In 1530, Luther published his Open Letter on Translation.Footnote ¹ In it he responded to critics of his German Bible translation, focusing in particular on two passages. The first was Romans 3: 28, which in his German New Testament, published in 1522 (the so-called Septembertestament), Luther had translated: ‘So halten wir es nun, daß der Mensch gerecht werde ohne des Gesetzes Werke, allein durch den Glauben’ – ‘We hold, therefore, that a man is justified without the work of the law, by faith alone’. The second was his rendering of Luke 1: 28, the angel's greeting to Mary: ‘Gegrusset seystu holdselige, der herr ist mit dyr, du gebenedeyte vnter den weyben’ – ‘Greetings to you, sweet Mary, the Lord is with you, you who are blessed amongst women’. In his Open Letter, Luther vigorously defended his choice of the adjective holdselig, literally `blessed by grace' but in English probably best rendered `gracious', ‘sweet’ or ‘lovely’, to describe Mary in preference to the Vulgate's voll Gnaden, ‘full of grace’. He also laid out his reasons for introducing the word allein, ‘only’, into his translation of Romans 3: 28, insisting: ‘I knew very well that the word solum is not in the Greek or Latin text; the papists did not have to teach me that. It is a fact that these four letters s o l a are not there.’Footnote ² However, he assured his readers, the inclusion of allein reflected his desire to produce a fluent German translation, for German functioned differently from Latin. It was, Luther claimed, a matter of language rather than theology.

Is Luther's assertion to be believed? There can be no question but that his use of allein – allein durch Glauben (‘by faith alone’) – in translating Romans 3:28 also makes a theological point, emphasizing the doctrine of justification by faith that is fundamental to Luther's theology. Is Luther's claim in his Open Letter on Translation specious? This article considers this question firstly by placing Luther's translation of Romans 3: 28 in the context of other contemporary translations of other New Testament passages and secondly by comparing it to his rendering of Romans 3: 28 elsewhere. In doing so, it highlights the challenge posed by Luther's task of translating Scripture at a time when the meaning of Scripture was itself contested, and translation was not only a question of textual accuracy but also of defining orthodoxy and heresy. There can be no doubt that Luther was indeed concerned to produce a fluent and accessible German translation of the biblical text, but how did his theological priorities shape his definition of a good translation?

The questions raised by Luther in his Open Letter on Translation remain pertinent to translators today.Footnote ³ All translators are faced with a challenge, as Birgit Stolt points out: ‘How true to the original must one remain; how freely may one formulate things? The free, adaptive method, oriented toward the language of translation, stands over against the “alienating” method, oriented toward the original language, with its emphasis on remaining true to the words.’Footnote ⁴ James Arne Nestingen observes: ‘Translation takes place in two dimensions. The first is a linguistic exchange, roughly equivalent language being substituted for the original. The second is cultural, the new language inevitably giving that which is being translated another hue, colored with its own specific assumptions.’Footnote ⁵ For Stolt, the ‘truly remarkable’ aspect of Luther's method of translating lies in his ability to attend both to the original text and to the language into which he was translating.Footnote ⁶ It is widely recognized that it was this ability to render the words of the Bible into a language which seemed familiar to those who spoke it that distinguished Luther's translations from earlier German translations of the Bible, which had generally sought to stay closer to the language of the original text, and thus had resulted in less idiomatic German.Footnote ⁷ Luther's linguistic aim, as Antoine Berman has pointed out, was to compose a text which was ‘not Latin, not a pure dialect, but a generalized popular speech’.Footnote ⁸

What, however, was the text being translated? The wider context of early modern biblical translation, and of Luther's Open Letter itself, throws into stark relief the complexity of establishing what it means to ‘[remain] true to the original’. This is in part because the Bible presented (and continues to present) particular difficulties when it comes to establishing what is meant by ‘the original’.Footnote ⁹ The first challenge for Luther was that of defining the source text.Footnote ¹⁰ Luther's German Bible was a new departure, not only in his efforts to produce an idiomatic German text, but also in his commitment to offering a translation (at least in theoryFootnote ¹¹) from Greek and Hebrew rather than from the Latin of the Vulgate.Footnote ¹² This decision was not value-neutral, but had far-reaching theological consequences. By the time Luther began his theological career, the humanist insistence on the importance of studying works in their original language had already generated an awareness that some aspects of medieval theology and church practice drew their rationale from passages in the Vulgate which humanists had come to see as inaccurate translations of the Greek text. Lorenzo Valla, Erasmus and Jacques Lefèvre d’Étaples, amongst others, questioned the accuracy of the Vulgate's Latin translation and suggested revisions to it. Indeed, the format of Erasmus's 1516 Novum Instrumentum – which presented the Greek text and his own Latin translation on facing pages, followed by his annotations discussing the relationship between the two – was, as Paul Botley observes, ‘largely inspired by Erasmus’ decision to publish his own Latin translation of the New Testament’, and his realization that this ‘required an edition of the Greek text on which it was based’.Footnote ¹³ Luther's decision to base his own translation of the New Testament on the Greek text, which he regarded as conveying the ‘true meaning’ of Scripture, therefore made a theological as well as a linguistic statement. For Luther, theological authority was rooted in the principle of sola scriptura; and the scriptura which he wished to make known in German was that found in the original languages, from which, he believed, readers would learn true theology and a better understanding of what the Church should be.

For Luther, then, a translation ‘[remained] true to the original’ when it presented the theology which he believed to be proclaimed in the Greek New Testament and Hebrew Bible. However, from 1520 that theology had been defined to be heretical. Consequently, his translation embodied and articulated precisely the theological assumptions that his opponents were seeking to suppress, and thus raised issues not just of language but of power. Lynne Long observes that ‘sacred text translation’ is particularly prone to ‘promot[ing] contention between the users of the texts involved’,Footnote ¹⁴ since, as Lefevere asserts, ‘it is not only the authority of the text that requires validation, it is also the authority of those whose power resides in the text’.Footnote ¹⁵ In the Reformation context, the challenge to authority posed by vernacular Bible translations was explosive, as Richard Duerdon highlights in relation to early modern English translations of the New Testament: ‘all about and through these texts swirl the perils and promises of conviction – both kinds – and of ideology, authority, and power’,Footnote ¹⁶ for ‘if Tyndale is a heretic, no amount of philological ability will make the text acceptable; ideology and language form a single inter-text’.Footnote ¹⁷ David R. Glowacki argues that, for the authors of early modern English Bible prefaces, ‘the economic forces, the political forces, and the effort of the translators are ultimately sanctioned by God’.Footnote ¹⁸ This was not a new phenomenon – Hermann Schüssler has shown that the scriptural authority and ecclesiology were already intimately entwined in late medieval definitions of doctrineFootnote ¹⁹ – but there can be no doubt that Luther's German New Testament, and later Bible, formed a nexus for complex processes which aimed to control both the biblical text and its interpretation.Footnote ²⁰ Luther himself, however, makes no explicit claims to divine (or, indeed, secular) authority regarding his own translation; his stated intention was to mediate an understanding of the ‘real nature of the gospel’: ‘“gospel” [evangelion] is a Greek word and means in German good message, good tidings, good news, a good report, of which one sings and tells and about which one is glad’.Footnote ²¹ For Luther the content of this message was everything: people were to understand that ‘faith in Christ overcomes sin, death, and hell, and gives life, righteousness, and salvation’.Footnote ²² But how was that faith acquired?

The theological questions raised for translators and interpreters by the return to Greek and Hebrew texts are exemplified in two texts identified by the humanists as key: Matthew 4: 17, which Luther did not discuss in his Open Letter on Translation, and Luke 1: 28, which he did. In the NRSV, Matthew 4: 17 is rendered into English: ‘From that time Jesus began to proclaim, “Repent, for the kingdom of heaven has come near.”’ The key phrase is Jesus's exhortation, which in Greek reads: μετανοεῖτε, ἤγγικεν γὰρ ἡ βασιλεία τῶν οὐρανῶν. The Vulgate translation is paenitentiam agite adpropinquavit enim regnum caelorum, and the terminology paenitentiam agite (‘do penance’) helped to justify the practice of sacramental confession. Lorenzo Valla had recognized this as a problematic translation of the Greek term μετανοεῖτε, and Erasmus and Luther agreed.Footnote ²³ In English, from the time of Wyclif, μετανοεῖτε has been translated ‘repent’;Footnote ²⁴ however, neither German nor Latin has an equivalent to the English verb ‘repent’, and both Erasmus and Luther struggled to find a translation which did not carry the overtones of Jerome's paenitentiam agite. In the 1516 Novum Instrumentum, Erasmus opted for poeniteat vos; in 1519, the edition used by Luther for his 1522 Septembertestament, Erasmus tried resipiscite; in the 1522 edition he reverted to an amended version of the Vulgate: paenitentiam agite vitae prioris. Similarly, in 1522 Luther translated μετανοεῖτε with ‘Besser^t euch’ – ‘improve yourselves’ – but by 1534 he had opted for ‘Tut Buße’ – ‘do penance’, returning to the reading he had offered in 1517 in the first of the Ninety-Five Theses: ‘Dominus et magister noster Iesus Christus dicendo “Penitentiam agite &c.” omnem vitam fidelium penitentiam esse voluit’ (‘When our Lord and Master Jesus Christ said, “Do penance etc.”, he willed the entire life of believers to be one of repentance’).

The translation of Matthew 4: 17 demonstrates the complex interplay between theological meaning and language, but it also shows the way in which the constraints of the target language – in this case German – determine possible translations and meanings. A second problematic Vulgate translation, which Luther discussed at some length in his Sendbrief, was the angel's greeting to Mary at the annunciation (Luke 1: 28), which in the NRSV reads: ‘And he came to her and said, “Greetings, favoured one! The Lord is with you.”’ Here too, it was the spoken words which presented the translation challenge. In Greek, the angel's words read: Χαῖρε, κεχαριτωμένη, ὁ κύριος μετὰ σοῦ; this was translated by the Vulgate: Ave gratia plena Dominus tecum. Some form of this phrase was well known to many late medieval and early modern Christians as the opening of the Ave Maria, one of the texts they were expected to know by heart in Latin or the vernacular, or both.Footnote ²⁵ The problem, as humanist scholars identified it,Footnote ²⁶ was that the Latin gratia plena – ‘full of grace’—implied that grace was a measurable commodity, rather than describing the quality of the relationship between God and Mary, as the Greek κεχαριτωμένη seemed to do. Both Erasmus and Luther found solutions which they went on to use consistently. Erasmus in his Novum instrumentum translated the angel's greeting: Ave gratiosa, dominus tecum! Luther chose to use the term holdselig – ‘lovely’, ‘sweet’, ‘gracious’: Gegrusset seystu holdselige, der herr ist mit dyr.Footnote ²⁷

Luther's explanation for his decision centres on his view that κεχαριτωμένη expressed Mary's relationship to God:

Beneath the polemic, which witnesses to the conflicts in which Luther and his New Testament translation had by 1530 become deeply embroiled, Luther's linguistic point was that a literal translation of the Vulgate encouraged German-speakers to misunderstand the nature of grace. However, he also held that the Vulgate had misunderstood the Greek term, which he believed to be the translation of a Hebrew phrase.Footnote ²⁹ Luther looked for a model in angelic greetings in the Old Testament, and found one in the angel's greeting to Daniel as ‘greatly beloved’ (Daniel 10: 11, 19 NRSV):

Grace, as Luther had come to believe by 1519, is not a commodity to be bought or sold, but is manifested in ‘the righteousness of Christ my God which becomes ours through faith and by the grace and mercy of God’.Footnote ³¹ Here too, therefore, Luther's point was not only linguistic but deeply theological.

Reflecting on the hostile reception of Erasmus's translation of the opening of John's gospel in the 1519 edition of his Novum Testamentum (as it was now titled), which he rendered In principio erat sermo, rather than verbum, Long comments that ‘the overlay of tradition and exegesis’ had ‘crystallise[d] the text into an original to which any alteration was regarded with aggression and suspicion’.Footnote ³² Luther's rejection of voll gnaden, the German equivalent of the Vulgate's plena gratia, in favour of holdselige received a similarly negative response. And yet, when in 1523 Hieronymus Emser published a version of the German New Testament which attempted to render Luther's translation doctrinally acceptable to Catholics, he retained Luther's use of holdselige, even though, as he commented in a marginal note, he affirmed that Mary ‘is called by the angel full of grace’.Footnote ³³ Indeed, elsewhere Emser explicitly rejected Luther's interpretation:

Stolt argues that in the view of Emser and of others who objected to Luther's translation, ‘the church had established once and for all how this passage was to be interpreted, namely, in harmony with dogma and typology, and any questioning of this reading came close to blasphemy’.Footnote ³⁵ Emser wished to maintain this traditional theology, but his German translation nonetheless retained Luther's use of holdselige. A new or revised translation did not always give rise to a new theology.Footnote ³⁶

The Vulgate's translation of Matthew 4: 17 and of Luke 1: 28 had already been identified as problematic by humanist scholars. No such questions had been raised about Roman 3: 28, which in the Greek in Erasmus's 1516 Novum Instrumentum reads: λογιζόμεθα οὖν πίστει δικαιοῦσθαι ἄνθρωπον χωρὶς ἔργων νόμου.Footnote ³⁷ This was translated in the Vulgate as arbitramur enim iustificari hominem per fidem sine operibus legis. Erasmus introduced minor changes which reflected the word-order of the Byzantine codex on which his edition was based, emphasizing the place of faith: arbitramur igitur fide iustificari hominem absque operibus legis.Footnote ³⁸ Luther's 1522 translation, which remained unchanged in all subsequent editions, placed even stronger emphasis on the role of faith by introducing the word allein – ‘only’: So halten wir es nun, daß der Mensch gerecht werde ohne des Gesetzes Werke, allein durch den Glauben (‘a person is justified without the works of the law, through faith alone’).Footnote ³⁹

In his Sendbrief, Luther protested at the criticism being levelled at him by ‘these blockheads’: ‘I knew very well that the word solum is not in the Greek or Latin text; the papists did not have to teach me that. It is a fact that these four letters s o l a are not there.’Footnote ⁴⁰ Here too, he justified his use of allein on the basis of German usage:

The use of allein, he claimed, followed from his commitment to producing a German translation which was recognizably German and not Latin or Greek: ‘I wanted to speak German, not Latin or Greek, since it was German I had undertaken to speak in the translation. . . . For the literal Latin is a great hindrance to speaking good German.’Footnote ⁴² Consequently, he emphasized:

What ‘these blockheads’ had failed to understand, he complained, was that his use of allein ‘conveys the sense of the text; it belongs there if the translation is to be clear and vigorous’.Footnote ⁴⁴

The problem, as Waldtraut-Ingeborg Sauer-Geppert notes, is that this ‘apparently simple linguistic aid’ achieves ‘an exclusivity which the original text can have, but which it does not have to have’.Footnote ⁴⁵ Key here, therefore, is Luther's understanding of ‘the sense of the text’. By 1522, he had come to emphasize that justification was by faith alone – sola fide. Bluhm claims that Luther uses sola or allein in ‘a series’ of quotations of Romans 3: 28 before the publication of the Septembertestament, but that none predates the 1515 lectures; however, the first direct evidence he cites is from 1518, in Luther's ‘Sermon on the proper preparation for receiving the sacrament’: Apostolus . . . clamat omnes esse peccatores et sola iustificandos fide – ‘The apostle . . . proclaims that all are sinners and are justified only by faith’.Footnote ⁴⁶ Bluhm is correct that the conviction that justification is received solely by grace through faith was beginning to emerge in Luther's Lectures on Romans. He commented on Romans 1: 17: ‘Only in the Gospel is the righteousness of God revealed [in solo evangelio revelatur Iustitia Dei] . . . by faith alone [per solam fidem], by which the Word of God is believed.’Footnote ⁴⁷ However, the Lectures on Romans did not yet include sola in the discussion of Romans 3: 28. Luther's gloss explained: ‘that a man is justified, reckoned righteous before God, whether Greek or Jew, by faith, apart from works of the Law, without the help and necessity of the works of the Law’.Footnote ⁴⁸ In the Schola, he distinguished between ‘works of the law’ and ‘works of grace’:

Luther did not refer explicitly to Romans 3 either in the 1516 disputation on the extent of human power and will without grace, or in his 1518 sermon on indulgences and grace.Footnote ⁵⁰ Although Luther's Lectures on Romans show that Luther was wrestling with the relationship between justification and grace, Romans 3: 28 had not yet emerged as a foundational text for his theology.

Luther's first explicit statement of the doctrine of justification by faith alone arguably occurred in the Heidelberg Disputation, in thesis 25: ‘He is not righteous who does much, but he who, without work, believes much in Christ [non ille iustus est, qui multum operator, sed qui sine opere multum credit in Christum].’ Luther explained: ‘I wish to have the words “without work” understood in the following manner: Not that the righteous person does nothing, but that his works do not make him righteous, rather that his righteousness creates works. For grace and faith are infused without our works. After they have been imparted the works follow.’Footnote ⁵¹ Here he cites Romans 3: 28 to substantiate this point, reproducing the Vulgate text (i.e. not that found in Erasmus's Novum Instrumentum, which he had probably not yet read): arbitramur enim iustificari hominem per fidem sine operibus legis – ‘for we hold that man is justified by faith apart from works of law’.Footnote ⁵² By the time he came to dispute with Cajetan in Augsburg in autumn 1518 Luther was relating the language of sola fide more explicitly to justification: ‘Through no attitude on your part will you become worthy, through no works will you be prepared for the sacrament, but through faith alone, for only faith in the word of Christ justifies, makes a person alive, worthy, and well prepared [sola fides verbi Christi iustificat, vivificat, dignificat, praeparet].’Footnote ⁵³ Two years later, in The Freedom of a Christian, he similarly asserted: ‘it is clear that, as the soul needs only the Word of God for its life and righteousness, so it is justified by faith alone and not any works [ita sola fide et nullis operibus iustificantur]’.Footnote ⁵⁴ However, Romans 3: 28 is not explicitly cited in either case, although his words in The Freedom of a Christian could imply that he had this verse in mind.

The first instance of the use of allein explicitly associated with Romans 3: 28 occurs in a sermon preached at Epiphany 1521, in which Luther affirmed ‘that we do not become godly [fromm] through our own works, but only through faith in Christ, as Paul says to the Romans in the third and to the Galatians in the second chapter’.Footnote ⁵⁵ The Kirchen- or Weihnachtspostille, completed while Luther was at the Wartburg in 1521 and published in 1522, cites Paul, presumably implying Romans 3: 28, in the same terms that Luther would use in the Septembertestament: ‘so you see here how St Paul teaches in all places that justification does not come through works, but only from faith without any works [das die rechtfertigung nit durch werck, ßondernn alleyn auß dem glawben ohn alle werck kome]’.Footnote ⁵⁶

It is apparent, then, that, as Bluhm points out, ‘Luther the interpreter preceded Luther the translator’.Footnote ⁵⁷ By 1522 Luther had come to believe that the true meaning of Romans 3: 28 was that justification occurred not by works, but by faith alone. His inclusion of the word allein was intended to make this meaning – which for Luther was the true meaning of the biblical text – clear to the German reader. It was probably also intended to counterbalance the only use of the phrase ‘faith alone’ – in this case fide tantum – in the Vulgate, which occurs in James 2: 24 and inconveniently contradicts Luther's theology: ‘ex operibus justificatur homo et non ex fide tantum’ [‘a person is justified by works and not by faith alone’], or, in Luther's translation, ‘das der mensch durch die werck gerecht wird / nicht durch den glauben alleine’.Footnote ⁵⁸ In his preface to the epistle, Luther complained that James contradicted the theology presented in Romans, which, he asserted, clearly taught ‘that Abraham was justified without works, through his faith alone’.Footnote ⁵⁹ Romans, he insisted, was ‘the chief part of the New Testament, . . . truly the purest gospel’,Footnote ⁶⁰ and within that epistle, 3: 19–28 – or possibly 3: 23–4 – was

Luther's translation of Romans 3: 28 supported – or, perhaps better, emphasized – this reading, seeking to employ the German language so as to ensure that the passage would be read with the meaning he believed it was meant to have.

By the time Luther began lecturing on Genesis in 1535, he was deeply aware that his theological position had come to be characterized by the phrase sola fide. Affirming, quite explicitly, that ‘God wants to teach us that we are saved by grace alone or by faith alone’,Footnote ⁶² he offered a spirited defence against the criticisms of those who ‘call us “solafideists,” because we attribute righteousness to faith alone’ [nos vocant solarios, propterea quod soli fidei tribuimus iustitiam].Footnote ⁶³ They were, he thought, wrong, and the German Bible should tell them so. As Beutel has observed, Luther's Bible translation was intended ‘to renew, not the letter of Scripture, but its spirit’.Footnote ⁶⁴ That is, the language of the translation was intended to claim scriptural authority for his own interpretation of Scripture. Moreover, whilst the meaning of a text expounded in a sermon was explicated by the preacher,Footnote ⁶⁵ the Bible translation had to speak for itself.

Long suggests of sixteenth-century Bible translation that ‘a vernacular version made the text available to the literate Christian without the intervention of the priest. The intervention of the translator was not considered to be intrusive.’Footnote ⁶⁶ She is certainly right that Luther's intention was to provide direct access to Scripture. He wanted, as he put it in To the Christian Nobility of the German Nation, to give to all people ‘the power to test and judge what is right or wrong in matters of faith, . . . to become bold and free on the authority of all these texts, and many others, . . . and test all that [the Romanists] do, or leave undone, by our believing understanding of the Scriptures’.Footnote ⁶⁷ However, the translator was not neutral in this process. Luther's Bible translation was intended, as Volker Leppin concludes, to offer ‘not so much a popularisation, but an authoritative meaning of the Bible’.Footnote ⁶⁸ Consequently, in some sense, as Alec Ryrie argues, for the reader of Luther's Bible translation, ‘[s]ola fide is logically and chronologically prior to sola scriptura.’Footnote ⁶⁹ It was the recognition that justification came about sola fide, which had so profoundly informed, and fundamentally changed, his own experience of God, that Luther intended his translation to communicate to its readers.Footnote ⁷⁰

Luther's theology was, therefore, key to determining the shape of his translation of the Bible, since it defined the ‘true’ meaning of the text which he wished to articulate in German. Stolt argues that his theology determines his translation technique, and ‘dictate[s] his decision when to remain true to the text and when to translate more freely’.Footnote ⁷¹ Luther did not believe that complete mastery of the interpretation of a text was possible, but he did think that ‘the right text’ would lead to a better understanding,Footnote ⁷² and also asserted that ‘no false Christian or trouble-maker can faithfully translate [the Scriptures]’.Footnote ⁷³ His achievement, as Berman has put it, was ‘to create a work accessible to the German people, capable of providing a solid base for the religious sensibility of the Reformation’.Footnote ⁷⁴ That ‘religious sensibility’ was significantly different to that of the late medieval Church, and Luther's translation of the New Testament both reflected and helped to define that difference. Lawrence Venuti observes that any translation ‘creates values in social formations at specific historical moments’; consequently, ‘retranslations reflect changes in the values and institutions of the translating culture, but they can also produce such changes by inspiring new ways of reading and appreciating the source texts’.Footnote ⁷⁵ This was precisely Luther's intention. His translation – itself a retranslation – was intended to purvey a particular understanding of the central message of the gospel, with the expectation that those who read it would amend their faith, and with it their religious beliefs and their moral and ethical behaviour. To this extent, then, Long is wrong to suggest that ‘the translator was not considered to be intrusive’.Footnote ⁷⁶ The controversies that arose around Luther's translation – and indeed other translations of this periodFootnote ⁷⁷ – indicate that Luther's contemporaries were very aware that the translator was creating a text which was intended to guide its readers to, and support them in, a particular theological position. Luther's translation of the New Testament illustrates vividly the ways in which the translator was – and is – not neutral, and it shows that, and illustrates how, theology and language are intimately entwined.