiescent, e.g. לֹא (loˀ).
  • It represents matres lectionis (י, ן, ה) as superscripts (see table).
  • It represents א (ʾā́lep̄) as a superscript when quiescent, thus marking a distinction in cases like קָרָא (qɔrɔˀ, he called) and קָרָאת (qɔrɔˀṯ, you called) VS קָֽרְאָה (qɔrəʔɔʰ, she called) and קָֽרְאוּ (qɔrəʔuʷ, they called).

Pros edit

  1. Transliteration is never ambiguous.

Cons edit

  1. Superscript letters are visually noisy (take up horizontal space).

Author's Comment edit

The superscript romanization attempts to consistently show it when a letter is not pronounced as a consonant, mainly the letters אהוי. These letters are sometimes pronounced as the consonants /ʔhwy/ and sometimes are not pronounced as a consonant, serving as a marker for a vowel. In cases where they are not pronounced as /ʔhwy/ they are referred to here as non-consonantal. This includes cases where they are traditionally called matres lectionis and cases where they are called quiescent. (There are rare cases of other letters being non-consonantal, as with ש in יִשָּׂשכָר‎.) Writing the status of a letter makes it considerably easier for newbies to understand the intention of the Hebrew spelling.

Hebrew spelling consists of consonants with diacritics added later by the Tiberian Masoretes to indicate vowel qualities, consonant doubling, and whether letters were consonantal. These rules involve the presence of a vowel before or after the letter. Most other romanizations do not consistently show non-consonantality; the tendency is that matres lectionis are indicated with diacritics and quiescent letters with consonant letters. Typical romanizations work well for more conventional spellings where the vowels [ɛ e i] are written with י and [o u] with ו and word-final [ɔ] with ה. They are less helpful for unconventional spellings like חַטֹּאות‎ (superscript romanization ḥaṭṭoˀʷṯ), whose typical romanization ḥaṭṭōʾwṯ doesn't write או ʾw differently to indicate that it is quiescent and not pronounced /ʔw/; you can only infer that if you know the rules for quiescent letters.

But the superscript romanization always writes non-consonantal letters as superscript letters, which indicates both that they are present in the Hebrew spelling and that they are not pronounced as consonants. With this romanization, readers only have to learn the meaning of the superscript letters, which is relatively simple; they do not have to learn the rules for when a letter is a mater lectionis or quiescent, which are somewhat complex.

The disadvantage of the superscripts is that, like the original non-consonantal Hebrew letters, they are visually noisy (take up horizontal space).

Public comments edit


Sartma 1 edit

Description edit

  • Seven-vowel system (no distinction between qamats qatan and qamats gadol).
  • Transliterates vowels spelt with matres lectionis י (yôḏ) and ו (wāw) with a macron.
  • Transliterates word-final ה (hēʾ) with a circumflex.
  • Transliterates quiescent aleph (as traditionally done for Biblical Hebrew).
  • Indicates the accent only when non-final.

Pros edit

  1. Transliteration is never ambiguous. Compared to Sartma 2, it can clearly show the difference in the spelling of the last syllable between אַתָּה (ʔattɔ̂, you (m)) and הָלַכְתָּ (hɔláḵtɔ, you (m) walked).

Cons edit

  1. Extremely rare, irregular spellings of some proper nouns and inflected words (but never headwords) could give "odd" transliterations that would need to be manually normalised:

Author's Comment edit

We have chosen the seven-vowel system not only because it is the one nowadays most used in the academia when transliterating Biblical Hebrew words, but also because it uses simple vowels without macrons.
Traditionally, the circumflex accent has been used to indicate any vowel spelt with a mater lectionis, but this unavoidably produced some ambiguous transliterations (see notes in the table above). To avoid those ambiguities, we repurposed the macron sign to be used in a similar fashion as the circumflex (i.e. to show vowels spelt with matres lectionis), but limited to class i vowels (i, ɛ, e: spelt with mater lectionis י (yôḏ)) and class u vowels (u, o: spelt with mater lectionis ו (wāw)), i.e. those matres lectionis always used within their vowel class and in any position.
The use of the circumflex was then restricted to mater lectionis ה (hēʾ), which, unlike the previous two, can be used with vowels of any class (class a, i and u) but only in word-final position.
(Incidentally, such a use of the circumflex doubles as etymological in words where a final ה (hēʾ) indicates the disappearance of an original final consonant, like the final -t of feminine nouns (סוּסָה (sūsɔ̂), אִשָּׁה (ʔiššɔ̂), etc.) or the final -y or -w of lameḏ he' verbs (גָלָה (gɔlɔ̂), בָּנָה (bɔnɔ̂), etc.), in a similar way to how the circumflex is used in French, for example, to show the etymological disappearance of a consonant, like in hôtel or in île.)
This method solves all the ambiguities of the traditional scholar transliteration system.

Public comments edit


Sartma 2 edit

Description edit

  • Same as Sartma 1, but:
    transliterates final vowels followed by ה (hēʾ) as simple vowels (with no diacritic sign), on the basis that final simple vowels in BH are generally followed by a ה (hēʾ) (with a handful of exceptions, like the second-person masculine singular personal pronoun suffix סוּסְךָ (sûsəḵā, your horse), the second-person masculine singular past הָלַכְתָּ (hāláḵ, you (m) walked), etc.).

Pros edit

  1. It has one less diacritic (the circumflex) than Sartma 1.
  2. It allows to easily indicate the accent when it falls on the last syllable (i.e., without any conflict with the circumflex)

Cons edit

  1. Transliteration is ambiguous when it comes to the exceptions mentioned in the description. E.g., it doesn't show the difference in the spelling of the last syllable in cases like אַתָּה (ʔattɔ, you (m)) vs הָלַכְתָּ (hɔláḵtɔ, you (m) walked).

Author's Comment edit

Since a final simple vowel in BH is generally followed by ה (hēʾ) (a handful of exceptions excluded), we could omit the circumflex proposed on Sartma 1 without too much definition loss. This might also make it easier to show word-final accent with an acute accent sign (without it conflicting with the circumflex), in case we wanted to do so: compare Sartma 1 מַמְלָכָה (mamlɔḵɔ̂) VS Sartma 2 מַמְלָכָה (mamlɔḵɔ or mamlɔḵɔ́).

Public comments edit


Examples edit

When the transcription is identical in two or more systems, the cells have been merged to highlight differences and make comparison easier. Traditional transliteration is given throughout.

Transliterations Normalizations
Word Traditional Superscript Sartma 1 Sartma 2 Traditional Modern
בָּרָא bārā́ʾ bɔrɔˀ bɔrɔʔ bārā bɔrɔʔ
אַרְגָּמָן ʾargāmā́n ʔargɔmɔn ʾargāmān ʔargɔmɔn
תַּ֫חַת táḥaṯ táḥaṯ táḥaṯ
עַ֫יִן ʿáyin ʕáyin ʿáyin ʕáyin
כֹּהֵן kōhḗn kohen kōhēn kohen
אֵל ʾēl ʔel ʾēl ʔel
אֶל ʾel ʔɛl ʾel ʔɛl
נֶ֫פֶשׁ nép̄eš nɛ́p̄ɛš nép̄eš nɛ́p̄ɛš
אִם ʾim ʔim ʾim ʔim
יִשְׂרָאֵל yiśrāʾēl yiśrɔʔel yiśrāʾēl yiśrɔʔel
קֹ֫דֶשׁ qṓḏeš qóḏɛš qṓḏeš qóḏɛš
רֹב rōḇ roḇ rōḇ roḇ
נְאֻם nəʾúm nəʔum nəʾum nəʔum
שֻׁלְחָן šulḥān šulḥɔn šulḥān šulḥɔn
אָנֹכִי ʾānōḵî́ ʔɔnoḵiʸ ʔɔnoḵī ʾānōḵī ʔɔnoḵi
אֲנָשִׁים ʾănāšî́m ʔănɔšiʸm ʔănɔšīm ʾănāšīm ʔănɔšim
בֵּין bên beʸn bēn bēn ben
תֵּימָן têmā́n teʸmɔn tēmɔn tēmān temɔn
יְהוּדִי yəhûḏî́ yəhuʷḏiʸ yəhūḏī yəhūḏī yəhuḏi
שְׁמוּאֵל šəmûʾḗl šəmuʷʔel šəmūʔel šəmūʾēl šəmuʔel
גָּדוֹל gāḏốl gɔḏoʷl gɔḏōl gāḏōl gɔḏol
שָׁלוֹם šālốm šɔloʷm šɔlōm šālōm šɔlom
מַצֵּבָה maṣṣēḇấ maṣṣeḇɔʰ maṣṣeḇɔ̂ maṣṣeḇɔ maṣṣēḇā maṣṣeḇɔ
מֶרְכָּבָה merkāḇâ mɛrkɔḇɔʰ mɛrkɔḇɔ̂ mɛrkɔḇɔ merkāḇā mɛrkɔḇɔ
מַה mah maʰ ma ma
יָפֶה yāp̄éh yɔp̄ɛʰ yɔp̄ɛ̂ yɔp̄ɛ yāp̄e yɔp̄ɛ
מֹשֶׁה mōšéh mošɛʰ mošɛ̂ mošɛ mōše mošɛ
אַרְיֵה ʾaryḗh ʔaryeʰ ʔaryê ʔarye ʾaryē ʔarye
אַיֵּה ʾayyḗh ʔayyeʰ ʔayyê ʔayye ʾayyē ʔayye
כֹּה kōh koʰ ko ko
פֹּה pōh poʰ po po
מָה mɔʰ mɔ̂
מֶה meh mɛʰ mɛ̂ me
מַה mah maʰ ma ma ma
כְּנַ֫עַן kənáʿan kənáʕan kənáʿan kənáʕan
נְחֹ֫שֶׁת nəḥṓšeṯ nəḥóšɛṯ nəḥṓšeṯ nəḥóšɛṯ
מַחֲשָׁבָה maḥăšāḇấ maḥăšɔḇɔʰ maḥăšɔḇɔ̂ maḥăšɔḇɔ maḥăšāḇā maḥăšɔḇɔ
יַעֲקֹב yaʿăqṓḇ yaʕăqoḇ yaʿăqōḇ yaʕăqoḇ
אֱדוֹם ʾĕḏốm ʔĕḏoʷm ʔɛ̆ḏōm ʾĕḏōm ʔɛ̆ḏom
אֱלֹהִים ʾĕlōhî́m ʔĕlohiʸm ʔɛ̆lohīm ʾĕlōhīm ʔɛ̆lohim
מָרְדֳּכַי mordŏḵái mɔrdɔ̆ḵay mordŏḵáy mɔrdɔ̆ḵay
עֳנִי ʿŏnî́ ʕŏniʸ ʕɔ̆nī ʿŏnī ʕɔ̆ni
כֹּחַ kṓᵃḥ kóaḥ kṓaḥ kóaḥ
רֵיחַ rếᵃḥ reʸaḥ rḗaḥ rḗaḥ réaḥ
יֵשׁוּעַ yēšû́ᵃʿ yešuʷaʕ yešū́aʕ yēšū́aʿ yešúaʕ
רֶ֫גֶל réḡel rɛ́ḡɛl réḡel rɛ́ḡɛl
חֹ֫רֶשׁ ḥṓreš ḥórɛš ḥṓreš ḥórɛš
אֵ֫פֶר ʾḗp̄er ʔép̄ɛr ʾḗp̄er ʔép̄ɛr
כֹּל kōl kol kōl kol
כּוֹל kôl koʷl kōl kōl kol
נֹ֫גַהּ nṓḡah nóḡah nṓḡah nóḡah
עָלֶה ʿāléh ʕɔlɛʰ ʕɔlɛ̂ ʕɔlɛ ʿāle ʕɔlɛ
מַרְפֵּא marpḗʾ marpeˀ marpeʔ marpēʾ marpeʔ
אֵי ʾê ʔeʸ ʔē ʾē ʔe
אֹפֵּה ʾōppḗh ʔoppeʰ ʔoppê ʔoppe ʾōppē ʔoppe
מְנַשֶּׁה mənaššéh mənaššɛʰ mənaššɛ̂ mənaššɛ mənašše mənaššɛ
פֹּה pōh poʰ po po
פּוֹ poʷ po
פֹּא pōʾ poˀ poʔ po

Genesis edit