Translating from Greek as Source Language? The Lasting Influence of Latin on New Testament Translation

Abstract

Translational choices in New Testament translation appear to be influenced far more strongly by the Latin tradition and Martin Luther’s towering translation than hitherto acknowledged. This contribution uses examples from the synoptic gospels to trace the influence of Martin Luther, the Vulgate, Erasmus and the Old Latin version of the New Testament in current dictionaries like the Bauer/Aland and BDAG.

Introduction

It almost goes without saying that the aim of NT translations is to translate the authoritative Greek source text into modern target languages correctly and faithfully. This aim is supported by a host of linguistic tools dealing exclusively with the Greek of the NT, designed as they are to foster a correct understanding of the Greek text. Thus, the field of NT Greek has seen more attention than any other extant Greek text or text corpus from antiquity. This abundance of research and the high number of publications connected with it has made it viable to write a history of NT lexicography (Lee 2003). Thus, the suggestion to be put forward here, that these dictionaries and their companions – the grammars of the NT – may support translational choices which are not after all fully justified by the Greek text of certain passages of the NT, challenges a firm conviction which is now close to 500 years old. Luther’s so called ‘Septembertestament’, published in 1522, is the first NT in Northern Europe to translate from Greek as the source text. There is a widespread conviction that he was able to do this ‘faithfully’ ¹ and his translation thus became a model. Older translations into the various vernaculars of medieval Europe, by contrast, used the Latin text of the Vulgate as the source text. Martin Luther’s translation (or at least its proclaimed method) ² represents the direction which has been taken ever since: in order to base their translations on a reliable source text, scholarly translations of the NT use the critical editions of the Greek NT, which are generally recognized as the authoritative sources. The critical edition of the Greek NT in the version of the venerable Nestle-Aland has by now reached its 28th edition, which in itself shows how much and how long lasting research has been spent on the text of the NT, its transmission and its linguistic peculiarities.

In light of this monumental research, it may seem a daring suggestion that Latin and also Martin Luther’s translational choices (some of which disagree with both the Vulgate and the Greek text) are still influencing modern translations of the NT. This suggestion contains two implicit claims: (1) the Latin version is a translation which can be shown to be less faithful to its source text than hitherto taken for granted. However, Latin influence comprises influential Humanists like Erasmus who wrote in Latin. (2) Martin Luther took liberties with the text in translating it, which need to be understood better. In both cases, it would appear that some of the translational choices introduce or intensify anti-Judaism in translation.

In order to validate these claims the following steps will be taken. Step 1: The history of NT lexicography indicates that NT lexicography is still deeply rooted in Latin semantics, even if these dictionaries seemingly bypass Latin. Indeed, modern NT dictionaries offer Greek words from the NT and glosses in modern languages. The semantics of these glosses, however, depends in numerous cases on a Latin understanding of the Greek words. Step 2: Martin Luther translated virtuously – but sometimes freely – into German. The first example will show how a translational choice of Martin Luther found its way into a Greek-Latin dictionary in the nineteenth century and from there into modern dictionaries. Step 3: Erasmus and his Latin translation are also still of influence concerning the semantics of certain translational equivalents in current NT dictionaries. Step 4: The influence of the Vulgate is acknowledged by the introduction to Grimm’s dictionary. The use of the Vulgate has consequences for the perception of the semantics if compared to the semantics of the Greek word. Step 5: This step will introduce evidence that the Old Latin version has also been used to understand the semantics of a Greek word in modern NT lexicography. Step 6: These translational choices are of exegetical importance. The question of paying taxes in the version of Luke (Lk. 20[19].20-26) is used to demonstrate the potential for transforming the understanding of a passage in the case that the translation is revised according to Greek semantics. Step 7: A summary will conclude this contribution by outlining possible research perspectives as a consequence of the research results presented here.

1. New Testament Lexicography and Latin Semantics

Whilst anti-Judaism is acknowledged as a problem of NT interpretation, the possible distortion of the text via translational choices has in most instances escaped detection. A possible explanation for this oversight might be that the lexical equivalents in one of the most important tools of NT philology – the famous Bauer/Aland and its English translation BDAG – are deeply rooted either in Latin semantics or in Martin Luther’s translational choices. At the same time, Luther and other ground-breaking and authoritative translations like the King James Version (KJV) or Tyndale translations were inevitably deeply influenced by Latin as the then-prevalent language of academic discourse. The dictionaries available at the time were Greek-Latin dictionaries which drew heavily on the translational equivalents of the Vulgate. Thus, semantics of Latin equivalents would and could not be understood as differing from Greek semantics.

The way of Latin into modern tools of NT studies is actually straightforward: Walter Bauer’s first edition of the Wörterbuch was published in 1928. As to its semantics, it strongly relies on Preuschen’s dictionary (1910). The title of Bauer’s dictionary makes this dependence clear (Bauer 1928). Bauer mentions in his preface that he used (among others) Wilke-Grimm’s Clavis Novi Testamenti philologica (3rd edn, 1888) and its English rendering (Thayer 1886). The evidence adduced below will demonstrate that he made far greater use of these sources than hitherto appreciated. Bauer in fact relied heavily on Wilke-Grimm. In turn, Wilke-Grimm acknowledged in its preface its strong dependence on the Vulgate: Omnino latinae translationis, quae Vulgata dicitur, longe saepius rationem habui quam qui ante me lexica N. Ti adornaverunt (Grimm 1868: VIII), ‘I have made use of the Latin translation, commonly called the Vulgate, far more often than those who made lexicons of the New Testament before me’. Among the other research tools mentioned in Bauer’s preface, the lexical equivalents in Zorell’s Novi Testamenti Lexicon Graecum (1911) are also deeply rooted in Latin tradition. In his preface, Preuschen also acknowledges his dependence on Grimm and Thayer. ³ Preuschen is therefore the most important source for the lexical equivalents exhibited in Bauer and thus introduces Latin interpretation into Bauer. As a result, these possibly interpretative lexical equivalents also found their way into the English-language BDAG: Frederick William Danker’s 3rd edition of the Lexicon of the New Testament (first published by Arndt and Gingrich in 1957) was published in 2000 (Danker 2000). Arndt and Gingrich translated Walter Bauer’s 4th edition. A thorough and systematic revision of the semantics implied in the lexical equivalents did not take place. This even holds true also for the ‘6th and fully revised edition’ of Bauer, published by Kurt Aland and Barbara Aland. The editors of the ‘fundamentally revised’ Bauer/Aland state that a revision of the lexical choices ‘would have produced a totally new dictionary and would not have been any more “the Bauer”’. (cf. Bauer 1988: vii).

In other words, the lexical equivalents and interpretative comments offered by Walter Bauer have never come under systematic and rigorous scrutiny despite the number of editions and revisions. Thus, the newest revision of Bauer/Aland is – as far as the semantics of the lexical equivalents is concerned – still just ‘the Bauer’ of 1928. ⁴ As we have seen, following the chain further backwards via Preuschen and Wilke/Grimm, these dictionaries therefore implicitly introduce Latin translational choices of the Vulgate as seemingly authoritative interpretation of the NT. At the same time, the dictionaries by Bauer/Aland and Danker (also known in its early printing as BAGD [i.e. Bauer/Arndt/Gingrich/ Danker] and later as BDAG [i.e. Bauer/Danker/Arndt/Gingrich]) are highly authoritative and trusted tools and are employed as standard dictionaries of NT studies. John A.L. Lee describes the problem of such trusted tools aptly: ‘Yet this trust is misplaced. The concise, seemingly authoritative statement of meaning can, and often does, conceal many sins – indecision, compromise, imperfect knowledge, guesswork, and, above all, dependence on predecessors’ (Lee 2004: 66). In the case of Bauer/Aland and BDAG, this dependence on predecessors embeds the Vulgate and also Martin Luther as interpretative tools for the Greek vocabulary of the NT. The following examples will show how rigorous attention to Greek semantics may lead to change in the understanding of certain passages, among them some containing anti-Jewish wording.

2. Martin Luther’s Translational Choice in Matthew 12.14

The first example will discuss the meaning of Mt. 12.14. Common translational practice introduces a ‘murderous’ or at least ‘harmful’ intent of the Pharisees in Mt. 12.14. ⁵ This example shows Martin Luther’s lasting influence on NT dictionaries. Matthew 12.14 reads: ἐξελθόντες δὲ οἱ Φαρισαῖοι συμβούλιον ἔλαβον κατ᾿ αὐτοῦ ὅπως αὐτὸν ἀπολέσωσιν. The Good News Translation (GNT) offers: ‘Then the Pharisees left and made plans to kill Jesus’. The New American Bible (Revised Edition; NABRE ⁶ ) renders the text thus: ‘But the Pharisees went out and took counsel against him to put him to death’. Translated thus, there is apparent intent and premeditation to kill Jesus. As conscious planning is involved, one could argue that this was a conspiracy to murder Jesus. Such a conspiracy to murder is all the more troubling, since it takes place among the Pharisees. The Pharisees have trained legal minds, and their training concerns what NT texts usually call ὁ νόμος, ‘the Law’, denoting the Torah. It would be immediately apparent that such a conspiracy to murder is not in line with the divine commandments as contained in the Torah. It is apparent that earlier translations from the Latin do not introduce the direct intent to kill. One could adduce, for instance, the Douay-Rheims translation, which is an English translation from 1582 aimed at giving a faithful rendering of the Vulgate: ‘And the Pharisees going forth made a consultation against him, how they might destroy him’. It appears that a possibly judgmental translational choice has been based on the semantics of perdere. The verb perdere is translated here as ‘to destroy’, whereas the very same word is translated in Lk. 15.4 as ‘to lose’ (Douay Rheims 1582): ‘What man of you having an hundred sheep: and if he hath lost one of them, doth he not leave the ninety-nine in the desert, and goeth after that which was lost, until he find it?” ⁷ It appears that the semantics of perdere allow for both ‘to lose’ and ‘to destroy’. Consequently, one is tempted to translate the Latin rendering of Mt. 12.14 as follows (adapted from Douay Rheims): ‘And the Pharisees going forth made a consultation against him, how they might lose him’. There are many ways to lose a person. Most of them do not mean ‘to destroy’ or ‘to kill’. The semantics of this translation are less negative if compared to the original Douay Rheims version of this verse. Since the semantics of the verb perdere include ‘to lose’, it is necessary to justify why negative semantics are chosen in the context of the Pharisees acting against Jesus.

Before a semantic analysis of the word ἀπόλλυμι raises questions of the philologically correct understanding of this word, a short excursus will describe here how the lexical equivalent ‘to kill’ (which is different from ‘to destroy’) found its way into NT dictionaries. It starts with Martin Luther. He renders Mt. 12.14 (Martin Luther Last Hand/1545): ‘Da giengen die Phariseer hin aus / vnd hielten einen Rat vber jn / wie sie jn vmbbrechten’. This translates into English thus: ‘And the Pharisees going forth made a consultation concerning him, how they might kill him’. The semantics of ‘umbringen’ definitely concern a murderous intent. The text of the Vulgate has: exeuntes autem Pharisaei consilium faciebant adversus eum quomodo eum perderent. It appears that the Latin verb perdo does not necessarily mean ‘to kill’. According to Bergren’s index (Bergren 2018: s.v.), in the Vulgate of the NT perdere is a translational choice for either ἀποκτείνω or ἀπόλλυμι. This seems to support Martin Luther’s translational choice. Schmoller’s concordance (Schmoller 1989) (which constitutes the basis for Bergren) indicates that all instances for the active voice of ἀπόλλυμι are translated with perdere, but only one instance of the multiple occurrences of ἀποκτείνω is rendered as perdere. The verb ἀποκτείνω is commonly translated with interficere. The passive voice of ἀπόλλυμι is translated with perire and deperire. According to Bergren, perire is also used to translate φθείρω. ⁸ Thus, the Vulgate of the NT appears to confirm that ἀπόλλυμι is not rendered as ‘to kill’. Perdere is not a terminus technicus for ‘killing’. The only instance where the Vulgate uses perdere as translational equivalent for ἀποκτείνω is a context where one is tempted to supply such a translational equivalent (Mk 3.4b): ἔξεστιν τοῖς σάββασιν ἀγαθὸν ποιῆσαι ἢ κακοποιῆσαι, ψυχὴν σῶσαι ἢ ἀποκτεῖναι; This is translated as: licet sabbatis bene facere an male animam salvam facere an perdere. One could argue – and this seems actually plausible – that the synoptic parallel in Lk. 6.9 might have influenced the translational equivalent of ἀποκτείνω in this context: ἐπερωτῶ ὑμᾶς εἰ ἔξεστιν τῷ σαββάτῳ ἀγαθοποιῆσαι ἢ κακοποιῆσαι, ψυχὴν σῶσαι ἢ ἀπολέσαι; This is translated as: interrogo vos si licet sabbato bene facere an male animam salvam facere an perdere. It appears plausible that the translational equivalent of ἀποκτείνω in Mk 3.4 might be influenced by the translational equivalent of ἀπόλλυμι in Lk. 6.9: as is apparent, both are rendered with perdere. Consequently, the only definite attestation of perdere meaning ‘to kill’ is an instance where the synoptic gospels might have influenced each other. Thus, from a philological perspective, the example is weak and questionable.

The next step is now to trace how ‘to kill’ might have become a ‘correct’ meaning of ἀπόλλυμι. Luther is using ‘umbringen’ as a translational choice. His lexical tools in translating were either Aleander’s (1512) or Curio’s (1519) dictionary. ⁹ Aleander (1512: 46) and Curio (1519: 21v) offer as lexical equivalents for ἀπόλλυμι: pereo, perdo, destruo. Thus, Luther’s translation is not supported by either the Vulgate or the equivalents given in his lexical tools. Consequently, to translate ἀπόλλυμι as ‘kill’ in Mt. 12.14 is a translational choice. It is apparent that Luther’s translational choice for ἀπόλλυμι (‘umbringen’) is not a faithful rendering of the Greek text, but rather an interpretative translational choice possibly introducing anti-Judaism in translation. It is with Luther that ‘umbringen’/‘to kill’ entered the perceived semantics of the Greek word ἀπόλλυμι.

It appears that Grimm’s dictionary did not only use the Vulgate (as claimed in the introduction) but also Luther’s translational choices to offer lexical equivalents. Grimm offers interficio for ἀπόλλυμι at Mt. 2.13; 12.14; Mk 9.22; 11.18; Jn 10.10 (suggesting that there are even more instances for this meaning of ἀπόλλυμι). It does not come as a surprise that Luther (1545) offers in all these instances ‘umbringen’ as his translational choice. This is clear evidence that Grimm’s lexical choice has been influenced in this instance by Luther. Preuschen offers the lexical equivalent ‘umbringen’ for Mt 2.13; 12.14; 27.20 (Luther in these instances: ‘umbringen’); Mk 3.6 (Luther also: ‘umbringen’) etc. These verses not quoted in Grimm are further examples of Martin Luther using the German equivalent of the verb ‘to kill’ to translate ἀπόλλυμι. It appears that Bauer decided against Zorell, who does not offer interficio for the active voice of ἀπόλλυμι, and notes ‘umbringen’, offering the quotes as given in Preuschen. Bauer/Aland offer ‘umbringen’ too, and BDAG gives ‘kill, put to death’ as appropriate meaning for these instances. In light of the fact that even the Vulgate refrains from translating ἀπόλλυμι with interficio, one has to conclude that ‘to kill’ represents a translational choice that risks misrepresenting the Greek text and that this has been introduced by Martin Luther.

The semantics of ἀπόλλυμι appear to be similar to those of ἐκβάλλω. In Deut. 33.27 LXXb the following is stated: καὶ ἐκβαλεῖ ἀπὸ προσώπου σου ἐχθρὸν λέγων· Ἀπόλοιο. The New English Translation of the Septuagint (NETS) offers for this: ‘And he will drive out the enemy before you, saying, “May you perish!”’ One is tempted to suggest an alternative translation, one which might be closer to the semantics of the Greek source text: ‘And he will expel the enemy before you, saying, “Get lost!”’ This is in accordance with Eccl. 3.6: καιρὸς τοῦ ζητῆσαι καὶ καιρὸς τοῦ ἀπολέσαι, καιρὸς τοῦ φυλάξαι καὶ καιρὸς τοῦ ἐκβαλεῖν. NETS suggests: ‘a right time to seek, and a right time to lose; a right time to keep, and a right time to throw away’. The impression that this may be a better and less judgmental understanding of the semantics of ἀπόλλυμι is strengthened by Jn 6.37 and 6.39. John 6.37 has: πᾶν ὃ δίδωσίν μοι ὁ πατὴρ πρὸς ἐμὲ ἥξει, καὶ τὸν ἐρχόμενον πρὸς ἐμὲ οὐ μὴ ἐκβάλω ἔξω. And John 6.39 reads: τοῦτο δέ ἐστιν τὸ θέλημα τοῦ πέμψαντός με, ἵνα πᾶν ὃ δέδωκέν μοι μὴ ἀπολέσω ἐξ αὐτοῦ, ἀλλ᾿ ἀναστήσω αὐτὸ [ἐν] τῇ ἐσχάτῃ ἡμέρᾳ. The parallels of the phrases in these two verses are apparent. It does seem to overstretch the semantics of ἀπόλλυμι to translate it with ‘to kill’. In light of these results, one is tempted to translate Mt. 12.14 thus: ‘Then the Pharisees went out and held council against him, how they might get rid of him’. This translation seems to be closer to Greek semantics, but it appears also less narrowly homicidal which – in light of the use of the motive of a Christocide by ‘the Jews’ or ‘the Pharisees’ in the history of Christian anti-Judaism – would seem to be an opportunity to correct a historically problematic translation. The result of the analysis of ἀπόλλυμι in Mt. 12.14 is that both Bauer/Aland and BDAG follow the translational semantics as introduced by Martin Luther and that Luther’s understanding of ἀπόλλυμι appears in this case to be outside the semantic boundaries of the Greek word as used in the Septuagint and the NT. ¹⁰

3. Erasmus’s Influence: The Word πανουργία in Luke 20.23

The motive that the unlawful killing of Jesus Christ was premeditated and without justification is part of European cultural heritage. Keith W. Blinn comments on the difference between murder and manslaughter: ‘The distinguishing characteristic between murder and manslaughter was the existence of malice aforethought in the former and its absence in the latter’ (Blinn 1950: 729). Our philological discussion of the semantics of πανουργία will examine the question of the intent of the Jewish authorities concerning Jesus and his death.

The pericope of the question of paying taxes to the emperor is attested in Lk. 20.20-26 and also in the other two synoptic gospels (Mt. 22.15-20; Mk 12.13-17). It seems to indicate malicious intent on the part of the Jewish authorities. The Greek text of Lk. 20.23 reads: κατανοήσας δὲ αὐτῶν τὴν πανουργίαν εἶπεν πρὸς αὐτούς. This is translated in KJV (King James Version): ‘But he perceived their craftiness, and said unto them, Why tempt ye me?’ The Oxford English Dictionary ¹¹ defines ‘craftiness’ as ‘artfulness in deceiving or overreaching’. Consequently, the concept of craftiness implies the willingness to use unethical means to achieve a goal. This translation seems to imply a malicious intent on the part of the Jewish authorities in asking their question. The ultimate goal of the Jewish authorities is to hand Jesus over to the ‘administrative authority of the governor’ (Lk. 20.20c: παραδοῦναι αὐτὸν τῇ ἀρχῇ καὶ τῇ ἐξουσίᾳ τοῦ ἡγεμόνος). ¹² On the face of it, it appears as if the Jewish authorities plan judicial murder. However, the semantics of πανουργία does not make this the only possible translational choice.

The first step is to trace this translational choice. The tools of trade in use at the beginning of the sixteenth century offer the following lexical equivalents for πανουργία: Valentin Curio (1512: 309) and Hieronymus Aleander (1519: 112v ) both suggest astutia and calliditas. At least astutia is a translational choice directly taken from the Vulgate (1 Cor. 3.19) and calliditas derives from the translation of πανοῦργος as callidus. Both words may be understood either in a positive or in a negative way. The Vulgate text of Lk. 20.23 uses another word as contained in this passage: considerans autem dolum illorum dixit ad eos quid me temptatis. In Roman law the term dolus malus (!) is used to describe negative or evil intent, consequently dolus in itself describes rational and intentional behaviour (cf. Baier 2013a: 1758). If one compares Erasmus’s version of Lk. 20.23 with the Vulgate, one sees that he replaced the translational equivalent offered in the Vulgate (dolus) with versutia. The notion of deceit and falsehood appears to be far stronger in versutia than in dolus. Thus, the translational equivalent dolus is non-judgmental, while versutia is definitely judgmental. It describes ‘craftiness’ or ‘trickery’ (Baier 2013b: 4985). It thus appears probable that Erasmus influenced Bible translations in this case. Grimm (1868: 325) offers the lexical equivalents vafritia and astutia (vafritia is far more negative than astutia) for Lk. 20.23. He mentions, however, that this word may have a positive meaning in the Septuagint – which should be seen as a strong caveat against the traditional lexical equivalents used in Lk. 20.23. This caveat is lost in Preuschen (1910: 854), who offers ‘Tücke, Verschlagenheit, Schlauheit’ as possible lexical equivalents to be used in Lk. 20.23. He mentions that πανουργία translates Hebrew òÈøÀîÈä but does not give German lexical equivalents here. Thus, only those with extensive knowledge of Hebrew can understand that there might also be something positive in the semantics of the word. Zorrel (1931: 978) supports a rather negative understanding of the term, offering calliditas, astutia and vafritia as lexical equivalents. Calliditas and astutia might support a positive understanding of the word. This is cancelled by adding vafritia and not explicitly mentioning any positive semantic aspects this word may have. It is interesting to note that Thayer (1889: 476) still offers the positive meaning of this lexeme in the Septuagint, giving ‘prudence, skill in undertaking and carrying out affairs’ as possible lexical equivalents. Bauer (1928: 969) claims that this word is exclusively used sensu malo in the NT. Bauer/Aland (1988: 1230) merely keep this interpretation. Danker (2000: 754) offers for πανουργία the translational glosses ‘cunning, craftiness, trickery’, following Bauer’s and Bauer/Aland’s decision to see this as a thoroughly negative term. The potential positive meaning still to be found in Grimm thus evaporated from the lexical tools of the NT.

The Greek word might actually describe something very positive in the Septuagint. The father will use proverbs to teach his innocent children πανουργία for them to survive in this world and to become righteous and God-pleasing persons. ¹³ There is no Jewish father who would read the Septuagint in antiquity and intentionally wish his children to become scheming and crafty deceivers. In light of this use of πανουργία, one is tempted to see its use in Lk. 20.23 as a possible positive acknowledgment that the question posed is a difficult and challenging problem which Jesus masters with an insightful answer. Such an understanding is precluded if Erasmus has his way. NABRE’s translation of Lk. 20.23 (κατανοήσας δὲ αὐτῶν τὴν πανουργίαν εἶπεν πρὸς αὐτούς) is closer to Erasmus than to the Greek text: ‘Recognizing their craftiness he said to them’. One may therefore conclude that not only the textus receptus but also the Latin version of the NT offered by Erasmus ‘is characterized by the errors of Erasmus’ (Borger 2004: 43). This translation was used by many theologians in the sixteenth century. Martin Luther and the other translators of the NT would have been hard pressed to understand πανουργία as something other than ‘malicious intent’.

Concerning the Sitz im Leben of this pericope, one might turn to Jewish educational practice as may be seen in rabbinical Judaism: it was common for learned persons of ‘the Law’ to ask each other difficult questions. ¹⁴ This was something of a touchstone by which to judge the legal knowledge of a rabbi. Consequently, a more positive translation of πανουργία might place this discourse between representatives of Jewish authorities and Jesus, right in the middle of Jewish culture: both parties ask each other difficult questions and – in the case of the question of paying taxes – there is an underlying strategic thinking on the part of the Jewish authorities. They are in danger of losing their grip on the situation (cf. Lk. 20.6: they see themselves in danger of being stoned). To have the Roman occupying power help them appears to be an additional bonus, in the event that Jesus answers as they hope. Suffice to mention that πανουργία could also be used to describe the strategic thinking of military leaders (cf. Montanari 2015: 1535). Thus, it seems to be possible to suggest as a translation for Lk. 20.23 the following less judgmental way of phrasing: ‘But understanding their strategy he said to them’. In conclusion, Erasmus’s interpretation appears to have influenced NT dictionaries.

4. The Vulgate’s Influence: The Word πονηρία in Matthew 22.18

The question of possible malicious intent is touched upon also by the synoptic parallel to the question of paying taxes in Mt. 22. It appears that this is a passage where one might be willing to agree that malicious intent is to be understood as the meaning of the text. The word πονηρία in Mt. 22.18 seems to be above suspicion. Matthew 22.18 has: γνοὺς δὲ ὁ Ἰησοῦς τὴν πονηρίαν αὐτῶν εἶπεν· τί με πειράζετε, ὑποκριταί; The Vulgate offers here: cognita autem Iesus nequitia eorum ait quid me temptatis hypocritae. KJV has: ‘But Jesus perceived their wickedness, and said, Why tempt ye me, ye hypocrites?’ Evidently, KJV uses ‘wickedness’ as the translation for πονηρία. ¹⁵

The lexical equivalents in Aleander (1512: 342) and Curio (1519: 123r) are nequitia and pravitas. Both Latin equivalents have negative semantics. This can be observed also in Grimm (1868: 364), offering pravitas, improbitas, malignitas, malitia and explicitly mentioning Mt. 22.18 as one of the instances where these lexical equivalents can be used. Preuschen (1910: 945) offers ‘Schlechtigkeit’ and mentions with explicit reference to Mt. 22.18 that this concerns the ‘Gesinnung’. Thus, Preuschen supports an understanding which would lead a legal evaluation of this passage to the conclusion that ‘malicious intent’ is the motivating factor for the Jewish authorities in their way of dealing with Jesus. Zorell (1931: 1103) offers malitia and pravitas. Bauer (1928: 1107) offers that this concerns ‘den sittlichen Sinn’. The lexical equivalents given there (‘Schlechtigkeit’, ‘Bosheit’, ‘Gemeinheit’, ‘Sündhaftigkeit’) have negative semantics. It is worth noting that also ‘Sündhaftigkeit’ (i.e. ‘sinfulness’) is mentioned as a possible lexical equivalent by Bauer. Danker (2000: 851) follows Bauer when he offers: ‘in our lit. only in the ethical sense: state or condition of a lack of moral or social values, wickedness, baseness, maliciousness, sinfulness’. ¹⁶

If one undertakes a semantic analysis, this may raise questions as to the lexical equivalents attested in the dictionaries: the lexical equivalent ‘wickedness’ is used also in NETS for translating the word πονηρία in Exod. 10.10, where Pharaoh discusses with Moses who may or may not leave Egypt. The text has: καὶ εἶπεν πρὸς αὐτούς Ἔστω οὕτως, κύριος μεθ᾿ ὑμῶν· καθότι ἀποστέλλω ὑμᾶς, μὴ καὶ τὴν ἀποσκευὴν ὑμῶν; ἴδετε ὅτι πονηρία πρόκειται ὑμῖν ‘And he said to them, “Let the Lord so be with you. According as I send you, I should not also send your chattels, should I? Take note that wickedness lies before you.”’ The tenth chapter of Exodus describes the locusts being announced and then devastating Egypt. Before the locusts arrive, Moses requests that the Israelites and all their people and possessions may be released to ‘serve the Lord’. Pharaoh denies the request. Pharaoh expects the Israelites to leave without their families and without their chattels, and the fact that they want to take their families with them is considered as ‘wickedness’. The question has to be asked: Is it appropriate to refer to the request to leave with all belongings (including wives, sons and daughters) as ‘wickedness’? Πονηρία may in fact also denote something difficult and/or challenging. Consequently, one should consider translating this as follows: ‘And he said to them: May it be so. The Lord is with you in the way I send you. Should I send also your chattels? ¹⁷ Look out: Difficulties lie before you.’ This translation is supported by the fact that in Exod. 10.11 Pharaoh denies the request to let them go with their chattels. ¹⁸

If this is taken into consideration, a new translation of Mt. 22.18 seems possible: ‘Jesus perceived the difficulty of their question and said to them: Why are you putting me to the test, you actors?’ It is apparent that πειράζω is translated – against standard practice – as ‘to put to the test’ and the noun ὑποκριτής as ‘actor’. Suffice to mention that Jesus is also ‘tempting’ his disciple in Jn 6.6: τοῦτο δὲ ἔλεγεν πειράζων αὐτόν· αὐτὸς γὰρ ᾔδει τί ἔμελλεν ποιεῖν. KJV translates this as: ‘And this he said to prove him: for he himself knew what he would do’. The lexical equivalent ‘put to the test’ instead of ‘to prove him’ may describe aptly what Jesus is doing here (to use the verb ‘test’ as translational equivalent in Jn 6.6 for πειράζω is widely attested in English translations of this verse). The question has to be asked whether it is not also appropriate to use a similar lexical equivalent in Mt. 22.18, refraining from a translational choice which could be seen as potentially judgmental.

This holds true also for the second word translated differently here. To translate ὑποκριτής as ‘actor’ would follow the genuine semantics of this Greek noun. An ‘actor’ is something less derogatory than a ‘hypocrite’. At the same time, the use of ὑποκριτής could actually point to what is going on between Jesus and the Pharisees in Mt. 22.18. One could argue that the Pharisees pose a question which does not concern them but only people following Jewish law rigorously. One could argue that in the narrative world of the synoptic gospels Pharisees have to ‘play-act’ and resort to hypothetical questions which do not pertain to them, since they have run out of difficult questions. Thus, in a narrative world where different schools pose difficult questions to each other, a group running out of possible questions to ask resorts to play-acting and becomes, as a consequence, a laughing stock. Thus, the term ὑποκριτής might far less point to a notion of deceit and far more to a notion of helplessness and, consequently, inferiority, if compared to the superior command of ‘the Law’ as exhibited by Jesus’ answer. ‘Wickedness’, ‘tempting’ and ‘hypocrite’ appear to introduce a notion of deceit and cunning which may be detrimental to the understanding of the passage. The alternative translation suggested above appears to be close to the Greek text and less judgmental than the traditional translation of Mt. 22.18, which might be closer to the Vulgate than to the original Greek.

5. Traces of the Old Latin and the Vulgate

The spies (ἐγκάθετος) mentioned in Lk. 20.20 provide evidence of how the Old Latin version and the Vulgate together form the basis for the understanding of the Greek word.

The Greek text of Lk. 20.20 is: Καὶ παρατηρήσαντες ἀπέστειλαν ἐγκαθέτους ὑποκρινομένους ἑαυτοὺς δικαίους εἶναι, ἵνα ἐπιλάβωνται αὐτοῦ λόγου, ὥστε παραδοῦναι αὐτὸν τῇ ἀρχῇ καὶ τῇ ἐξουσίᾳ τοῦ ἡγεμόνος. The Vulgate has: et observantes miserunt insidiatores qui se iustos simularent ut caperent eum in sermone et traderent illum principatui et potestati praesidis. Erasmus’s changes in Lk. 20.20 are small and not relevant here. ¹⁹ KJV offers as translation: ‘And they watched him, and sent forth spies, which should feign themselves just men, that they might take hold of his words, so that they might deliver him unto the power and authority of the governor.’ KJV’s translational choices are similar to those exhibited in Martin Luther 1545: ‘VND sie hielten auff jn / vnd sandten Laurer aus / die sich stellen solten / als weren sie frum / Auff das sie jn in der Rede fiengen / da mit sie jn vber antworten kundten der Oberkeit vnd gewalt des Landpflegers.’ NABRE offers: ‘They watched him closely and sent agents pretending to be righteous who were to trap him in speech, in order to hand him over to the authority and power of the governor.’

The first thing to note is that neither the Greek (παρατηρήσαντες) nor the Latin (observantes) participle at the beginning of the sentence has a direct object. Douay-Rheims acknowledges this, translating the beginning of the verse thus: ‘And watching, they sent spies’. Therefore, it is at least open to debate whether the Jewish authorities ‘watched him’ or ‘watched them/the crowds’. Taking into consideration that they actually feared the crowds (cf. Lk. 20.19) since they were – according to Luke’s narrative – ready to stone them (cf. Lk. 20.6), one is tempted to see ‘the crowds’ as a possible direct object. Such an understanding is precluded by the interpretative elaboration given in Danker’s dictionary which suggests for παρατηρέω in Lk. 20.20: ‘to watch maliciously, lie in wait for … watch one’s opportunity’. A semantic analysis of this word indicates that this slant found its way into the interpretation of this occurrence of παρατηρέω owing to the surrounding words and their semantics in the Latin translation. According to the dictionary of Liddell/Scott/Jones (Liddell, Scott and Jones 1996: s.v.) this word means first and foremost ‘watch closely’ or ‘observe carefully’. Further, it means – in combination with the word ἐνεδρεύειν – ‘to lie in wait for’, ‘watch one’s opportunity’. According to LSJ, in Lk. 6.7 ²⁰ the word παρατηρέω has the meaning ‘to lie in wait for’. Montanari’s dictionary (Montanari 2015: s.v.) changes the semantics of the lexical equivalents (‘to observe, survey, spy on’). It appears that ‘to spy on’ might actually describe an ‘observation from a distance, where the observer might not be visible to the one observed’. This appears to be Montanari’s understanding of Lk. 6.7: ‘They were watching him … to see if he would perform healing on the sabbath’. According to Muraoka’s dictionary, the occurrence of παρατηρέω in Dan. 6.12 (he uses the verse counting according to the Vulgate and quotes it as 6.11) is to be understood as ‘to watch closely with evil design’ (Muraoka 2009: 532). The Greek text of Dan. 6.12a (Theodotion) reads as follows: τότε ἄνδρες ἐκεῖνοι παρετήρησαν καὶ εὗρον τὸν Δανιηλ ἀξιοῦντα καὶ δεόμεον τοῦ θεοῦ αὐτοῦ. One could understand παρατηρέω as meaning ‘to watch from afar’ here as well. ²¹ And ‘watching from afar’ is not necessarily the same as ‘to spy’. Thus, Muraoka also seems to understand this word more negatively than necessary. Such an understanding is, however, supported by the words used with παρατηρέω in Lk. 20.20. Their lexical equivalents convey meanings which indicate a rather negative sense for παρατηρέω as well.

The next word which influences the understanding of Lk. 20.20 is ἐγκάθετος. This is a hapax legomenon in the Greek NT. LSJ translates it as ‘put in secretly, suborned’. Danker offers: ‘pert. to having the task of obtaining information secretly, hired to lie in wait’ and specifies for Lk. 20.20 the lexical equivalent as ‘spies’. The Vulgate has a rather interesting, formal, choice: insidiator. Jerome does here the same as with the word ἐπιούσιος in Mt. 6.11, which is translated as supersubstantialis. The translational choice panis supersubstantialis is called by Thayer a ‘barbarous phrase’ (Thayer 1889: 241); however, as far as their formal equivalence with the Greek words is concerned, both lexical choices are actually quite clever. While supersubstantialis might be close to incomprehensible, insidiator is formally a literal rendering of ἐγκάθετος, but also a distortion of the semantics of the Greek word. The Latin word describes a soldier who lies in wait and includes bandits and predators in its semantics (Baier 2013b: 2627). This does not hold true for the Greek word. LSJ, however, offers the same as Danker (‘put in secretly, suborned’). Montanari (who apparently did a linguistic analysis of the semantics of ἐγκάθετος) suggests ‘insinuated, introduced secretly’ but also ‘settler, resident’. In light of the rather different interpretations of the meaning of this word, it seems worthwhile to look briefly at two occurrences we may find of relevance in Josephus, both of which seem to have bearing on its use here. Further, Josephus is a Jewish author who is roughly contemporaneous with the author of Luke’s gospel. The word is used twice in the Jewish War. The first instance can be found in the accusatory speech of Antipater, using the word to describe the individuals who crowned Archelaos after the death of his father Herod. Thus, these people are not ‘introduced secretly’ but may have been secret adherents of Archelaos. By crowning him, they declare a strong allegiance to their leader – and, if such a coup does not succeed, the life expectancy of such followers is usually short. ²² The other instance is a situation where courage and strong loyalty are also a prerequisite. He mentions ‘prophets’ (characterized by Flavius as false prophets) who have the task of inspiring trust and hope among the citizens of Jerusalem while the Romans press hard. These persons are described as ἐγκάθετοι. ²³ Again, these are loyal and bold followers of their leaders. If these prophets were seen as simply trying to boost morale, they would probably face immediate death. If Lk. 20.6 and the danger of an upheaval in the temple are taken into consideration, those ἐγκάθετοι used by the Jewish authorities to pose a question to Jesus may well have been in danger of their lives. Thus, ‘agents’ (NABRE) or ‘spies’ (KJV) or the paraphrase of GNT (‘they bribed some men’) does not seem to do full justice to the semantics of the Greek word at hand. At the same time, one should mention that explorator would be the correct Latin word for ‘spy’. Thus, to translate insidiator with ‘spy’ seems to stretch the semantics of the Latin word. However, the Old Latin manuscript Hs r¹ reads suborn<a>tos. It appears probable that the gloss offered by BDAG is actually a conflation of the Vulgate and this Old Latin manuscript. It can be traced to Zorell (1931: 345: qui alqo. clam ac dolose mittitur, homo ad alqd. subornatus, vg. insidiator), who actually seems to paraphrase Grimm (1868: 112: ‘subornatus ad insidias faciendas; insidiator; dic. Lc. 20, 20 de eo, qui subornatus ab aliis aliquem dolosis verbis capturus est’). ²⁴

The translation of ὑποκρινομένους ἑαυτοὺς δικαίους εἶναι as ‘pretending to be righteous’ (NABRE) or as ‘to pretend they were sincere’ (GNT) may be viewed as underhand. These loyal adherents of the Jewish authorities play the role of strictly observant Jews (δίκαιος is used in the Seputagint to render the Hebrew öãé÷). Only such persons would have qualms about paying taxes to the Roman Emperor. The sub-clause introduced with ἵνα (ἵνα ἐπιλάβωνται αὐτοῦ λόγου, ὥστε παραδοῦναι αὐτὸν τῇ ἀρχῇ καὶ τῇ ἐξουσίᾳ τοῦ ἡγεμόνος) has been rendered by NABRE as follows: ‘who were to trap him in speech, in order to hand him over to the authority and power of the governor’. To lay a trap does again imply evil intent. In light of the fact that the Jewish authorities did fear for their lives and were therefore unable to lay their hands on Jesus, one is inclined to suggest the following translation for Lk. 20.20: ‘And watching from a safe distance they sent loyal servants play-acting as strictly observant Jews in order to obtain a statement from him so that they could hand him over to the administrative power of the governor’. They want an excuse for involving the Roman authority since their own authority has been challenged to the point of their being afraid of being stoned.

The negative slant given to ἐγκάθετοι in Lk. 20.20 and to πανουργία in Lk. 20.23 makes it virtually impossible to translate παρατηρέω in a non-judgmental manner. It appears that in case of the ἐγκάθετοι, Old Latin has also influenced the translational equivalents offered in standard dictionaries.

6. The Question of Paying Taxes According to Luke (Luke 20[19].20-26) ²⁵

It seems necessary to mention briefly what happened directly before the pericope of the question of paying taxes in Luke’s gospel. There is an encounter between Jesus and the representatives of the Jewish authorities within the temple. The representatives of the Jewish authorities had asked Jesus a question concerning on what authority Jesus taught. He answers in return with another question. The representatives of the Jews do not dare to answer Jesus’ question, fearing that they might be stoned (Lk. 20.6). This is a very tense situation: the representatives of the Jews police the temple, they may give orders to arrest people. The possibility that the assembled crowd may throw stones indicates a precarious situation. A riot may take place. Jesus adds fuel to the fire by telling the parable of the vineyard.

The end of the pericope containing this parable connects to what follows. Consequently, the translation of Lk. 20.19 is decisive for the perception of what is told in Lk. 20.20. Luke 20.19 is commonly translated in a way which might give the impression that the representatives of the Jewish authorities act sneakily (NABRE): ‘The scribes and chief priests sought to lay their hands on him at that very hour, but they feared the people, for they knew that he had addressed this parable to them’. ²⁶ The intention to lay hands on Jesus is not put into practice due to fear. This leads to the sending of envoys in Lk. 20.20. There are two translational choices which may change the meaning of the text. First, the verb ζητέω may well describe the attempt to do something. Second, the adversative translation of the copulative conjunction καί is not necessary. If the conjunction is understood as καί-consecutivum, the following translation is possible: ‘The scribes and chief priests made an attempt to lay their hands on him at that very hour. And then they feared the people. For they knew that he had addressed this parable to them.’ From this rendering of Lk. 20.19 the following situation emerges: the very authorities which police the temple are unable to apprehend Jesus. The crowds stand there and protect Jesus against the Jewish authorities. The riot is even closer than in Lk. 20.6. The reaction of the scribes and chief priests is to withdraw and deal with Jesus from a distance (Lk. 20.20): ‘And watching from a safe distance they sent loyal servants play-acting as strictly observant Jews in order to obtain a statement from him so that they could hand him over to the administrative power of the governor’. Then follows the actual dialogue (Lk. 20.21-22; NABRE): ‘They posed this question to him, “Teacher, we know that what you say and teach is correct, and you show no partiality, but teach the way of God in accordance with the truth. Is it lawful for us to pay tribute to Caesar or not?”’ Then comes Lk. 20.23, which may be translated thus: ‘But understanding their strategy he said to them’. The following two verses can and have been translated thus (Lk. 20.24-25; NABRE): ‘“Show me a denarius; whose image and name does it bear?” They replied, “Caesar’s”. So he said to them, “Then repay to Caesar what belongs to Caesar and to God what belongs to God.”’ Of importance is again the translation of the last verse (Lk. 20.26; NABRE): “They were unable to trap him by something he might say before the people, and so amazed were they at his reply that they fell silent’. ²⁷ There are some translational choices which are subject to debate. First, there is the verb παραλαμβάνω. This verb does not denote entrapment in Greek. The concept of entrapment conveys a negative meaning. Thus, this may be translated as follows: ‘They were unable to pin him down to a word before the people’.

It is rather telling that NABRE then elects to translate a participium coniunctum (θαυμάσαντες) as the main clause and turns the main clause into a subordinate clause (ἐσίγησαν). This may be described as paraphrasing translation. Further, the verb θαυμάζω in combination with the preposition ἐπί may be translated as ‘to admire’. Consequently, a philologically less interpretative translation is: ‘And since they admired his answer, they fell silent’. The implication of this translation is that the very envoys of the Jewish authorities start to shift their allegiance. The story starts in Lk. 20.20, and these loyal envoys pose a question which puts Jesus in a predicament. However, Jesus is able to answer in a way that he escapes the predicament. The consequence is admiration for this legally sophisticated mind. This is, according to Jeffrey Rubenstein, the very way how a Jewish teacher would ‘prove his mettle as a scholar’ (Rubenstein 2003: 40). A philologically careful translation, consequently, changes the story in a fundamental way. This passage is transformed from an ‘entrapment’ of Jesus to a dialectical discourse between Jesus and the Jewish authorities. Jesus emerges as winner of the match and acquires new followers.

7. Summary

Let me summarize the argument. Semantic analysis of the two sentences of Lk. 20.20 and Lk. 20.23 in context points to a different possible understanding of these two passages. The Jewish authorities might have been far less crafty and in far less control than implied by traditional translations. Such an understanding is only possible if dictionaries like Bauer/Aland or Danker are bypassed, since they inject the notion of deceit into the translational glosses offered as lexical equivalents, thereby following the Latin interpretation of the words in question. This indicates that the semantics of these Greek words has been neglected by research and that traditional translations like Martin Luther’s and their translational choices have influenced NT dictionaries, which in turn influence newer translations. The examples suggest that Latin translation and tradition deviate in their understanding of the semantics far more from their Greek source text than hitherto acknowledged. This semantic shift found its way into translations made from Greek into vernacular languages. One may therefore conclude that Lance Hewson’s comment also holds true for some aspects of NT translation, especially the semantics of translational choices which might depict the Jews in more negative terms than warranted by the Greek text: ‘Translations are fundamentally unpredictable texts that may embody interpretations that the critic has simply failed to foresee’ (Hewson 2011: 52).

Standard NT dictionaries like Bauer/Aland, Danker and TDNT seem to have made lexical choices which imply semantic decisions supporting anti-Jewish notions in translation. In these decisions they appear to follow the Latin understanding of the NT, and this ‘Latin understanding’ includes the translational decisions of the Vulgate and of the Old Latin translation of the NT as well as the Latin understanding of the Greek words as evidenced by standard dictionaries of the sixteenth century or by the Latin translation of the NT by Erasmus published as part of his edition of the Greek NT. The examples offered here suggest that Erasmus’s rather negative view of Jews and Judaism influenced his Latin rendering of the Greek NT. These examples also confirm an observation made already by John A.L. Lee: ‘Luther … relied heavily on the Vulgate, also using Erasmus’s version and German predecessors. So we have a chain of dependence here, leading back in the end to the Vulgate’ (Lee 2003: 39).

It appears that James Barr is correct in stating the following:

The Bauer dictionary gives a certain impression of a cheerful acceptance of and confidence in these formulations, an impression which is just as disconcerting as the more involved and tortured arguments of TWNT in its attempts at an intra-biblical system of understanding. I make these criticisms of the Bauer dictionary to show that I do not think it has used methods which are infallible either; in particular it has failed to be sufficiently self-critical in the semantic indications it has given (Barr 1983: 256).

This points to the necessity of revising standard NT dictionaries critically with the purpose of eliminating the Latin interpretation of the lexical choices in order to make NT translation adhere more closely to the semantics of the Greek text. Frederick William Danker seems to anticipate some translational choices which adapt the text for a certain target group: ‘Unlike a translation designed for public reading or for a particular kind of public, a lexicon also should not be party to ideological distortion’ (Danker 2004: 23). In light of the analysis presented here, it appears to be a reasonable suggestion that NT lexicography could profit from attention to the semantic details of Greek words. This appears to hold true especially for words used in contexts where the text seems to exhibit an anti-Jewish attitude. As argued, there may be anti-Judaism in NT texts. It is, however, open to debate whether every instance of its perceived anti-Judaism is intrinsically connected with the Greek text, or whether reception history has formed an anti-Jewish understanding of at least some such passages. At least in the case of the passages discussed here, the anti-Jewish reception of the texts is mirrored in translation and supported by the tools of NT philology. Taking into consideration the fact that Christian anti-Judaism has contributed to violence against Jews, it does seem to be a matter of not only academic accuracy but also of the urgent serving of justice to address these methodological problems of NT lexicography. Philip A. Cunningham describes appropriately what is at stake here:

The Gospels clearly contain passages that have been used in the past to promote animosity toward Jews and Judaism. That potential still exists. However, there is another reason why Gospel references to Judaism must be read carefully. Any distortion of the Jewish faith and tradition will inevitably result in a distorted picture of Christianity as well …(Cunningham 2016: 81). ²⁸