The Septuagint and the New Testament

Septuagint manuscript from The Schoyen Collection (#2649) containing a portion of Leviticus from second century Egypt - one of the oldest surviving manuscripts of this part of the Bible.

“It is written” – Quotations from the Old Testament in the New Testament (1)

There is a popular misconception that the earliest Christians used the Greek translation of the Hebrew Bible as their Scriptures, and that all the quotations from the ‘Old Testament’ in the New Testament are from this Greek translation, commonly known as “The Septuagint” [1] . In later posts I plan to look at some quotations which are more likely to be drawn directly from the Hebrew (then translated into Greek by the New Testament writer) or from some other source, but first I will deal with some of the quotations which are most likely from the Septuagint. My purpose is to give an overview of some of the difficulties we face when the New Testament quotation doesn’t match up with what we find in our own Bibles, and to provide some explanations for why this happens (and perhaps some solutions).

In this post I will look at just one quotation which illustrates the problems quite well. During the early church meeting commonly known as ‘the Jerusalem Council’ (Acts 15:1-35) James quotes from ‘the words of the prophets’ (vv15-18), specifically from Amos 9:11-12. This quotation is closer to the Septuagint (LXX) than to to the Hebrew Masoretic Text (MT) [2] , although there are actually eleven differences between the version in Acts and the LXX [3] .

The most significant differences are highlighted by the underlined words below:

Acts 15:15-18 (NRSV)

“After this I will return,
and I will rebuild the dwelling of David, which has fallen;
from its ruins I will rebuild it,
and I will set it up,
so that all other peoples may seek the Lord
even all the Gentiles over whom my name has been called.
Thus says the Lord, who has been making these things known from long ago.”

Amos 9:11-12 (NRSV)

On that day I will raise up
the booth of David that is fallen,
and repair itsbreaches,
and raise up its ruins,
and rebuild it as in the days of old;
in order that they may possess the remnant of Edom
and all the nations who are called by my name,
says the Lord who does this.

The difference is actually quite important in the context of the Jerusalem Council because it was specifically the words “all the Gentiles over whom my name has been called” to which James was appealing in accepting Gentiles into the church. However, these words are not in the Hebrew text although they are in the Septauagint. How, when and why were they changed? There are several possibilities:

  1. The translator who translated the Hebrew into Greek (the LXX) made a mistake, or a couple of mistakes, in reading the Hebrew and therefore made a wrong translation which James then uses.
  2. The translator was using a different Hebrew text to the one which eventually became the Masoretic Text (scholars call this ‘original’ text used for the translation the Vorlage). If so, there must have been at least two different versions of the Hebrew of Amos at the time when the LXX was translated. That is possible, and we have good manuscript evidence to indicate there were different Hebrew texts circulating in the first century CE, but which text then would be most faithful to the earliest or ‘original’ text?
  3. The translator deliberately altered the words to give it a different meaning for theological reasons.

For Christians who believe in the ‘inerrancy’ of the Bible any of these possibilities pose a problem: which text is ‘inspired’, the Hebrew Masoretic Text used for translating the Old Testament into English (which may include errors if we believe that the NT puts its stamp of approval on an alternative version), or the Greek Septuagint translation of a different Hebrew text (which is now lost), used by James and quoted in the New Testament? Either the MT is right, and the LXX and NT are wrong, or the LXX and NT are right and the MT is wrong. Do you see the problem?

So how could a translator or scribal copyist make a mistake in the first place? How does “that they may possess the remnant of Edom” become “all the Gentiles may seek [the LORD]”? Here is the Hebrew text of these words as they appear in the MT (without the vowel points, which were added later by the Masoretes between the 7th and 10th centuries CE):

למען יירשו את־שארית אדום

The difference rests mainly on two words (highlighted above). The Hebrew word for ‘Edom’ is אדום and the ‘o’ sound is represented by the letter ו. Strictly speaking written Biblical Hebrew is consonantal, and most of the vowels were added much later by the Masoretes. However, the letter ו is one of the four letters known as ‘Matres lectionis’ which were also used to indicate vowels and during the Biblical period a word could be written with or without it. It is possible that in some manuscripts the word for ‘Edom’ was written אדם. Now this presents a problem, because this is the same word as ‘mankind’ and ‘Adam’. In most cases there is no problem because the context will determine which of the three words is meant (although this is precisely the kind of problem that caused the Masoretes to later add marks to indicate the vowels and to avoid the confusion). So now we have two possibilities for how ‘Edom’ became ‘mankind’ (or Gentiles): the manuscript from which the LXX translator was working had אדם without the vowel ו and he read it as ‘mankind/Gentiles’ rather than as Edom; or, a scribe copying a manuscript at some point read אדם=mankind as ‘Edom’ and added the ו vowel to avoid confusion (but making a mistake in the process), and it is from this copy containing the error that our MT has come down to us. Take your pick.

Paleo-HebrewThe second difference is slightly more complicated. The word translated ‘seek’ in the LXX (ἐκζητήσωσιν) translates ירש which means ‘possess’. This would be the only place where the LXX translates ירש as ‘seek’. Twenty nine times the LXX uses this Greek word for ‘seek’ to translate the Hebrew word דרשׁ which differs only by the first letter, and seventy three times it uses it to translate the word בקשׁ which has only one letter in common. Some scholars suggest that the translator misread דרשׁ (seek) as ירשׁ (possess) which is possible especially if the copy from which he was working was worn, faded or smudged. McLay says it is less likely that he misread it as בקשׁ as this would require him to confuse two consonants which don’t look very similar[4]. However, and this is where I think it becomes even more interesting, we need to know that the Hebrew system of writing began to change after the Babylonian exile from ‘Paleo-Hebrew’ to the ‘block’ or ‘square’ Aramaic or Assyrian style that is still in use today (with some further modifications). The earliest copies of the scroll containing Amos (‘the Scroll of the Twelve Prophets’) would most likely have been in Paleo-Hebrew, and in this script the words בקשׁ and ירשׁ are much more similar. The chart shows these words in the square script which is now used, the Paleo-Hebrew in which Amos was probably first written, and the early form of the square script which is found in the Dead Sea Scrolls Isaiah manuscript to show how it may have been written if the Aramaic script was already in use. While ר and  ק are not similar in the square script they could easily be mistaken in Paleo-Hebrew. Similarly ד י and ב could also be mistaken, especially if the text was faded, smudged or damaged. Many scholars understand that the LXX text arose from a different version of the Hebrew text through a misreading of this kind [5] .

Where does this leave us? We can see how an error may have crept in somewhere and this explains why we have two versions of the words of Amos in our Bibles. But which one is likely to be original? Unfortunately there is no easy answer to this and scholars are divided on the question. Most scholars tend to prefer one version or the other depending on which one they feel best fits the context and until another ancient manuscript is discovered to shed more light on the subject we will have to be content with that.


[1] The ancient Greek translations of the Hebrew Bible are popularly called “the Septuagint”, although several different translations have all in fact been designated as such. There is probably no one translation which has a better claim to the designation than the others, so some scholars tend these days to speak of them as “Septuagints” (plural) and specify the particular manuscript to which they are referring, or collectively as the Greek Jewish Scriptures. Often when commentaries refer to “the Septuagint” (or by its abbreviation LXX) they mean the text and translation by Brenton (1844). For a good overview of current Septuagint scholarship see R.Timothy McLay The Use of the Septuagint in New Testament Research (Grand Rapids: Eerdmans, 2003).

[2] I. Howard Marshall, ‘Acts’ in G.K.Beale and D.A.Carson (eds.) Commentary on the New Testament Use of the Old Testament (Grand Rapids: Baker Academic, 2007) 589

[3] Gert J. Steyn Septuagint Quotations in the Context of the Petrine and Pauline Speeches of the Acta Apostolorum (Kampen, Netherlands : Kok Pharos, 1995) 252-3

[4] McLay, 23

[5] Marshall, 590