What can the Greek Septuagint tell us about ancient Hebrew?

In the backstory of the Ḥanukkah holiday, describing the persecution and forced assimilation led by Seleucid Hellenist ruler Antiochus IV Epiphanes, and the victorious revolt led by the Maccabee family, the Hellenists and their allies are depicted as inherently hostile to the Jews and the Jewish tradition and way of life.

But this wasn’t always the case. Alexander the Great, who conquered Israel during his wave of expansion twenty-three centuries ago, was viewed positively at the time, having removed the Persian imperial regime, which had become unpopular under its last ruler, Darius III. On arrival in Jerusalem, it is said that Alexander initially expected statues of him to be erected in the Temple, but was told that would be impossible, and graciously accepted the compromise offered by the Kohanim, that instead every Kohen child born that year would be called Alexander. The name retained its popularity. Even the Maccabees’ descendants, the Hasmonean dynasty of Jewish kings, included one bearing the name, Alexander Yannai.

Another example of positive exchanges between Hellenists and Jews was the project undertaken to translate the Torah, and subsequently the rest of the Tanakh, into Greek for inclusion in the Alexandria Library, initiated at the request of Ptolemy II Philadelphus,[i] a couple generations after Alexander’s conquests. This project culminated in the production of the Septuagint, a collection of translations done over several generations and of varying quality. (The Septuagint, together with the emergence of Jewish communities in the burgeoning Hellenistic cities of the Near East, in turn served as a catalyst for the emergence of Greek-speaking Jewish culture and philosophy developing over centuries, most famously personified by Philo of Alexandria.)

Through its transliteration of Hebrew names and other words, the Septuagint also serves as a source indicating how Hebrew was then pronounced, a thousand years before scholars in Northern Israel wrote out the full Tanakh using the Tiberian vowel and diacritic niqqūd system used until today. There are many instances, particularly in the later books, in which it seems that the translators either did not understand the Hebrew term or phrase or for some other reason thought it best to transliterate it into Greek letters instead of actually translating it into Greek.

Comparing the Septuagint’s Hebrew names and terms rendered in Greek with their Masoretic Hebrew rendering as preserved in Jewish tradition produces numerous observations of intriguing differences.

  1. Were there then two kinds of letter ‘ayin (ע) in Hebrew? In the Arabic language, Hebrew’s close cousin, there are two forms of the letter ‘ayin: the “normal” ‘ayin (ع), pronounced as a constriction of the throat, and an ‘ayin with a dot over it, called ghayin (غ), pronounced as an affricated or softened “g” sound, often compared to a French “r”. In the Septuagint the ‘ayin is often ignored (the constriction of the throat having no Greek parallel), but in some cases is transcribed as a Greek gamma (Γ). (Strengthening the suspicion that this indicates the presence of an archaic Hebrew ghayin in those words is the fact that, in modern Greek, gamma is pronounced much like Arabic ghayin.) The two different Greek renderings of an apparently otherwise identical Hebrew ‘ayin can sometimes be seen in close proximity to each other, even in the same verse. One such example is in Devarim / Deuteronomy 2:23, which notes that the pre-Philistine inhabitants of Israel’s southern coastal plain were a people called the Euaīoi (from the Hebrew word עַוּיִם, ‘Awwīm), with the usual ‘ayin. But the name of one of their cities is given as Gaza, as opposed to the traditional Hebrew form, ‘Azzåh (עַזָּה), seemingly indicating that the ‘ayin here was once a now-extinct Hebrew ghayin.[ii]

Similar ghost ghayins are seen in the name of Mount Ebal, Hebrew ‘Eivål (עֵיבָל), called Gaibál in the Septuagint (Devarim / Deuteronomy 11:29)[iii], or in the names Beit Pe‘ōr (בֵּית פְּעוֹר) and Ba‘al Pe‘ōr (בַּעַל-פְּעוֹר), called oíkou Phogór (the house of Phogór) and Beelphegòr in the Septuagint (Devarim / Deuteronomy 3:29 and 4:3, respectively). Queen ‘Athalyåhu (עֲתַלְיָהוּ) daughter of ‘Omrī (עָמְרִי) is called Gotholía daughter of Ambrí (II Melakhim / Kings 8:26), though both her name and her father’s begin with an ‘ayin, and a woman named ‘Azūvåh (עֲזוּבָה) is called Gazoubà in the Greek text (I Divrei haYamim / Chronicles 2:18).

The disappearance of a seemingly basic form of a letter so late in Hebrew history would be strange. Did the Masoretes still know of a harder pronunciation of the letter ‘ayin, but lack the tools to express it in writing (as a guttural, ‘ayin not being able to take a dagesh, the usual method of showing a harder pronunciation of a letter)? (The presence of a mappiq mark indicating a heavier hei would seem to indicate otherwise.) Did they consider the occasional pronunciation of ‘ayin as ghayin to be unimportant? Or was Hebrew ghayin already completely lost in the language’s pronunciation as recorded in northern Israel by the Masoretic scholars, despite the fact that the Arabic language, with its own letter ghayin very much alive, then dominated the region?

  1. Various differences in Hebrew vowel patterns and pronunciations, as seen in the names noted above. Some specific phenomena that stand out:

(a) It seems that the Septuagint’s Hellenized Hebrew would indicate that the sheva na‘ (שווא נע), generally pronounced as a short “eh” sound by Hebrew speakers today, was often closer to an “ah” or “ae” sound, similar to how many Yemenite Jews and some other communities traditionally pronounce it (and bringing it closer to its relative, the ḥataf pataḥ, ֲ  ). Though it is often transliterated as an epsilon (short “eh”), it is transliterated with an alpha (short “ah”) in many cases, such as in the name Menasheh (מְנַשֶּה), given as Manasse, in the word Refå’īm, written in Greek as Raphaïn (Devarim / Deuteronomy 2:20), and in “sons of – ”, Benéi in Hebrew (בְּנֵי), transliterated in the Septuagint as “Banèi – ” (I Divrei haYamim / Chronicles 7:35). In the same way, Shelōmōh (שְׁלֹמֹה) was written as Salōmōn. The case of Pe‘ōr (פְּעוֹר) being transliterated as Phogór, noted above, is also noteworthy for its rendering of a shvah na‘ as an omicron, or short “o”.

(b) Sheva naḥ (שווא נח), which is completely silent in traditional Hebrew, is occasionally paralleled by a vowel in the Septuagint, whether with an alpha or epsilon (short “a” and “e” sounds), as in transliteration of Par‘ōh (פַּרְעֹה) as Pharaò, or as an eta or ei combination (epsilon + iota), both indicating a long “ē” /”ei” sound, which is somewhat more surprising. It occurs often enough, though, to suggest that this does reflect a somewhat consistent pronunciation tradition for such words. Examples of this include Asēdoth for Ashdod (Devarim / Deuteronomy 2:23), sadēmoth for shadmot, Hebrew שַׁדְמוֹת, a word for “fields” (II Melakhim / Kings 23:4), and ammazeibī in II Melakhim / Kings 12:10, an attempt at transliterating the Hebrew ham-mizbē’aḥ (הַמִּזְבֵחַ), meaning “the altar”.

Pronunciation of the sheva ( ְ ), whether na‘ (pronounced) or naḥ (silent) according to received Hebrew tradition, has clearly been more fluid over the ages… Just as today’s Hebrew speakers seem to have a new set of rules determining when sheva is pronounced or silent.

(c) Apparent retention of an earlier “a” sound that later developed into an “e” or “i”. This can be seen above in the rendition of ham-mizbē’aḥ (הַמִּזְבֵחַ) as ammazeibī. This would bring it closer to the Arabic pronunciation of the same word, madhbaḥ (مَذْبَح). Similarly, the word מִנְחָה minḥåh (meaning “offering”) is written in Greek letters as manaa in II Melakhim / Kings 8:8, apparently indicating its pronunciation as manḥah.

A phrase meaning a “the staircase”, gérem ha-ma‘alōt (גֶּרֶם הַמַּעֲלוֹת) is half-transliterated into Greek letters as garem (II Melakhim / Kings 9:13), similar to Classical Hebrew pausal forms (likely preserving archaic pronunciation).[iv]

A nearly opposite phenomenon can be seen in an enigmatic Greek transliteration of the phrase קֳבָל-עָם, meaning “before the people”, as keblaàm (II Melakhim / Kings 15:10), though the Masoretic pronunciation indicated is qovol-‘åm. (Hebrew speakers today would more likely say something like qoval-‘am, or qval-‘am if the text is without niqqūd.) The pronunciation indicated in the Septuagint seems nonetheless to bring the first element in the phrase, the Hebrew word  קבל(meaning “before – ” or “in front of – ”), closer to the Classical Arabic pronunciation of the same word, which is qabla (قَبْلَ).

(d) The preference in many cases for a soft phi (φ) instead of a hard pi (π), even when the Hebrew word is generally to have a dagesh in the letter peh / feh (פ), indicating the harder pronunciation (as “p”). This is seen above in the examples of the name Pe‘ōr being rendered as either Phogór or –phegór and Pharaò for Par‘ōh (still influencing the pronunciation of the Egyptian royal title Per-‘āh / ‘Āh-per-ti as “Pharaoh” in European languages). Other examples include the names Pū‘åh (פּוּעָה)[v], transliterated as Phouá (Shemot / Exodus 1:15), and Pīneḥås (פִּינְחָס), transliterated as Phineés (Bemidbar / Numbers 25:7).

On the other hand, the Egyptian city of Pithōm (פִּתֹם) keeps its hard “p” in Greek as Peithò (Shemot / Exodus 1:11), and the consonants of the name Pōtīfar (פּוֹטִיפַר) are preserved in Petephrès (Bereishit / Genesis 39:1).

An interesting case is that of the Hebrew name Tsippōråh (צִפֹּרָה), meaning “bird”, rendered as Sepphóra (Σεπφόρα) in Shemot / Exodus 2:21, indicating an unexpected take on the doubling of the feh into a peh (by means of the dagesh), opposite to what Masoretic conventions would theoretically indicate. (Not only are the two sounds given separately, but with a hard “p” at the end of a syllable, and a soft “f” or “ph” sound at the beginning of a syllable.) Similarly, Joseph’s Egyptian name (given in the Hebrew text as צָפְנַת-פַּעְנֵחַ, Tsåfenath-Pa‘nēaḥ, understood in Hebrew as meaning something like “decipherer of codes”) is approximated in the Greek text of Bereishit / Genesis 41:45 with a phi at the beginning of its second element instead of a pi, as Psomthomphanéch (Ψονθομφανήχ).

The examples seen don’t allow for a clear rule to be formulated, so the overall impression is that in this matter as well there was a good amount of fluidity in the pronunciation of peh (פּ) versus feh (פ).

(e) Rival traditions of how to pronounce Hebrew place names apparently explain the transliterations of the Golån (גּוֹלָן), the Yardēn (יַרְדֵּן) and Gilgål into Greek as Gaulòn, Iordán and Golgól. Were they then גַּולֹן (“Gawlōn”, with a since-lost diphthong), י­­ֹרְדָּן (“Yordån”) and (גָּלְגֹּל) (“Golgōl”) in the eyes of the Septuagint translators?

The Septuagint has much to teach in terms of the lost Hebrew manuscripts of the Tanakh that were used to produce the ancient Greek texts, both in terms of their overwhelming similarity to those that survived as the Masoretic tradition, and in their occasional divergences, shedding light on some questions. Though in some cases it appears that the Septuagint’s translators just simply misread a word, as in the translation of I Melakhim / Kings 20:10, where it renders שְׁעָלִים (she‘ålīm), meaning “handfuls” of dust, which makes more sense, as “foxes” in Greek (which would be שׁוּעָלִים shū‘ålīm). Similar sounding names are also confused: Shifråh (in Bereishit / Genesis 1:15) is called Sepphóra, the proper transliteration of Tsippōråh, a figure introduced soon thereafter. The land of Gōshen (גֹּשֶן) is called the land of Gesem (in Bereishit / Genesis 47:1, 4), and the Septuagint even explicitly mistranslates Gōshen (גֹּשֶן) as the land of “Gesem of Arabia” in verse 45:10. An Arab chieftain called Geshem / Gashmu appears perhaps a millennium later in ancient Israel’s narrative, as Neḥemyåh rebuilt the city walls of Jerusalem, so his anachronistic appearance here in the Septuagint is most likely due to simple confusion based on the similarity of the names.

Illustrating how translating the Tanakh into Greek inevitably involved adapting its content to Greek culture, when Ya‘aqōv gives a guilt-trip to his sons, referring to his grave as She’ōl (Bereishit / Genesis 44:29), the Septuagint translated this into Greek as “Hades”.




[i] At around the same time, the Ptolemaic monarchy, and perhaps Ptolemy II Philadelphus as well, may have commissioned the work of Manetho documenting Egyptian narrative history. So it could be that the Septuagint was part of a wider effort to collect information and document ancient narrative histories related to the peoples in their domain.
The name of the Septuagint refers to the legend that it (or at least the translation of the Torah) was produced by a group of seventy Jewish scholars, and that all seventy independently – and miraculously – produced the same translation into Greek.

[ii] Further evidence of this possibility is seen in the fact that when Arab armies conquered the region, two centuries before the Tiberian manuscripts were written, they adopted the name of the city as Ghazzah (غَزَّة), indicating their awareness of the same Hebrew-Semitic pronunciation reflected in the Septuagint centuries before. I am assuming that later Arabic use reflects a tradition going back at least to the Islamic conquest. As Nabataean Arab traders did business in Gaza for roughly a millennium before the Arab conquest, it is likely that awareness of the pronunciation of the city’s Semitic name, with a ghayin and double z sound, had entered the Arabic language long before Arabs entered the Levant as conquerors.

[iii] In the case of Mount Ebal, the Arabic name follows the Masoretic Hebrew tradition, as ‘Aibāl (عَيباَل), with a normal ‘ayin, not a ghayin as the Septuagint suggests could have been in the pre-Masoretic Hebrew name.

[iv] In II Melakhim / Kings 6-19 the Septuagint nonetheless transliterates בֶּדֶק – meaning “checking”, in this case, renovating the Temple – in accordance with the normative Hebrew tradition as bedek.

[v] The Hebrew-Canaanite name Pū‘åh (פּוּעָה) is also known from the Late Bronze Age Ugaritic Epic of Aqhat, where it is given (in Shirly Natan-Yulzary’s edition) as פֻּעַ’תֻ (Pughatu), with a ghayin in place of the ‘ayin. So either this particular ghayin became an ‘ayin at some point early on, with the transition to Iron Age Israelite Hebrew, or was lost by the time the Septuagint’s translators transliterated the name as Φουά (Phouá), and not Φουγά (Phougá), assuming the gamma does indicate the presence of a ghayin.


About the Author
Daniel Kennemer is the founder of the Mount Carmel Arabic Immersion and the Jethro's Tent Initiative for Biblical Hebrew. He studied Archaeology and Geography at the Hebrew University of Jerusalem
Related Topics
Related Posts