Reading God’s Word in Translation

The Septuagint in historical, philological and theological perspective

Jan Joosten, Oxford

In our global and multicultural world translations are omnipresent. We see foreign heads of state on television and hear them speak in English, in the voice over. We watch Iranian movies in our own language. And when a Finnish friend posts a message on Facebook, a click suffices to get a translation – unfortunately so far usually an incomprehensible one.

The Bible too is read in translation. Written in Hebrew, Aramaic and Greek around 2000 years ago, the Bible has to this day been translated into some 3000 languages. It has been translated into English many times over: Tyndale, King James, Revised Standard Version, Living Bible, and many others. To read the Bible in translation seems natural to us.

It wasn’t always so. Translation is a risky business. Traduttore traditore says an Italian proverb: to translate is to betray. To translate from one language into another is hard to do, and translation always involves changes in meaning. Translation is not impossible. In daily life, we make do – we install a Korean printer by help of the manual. If an error causes problems it can be corrected. Think of foreign spokesmen claiming that their statements have been wrongly translated. But when it comes to eternal truths, the stakes are much higher. If perfectly faithful translation is impossible, how shall one deal with a divinely inspired word? The Talmud quotes a Rabbi Judah who said:

He who translates a biblical verse literally is a liar, and he who adds to it is a blasphemer (Tos. Meg. IV 41).

Translating scripture is problematic. In official Islamic doctrine, the Quran is considered to be untranslatable. What we would call translations of the Quran have the status of tafsir, commentary.

Questions regarding biblical translation must have confronted those who first ventured to make one in a particularly arresting way. In the present lecture, I would like to reflect on the notion of Bible translation from the vantage point of the first translation of scripture, to wit, the Septuagint.

I will address three distinct but interconnected topics. Firstly, I will review the historical context in which the Septuagint came into being and try to explain why, and how, that context was propitious to the idea that scripture could be translated. Secondly, I will show how translation into Greek changed the meaning of the text, globally and in many details; the risk taken in translating God’s word proved indeed to be a real one. Thirdly, I will mount a defence on behalf of the Septuagint translators – and other translators of the Bible – on theological grounds. Although translation is indeed never faithful, it is an integral part of the process of revelation.

1. Historical context: The western diaspora

Specialists agree that the Septuagint translation was begun in Egypt at the beginning of the third century BCE. The first part to be translated was the Pentateuch, which appears to have been in public use before the year 250 or so.

Most probably at this period some form of oral translation of the Torah was already practiced in Palestine. Oral translation leaves few tangible traces, but the later Jewish Targums, written down after the Jewish wars in 70 and 130 CE, contain many old elements, some of which appear to go back to the early Hellenistic period. It seems that in Palestine too readers and listeners struggled to understand the archaic Hebrew of the Pentateuch: a translation was needed. But the translation did not function in the same way in the diaspora and in the home country. In the land of Israel, where some form of Hebrew was still spoken, translations were not regarded as full equivalents of the source text. Translations into Aramaic never supplanted the Hebrew text, but accompanied it: one reads a verse in Hebrew, and then the same verse in Aramaic.

In the western diaspora – Alexandria, but also other places in Egypt – we find a different conception. The translation appears from the start as a written document, no different in this regard than the source text. In its style and translation technique, the Septuagint presents itself as a stand-in for the Hebrew text. The earliest manuscripts, dating from the beginning of the second century BCE, contain only the Greek text – not a bilingual edition as in the case of the Targums. From the end of the third century BCE, Jewish-Hellenistic authors such as Demetrius the Chronographer and Aristobulus quote the Greek text as scripture. A few centuries later, Philo will take a step further and qualify the translators as prophets inspired by God: like the Hebrew, the Greek version contains the word of God.

How can we explain this different attitude among Egyptian Jews? Where does their openness to the principle of translation come from? Is it simple necessity – have they definitively forgotten their Hebrew? This cannot be the sole reason: in Palestine, too biblical Hebrew is partly forgotten; and in the diaspora Hebrew is still known well enough to produce the translation.

Many scholars until today would answer our questions with a reference to the Letter of Aristeas. What explains the production of the Greek Pentateuch at the beginning of the Hellenistic age? King Ptolemy! For political, juridical, or cultural reasons, the Ptolemaic administration demanded a Greek version of Jewish scripture, overruling any religious reservations toward the principle of translation. For various reasons, this scenario is hard to accept. The Letter of Aristeas is hardly a dependable historical source. More to the point, the text of the Septuagint itself shows in many details that it is a work written by Jews for Jews. The Hebraisms, transcriptions of Semitic words, and typically Jewish terminology it contains would have made the Septuagint hard to understand – and impossible to tolerate – to a Greek readership, let alone the Lagide king and his court.

The decision to translate scripture was made by Egyptian Jews – not King Ptolemy. The diffidence toward translation expressed in Rabbinic sources reflects a view that remained dominant in the home country. But in the western diaspora, we find a different spirit. Why is this so?

A historical view will help to understand the specificity of Egyptian Judaism in this regard. By the time the Pentateuch was translated into Greek, Jews had been present in the Land of the Nile for a long time. Those who arrived, as soldiers or prisoners-of-war, in the early Hellenistic period, were added to Jewish groups settled in Egypt, some of them since the sixth century BCE or perhaps even earlier. The existence of Jewish colonies in Egypt in the Persian period is known from the Elephantine archives and other more sporadic documents. The Jews of Elephantine, in the fifth and fourth centuries BCE, resolutely claim to belong to the Judaism of their period: they celebrate the Pesach festival and observe the Sabbath; they practice purity laws; they worship the god Yaho – although the name differs, this is no doubt the same God who is designated with the tetragram in the Hebrew Bible – in their temple, and in the names they give to their children. They correspond regularly with the temple in Jerusalem. They designate themselves as Yehudaye, the Aramaic term for Judaeans/Jews.

Yet the “Jews” of Elephantine exhibit some surprising characteristics. Alongside the God Yaho they seem to recognize other Gods as well. In letters they sometimes use the habitual greeting formula of the time: “May the gods seek the welfare of my lord at all times…” A legal document shows that some of them accepted to take an oath by the name of other gods at the request of non-Jews: “You, Mibtahya (a Jewish lady) have sworn by the goddess Sati, and my heart is satisfied” – says Piya son of Pahi, an Egyptian man, in a legal document. We would be wrong to qualify all this as syncretism. Rather what we observe in the Elephantine documents is an advanced stage of acculturation. The Jews of Elephantine are in the service of the Persian Empire; they carry out transactions according to Aramaic law; they trade with their Aramaean and Egyptian neighbours; there are even some case of mixed marriages. In such a context, it is natural to show a certain tolerance, even openness, toward the religion of other groups. Although similar tendencies must have existed in Judaea and Samaria, the documents of which we dispose tend to show that this process of acculturation has progressed further in the diaspora, and has met with less opposition.

This openness and this movement toward other cultures goes hand-in-hand with a positive attitude toward translation. The Elephantine archives contain not a scrap of biblical texts, neither in Hebrew nor in translation. They consist almost exclusively of documents of day-to-day life: letters, contracts, lists. They show nevertheless that the community saw no problem in speaking of religious things in other languages than Hebrew. The Elephantine Jews write to Jerusalem about ritual matters in Aramaic. They discuss fine points of the Pesach festival in Aramaic. Pesach is referred to as Pascha – its Aramaic name – and the Sabbath as Shabbeta.

Many indications show that the western diaspora of the Hellenistic period continues certain tendencies characterizing the diaspora of the Persian period. Translating the entire Pentateuch into Greek was a radical step forward. But it was not unprepared. The possibility of translating the Jewish experience into languages other than Hebrew was already widely admitted. It was part of an open and positive attitude that had formed in dialogue with non-Jews over a very long period.

2. Translation as dialogue: philological considerations

The decision to translate the Pentateuch into Greek was made in a context of acculturation. The translational process itself proceeded along the same track. Translation implies a movement to and fro between cultures, a dialogue. The biblical message and Hellenistic culture enrich and modify one another. This is true even if, as I argued before, the Septuagint was at first addressed only to Jews, for those Jews were themselves integrated in the Hellenistic state. The dialogue at first happened within the Jewish community; in a second stage it extended further outward.

Let us look at a couple of examples illustrating this dialogue. I will focus on lexical equivalents. The Greek words selected to translate Hebrew words can be quite revealing. Ideas and conceptions are exported from the biblical culture to the Hellenistic one, and vice versa. Let’s take the export into Greek first.

Studying the Septuagint one realizes to what extent the Greek language, rich as it may be, lacked resources to render the Hebrew text. From the first verse of the Bible, the translators were confronted with terms and expressions that had no ready counterpart in Greek. “In the beginning God created the heavens and the earth” – how to render this notion of creation? The translator of Genesis at first uses simply the verb ποιέω “to make”. In chapter 14, however, where divine creation is again mentioned, he reaches for a different verb: κτίζω whose primary meaning is “to found a city”. From Gen 14 onward, this verb will be used to express the notion of creation in all the Greek Bible including the New Testament.

The translators of the Pentateuch put together a new terminology to express the strange content of their source text. Confronted with “untranslatable” words they sometimes simply transcribe them into Greek. The cherubim who guard the way to the garden of Eden become in Greek the χερουβιν. In other cases, the translators create neologisms such as ἀκροβυστία “foreskin” or θυσιαστήριον “legitimate altar.” Most often, however, they take existing Greek words and “press them into service”: Hebrew bara’ is rendered with κτίζω, and Hebrew torah “religious instruction” is rendered with νόμος “law”. A new meaning is forced upon the Greek words in each instance.

Let us look at another example, the notion of blessing. In the Pentateuch in many passages, God “blesses” his creatures, pronouncing beneficial and vivifying words. Humans too can bless one another or bless God. “To bless” – berek in Hebrew – is not simply a verb of speech; the speech effectuates something. When God blesses creation, it proliferates. The notion exists also in other Semitic languages, but Greek has nothing of the sort. In Greek culture, Gods may of course come to the help of human beings, but they do not “bless” by transferring to them a vivifying force. How then can this Hebrew verb be translated into Greek? The Septuagint translators solve the problem by recruiting the Greek verb εὐλογέω, whose meaning is “to speak well of, to praise.” The effect is at first a bit disorienting. To a Greek reader in the third century it would have sounded as if, in Gen 1:22, God “spoke well” of the great sea monsters, telling them to be fruitful, to multiply and to fill the oceans; in Gen 12:2 God promises Abram: “I will speak well of you and render your name great”; and in Gen 39:5 “The Lord praised the house of Potiphar because of Joseph, and the praise of the Lord was on all Potiphar’s household.”

In the longer run, however, the choice of the translators worked out. Being used in many passages where the required meaning was that of Hebrew berek, the Greek verb ended up absorbing some of that meaning. Dictionaries of the Septuagint are correct in attributing the meaning “to bless” to the verb εὐλογέω. New Testament writers and later authors use the Greek verb in its “Hebrew” meaning in non-translated texts.

In this way, Greek gained a new word (or at least, a new meaning). Moreover, the addition of this new word imports into Greek the biblical notion of efficacious speech.

These are only some examples, but they show how the Septuagint translation enriched the Greek language and Hellenistic culture. New concepts emerged. At first the creative impact of the translation affected only the Jewish community of readers, but in time the new ideas spread abroad and became known more widely.

But the exchange between biblical and Hellenistic culture was not a one-way street. Translation implies dialogue. In translating the Torah into Greek, Egyptian Jews also imported typically Greek ideas into their sacred scriptures. Let us look again at some of the lexical choices made by the translators.

When God makes a covenant with Abraham, in Gen 17:1, he tells him: “Walk before me, and be whole (Hb tamim).” The Hebrew word tamim “complete, whole, perfect” here seems to imply moral integrity, as it does in many other texts. Another possibility is that it refers to the unfailing faithfulness of a covenant partner (cf Deut 18:13). In other contexts, tamim is used to qualify the sacrificial animals, which are to be “without blemish.”

In the Septuagint of Gen 17:1, God demands that Abraham be ἄμεμπτος – “blameless, beyond reproach.” The translation is tolerably exact, but it suggests, unlike the Hebrew, that perfection is something one acquires in the public arena. Abraham’s perfection will be manifest in that no one will blame him (μέμφομαι).

A similar inflection of the sense can be observed in the translation of words meaning “honour” and “shame”. Hebrew kabod “honour, glory” derives from a root meaning “to be heavy”; the noun refers to a quality a person has in and of himself or herself. Against this, the Greek equivalents used in the Septuagint, τιμή “appreciation” and δόξα “reputation”, relate honour to the judgment of others. Similarly Hebrew qalon “dishonour” is associated with lightness (being “lightweight” – note the root qalah “to be light”), a quality again situated within the person. But the Greek equivalent, ἀτιμία “dishonour”, refers explicitly to the attitude of others.

We would be wrong to think the translators intentionally introduced these Greek nuances into the biblical text. Rather, the change is an inevitable corollary of translation. If one wants to speak of moral perfection, of honour and shame, in Greek, it is practically impossible to do so without bringing in the notion of one’s standing in the community. This notion is not in the forefront in the Hebrew words. The changes are subtle and were almost certainly made unconsciously. Nevertheless, they do affect the target text substantially.

Unconscious modification is not the only mechanism by which the Hellenization of the Septuagint came about. In other cases, it appears the translators wittingly and willingly corrected biblical expressions under the influence of Greek ideas.

A somewhat complicated example, but interesting nevertheless, is that of anthropomorphisms. As is well known, the Hebrew Bible often speaks of God in human terms. God has hands, eyes, ears and a mouth; he sits, stands up, comes down; he has a heart, and bowels, he is moved and gets angry. Some of these modes of discourse are rooted in a theological conception according to which the God of Israel really has a perceptible body. Admittedly, in other texts, the Hebrew Bible seems to deny the corporal idea of God. The Hebrew evidence deserves to be discussed in more detail, but presently we’re interested in what happens in the Greek translation.

Many anthropomorphisms of the Hebrew Bible are attenuated in the Septuagint. Instead of repenting of having created the world as the Hebrew has it (Gen 6:6), God is preoccupied according to the Septuagint, and he meditates; instead of having God dwell among the Israelites, the Greek text states that he is invoked among them (Exod 29:5); the mouth of God becomes the command of God (Exod 17:1). Many other anthropomorphisms are preserved, however: God sees, he is angry, he descends, he has eyes and a hand. Faced with the disparate data one wonders how the translators apprehended anthropomorphic discourse. Did they accept it, perhaps without thinking too much about the issue? Or did they consider it metaphorical and therefore unproblematic? Only the latter attitude would have been acceptable to Greek philosophers of their time.

An interesting clue allowing us to answer these questions has recently come to light. The clue is found in an article on the verb ἀποστέλλω “to send” prepared for the new Historical and Theological Lexicon of the Septuagint. As is noted in the article, this verb almost always corresponds to Hebrew shalah “to send.” But there are some exceptions: while ἀποστέλλω implies a separation, one sends away, shalah can be used even with objects that remain attached to the sender. In Hebrew one can “send” one’s hand, and a vine can “send out” its shoots. In Greek, the verb ἀποστέλλω cannot be used in this way, and the verb ἐκτείνω “to stretch out” is used instead. There are some exceptions, however. While humans always stretch out their hands, in four passages God is said to send – ἀποστέλλω – his hand. These passages are proof that, in the mind of the translators, God’s hand was not like a huan being’s hand: while a human hand is attached to the body and cannot be sent away, God’s “hand” is not so attached. We will not go far wrong if we state that, for the translators, God’s hand stands for God’s action or power.

Translations have been compared to plumbing or electric wiring. They are vital, indispensable, but they are at their best when they go unnoticed. As soon as they attract attention, their problematic nature stands revealed.

Greek ideas – cultural, political and philosophical – slipped into the Septuagint version. Although the translation is literal and ostensibly follows the Hebrew text word for word, Hellenistic notions are present in the language itself. The translation contains many connotations that are completely absent from the source text.

Of course, a translator can block out notions that are considered undesirable. The Septuagint translators seem to have avoided many words that they found to be too much tainted by pagan religion. But no translator can filter out all foreign connotations present in his target language. Language and thought go hand in hand, one cannot have the one and refuse the other.

3. A theological defence of translation

Our examples reveal the dialogical process underlying the Septuagint. This process might have developed differently, but it couldn’t have been avoided. Translators do not choose to open up to exchanges between the source language and the target culture: such exchanges flow automatically from the decision to translate. The die is cast from the moment the Jewish community decides to produce a translation of Scripture into the language they use in day-to-day life.

Some will consider this exchange a betrayal: the goal text does not mean the same thing as the source text, it is not faithful but perverts the words of God. Others may think that the exchange was a necessary step in the divine economy. The fiercely particularistic claims of the Hebrew Bible needed to be mitigated by means of Hellenistic rationalism and universalism. Some have gone as far as to say that the Septuagint is a praeparatio evangelica – a preparation for the message of the Gospel.

To me it seems more helpful to underscore what the Septuagint does qua translation. Translations have an in-built hermeneutical function. They permit, and require, to address the disjunction between the biblical message and its application in a new situation. Formerly, God spoke to the Israelites installed in their own country, with a functioning sanctuary in their midst; today, Jews of the diaspora are installed in another land, mingled among the nations, and their entire cultic practice is based in a text. If this disjunction cannot be addressed, scripture itself will become meaningless.

At this point I would like to introduce an idea from a somewhat out-of-the-way place. In the third part of his work The Star of Redemption (1921), the German-Jewish philosopher Franz Rosenzweig shows how reconciliation between humanity, the world, and God – a reconciliation that will be fully achieved only in the end of times – is experienced and verified already in the Jewish and Christian communities. For Rosenzweig, Judaism and Christianity are not competitors; they do not compare as one true and one false religion. Both are necessary, they complement one another. The view developed by Rosenzweig – a Jew who in his youth flirted with Christianity – although it is based on concrete experiences, remains to a large extent idealistic. Never mind: it is the idealistic aspect that interests us. Judaism and Christianity complement one another: Judaism, through it liturgy and observance, is oriented toward the “burning fire”; Christianity is turned to the world that needs to be brought back to God. If it weren’t for Christianity, the world would never be reconciled with God. But Christianity would loose its way if Judaism did not ever remind it of the burning fire, which justifies the whole process.

I find in Rosenzweig’s scheme a possible analogy of the relation between the Hebrew source text and the Septuagint as translation. The Hebrew text is close to the “burning fire”, revelation itself. The Hebrew text is like an oracle: mysterious, hard to grasp, and never exhausted. Those who study the Hebrew text are oriented toward the centre, the essence. In contrast, the Septuagint, as a translation of scripture, is turned outward an facing the world (I use this term in dependence on Rosenzweig). The Septuagint incarnates the desire to bring the word of God to those who do not know it yet. Both, Hebrew and Greek, are necessary. They complement one another. There is no reason to choose between them, they are two sides of the same coin.

The dynamic of source and translation does not apply to the Septuagint alone, but to any translation of scripture. I have held up the Septuagint as an example, not an exclusive norm. The merit of the very first “Bible” translators – I put Bible in quotation marks, because when the Septuagint is begun, the Bible doesn’t exist yet – the merit of the Septuagint translators is to have created a model that is followed until today: the model of reading scripture in translation.


The Hebrew Bible has a lot to offer. In Christian perspective this goes without saying, but it is true too in the perspective of cultural history. The notions of divine monarchy, of linear time, of the primacy of ethics, and many subsidiary ones, were meant to – and did in fact – impact the history of civilization in crucial ways. However, in order to reach beyond the limited sphere of Judaism – in fact, beyond the limited sphere of those who kept the knowledge of ancient Hebrew alive – the Hebrew Bible needed to be translated. The Septuagint is the first documented step in the history of the Bible’s translation.

As I have tried to show in the present lecture, the idea to translate scripture into a different language, and to read scripture in translation was anything but self-evident. Translation was around. Jews spoke many different languages, and translating the ancient Hebrew was a school exercise of choice. But such translation always remained in the shadow of the Hebrew text. To produce a full translation in writing, and to relate to the written version as scripture, was something else altogether. To translate is to betray, and Jews were wary of betraying the word their God had entrusted them with. As I have argued, it is no accident if the translation paradigm emerged in the periphery, where Jews had learnt over many generations to open up to other languages and cultures. In Egypt, the translation of scripture broke loose from the Hebrew source text and established itself as a new reference.

The results were as far-reaching as one could have feared. Even the successes of the Septuagint are often approximations only. And if the criterion of successful translation is faithfulness to the source, it must be admitted that much of the Septuagint is a failure. The Greek text diverges from the Hebrew in many details, but not only in details. The adaptation of the source text to Hellenistic thought imprints upon the Bible many ideas that go beyond the Hebrew, or are contrary to it. Until today, biblical scholars continue to struggle with the Hellenistic baggage that was imported into the interpretation of the Bible although it did not belong there originally.

In the end, however, the enterprise was justified. Translation is tricky, the results may be messy, but as I have argued translation is also necessary. More than that: translation and scripture belong together intrinsically. Source text and translation are two sides of the same coin: the oracles of God need to be interpreted in new situations that differ radically from the circumstances of their original revelation.


Barthélemy, Dominique Études d’histoire du texte de l’Ancien Testament, Fribourg, Academic Press/Göttingen, Vandenhoeck & Ruprecht, 1978.

Bickerman, E. Der Gott der Makkabäer. Untersuchungen über Sinn und Ursprung der makkabäischen Erhebung, Berlin, Schocken, 1937.

Cowley, A. E. Aramaic Papyri of the Fifth Century B.C., Oxford, Clarendon, 1923.

Dion, P.-E. “La religion des papyrus d’Éléphantine : un reflet du Juda d’avant l’exil,” in U. Hübner, E. A. Knauf, Kein Land für sich allein, Fribourg, Academic Press/Göttingen, Vandenhoeck & Ruprecht, 2002, p. 243-154.

Dupont-Sommer, A. “Les dieux et les hommes en l’île d’Éléphantine, près d’Assouan, au temps de l’empire des Perses,” Comptes-rendus des séances de l’Académie des Inscriptions et Belles-Lettres 122/4, 1978, p. 756-772.

De Prémare, A. L. “Coran et langue arabe : quelques reflexions,” in Béatrice Bakhouche, Philippe Le Moigne (dir.), « Dieu parle la langue des hommes ». Études sur la transmission des textes religieux (Ier millénaire), Lausanne, Zèbre, 2007 (Histoire du Texte Biblique 8), p. 93-100.

Fritsch, C. The Anti-Anthropomorphisms of the Greek Pentateuch. Princeton, Princeton University Press, 1943.

Joosten, J. “The Aramaic Background of the Seventy: Language, Culture and History,” BIOSCS 43, 2010, p. 53-72.

Joosten, J. “Le dieu Iaô et le tréfonds araméen des Septante,” in Mireille Loubet, Didier Pralon (dir.), Eukarpa. Études sur la Bible et ses exégètes en hommage à Gilles Dorival, Paris, Cerf, 2011, p. 115-124.

Joosten, J. “Hebrew thought and Greek thought in the Septuagint. Fifty Years after Barr’s Semantics,” in Richard A. Taylor, Craig E. Morrison (eds.), Reflections on Lexicography. Explorations in Ancient Syriac, Hebrew, and Greek Sources, Piscataway, Gorgias, 2014, p. 149-157.

  1. Joosten, Th. Legrand (eds.), The Targums in the Light of Traditions of the Second Temple Period, Leiden, Brill, 2014.

Porten, B. “Elephantine Aramaic contracts and the priestly literature,” in M. Brettler and M. Fishbane (eds.), Minhah le-Nahum: Biblical and Other Studies Presented to Nahum M. Sarna in Honour of his Seventieth Birthday, Sheffield, JSOT, 1993,  p. 257-271.

Schenker, A. “L’Écriture Sainte subsiste en plusieurs formes canoniques simultanées,” in L’interpretazione della Bibbia nella Chiesa, Città del Vaticano, 2001, p. 778-786.

Schattner-Rieser, U. “La bibliothèque araméenne de la communauté juive d’Eléphantine,”  Tsafon 56, 2009, p. 13-27.

Siegert, F. Zwischen hebräischer Bibel und Altem Testament. Eine Einführung in die Septuaginta, Münster, LIT, 2001.

Tov, E. “Reflections on the Septuagint with Special Attention Paid to the Post-Pentateuchal Translations,” in Wolfgang Kraus, Martin Karrer, Martin Meiser, Die Septuaginta – Texte, Theologien, Einflüsse, Tübingen, Mohr (Siebeck), 2010, p. 3-22.


Read about Jan Joosten on Wikipedia.