Lead us not into temptation, but deliver us from the Evil One

If you grew up anywhere in the English-speaking world and have had even the slightest exposure to organized religion at some stage of your life, there is an excellent chance that you are familiar with The Lord's Prayer as it appears in the King James Version. You have almost certainly heard it; you have probably recited it; and you may well have memorized it.

The ending of the prayer runs like this: "And lead us not into temptation, but deliver us from evil: For thine is the kingdom, and the power, and the glory, for ever. Amen." (Matthew 6:13 KJV)

It might surprise some readers to know that analysis of ancient manuscripts by textual critics has revealed that the prayer in Matthew probably originally ended with 'deliver us from evil'; the rest was likely added at a later stage because "it was felt that this ending was too abrupt and negative."¹ (See also the abrupt ending preserved in Luke 11:1-2).

Our main focus here, however, is on the line, "And lead us not into temptation, but deliver us from evil." If you read this text in various modern translations, you will see that the majority of them translate the latter clause, "but deliver us from the evil one" (NKJV; NRSV; NIV; NET; NLT; HCSB). Some of these include 'from evil' as a marginal rendering. The ESV and NASB retain the KJV rendering 'from evil' but include 'from the evil one' in the margin.

Why is it that the majority of translators and commentators² today favour the reading, 'the evil one'? Are they simply biased by their belief in a personal devil, as a Christadelphian recently suggested to me? Or does this translation reflect the results of careful, responsible scholarship?

Ayo notes that "virtually all the Greek patristic writers" (that is, Christian writers in the earliest period of the church after the apostles) saw a reference to the devil in this verse.³ Now, post-apostolic church tradition generally carries no weight with Christadelphians, and might be discounted as 'biased' in the same way as modern translations. However, before we thrust aside the patristic testimony, we ought to remember that we are much further removed from New Testament Greek than they were.

From a grammatical point of view, both translations are possible. Matthew 6:13b in Greek reads, alla rhusai hemas apo tou ponerou. The crucial grammatical observation is that poneros ('evil') has the definite article, the Greek version of the word 'the'. The definite article in ancient Greek does not correspond exactly to the word 'the' however. It can perform many different functions; Wallace devotes 86 pages to this part of speech in his Greek grammar!⁴ One of these functions is that of a "substantiver" meaning that it can transform various parts of speech into nouns. Poneros is properly an adjective; yet in this verse the article transforms it to be a noun. But what sort of noun is it? Is it a definite noun referring to a particular individual called 'the evil one'? Is it a generic noun referring to any 'evil one'? Or is it an abstract noun referring to 'evil'? The answer is that grammatically, it could be any of these; and if it was the latter, we would best omit the definite article from our translation. Hence, "the Evil One", "the [generic] evil one" and "evil" are all, grammatically speaking, legitimate translations of this Greek phrase.⁵

There is an interesting literary parallel in a rabbinic prayer recorded in the Talmud which reads thus:

May it be Thy will, O Lord our God, and God of our fathers, to deliver us from the impudent and from impudence, from an evil man, from evil hap, from the evil impulse, from an evil companion, from an evil neighbour, and from the destructive Accuser, from a hard lawsuit and from a hard opponent, whether he is a son of the covenant or not a son of the covenant! (b. Ber. 16b)

In this prayer all three possibilities are explicitly mentioned: various types of evil men, two more abstract concepts (evil hap and the evil impulse), and the destructive Accuser (i.e. Satan). This serves as a reminder that the various interpretive options are not necessarily mutually exclusive, and that within an ancient Jewish context it was customary to pray for deliverance from various kinds of evil. Which of these possibilities, however, is most likely the primary referent of ho poneros in Matthew 6:13?

There is one other grammatical point which is helpful although not decisive. The verb rhuomai (to deliver or rescue) occurs 12 times in the New Testament with an indirect object from which someone is delivered. It is always linked to its indirect object by means of one of two Greek prepositions, ek or apo, which in this case both mean 'from'. In Matthew 6:13 the preposition is apo. If we look at the other 11 instances in the New Testament, we find that ek is usually used for deliverance from things (Romans 7:24; 2 Corinthians 1:10; Colossians 1:13; 2 Timothy 3:11; 2 Timothy 4:17; 2 Peter 2:9) whereas apo is usually used for deliverance from persons (Romans 15:31; 2 Thessalonians 3:2). There are two possible exceptions to this pattern and one clear exception. In Luke 1:74 ek is used where the indirect object is "the hand of our enemies." This is personal but it is not necessarily an exception because the immediate referent is 'hand', a thing. In 1 Thessalonians 1:10, the indirect object is a thing, "wrath to come." There is textual uncertainty about the preposition here because some manuscripts have ek while others have apo. If apo were the original reading then this would be an exception to the pattern. The one clear exception is in 2 Timothy 4:18, where the indirect object is "every evil deed" but the preposition is apo. There is nothing like a rule of grammar stating that rhuomai + apo implies a personal indirect object, but the pattern of usage nudges the balance of probabilities toward the reading, 'the evil one.'⁶

The decisive factor, however, must be the context. If we look at the broader context of Matthew there is no clear use of ho poneros (with article) to refer to evil as an abstract noun (either as in 'evil' generally or 'the evil impulse' specifically). Such a reading is possible in Matthew 5:37, but the meaning of ho poneros is disputed here too with most modern translations opting for 'the evil one.' ho poneros does refer to evil abstractly elsewhere in the New Testament (Luke 6:45; Romans 12:9). Within Matthew there is one use of ho poneros which most likely refers to 'an evildoer' in a generic sense (Matthew 5:39), though here too some scholars take it as a reference to the evil one.⁷ There are at least two uses of ho poneros to refer to 'the Evil One' in an individualized sense (Matthew 13:19; 13:38). That ho poneros in Matthew 13:19 means 'the Evil One', an epithet for the devil or Satan, can be ascertained by comparing it with the parallel accounts of the parable of the sower (Mark 4:15; Luke 8:12). In Matthew 13:38 ho poneros is identified as 'the devil' in the immediate context (v. 39). In both of these texts 'the Evil One' is evidently either a person or a personification, and not a generic or abstract noun, since the Evil One is described as 'coming and snatching' the word away, and as a father to the wicked. This "Evil One" thus figures prominently in Jesus' teaching in the Gospel of Matthew.

There are a number of other New Testament instances of ho poneros which virtually all modern translations (including NASB and ESV) render 'the evil one' (John 17:15; Ephesians 6:16; 2 Thessalonians 3:3; 1 John 2:13-14; 3:12; 5:18-19). The KJV follows an abstract reading in Ephesians 6:16, 2 Thessalonians 3:3, John 17:15 and 1 John 5:19, but in the other passages even it reads 'the wicked one', indicating virtual unanimity among major English translations. John 17:15 is particularly noteworthy since, like Matthew 6:13, it occurs in a prayer of Jesus. Note also that the context of both Ephesians 6:16 (v. 11) and 1 John 3:12 (vv. 8, 10) imply an identification of ho poneros with ho diabolos, the devil.

The decisive point supporting the rendering 'the evil one' is the need to read the Lord's Prayer in light of the wilderness temptation accounts. The petition, "Lead us not into temptation, but deliver us from [the] evil [one]" draws on Matthew 4:1-11 in which Jesus was led by the Spirit into the wilderness to be tempted by the devil. Chase argued that every clause of the Lord's Prayer is clarified when read in light of the temptation narrative, and that "It is difficult to imagine that the analogy between the two breaks down in the last clause, and that the prominence of the tempter in the history has no counterpart in the Prayer."⁸

In view of the above evidence, it is best to understand ho poneros in Matthew 6:13 as synonymous with ho diabolos just as in Matthew 13:19 and 13:38-39. The grammatical and contextual evidence leads us to the conclusion that Jesus actually instructed us to pray for deliverance from the individualized 'Evil One.' This leaves open the question of whether he understood the devil to be an actual personal being or a personification. In my papers The Devil in the Wilderness and The Enemy is the Devil I have argued at length that the Gospel writers (and Jesus himself) intended to convey to their audiences that the devil is a supernatural personal being. This conclusion is supported by an overwhelming majority of modern New Testament scholars.

If we don't share Jesus' view that there is a real, external enemy who seeks to turn us away from God, it seems inevitable that we will fall short of his admonition to "pray in this way" (Matthew 6:9). On the other hand, if we appreciate the reality of this enemy and all the forms of evil that characterize his dark domain, we will pray earnestly for deliverance, taking heart in the fact that Jesus has already overcome him and broken his power (Hebrews 2:14).

---

1 Bruner, F.D. (2004). Matthew: The Christbook, Matthew 1-12. Eerdmans, p. 315.^↩
2 See for example Bruner (op. cit., p. 314); Carson, D.A. (2010). Matthew. In The Expositor's Bible Commentary, Matthew-Mark. T. Longman and D. Garland (Eds.). Zondervan, p. 208; Evans, C.E. (2012). Matthew. Cambridge University Press, p. 148.^↩
3 Ayo, N. (2002). The Lord's Prayer: A Survey Theological and Literary. Rowman and Littlefield, p. 95.^↩
4 Wallace, D.B. (1996). Greek Grammar Beyond the Basics: an exegetical syntax of the New Testament. Zondervan, pp. 206-291.^↩
5 Wallace himself argues based on the relationship between the wilderness temptation narrative and the prayer that tou ponerou here refers to the evil one, the devil (op. cit., p. 233).^↩
6 This analysis has been made by scholars such as Zerwick, who notes that the same general pattern holds in the Septuagint (Zerwick, M. (1963). Biblical Greek: illustrated by examples. Gregorian Biblical Bookshop, p. 29).^↩
7 Bruner, op. cit., pp. 248-250.^↩
8 Chase, F.H. (1891). The Lord's Prayer in the early church. Cambridge University Press, p. 105.^↩