Abstract
Granted that God does not ‘tempt’ anyone to sin, and that a reference here to the end-time tribulation is unlikely, this article revives a neglected suggestion that the peirasmos ‘testing’ that Jesus wants his followers to avoid is putting God to the test, i.e. disobeying or rebelling against God. Being tested by God is hard; putting God on trial is sin, as prohibited at Deut. 6.16. Alongside Meribah (‘strife’) at Exod. 17.1–7 Massah (‘testing’) is a new name, signifying and symbolizing Israel’s repeated (Num. 14.22) disobedience and rebellion against God, as at the Exodus. That testing God was remembered in penitence in Ps. 78.18, 41, 56; 95. 8f.; 106.14 as the Exodus was recalled. The Septuagint always translates nsh (to test) peirazein, and its cognate massah, the place that memorializes it in the Pentateuch and Psalter, peirasmos. The specific topographical echo was lost in translation into Greek and the petition gained its general meaning. Later Syriac, Hebrew and Aramaic versions translated Matthew’s Greek, not the original, and accepted Matthew’s interpretation which lacked the symbolism of massah. Jesus himself perhaps combined a positive request that God lead us, with a warning: not to that place and now symbol of sinful disobedience. This proposal illustrates how historical Jesus suggestions can sometimes enrich the reception of a biblical text without displacing the evangelists’ meanings.
Exegetical problems and multiple interpretations
The Lord’s Prayer in its mostly Matt.6 9–13 (and Didache 8) form is repeated every day by millions unaware or unworried by its difficulties. Most of the questions it raises pass unnoticed in English until Matthew or the Prayer Book are compared with Luke’s version, and some of them remain invisible even then. The ‘daily’ bread puzzle of the Greek adjective epiousios remains unsolved. 1 The theological and philosophical debates about ‘as we forgive’ have also been much discussed. 2 The temptation petition is more perplexing than the preceding two. It is axiomatic that God does not entice to evil 3 , so why ask God not to lead us into moral jeopardy?
Both the Greek verb and the Hebrew root nsh support the alternative translation ‘testing’. But moral testing is what the Bible says God sometimes does and, however hard, it is positive and intended ultimately for our own good (Deut. 8.16). At Rom 5.3f. the synonym dokimē means tested character. So why ask to be spared what is normal (1 Cor. 10.13) and may be beneficial? Hard testing can be endured and overcome (1 Pet. 1.6; 4.12; James 1.2–4, 12). Heb. 12.5f quotes Prov. 3.11 which like Prov. 9.8, Ecclus., 18.13, 1 Cor. 11.32 and Rev. 3.19 associates discipline with God’s fatherly correction. Prayers such as Ps. 26.2, ‘Examine me O Lord and prove (nsh) me. . .’ and 17.3 suggest that if testing is the meaning, the Lord’s Prayer could be positive: lead us into it, 4 or bring it on! It is not surprising that this natural translation and interpretation of the Greek as testing rather than temptation was barely considered until modern times. Asking God not to test us makes as little sense as asking God not to tempt us to do wrong.
Few who pray it as ‘temptation’ notice the implication that God might lead us towards evil unless persuaded not to. Matthew’s explanatory addition softens it. This positive request for deliverance (by God) from evil, or the evil one, absent from the original Luke 11.4, neutralizes that previous implication. If challenged about the incongruity, thoughtful Christians usually offer a solution familiar to Marcion and Tertullian, and similar to the Epistle of James (1.13f.), perhaps the earliest indication of trouble with this petition. Rather than changing the request that God tempt us not (to evil) into God not testing us, as most modern Bible translations except NIV do, they change the agent or cause that might lead us astray, from God to another, whether the devil, or our human weakness, or other people. They assume that God only permits it, as he permitted the Satan to torment Job, and the devil to tempt Jesus.
Tertullian explains the petition ‘. . .that is, do not suffer us to be led by the one who certainly does tempt’, 5 i.e. the devil (cf. ‘the tempter’ at Matt. 4.3). James 1.14 substitutes human moral weakness (the evil inclination) for the devil. Liberals who dislike original sin as much as myth may ask protection from other people tempting them. All agree that God cannot be the agent if peirasmos is translated ‘temptation’, entice to evil. That is axiomatic for biblical Christianity (n. 3 above), unlike gnostic dualism. The alternative translation ‘testing’ depends less on the permissive interpretation but can draw on it to make God’s most severe testing more palatable. 1 Cor. 10.13 assures us that while testing (NIV ‘temptation’) is inevitable, God will not allow us (permissive) to be tested/tempted beyond what we can with God’s help endure. More generally, God permits evil, and uses it to educate us, as he did Job. Even Jesus ‘learned obedience from what he suffered’ (Heb. 5.8).
However, the permissive explanation is open to a moral objection which raises questions about it, even if we can explain its deviation from the plain meaning of the gospel text. Permitting or failing to stop evil or suffering when one could (as an omnipotent God can) is almost as bad as doing it. Christians reply that the gospel is about how God has dealt and deals with evil, enabling us to cope, not with theoretical questions about why God permits evil. However, God’s answer to Job does not satisfy all sufferers, let alone the philosophers worth a nod in our conclusions.
Faced with the moral problem of God permitting and so accepting responsibility for what some think a loving God should stop, French translations adopt a more radical strategy. They blunt God’s not leading us astray into a request not to permit us to succumb: et ne nous laissez pas succomber à la tentation. Like Tertullian’s permissive interpretation, that makes something other than God responsible for evil. The plain sense of the gospels as repeated in private and public prayer remains a challenge. We may justify the permissive explanation by noting the difficulty of saying ‘cause to be led’ in Hebrew and Aramaic. For leading people or bringing things the Hebrew Bible normally uses the hiph‘il of bo’, ‘cause to come’, an intransitive verb. A permissive meaning may therefore have been intended by Jesus but not expressed. The moral problem of permitting evil (unless for the sake of a greater good?) still niggles. And if the pre-Matthean Greek translator thought this is what Jesus intended he could have put his Greek transitive verb into the passive. ‘Let us not be brought’.
What Jesus meant remains an open question, but that Matthew understood the petition as God permitting temptation to evil is supported by his preceding account of Jesus’ temptation. That implied double agency. God tests his Son; the devil tempts him to evil. But God is responsible for the devil’s action, as in Job 1.6–12. 6
Another indication of how the Jewish-Christian Matthew, in some respects closer to early rabbinic Judaism than Jesus, understood the petition is a much cited parallel from the Babylonian Talmud (c.400-500 ce, but containing older material). This synagogue morning prayer (b. Ber. 60b; see also b. Sota 107a, 40.27), like Matthew promoting personal devotion, included ‘Do not lead or bring me (hiph‘il of bo’, cause to come) into the hands (i.e. power) of sin. . ., of guilt. . ., of temptation/testing. . .’. This too invites a permissive explanation.
Its Hebrew word nissayon (root nsh) implies testing, but neither this prayer nor Matthew (nor old English 7 ) distinguished sharply between moral testing and tempting. By contrast, modern English does distinguish them, even in non-religious contexts where sin and evil are excluded (chocolates are tempting, temptation naughty but nice). In the synagogue prayer the dark content to be avoided includes temptation arising from human weakness - its next sentence refers to the ‘evil inclination’. Matthew also means by testing temptation to evil, probably by the devil (cf. 4.1,3 and his addition 6.13b).
Then God cannot be the agent. The plain sense of the petition has to be given a permissive meaning. But that too is morally questionable. The weaker understanding of testing is less of a problem. God does test (n. 8 below). Perhaps the petition does not refer (as 1 Cor. 10.13a does) to life’s ordinary challenges. Luke 9.23 and 14.27 recognize the cost of discipleship without the word peirasmos, but at 9.23, Mark 8.34f. there Mark 8.34f. is made less dark by Luke’s addition of ‘daily’. In the Prayer he probably takes peirasmos, testing, to refer to more desperate situations (war, famine, martyrdom etc.) that we naturally ask to be spared despite Jesus’ warning (21.12–19). His redactional insertion at 8.13 of a time of peirasmos causing some to defect suggests that at 11.4 too he is thinking of persecution inducing apostasy. Humans do that evil, but according to 1 Pet. 3.17 God permits, and may even will (surely not!) innocent suffering like Job’s. Luke is not troubled by the moral or metaphysical issue. Like Job (42.3) and Wesley, he would accept the limits to human understanding of God’s ‘strange design’. For him too the life, death and resurrection of Jesus reveal God involved in the world’s suffering, healing individuals and overcoming divisions and exclusions, and so saving the world, as in the rather Lukan (early catholic) Ephesians and 1 Peter. His Christian hope (21.19) is confirmed in the experience of some, and taken on trust by others. But as for Matthew (and Jesus) peirasmos in this context (unlike 1 Cor. 10.13a) is a bad thing, inviting prayer for help in avoiding it. ‘Temptation’ (to the evil of lapsing) is therefore an appropriate English translation of Luke 11.4c, as of Matt. 6.13a.
Like Matthew, but differently, Luke will have understood Mark 14.28 where Jesus prays for his disciples not to come into peirasmos (Luke 22.40, 46—twice for emphasis) in the same way as he understood the Lord’s Prayer, faintly echoed here as more clearly at Matt. 26.39. Gethsemane thus throws light on how each evangelist understood the Lord’s Prayer. Luke’s addition of peirasmoi into his version of the Last Supper (22.28) and his use of the plural also at Acts 20.19 again illustrate the cost of discipleship that does not fail (as Jesus and Paul do not). Michael Wolter comments on Luke 11.4a that the evangelist understands that ‘God sometimes leads his chosen and pious ones in order to put them to the test . . .’. 8 Granted that ‘the concern is not with their temptation to evil’, lapsing remains a possibility needing a warning.
These different nuances get discussed in assessing the different backgrounds and contexts of each evangelist, but have no importance for most Christians praying Matthew’s version. Some might, however, be interested in suggestions that Jesus could have meant something different, even though such speculations cannot replace the evangelists’ meanings. At most they can offer further layers of meaning which may be incorporated into our reception of familiar texts, and in this case offer a possible way out of a difficulty.
Matthew, like the synagogue, sees diabolical tempting to do evil always a threat on account of humans’ evil inclination. Luke is closer to the moral testing of 1 Cor. 10.13a, but knows it could involve extreme suffering and so become a temptation to the evil of apostasy. Anyone who sees different nuances here can embrace both. The liturgical revisers who retained ‘temptation’ were true to Matthew, and the modern biblical translators who prefer testing or trials may be closer to Luke.
Subsequent Christianity followed Matthew. His meaning was further darkened by a sometimes morbid preoccupation with sin in the ‘introspective conscience of the West’, monastic spirituality, penitential practice, modern psychology etc. Churches followed Anselm’s criticism of Gaunilo (nondum considerasti . . .) for having not yet considered the full weight of sin. Peirasmos gained religious and moral weight more like Matthew’s dark nuance than Luke’s more general Greek meaning. The Western church’s tentatio/tentare or temptare was followed by Wyclif and (with Luther’s Versuchung) by Tyndale, BCP and AV. The noun and verb could then (n. 7 above) mean either tempt or test, but now English translators and liturgical revisers must choose between the two words, as in English they have different meanings.
The possibility that Jesus may have meant something different from either gospel, first argued on other issues by 18th-century Deists, 9 added new options. A 20th-century proposal refers ‘testing’ to the eschatological tribulations expected by some Jews at the end of this present evil age. This suggestion was a by-product of the discovery how important eschatology was for Jesus’ mission and message, and the possible relevance of Jewish apocalypses for Jesus calling himself the son of man. Albert Schweitzer, Lohmeyer (n. 16 below), Jeremias and others stretched Johannes Weiss’ 1892 eschatological interpretation of Jesus’ Proclamation of the Kingdom of God (English 1971) to explain peirasmos as the expected birth-pangs of the age to come.
By contrast, Bultmann agreed with Weiss about the future kingdom but accepted Luke’s meaning of peirasmos and concluded that it ‘could scarcely have come from Jesus himself but had its origin in the church.’ 10 Neither suggestion is likely, but both illustrate the critical judgments involved: exegesis of the gospel texts, historical constructions of the gospel tradition, and of Jesus himself.
Others accept the importance of futurist eschatology for Jesus, the early church, and subsequent Christianity, but think Schweitzer’s total construction of Jesus and what followed at best one-sided. They have not been persuaded by this explanation of peirasmos. 11 Despite Rev. 3.10 neither peirasmos not massah are technical terms for the eschatological tribulations in Jewish apocalypses or in the New Testament. Secondly, if Jesus understood his own and his followers’ suffering as part of the birth-pangs of the new age, how much room does that leave for separate future woes? 12 Thirdly, although the Greek article is unpredictable and the Prayer’s formulations brief, we would expect a definite article if a specific event was in mind, as at Rev. 3.10. (NRSV, unlike its margin and RSV, adds one in).
This apocalyptic explanation of peirasmos is historically doubtful, and certainly not how the evangelists understood it, but that Jesus may well have understood what he said differently from the evangelists and subsequent Christian tradition is now a commonplace of biblical criticism. On a devotional level some Christians include the apocalyptic interpretation of the petition among the multiple meanings they find in the Lord’s Prayer. We may not be spared tribulations at the end of our life (or our world) but a prayer that these will not cause us to lose trust in God (cf. Luke 8.13; 21.36; cf. also 18.8) is true to the gist of Jesus’ message and compatible with the evangelists’ understandings of his prayer. Jesus warned his disciples of both persecution and eschatological tribulation. They can pray to be spared the worst. The optimism inherent in resurrection faith does not preclude the realism about evil signalled by the cross. Apocalyptic warnings are an element in Christian preaching perhaps sometimes ignored by now wealthy, well-fed, happy, and even respected theologians (cf. Luke 6.24–26). If the apocalyptic account of peirasmos is exegetically and historically doubtful, but religiously valuable, this shows how critical uncertainties can be hermeneutical opportunities. However, the wide gap that the apocalyptic hypothesis sees between the gospels and Jesus emboldens a different ‘historical Jesus’ proposal closer to the gospels’ moral meanings. What follows has two parts, the second more speculative than the first. Both parts depend on the same antecedent probability of the importance of the Exodus traditions for Jesus as for other Jews. Each part will therefore reinforce the other.
Against putting God to the test
There can be little doubt that the pre-gospel Greek translators of the Lord’s Prayer took the rare word peirasmos from the Septuagint. 13 The three occurrences found in secular Greek are all later, and the 19 elsewhere in the NT may all stem indirectly from familiarity with this translation of Jesus’ Hebrew or Aramaic word. In the Septuagint the more common verbal form peirazein always translates nsh, to try, test. That explains massah being behind peirasmos in the Exodus traditions of the Pentateuch and Psalter. Despite its pain (Deut. 8.2; 13.3) and collateral damage (Deut. 4.34; 7.19; 29.1f.), this testing is ultimately (Deut. 8.6) good. But testing is very bad when people put God to the test or on trial by their disobedience to his command, or their grumbles (Exod. 16–17; Num. 14) leading soon to idolatry and apostasy (Exod. 32), as Job’s protest did not.
Our contention is that only the bad testing explains what Jesus said in a warning signalled by his symbolically loaded word massah. 14 In the Hebrew Bible that noun is rare, but the related verb more common. God tests people at the Exodus (Pharaoh x3; Israel x10). Job 9.23 may be a symbolic use of massah. But our argument depends on where the Israelites test God and nsh/massah refers explicitly or by implication to that eponymous place, providing a symbol and reminder of the primal sin of denying the Godness of God by putting God on trial. Disobedience to God’s command (e.g. Exod. 16.20), failures of trust (Num. 20.12; Ps. 78.8, 22) and the grumbling that was followed by rebellion or turning away from God were features (Ps. 78.40f.) of the Exodus wanderings remembered with shame by future generations. The absurdity of all this is underlined by its place in that larger narrative of God’s saving work: leading his people out of slavery, feeding them, forgiving them (up to a point), and leading them through painful testing towards the promised land where they would flourish if they lived under God’s just and merciful rule, or perish if they broke the covenant and behaved as badly as other nations (Deut. 8). That choice between life and death (Deut. 30.15-20) remains on the table. The subsequent history of failure (Deut. 30.29; Ps. 78.57f.) eventually evoked the eschatological hope of God’s coming rule prayed for in the 18 Benedictions and by Jesus in the second petition.
The weighty reminder of Massah was lost in translation as the peirasmos petition gained from that Greek word the general moral reading it has in Ben Sira. The religious and theological horror of Massah was already lost in post-biblical Hebrew writings which emphasized the blessings of the Exodus and ignored the repeated failures of the whole people. The rebellion by Korah’s band and Dathan and Abiram (Ecclus. 45.18f.) is mentioned briefly but only to glorify Moses, not in national repentance. Massah yields to nyssyn in the Hebrew fragment of Ecclus. 6.7. The issue is sensible moral testing, not the outrage of testing God. Even where testing God is mentioned (18.23) it is only trying the Lord’s patience, not the primal sin of Massah, adumbrated in Eden. Even Hebrew Ben Sira has already lost the dreadful weight of the people’s testing God ten times (Num. 14.22) i.e. repeatedly in the Exodus wanderings. By contrast, the topographical reference to Massah was obvious in Hebrew where testing God was the issue, and was recognizable also in Aramaic if the later Targums word (mssitha’) was what Jesus said.
Neither the noun at Deut. 4.34; 7.19; 29.3(2) nor the verb at Exod. 20.20; Deut. 4.34; 8.2; 13.3 are relevant to our understanding Jesus’ word which has to refer to something evil, but God’s good testing at Exod. 15.25 and 16.4 sets the scene for the people’s disobedience in the face of God’s feeding, forgiving and leading the people towards the future rule of God which Jews and Christians continue to pray will come soon. Hezekiah at Isa. 7.12 at least knows the Deut. 6.16 ground-rule of divine-human relations.
The relevant Massah passages occur with the noun at Exod. 17.7; Deut. 6.16; 9.22; Ps. 94 (95). 8 and in the related verb nsh referring to that eponymous event at Exod. 17.2, 7; Num. 14.22; Deut 6.16; Pss. 78.18, 41, 56; 95.9; 106.14. 15 This Massah event, coupled with the Meribah grumbling, and linked at Ps. 106 and by Paul in 1 Cor. 10 with the idolatry and apostasy of the golden calf episode (Exod. 32) was regularly recalled in Israel’s worship in which the Pentateuch and Psalter were central. The evil of putting God to the test was taken for granted in early Christianity (Acts 5.9 and 15.10, and especially at Heb. 3 which compares Jesus with Moses and quotes Ps. 95 at length). The prohibition against testing God is quoted in the temptation stories. It is applied to not testing Christ (or the Lord) by Paul at 1 Cor. 10.9. Rejecting Moses was rejecting God (Exod. 17.2; Num. 14.9)—‘in spite of all the signs I have done among them.’ Cf. Matt. 11.21. The ‘prophet like Moses’ christological motif might have roots in how Jesus understood his mission as a (messianic?) prophet and teacher.
The familiarity and importance for Jesus and other Jews of the Exodus events, including its negative episodes recalled as a warning in scripture and at worship, is unquestioned, but this explanation of what Jesus originally said must be weighed against an alternative possibility, closer to the evangelists than the apocalyptic hypothesis. The general moral meaning of peirasmos in Ben Sira is at least once (6.7) a translation of ‘inyon, perhaps influenced by Hebrew (and earliest Greek) Ecclesiastes. If evidence emerges that Jesus said ‘inyon, then he made no link to Massah in the Exodus or the Psalter, and our hypothesis would be falsified.
Jesus did value wisdom traditions and could have known Hebrew Ben Sira without placing either on a par with Torah, the prophets, or the Psalms, but the antecedent probability in favour of their word massah is strong. The many echoes of the Exodus wanderings in the gospels, possibly even in the Lord’s Prayer itself, suggest some basis in the ministry and teaching of the son of man with nowhere to lay his head. The evangelists’ interest in these traditions led to further developments (e.g., John 3.14) but even the temptation, feeding, and transfiguration narratives may have some historical basis. The demand for a sign and opponents ‘testing’ ‘Jesus also recall the wilderness wanderings, and every word in the Lord’s Prayer can be elaborated in Exodus terms. If the Gospel of the Hebrews was right to link the bread petition to ‘just enough’ food (n. 1) the gift of manna (maḥar, for tomorrow) that event is narratively and also thematically (Exod. 16.20) close to Massah. Further, the adjacent plea for forgiveness is sharpened by the reminder of Massah.
The quantity and variety of allusion is striking. However, it is likely (and more verifiable) that Ben Sira provides the key to how Matthew understood the word peirasmos. His Jesus interpreting the law shares Ben Sira’s respect for the covenant (15 times) and its commands (Ecclus. 28.6b), cf. Matt. 28.20. Both share the keyword righteousness (Ben Sira x5, plus dikaios x10 ; Matthew x7; adjective x10 against Mark 0 plus2). He echoes Ecclus. 51.23, 26 at 11.28f. and perhaps sees Jesus as himself divine wisdom (cf. 11.19; 23.34). Ecclus. 33.1 connects divine rescue to peirasmos, which may account for Matthew’s 6.13b addition explaining peirasmos with a clearer word for rescue.
Matthew’s loss of Massah can be explained and even defended. Understanding peirasmos with Ben Sira rather than the Psalter yielded a meaning tolerably close to the solemn Deut. 6.16 warning that Jesus may have intended. It would also be more intelligible to Greek speakers, especially Gentiles. NT meanings are what believers absorb, but historical study of Jesus generates hypotheses worth considering. Some can be incorporated and enrich believers’ faith-pictures of their Lord.
The vivid warning against rejecting God’s authority, recalling Meribah and Massah like Ps. 95.8f., was explicit if Jesus said Massah or a recognizable Aramaic equivalent. The pivotal petition for forgiveness was already adumbrated in the bread petition if that contains a secondary allusion to the manna testing and disobedience (Exod. 16.4, 20). The translator who followed the Septuagint in rendering massah by peirasmos can have seen no great difference in his general moral meaning because the disobedience at Massah was itself a moral failure, even if Jesus meant more. But out of its Exodus context the general noun changed the emphasis of the petition from the honour and demand of God to human weakness. The recovery of Deut. 6.16 perhaps intended by Jesus redirects attention to God who is tested rather than the devil who tempts, or humans tempted. Matthew’s anthropocentric meaning was not wrong, and Luke’s general moral testing is also true in itself. Jesus, however, perhaps wanted his followers to let God be God and so signalled the weightier theological implications of Massah, whether or not he was also interested in their psychological aspect.
Lead us . . . not to Massah
The Massah explanation of peirasmos has not resolved the problem of God leading us (unless persuaded otherwise) into moral hazard. It has made it more acute because putting God to the test is far worse than ordinary moral risk and failures. But the Exodus is all about God leading his people in a positive direction and this backcloth may offer a way out of that difficulty.
The Greek translator was guided by the Septuagint but his choice of eispherein, bring or carry things, rather than its normal eisagein (lead) for bringing persons is surprising. 16 It was presumably prompted by the negative action and evil destination to be avoided (cf. Luke 12.11). God ‘leading’ his people would sound odd in a negative context. The Septuagint once used pherein negatively for God bringing fear on Israel’s enemy (Jer. 49.5, 10), and once for the elite being taken (not led) to Babylon (Ezk. 19.9, cf. 17.4).
In contrast to the gospels’ eispherein, the Vulgate and dependent Western translations reverted to the usual translation of bo’ for God leading people: ducere. Tertullian’s Old Latin already had inducere. Most Western Christians prefer to pray and sing ‘lead us’ (heavenly Father, lead us; Guide me O thou great Jehovah etc.). A hypothetical construction of Jesus’ words could support the Western change, and also show how the gospels’ verb followed from their change of peirasmos from the Exodus to Ben Sira’s meaning.
An anticipation of our proposal justifying the positive word ‘lead’, against the gospels’ negative ‘bring’ can be read into Luther and Tyndale’s divergence from the Vulgate. They had rhetorical reasons for changing the word-order, but possibly also an intuition about Jesus’ positivity. The Vulgate follows the gospels in attaching the negative particle to the verb: ‘Do not lead us . . .’. Tyndale followed Luther and was followed by Cranmer, AV, NIV in moving the negative from clearly qualifying the verb into a more ambiguous position. It can be read like the gospels, but more naturally as qualifying the noun: ‘lead us—not into temptation!’. A pause would make that clearer, as would a single letter in Hebrew, an adversative waw—‘but (not into)’. If Jesus did say w (and the proposal does not require it) the translator would have had to delete it, having moved the negative from the noun to the verb—a decision itself conditioned by his giving massah a general moral meaning in place of Jesus’ symbolic topographical allusion.
Whether those who say the Prayer in English (or German) think they are negating the verb or the noun or both, the pause between lead us and not into temptation gives the line two points: lead us, as in the Exodus, Ps. 23 etc., and then the warning. This fits the probable Hebrew (bo’) or Aramaic (acel) of Jesus, and retains both the negative destination of the warning and the positive sense of God leading his people. Leading people to one place not another sounds more natural than not bringing them into a state of mind.
Our positive reading of Jesus’ fourth petition, prior to the appended warning, matches the preceding appeals to God’s grace: give, forgive, lead. The sudden switch to a negative petition grates, as an appended warning does not. Jesus does warn of historical disasters and eschatological judgment but his gospel, like deutero-Isaiah’s, is more joyfully positive than Jeremiah, Ezekiel, or John the Baptist. These prioritized judgment before offering hope. Jesus reverses the emphasis, proclaiming God’s coming rule which even now some are entering. But on our pilgrim journey through a wilderness world we still need the warning to avoid Massah and all it stands for.
The Massah hypothesis implies the Exodus backcloth and reinforces the theocentric character of the Prayer, as of all Jesus’ teaching. His talk from and of God addresses human needs but resists our tendency to invert the order and make it firstly about ourselves. Its moral and eschatological dimensions are analytic to talk of God, but the Greek translation’s loss of Massah weakens the theocentric (Deut. 6.16) focus of Jesus on the holiness of God, in the last as in the first petition. Matthew’s moral emphasis on God’s will is true to Jesus’ proclamation of what is required of us in face of God’s coming rule (Ps. 47.6–10; 96; 97.1f., 10-12; 98–99), but risks unbalancing eschatology and ethics. It contributed to the loss of this-worldly eschatology in Christian history, as did John 18.36. Jesus knew as well as Moses that God was already at work in his world through his work—by the finger of God. The Luke 11.20 echo of Exod. 8.15 was (arguably) not Lukan embellishment of an earlier tradition, as we might suspect had he written ‘spirit’, and Matthew ‘finger’. Luke’s access to other versions of what Jesus said enabled him to correct Matt. 12.28 historically even though he agreed with it theologically.
For Jesus, God’s rule was firstly about God. Participation could already be anticipated (Matt. 5.3, 10). Like Moses, Jesus knew he would himself die before its consummation but he expected his disciples to join him in its future (Mark 14.25).
Matthew’s redactional addition ‘with you’ at 26.30 is true to both Jesus himself at the Last Supper and to the disciples’ post-resurrection experience. His mission was to all Israel, especially its lost sheep. Its failures (Matt. 11.20–24) perhaps recalled the wilderness rebellions, but like these were subsumed in the positive liberating purpose of God who leads his people as a shepherd. The followers of Jesus acknowledged the identity of will (Mark 9.37; Matt. 10.40; Luke 10.16; John 5.19; 13.20 and passim) later expressed christologically.
Our transposing the word-order with Luther, Tyndale and current liturgical practice protects the positive Exodus resonance of God leading us, and confirms the authenticity of the petition (contra n. 10 above). Like Paul in principle (1 Cor. 7.19) and Matthew in practice (5.19) Jesus saw obedience to God’s commands, and the reality of divine judgment, inseparable from the good news of God that they all proclaimed and lived, confidently expecting its consummation. Jesus’ final warning at the end of the original (Lukan) version was more than the synagogue prayer for protection from temptation. Massah recalls the primal sin of reversing the divine order by putting our desires first. That place does not today carry the symbolic weight of Auschwitz, but its unfamiliarity opens doors to fresh reflections.
Consequences
These need separate articles, but historical suggestions that cannot be verified or falsified call for some other justification. Historians demur at unlikely hypotheses, but theologians might draw benefits from even fanciful ideas. The linguistic arguments can be followed up elsewhere, 17 but there may be hermeneutical, theological and practical gains in reflecting on a neglected suggestion.
Hermeneutical
Advocacy of multiple interpretations plunges into philosophical debates about textual indeterminacy and the nature of interpretation. Where the identity of Christianity is at stake (as it is not in the present discussion) and scripture is seen as a norm, theology needs textual stability. NT theology, like most biblical scholarship, has usually taken that for granted. However, the churches’ engagement with scripture has a wider scope than linguistic and historical research. Whether NT theology can learn from spiritual exegesis without breaking its own scholarly conventions remains a question, but there is merit in leaving different historical hypotheses on the theological table. Some bear fruit.
NT theologians trust the spirit to work through their scholarship. Other sources of scriptural knowledge are not their business; even Barth was dismissive of ‘pneumatic exegesis’. But religious aims are intrinsic to ‘theology’ rightly understood, and most NT theologians admit them without allowing them to distort their rational judgments. Their independent study of ancient languages and history pays dividends in apologetics and in churches’ struggles for consensus. Its critical task of discrediting silly ideas is limited to assessing scriptural warrants. It does not control the theological imagination, much less private prayer. It may, however, protect congregations from some of what passes as biblical preaching when theological education neglects historical exegesis.
Theological
The relationship and relative weight of historical Jesus research and the gospel portraits of Jesus (the Kähler question) invites further reflection. Reimarus and his British Deist predecessors opened a can of worms, but some of them have proved to be surprisingly edible. Thoughtful believers test what they hear in scripture and in sermons against their own understandings of God in Jesus. Faith-images of Jesus constructed from their tradition and experience whisper what Jesus ‘must have meant’, and Luther’s willingness to ‘urge Christ against scripture’ has become widespread in liberal congregations. Its scope for error and danger of closing ears to unwelcome challenges can be minimized by allowing different views to remain on the table, and to disagree well.
In the two words of the Lord’s Prayer discussed here the Exodus backcloth proposed underlines a textual fact of some significance. Matthew’s 9 plurals (6 in Luke) present a contrast to the singulars of b. Ber. 60b (above). Those nourish private devotion. The prevalent use of the Lord’s Prayer too in private contexts recommends the plurals’ reminder, reinforced by the possible Exodus motifs, that Christianity like Judaism is a corporate matter. Individuals’ faith is rooted in a tradition which allows it to be expressed in the community’s worship and reflected on in its theology.
The Exodus backcloth also highlights the people of God motif in Christian theology. The centrality of that in 20th-century ecclesiology, especially in the ecumenical movement and at Vatican II, is not without dangers. Does it encourage sectarianism or a ghetto mentality, altering the necessary balance between God’s universal saving intent (1 Tim. 2.4) and care for all creation, and the responsibilities of a chosen people and individuals? 18
God’s election of Israel remains foundational to Christianity, and the centrality and proper use of the Old Testament has received renewed attention since its neglect in rationalist theologies. However, the people of God motif refers firstly to God’s covenant people Israel. Its appropriation was natural when most Christians were Jews, conscious of belonging to a minority that recognized the Messiah, and so being a righteous remnant of that chosen nation (1 Pet. 2.9). In Gentile churches the motif is open to Christian anti-Judaism, and also to triumphalism. God’s election is to service. Calling ourselves ‘the elect’ risks making us feel more special than we are. Some would gladly return this motif to God’s ancient and faithful covenant people, trusting them to resist its secular nationalist perversions.
Other questions posed by the command not to test God, and God’s answer to Job, include the necessity and limits of rational enquiry in face of the mystery of the universe we call ‘God’, and the mystery of evil. Related to that, the doctrine of atonement, and the necessity of revelation, may be clarified from this perspective.
Liturgical
Our defence of ‘temptation’ and ‘lead us’ by appeal to Matthew in the first case, and the proposed meaning of Jesus in the second, has recommended conservatism in liturgical revision. There is no suggestion here of introducing Massah in a new version of the Prayer, only of allowing its meaning to enrich our reading of the traditional word. Our discussion has taken for granted both critical biblical scholarship and the theological imagination. The diversity in the NT extends to single words, not only NT authors’ theologies. How scripture can still be a norm for Christian belief and practice calls for separate treatment. The four-fold gospel and church history indicate that Christians have always lived with multiple interpretations of the crucified and risen Jesus as the decisive revelation of God. In worship believers share words that they sometimes understand differently. Theological disagreement sometimes necessitates liturgical ambiguity and tolerance. As in some medieval exegesis, the ‘literal sense’ has priority in theological argument, without this quenching the Spirit in prayer and preaching. Knowledge of the biblical texts is entailed by the first commandment, but biblical preaching requires more, including enough exegetical skill to know when our reflections legitimately go beyond the textual meanings.
Devotional
Accepting the possible allusion by Jesus to all that this symbol of the primal sin can include recognizes the partly shameful history of the Abrahamic religions. Apostasy was named in Luke’s interpretation of peirasmos and remains a threat where Christians are persecuted or led astray by secularism. Idolatry is the most prominent temptation for individuals in materialistic societies, and churches are not exempt. Its consequences for God’s world are eliciting theological and other remedies from a variety of perspectives. Many share the Christian hope for a better world and adopt some religious practices. Even Nietzsche’s crazy man recognized the enormity of what ‘we’ have done in nullifying God’s help to by-pass Massah, a bad place illuminated also by other psychologists. He too had read Deut. 8, but those who repent and firmly resolve to do better repeat Matthew’s addition and pray ‘deliver us from evil’.