It's well-known that Classical Hebrew had phonemic length distinctions in the stops, since geminated stops didn't turn into fricatives: compare רַב raβ "rabbi" against רַבִּי rabbī "my rabbi".

But I've noticed a curious pattern with Hebrew and Aramaic names transcribed in Greek (in the LXX and New Testament): the letter ש (šin) is transcribed with a single sigma after a long vowel, but a double sigma after a short one:

  • מְשִׁיחַ‎ məšīaħ → μεσσίας messias
  • יֵשׁוּעַ jēšūaʕ → ἰησοῦς iēsūs
  • אַבְשָׁלוֹם 'aβəšālōm → ἀβεσσαλωμ abessalōm
  • הוֹשֵׁעַ hōšēaʕ → ὡσηέ hōsēe

This isn't a phonological rule of Koine Greek, which makes me wonder: is it just a transcription convention? Or did Classical Hebrew and Aramaic actually lengthen their continuants after short vowels?

(P.S. Ancient Greek did have phonemic length distinctions on /a/, /i/, and /y/, but didn't mark them orthographically. In other words, the final a in Abessalōm was presumably long, but there's no way to know from the written form. Koine eventually lost these distinctions, but not by the time of the LXX.)

  • Is there a difference in the Koine realization of epsilon in μεσσίας vs. μεσίας, for instance? In other words, could doubled consonants in the transcription be used to specify length of the preceding vowel? (There is something similar in modern Dutch, e.g. bakken vs. baken.) – Keelan Jun 29 '19 at 6:56
  • @Keelan Interesting thought! For some vowels, potentially, but for /e/ in particular the Koine orthography had a way to indicate vowel length (using epsilon versus eta). – Draconis Jun 29 '19 at 17:03

I don't know anything significant about Ancient Hebrew. Since there were different varieties of Hebrew and Aramaic in ancient times, I can't be sure whether information that I find in documents online applies to all varieties.

With that caveat, the title and first sentence of this question seem a bit strange to me, because my understanding is that Ancient Hebrew had phonemic (not allophonic) length distinctions for most consonants, not just the stop series. As you say in your answer, certain consonants were supposed to be non-geminatable in Biblical Hebrew: the Wikipedia article "Biblical Hebrew" says

Geminate consonants are phonemically contrastive in Biblical Hebrew. In the Secunda /w j z/ are never geminate. In the Tiberian tradition /ħ ʕ h ʔ r/ cannot be geminate; historically first /r ʔ/ degeminated, followed by /ʕ/, /h/, and finally /ħ/, as evidenced by changes in the quality of the preceding vowel.

But I haven't found any source that lists /s/ as a consonant that cannot appear as a phonological geminate in Biblical Hebrew. So the implication of the sources that I've read seems to be that singleton /s/ was phonologically contrastive with geminate /ss/ in Ancient Hebrew.

Phonologically contrastive consonant length was not written in the Hebrew alphabet. There are later traditions of marking length (or for plosives, the absence of spirantization) with diacritics, but these don't necessarily correspond to the distribution of length in earlier time periods. I'm not sure what evidence experts use to establish the existence of geminate consonants in particular words in Ancient Hebrew.

In terms of possible conditioning factors for a process of consonant lengthening, stress seems to be important. I found a web page that refers to "Pretonic Vowel Lengthening or Equivalent Consonant Gemination" and says that "pretonic gemination at times substitutes for pretonic lengthening" ("Phonemic Structure of Pre-Exilic, Tiberian and Israeli Hebrew Contrasted", History of the Ancient and Modern Hebrew Language, by David Steinberg). "Pretonic" in this context seems to refer specifically to the syllable immediately before the stressed syllable. Also, I don't know how schwa relates to the process described there.

I don't know the original position of the stress in the words that you mention, but I would recommend studying that.


Having confirmed with a Hebraist, this is indeed a phonological rule of Hebrew (and Aramaic). In his words:

in certain positions, consonants geminate after historically short vowels

Some of these geminations didn't survive into the modern day, especially with the loss of contrastive vowel length; those that are still relevant are marked with a dageš ħazaq (a dot in the middle) in pointed texts.

This lengthening process applied consistently to all consonants, not just sonorants, except for aleph, he, ħeth, ʕayin, and reš: these four could not be geminated, so they lengthened the vowel instead.

(In fact, this phonological rule is the reason why raβ + -i gives rabbi "my rabbi" rather than *raβi, as mentioned in the question: the beth comes after a short vowel and is before another vowel, so it gets geminated and remains a stop. Alternately, raβrabb due to the short vowel, but the bb degeminates at the end of a word.)

  • Agrees with this answer to Why Did Koiné Greek Still Write Double Letters? – Michaelyus Jun 28 '19 at 10:07
  • 3
    I don't think the example is correct. rabbi comes historically from rabb + i. The sound change here is rabb > rab (loss of gemination at the end of a word), not rabi > rabbi. I'm not sure about the general rule "consonants geminate after historically short vowels" either (and if it were true, why does this show up in Greek only for the letter shin?) – b a Jun 28 '19 at 12:21
  • 1
    @Draconis The difference is that there was never a specific point in time at which there was a "lengthening process" that made non-geminate consonants geminate. If I understand your answer correctly, you use this sound change to explain the gemination of s in Greek transcriptions, but I don't think such a rule existed, at least not in the example of rabbi. – b a Jun 28 '19 at 16:03
  • 1
    If I am understanding what your Hebraist is referring to, I would describe the sound change (for Tiberian Hebrew) as: all historically short vowels stay short before geminate consonants, and become ultra-short before non-geminates. All of your examples of Greek transcriptions of ss after short vowels are the latter category, so I don't think this particular sound change applies. Maybe there was another sound change. If the ss is actually gemination, it would have to be a feature of a different dialect anyway – b a Jun 28 '19 at 16:06
  • 1
    I'm not sure this rule is relevant. First, I don't think "historically short vowels" include schwa; and second, if the rule did apply in these words, the gemination would presumably be indicated with a dagesh, which it isn't. – TKR Jun 29 '19 at 22:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.