The New Perspective on Paul

By James D.G. Dunn

Acknowledgements: “The New Perspective on Paul” by James D.G. Dunn originally appeared in the Bulletin of the John Rylands Library, Vol. 65, 1983, pp. 95-122. It was included in Dunn, Jesus, Paul and the Law: Studies in Mark and Galatians (London: SPCK), 1990, pp. 183-214; more recently in Dunn, The New Perspective on Paul: Collected Essays (WUNT 185; Tübingen: Mohr/Siebeck), 2005; now available from Eerdmans: James D.G. Dunn, The New Perspective on Paul: Revised Edition © 2008 Wm. B. Eerdmans Publishing Company, Grand Rapids, Michigan. Reprinted by permission of the publisher; all rights reserved. Visit for ordering information, or call 1-800-253-7521 in the U.S. The “Additional Note” following the essay is unique to the SPCK volume and has been included at the author’s request with the permission of SPCK.


The Manson Memorial Lecture delivered in the University of Manchester on 4 November 1982. Subsequently delivered in modified form as one of the Wilkinson Lectures in the Northern Baptist Theological Seminary, Illinois, under the title ‘Let Paul be Paul’.




When chatting with other New Testament specialists I occasionally mention the fact that I am engaged in writing a commentary on Paul’s letter to the Christians in Rome. The most frequent response is one of surprise, sometimes even amazement – ‘Not another commentary on Romans!’ The underlying implication is that we have had quite sufficient commentaries on Romans, that surely there can be nothing new or novel to say on such a well-worked document, that a new commentator is bound to spend most of his time simply repeating the thoughts of his predecessors. I cannot say that I am particularly taken aback by such responses, because when I was first invited to write a commentary my own reaction was more or less the same – a rather stultifying sense that it had all been said before, that interpretation of Pauline theology had lost a lot of steam, and that the really interesting and challenging frontiers in New Testament studies were to be found elsewhere.


I do not for a moment want to suggest that a commentator should refrain from re-expressing the old truths and rich insights of former days and previous commentators on Paul. Mere novelty is not of itself a mark of merit, and novelty for its own sake should certainly not be encouraged in an interpreter or expositor of any text. As students of Paul we all would be the poorer if scholars like F. F. Bruce or Otto Kuss or Heinrich Schlier had refused to distil their lifetime’s study of Paul into single volumes, simply because they did not have some revolutionary new theories to put forward.1 Nor do I wish to imply that fresh thought on particular points of Pauline theology or lively debate on particular passages within the Pauline corpus has been lacking. If we think only in terms of the last few years, for example, there has been more than one controversial reconstruction of Pauline chronology.2 Older emphases on the significance of Paul’s conversion for his subsequent theology, and on the importance of the apocalyptic aspect of his teaching, have been strongly and fruitfully revived.3 There has been a challenging reappraisal of the way in which Paul was regarded in the ancient church.4 Interesting new hypotheses on the development of Paul’s thought between his writing of Galatians and his writing of Romans have been formulated,5 and the posing of sociologically inspired questions has thrown up some important new insights.6 The old introductory questions as to the occasion for and situation addressed by particular letters still provokes heated controversy,7 and we can even say that a new subdivision of the literary criticism of the letters has recently been opened up – rhetorical criticism.8 As a final example, perhaps I could be forgiven for hoping that one or two useful comments on Paul’s religious experience, ecclesiology and Christology have flowed from my own pen.9


In none of these cases, however, could I confidently say that I have been given (I speak personally) what amounts to a new perspective on Paul. In some cases the old pattern has been shaken up somewhat and the pieces have fallen a little differently. In other cases particular aspects of Paul’s writing and thought have received fuller illumination or previous conclusions have had a question mark appended to them. In others I strongly suspect red herrings have been drawn in and wild geese chased. But none have succeeded in, to use a contemporary phrase, ‘breaking the mould’ of Pauline studies, the mould into which descriptions of Paul’s work and thought have regularly been poured for many decades now. There is, in my judgement, only one work written during the past decade or two which deserves that accolade. I refer to the volume entitled Paul and Palestinian Judaism by E. P. Sanders [formerly] of McMaster University in Canada.10


Sanders’ basic claim is not so much that Paul has been misunderstood as that the picture of Judaism drawn from Paul’s writings is historically false, not simply inaccurate in part but fundamentally mistaken. What is usually taken to be the Jewish alternative to Paul’s gospel would have been hardly recognized as an expression of Judaism by Paul’s kinsmen according to the flesh. Sanders notes that Jewish scholars and experts in early Judaism have for long enough been registering a protest at this point, contrasting rabbinic Judaism as they understand it with the parody of Judaism which Paul seems to have rejected. Thus, for example, Solomon Schechter: ‘Either the theology of the Rabbis must be wrong, its conception of God debasing, its leading motives materialistic and coarse, and its teachers lacking in enthusiasm and spirituality, or the Apostle to the Gentiles is quite unintelligible’; or a few lines later, James Parkes: ‘…if Paul was really attacking “Rabbinic Judaism”, then much of his argument is irrelevant, his abuse unmerited, and his conception of that which he was attacking inaccurate’.11 But such protests seem to have fallen for the most part on deaf ears. For a hundred years now, as Sanders observes, the majority of New Testament scholars have maintained a fundamental antithesis between Paul and Judaism, especially rabbinic Judaism, and have seen this antithesis as a central factor, usually the central factor, in understanding Paul the Jew-become-Christian.12


The problem focuses on the character of Judaism as a religion of salvation. For rabbinic specialists the emphasis in rabbinic Judaism on God’s goodness and generosity, his encouragement of repentance and offer of forgiveness is plain. Whereas Paul seems to depict Judaism as coldly and calculatingly legalistic, a system of ‘works’ righteousness, where salvation is earned by the merit of good works. Looked at from another angle, the problem is the way in which Paul has been understood as the great exponent of the central Reformation doctrine of justification by faith. As Krister Stendahl warned twenty years ago, it is deceptively easy to read Paul in the light of Luther’s agonized search for relief from a troubled conscience.13 Since Paul’s teaching on justification by faith seems to speak so directly to Luther’s subjective wrestlings, it was a natural corollary to see Paul’s opponents in terms of the unreformed Catholicism which opposed Luther, with first-century Judaism read through the ‘grid’ of the early sixteenth-century Catholic system of merit. To a remarkable and indeed alarming degree, throughout this century the standard depiction of the Judaism which Paul rejected has been the reflex of Lutheran hermeneutic. How serious this is for New Testament scholarship may be seen when we recall that the two most influential New Testament scholars of the past two generations, Rudolf Bultmann and Ernst Käsemann, both read Paul through Lutheran spectacles and both made this understanding of justification by faith their central theological principle.14 And the most recent full-scale treatment of this area of Pauline theology, on Paul and the law, still continues to work with the picture of Paul as one who rejected the perverted attempt to use the law as a means of earning righteousness by good works.15


Sanders, however, has built up a different presentation of Palestinian Judaism at the time of Paul. From a massive treatment of much of the relevant Jewish literature for that period, a rather different picture emerges. In particular, he has shown with sufficient weight of evidence that for the first-century Jew, Israel’s covenant relation with God was basic, basic to the Jew’s sense of national identity and to his understanding of his religion. So far as we can tell now, for first-century Judaism everything was an elaboration of the fundamental axiom that the one God had chosen Israel to be his peculiar people, to enjoy a special relationship under his rule. The law had been given as an expression of this covenant, to regulate and maintain the relationship established by the covenant. So, too, righteousness must be seen in terms of this relationship, as referring to conduct appropriate to this relationship, conduct in accord with the law. That is to say, obedience to the law in Judaism was never thought of as a means of entering the covenant, of attaining that special relationship with God; it was more a matter of maintaining the covenant relationship with God. From this Sanders draws out his key phrase to characterize first-century Palestinian Judaism – ‘covenantal nomism’. He defines it thus:


covenantal nomism is the view that one’s place in God’s plan is established on the basis of the covenant and that the covenant requires as the proper response of man his obedience to its commandments, while providing means of atonement for transgressions … Obedience maintains one’s position in the covenant, but it does not earn God’s grace as such … Righteousness in Judaism is a term which implies the maintenance of status among the group of the elect.16


If Stendahl cracked the mould of twentieth-century reconstructions of Paul’s theological context, by showing how much it had been determined by Luther’s quest for a gracious God, Sanders has broken it altogether by showing how different these reconstructions are from what we know of first-century Judaism from other sources. We have all in greater or less degree been guilty of modernizing Paul. But now Sanders has given us an unrivalled opportunity to look at Paul afresh, to shift our perspective back from the sixteenth century to the first century, to do what all true exegetes want to do – that is, to see Paul properly within his own context, to hear Paul in terms of his own time, to let Paul be himself.


The most surprising feature of Sanders’ writing, however, is that he himself has failed to take the opportunity his own mould-breaking work offered. Instead of trying to explore how far Paul’s theology could be explicated in relation to Judaism’s ‘covenantal nomism’, he remained more impressed by the difference between Paul’s pattern of religious thought and that of first-century Judaism. He quickly – too quickly in my view – concluded that Paul’s religion could be understood only as a basically different system from that of his fellow Jews. In Christianity a quite different mode of righteousness operated from that in Judaism, righteousness which is through faith in Christ, ‘from God’ and not ‘from the law’ (Phil. 3.9). Paul had broken with the law for the simple reason that following the law did not result in his being ‘in Christ’. Christ was the end of the law (Rom. 10.4). It was this change of ‘entire systems’ which made it unnecessary for Paul to speak about repentance or the grace of God shown in the giving of the covenant.17


But this presentation of Paul is only a little better than the one rejected. There remains something very odd in Paul’s attitude to his ancestral faith. The Lutheran Paul has been replaced by an idiosyncratic Paul who in arbitrary and irrational manner turns his face against the glory and greatness of Judaism’s covenant theology and abandons Judaism simply because it is not Christianity. It may be, of course, that Paul was totally bowled over by his encounter with the risen Christ outside Damascus, and this experience gave him a jaundiced and unfairly prejudiced view of his erstwhile faith from that time on. But Paul was by no means the only Jew who became a Christian, and it is difficult to see such an arbitrary jump from one ‘system’ to another commending itself quite as much as it in the event obviously did to so many of his fellow Jews.


The critiques of Sanders which inevitably followed have also failed in greater or less measure to capitalize on the new perspective opened up by Sanders, either because they dispute the main thrust of Sanders’ thesis, or because they do not know quite what to make of Paul when viewed from that perspective. Hans Hübner, for example, continues to operate largely within the classic Reformation categories, criticizing Sanders for failing to see Paul’s attack on ‘legalistic works-righteousness’ as central for Paul’s theology.18 On the other hand, Heikki Räisänen accepts Sanders’ strictures on Paul: Paul does misrepresent and distort the Judaism of his own day. He has separated law from covenant and adopted a Gentile point of view. Having ‘become internally alienated from the ritual aspects of the law’ over the years, he has branded ‘the covenantal theology of his Jewish-Christian opponents as salvation by works of the law’, thus attributing to the law a different role than the Jewish Christians themselves did.19 And Morna Hooker points out the oddity of Sanders’ conclusion, that the ‘pattern of religion’ which emerges from Sanders’ study of Palestinian Judaism bears a striking similarity to what is commonly believed to be the religion of Paul, but then struggles with only little more success than Sanders to explain why it was in that case that Paul felt the need to distance himself from that Judaism.20


Sanders himself has returned to the subject in a monograph entitled Paul, the Law, and the Jewish People, the manuscript of which he kindly permitted me to read. In it he broadens out the perspective on Paul from the narrower question of ‘getting in and staying in’ the covenant, which was the preoccupation of Paul and Palestinian Judaism, and restates his position in more detail. The picture of Judaism which emerges from this fuller study of Paul does correspond to Judaism as revealed in its own literature. Paul attacks covenantal nomism, the view that accepting and living by the law is a sign and condition of favoured status. It was never God’s intention, so Paul argues, that one should accept the law in order to become one of the elect. ‘His real attack on Judaism is against the idea of the covenant … What is wrong with the law, and thus with Judaism, is that it does not provide for God’s ultimate purpose, that of saving the entire world through faith in Christ …’21 But he still speaks of Paul breaking with the law, he still has Paul making an arbitrary jump from one system to another and posing an antithesis between faith in Christ and his Jewish heritage in such sharp, black-and-white terms, that Paul’s occasional defence of Jewish prerogative (as in Rom. 9.4-6) seems equally arbitrary and bewildering, his treatment of the law and its place in God’s purpose becomes inconsistent and illogical, and we are left with an abrupt discontinuity between the new movement centered in Jesus and the religion of Israel which makes little sense in particular of Paul’s olive tree allegory in Romans 11.22


I must confess that I find Sanders’ Paul little more convincing (and much less attractive) than the Lutheran Paul. I am not convinced that we have yet been given the proper reading of Paul from the new perspective of first-century Palestinian Judaism opened up so helpfully by Sanders himself. On the contrary, I believe that the new perspective on Paul does make better sense of Paul than either Sanders or his critics have so far realized. And, if I may, I would like in what follows to make a beginning to an exegesis and description of Paul’s theology from this perspective.




Let me attempt to demonstrate my case by focusing particularly on one verse and attempting to set it as fully as possible into its historical context. I refer to Galatians 2.16. This is the most obvious place to start any attempt to take a fresh look at Paul from our new perspective. It is probably the first time in the letters of Paul that his major theme of justification by faith is sounded. As such, the way in which it is formulated may well tell us much, not only about the theme itself, but about why it meant so much to Paul. We are encouraged in this hope by the fact that this first statement seems to grow out of Paul’s attempt to define and defend his own understanding of justification, over against whatever view was held by his fellow Jewish Christians from Jerusalem and Antioch; and also that it seems to form the basic statement of his gospel on which he builds his plea to his Galatian converts to hold steadfast to the gospel as he first proclaimed it to them.


It will perhaps be helpful if I sketch out the immediate preceding context of this important verse more fully. Paul has been recalling the unhappy incident at Antioch some time previously. At Antioch Gentiles had been accepted fully into the circle of those Jews who believed that Jesus was God’s Anointed and that, though rejected by the leaders of his own people, God had raised him from the dead. The leading apostles at Jerusalem had already agreed that such Gentiles need not be circumcised in order to be counted as fellow believers (Gal. 2.1-10). At Antioch the custom was for all those who had been baptized in this faith in Jesus the Christ to share a meal in common when they met – Jews together with Gentiles. But then ‘certain individuals’ had arrived from James in Jerusalem (2.11), and evidently they had found it unacceptable that the Jewish Christians should act in such disregard for the food laws laid down by Moses – the laws on clean and unclean foods, the laws on the proper slaughter of animals for meat, and probably also the various regulations governing tithing, ritual purity and avoidance of idol food already current among the more devout Jews. Whatever the men from James said or however they acted, it had an effect. Peter and all the other Jewish believers, including even Paul’s associate Barnabas, withdrew from the fellowship meals, presumably in order to demonstrate their continuing loyalty to their ancestral faith – to demonstrate that believing in Jesus did not make them any less devout Jews (2.12-13). But Paul had confronted Peter and accused him of hypocrisy, of not following the straight path of the gospel. In front of the whole community of believers he appealed to Peter: ‘If you, a Jew, live like a Gentile and not like a Jew, how can you compel the Gentiles to judaize?’ – that is, to observe the food laws and table regulations drawn out from the law by the devout Jews (2.14).23 Then Paul goes on, probably not repeating the precise words he used to Peter at Antioch, but probably echoing the line of argument which he tried to develop on that occasion,24 ‘We who are Jews by nature and not Gentile sinners, know that a man is not justified by works of law except through faith in Christ Jesus. And we have believed in Christ Jesus, in order that we might be justified by faith in Christ and not by works of law, because by works of law shall no flesh be justified’ (2.15-16) – the last clause echoing Psalm 143.2.


What precisely was Paul arguing here? What were the nuances and overtones which his fellow Jewish Christians would have recognized and appreciated? A careful analysis may well yield fruitful results.


(a) First, then, how did Paul mean to be understood by his sudden and repeated talk of ‘being justified’? – ‘Knowing that a man is not justified by works of law … in order that we might be justified by faith in Christ … by works of law shall no flesh be justified’. The format of his words shows that he is appealing to an accepted view of Jewish Christians: ‘we who are Jews … know …’25 Indeed, as already noted, Paul is probably at this point still recalling (if not actually repeating) what it was he said to Peter at Antioch. Not only so, but his wording shows that he is actually appealing to Jewish sensibilities, we may say even to Jewish prejudices – ‘we are Jews by nature and not sinners of the Gentiles’. This understanding of ‘being justified’ is thus, evidently, something Jewish, something which belongs to Jews ‘by nature’, something which distinguishes them from ‘Gentile sinners’.26 But this is covenant language, the language of those conscious that they have been chosen as a people by God, and separated from the surrounding nations. Moreover, those from whom the covenant people are thus separated are described not only as Gentiles, but as ‘sinners’. Here, too, we have the language which stems from Israel’s consciousness of election. The Gentiles are ‘sinners’ precisely in so far as they neither know nor keep the law given by God to Israel.27 Paul therefore prefaces his first mention of ‘being justified’ with a deliberate appeal to the standard Jewish belief, shared also by his fellow Jewish Christians, that the Jews as a race are God’s covenant people. Almost certainly, then, his concept of righteousness, both noun and verb (to be made or counted righteous, to be justified), is thoroughly Jewish too, with the same strong covenant overtones – the sort of usage we find particularly in the Psalms and Second Isaiah, where God’s righteousness is precisely God’s covenant faithfulness, his saving power and love for his people Israel.28 God’s justification is God’s recognition of Israel as his people, his verdict in favour of Israel on grounds of his covenant with Israel.


Two clarificatory corollaries immediately follow.


1. In talking of ‘being justified’ here Paul is not thinking of a distinctively initiatory act of God. God’s justification is not his act in first making his covenant with Israel, or in initially accepting someone into the covenant people. God’s justification is rather God’s acknowledgement that someone is in the covenant – whether that is an initial acknowledgement, or a repeated action of God (God’s saving acts), or his final vindication of his people. So in Galatians 2.16 we are not surprised when the second reference to being justified has a future implication (‘we have believed in Christ Jesus in order that we might be justified …’), and the third reference is in the future tense (‘by works of law no flesh shall be justified’). We might mention also Galatians 5.5, where Paul speaks of ‘awaiting the hope of righteousness’. ‘To be justified’ in Paul cannot, therefore, be treated simply as an entry or initiation formula;29 nor is it possible to draw a clear line of distinction between Paul’s usage and the typically Jewish covenant usage. Already, we may observe, Paul appears a good deal less idiosyncratic and arbitrary than Sanders alleges.


2. Perhaps even more striking is the fact which also begins to emerge, that at this point Paul is wholly at one with his fellow Jews in asserting that justification is by faith. That is to say, integral to the idea of the covenant itself, and of God’s continued action to maintain it, is the profound recognition of God’s initiative and grace in first establishing and then maintaining the covenant. Justification by faith, it would appear, is not a distinctively Christian teaching. Paul’s appeal here is not to Christians who happen also to be Jews, but to Jews whose Christian faith is but an extension of their Jewish faith in a graciously electing and sustaining God. We must return to this point shortly, but for the moment we may simply note that to ignore this fundamental feature of Israel’s understanding of its covenant status is to put in jeopardy the possibility of a properly historical exegesis. Far worse, to start our exegesis here from the Reformation presupposition that Paul was attacking the idea of earning God’s acquittal, the idea of meritorious works, is to set the whole exegetical endeavor off on the wrong track. If Paul was not an idiosyncratic Jew, neither was he a straightforward prototype of Luther.


(b) What then is Paul attacking when he dismisses the idea of being justified ‘by works of the law’? – as he does, again, no less than three times in this one verse: ‘… not by works of law … not by works of law … not by works of law …’ The answer which suggests itself from what has already been said is that he was thinking of covenant works, works related to the covenant, works done in obedience to the law of the covenant. This is both confirmed and clarified by both the immediate and the broader contexts.


As to the immediate context, the most relevant factor is that Galatians 2.16 follows immediately upon the debates, indeed the crises, at Jerusalem and at Antioch which focused on two issues – at Jerusalem, circumcision; at Antioch, the Jewish food laws with the whole question of ritual purity unstated but clearly implied. Paul’s forceful denial of justification by works of law is his response to these two issues. His denial that justification is from works of law is, more precisely, a denial that justification depends on circumcision or on observation of the Jewish purity and food taboos. We may justifiably deduce, therefore, that by ‘works of law’ Paul intended his readers to think of particular observances of the law like circumcision and the food laws. His Galatian readership might well think also of the one other area of law observance to which Paul refers disapprovingly later in the same letter – their observance of special days and feasts (Gal. 4.10). But why these particular ‘works of the law’? The broader context suggests a reason.


From the broader context, provided for us by Greco-Roman literature of the period, we know that just these observances were widely regarded as characteristically and distinctively Jewish. Writers like Petronius, Plutarch, Tacitus and Juvenal took it for granted that, in particular, circumcision, abstention from pork, and the sabbath, were observances which marked out the practitioners as Jews, or as people who were very attracted to Jewish ways.30 These, of course, were not all exclusively Jewish practices – for example, not only Jews practiced circumcision. But this makes it all the more striking that these practices were nevertheless widely regarded as both characteristic and distinctive of the Jews as a race – a fact which tells us much about the influence of Diaspora Judaism in the Greco-Roman world. It is clear, in other words, that just these observances in particular functioned as identity markers, they served to identify their practitioners as Jewish in the eyes of the wider public, they were the peculiar rites which marked out the Jews as that peculiar people.


When we set this alongside the Palestinian Judaism illuminated by Sanders, the reason for this becomes clearer, we can see why just these observances were regarded as so distinctively Jewish. The Jews regarded them in the same way! This strong impression of Greco-Roman authors, as to what religious practices characterize the Jews, was simply a reflection of the typical, the dominant, attitude of the Jews themselves. These identity markers identified Jewishness because they were seen by the Jews themselves as fundamental observances of the covenant. They functioned as badges of covenant membership. A member of the covenant people was, by definition, one who observed these practices in particular. How could it be otherwise, since precisely these practices belong so clearly to the basic ground rules of the covenant?


If we think of circumcision, no loyal Jew could ignore the explicit stipulations of Genesis 17:


And God said to Abraham, ‘As for you, you shall keep my covenant, you and your descendants after you throughout their generations. This is my covenant, which you shall keep, between me and you and your descendants after you: Every male among you shall be circumcised. You shall be circumcised in the flesh of your foreskins, and it shall be a sign of the covenant between me and you … So shall my covenant be in your flesh an everlasting covenant. Any uncircumcised male who is not circumcised in the flesh of his foreskin shall be cut off from his people; he has broken my covenant.’ (Gen. 17.9-14)


What could be clearer than that? There are some indications that a few Diaspora Jews avoided the literal force of this command by spiritualizing it,31 but they are noteworthy precisely as being so exceptional. Circumcision remained an identification marker of Jewishness, of membership of the Jewish people, in the eyes both of the Gentiles and of the Jews themselves.


The laws on clean and unclean foods do not hold such a central place in the Torah (Lev. 11.1-23; Deut. 14.3-21). But we know that at least from the time of the Maccabees they had assumed increasing importance in Jewish folklore and Jewish self-understanding. The Maccabean martyrs were remembered precisely as those who ‘stood firm and were resolved in their hearts not to eat unclean food’ and who ‘chose to die rather than to be defiled by food or to profane the holy covenant’ (1 Macc. 1.62-3). And the heroes of the popular tales beloved by several generations of Jews, Daniel, Tobit and Judith, had all shown their faithfulness to God precisely by their refusal to eat ‘the food of the Gentiles’ (Dan. 1.8-16; Tob. 1.10-13; Judith 10.5; 12.1-20). Without question, then, the devout Jew of Paul’s day would regard observance of the laws on clean and unclean foods as a basic expression of covenant loyalty. Moreover, from what we now know of the Pharisees at the time of Paul, not to mention also the Essenes at Qumran, the maintenance of ritual purity, particularly the ritual purity of the meal table, was a primary concern and major preoccupation.32 No wonder then that the men from James were so upset by the slackness of Peter and the other Jewish Christians at Antioch on these matters. And no wonder that Peter and Barnabas could not resist this strong appeal to national identity and covenant faithfulness precisely with regard to these items of the law, these practices of the covenant.


As to the observance of special days, particularly the sabbath, we need only recall that the Jewish Scriptures treat the sabbath as a fundamental law of creation (Gen. 2.3), that the sabbath was the only feast day to be stipulated in the decalogue (Exod. 20.8-11; Deut. 5.12-15), and that it was explicitly linked by Isaiah with the covenant as a determinative expression of covenant loyalty which would provide the basis on which Gentiles would unite with Jews in the last days in a common worship of the one God (Isa. 56.6-8). Here, too, was a work of the law which had the same basic character of defining the boundaries of the covenant people, one of these minimal observances without which one could hardly claim to be a good Jew, loyal to the covenant given by God’s grace to Israel.


Given this almost axiomatic tie-up between these particular regulations of the law and covenant membership, it is no exaggeration to say that for the typical Jew of the first century AD, particularly the Palestinian Jew, it would be virtually impossible to conceive of participation in God’s covenant, and so in God’s covenant righteousness, apart from these observances, these works of the law. If it helps, some may like to compare the role of the sacraments (baptism and the Lord’s Supper) in Christianity today. These have very much the same fundamental role in Christian self-understanding as circumcision, table regulation and sabbath had in the Jewish self-understanding of Paul’s day. Even though we acknowledge the Quakers and the Salvation Army as Christian bodies, even so any attempt to define the boundary markers which identify and distinguish Christians as Christians will almost certainly give a primary place to baptism and the Lord’s Supper. If an unbaptized Christian is for most of us a contradiction in terms, even more so was a Jew who did not practice the works of the law, circumcision, table regulations and sabbath.


The conclusion follows very strongly that when Paul denied the possibility of ‘being justified by works of the law’ it is precisely this basic Jewish self-understanding which Paul is attacking33 – the idea that God’s acknowledgement of covenant status is bound up with, even dependent upon, observance of these particular regulations – the idea that God’s verdict of acquittal hangs to any extent on the individual’s having declared his membership of the covenant people by embracing these distinctively Jewish rites.


Two clarificatory corollaries again follow.


1. ‘Works of law’, ‘works of the law’ are nowhere understood here, either by his Jewish interlocutors or by Paul himself, as works which earn God’s favour, as merit-amassing observances. They are rather seen as badges: they are simply what membership of the covenant people involves, what mark out the Jews as God’s people; given by God for precisely that reason, they serve to demonstrate covenant status. They are the proper response to God’s covenant grace, the minimal commitment for members of God’s people. In other words, Paul has in view precisely what Sanders calls ‘covenantal nomism’. And what he denies is that God’s justification depends on ‘covenantal nomism’, that God’s grace extends only to those who wear the badge of the covenant. This is a historical conclusion of some importance, since it begins to clarify with more precision what were the continuities and discontinuities between Paul, his fellow Jewish Christians and his own Pharisaic past, so far as justification and grace, covenant and law are concerned.


2. More important for Reformation exegesis is the corollary that ‘works of the law’ do not mean ‘good works’ in general, ‘good works’ in the sense disparaged by the heirs of Luther, works in the sense of self-achievement, ‘man’s self-powered striving to undergird his own existence in forgetfulness of his creaturely existence’ (to quote a famous definition from Bultmann).34 The phrase ‘works of the law’ in Galatians 2.16 is, in fact, a fairly restricted one: it refers precisely to these same identity markers described above, covenant works – those regulations prescribed by the law which any good Jew would simply take for granted to describe what a good Jew did. To be a Jew was to be a member of the covenant, was to observe circumcision, food laws and sabbath. In short, once again Paul seems much less a man of sixteenth-century Europe and much more firmly in touch with the reality of first-century Judaism than many have thought.


(c) In contrast to righteousness understood in terms of works of the law, Paul speaks of righteousness through faith in Jesus Christ – not just faith as such, but faith in Jesus Christ, Jesus Messiah. We are at once reminded that this is an internal Christian debate – between Paul and Peter, two Jews, but Jews who are also believers in Jesus. Paul appeals to what was obviously the common foundation belief of the new movement. What distinguishes Peter, Paul and the others from their fellow Jews is their belief in Jesus as Messiah.


But here we must be sure of what we are saying. Is it in fact this faith in Jesus (as) Messiah which marks them off from their fellow Jews, or is it their belief in justification by faith, as has so often been assumed? In the light of Sanders’ findings, as we have already noted, it is much less obvious than once appeared that the typical first-century Jew would have denied justification by faith. The emphasis on God’s electing grace, his covenantal mercy and loving-kindness, the very fact that one of Paul’s key terms, ‘the righteousness of God’, is drawn directly from the Old Testament in form and content – all this raises the question, What is the point at issue here? If not ‘justification by faith’ as God’s initiative in declaring in favour of men, if not ‘works of law’ as merit-earning good works, then what? What precisely is involved in Paul’s contrast between being justified by works of law and being justified by faith in Jesus Messiah?


Our verse suggests one answer: Paul’s point is precisely that these two are alternatives – justification by works of law and justification by faith in Jesus are antithetical opposites. To say that God’s favourable action towards anyone is dependent in any degree on works of the law is to contradict the claim that God’s favour depends on faith, faith in Jesus Christ. Indeed it is quite likely that Galatians 2.16 reflects the step by which Paul’s thinking hardened these two propositions into a clear-cut antithesis. Let me try to explain how I reach this conclusion.


According to verse 16a the common ground (between Peter and Paul) is that ‘a man is not justified by works of law except through faith in Jesus Christ’. Notice how he expresses the last phrase – ‘except through faith in Jesus Messiah’. According to the most obvious grammatical sense, in this clause faith in Jesus is described as a qualification to justification by works of law, not (yet) as an antithetical alternative. Seen from the perspective of Jewish Christianity at that time, the most obvious meaning is that the only restriction on justification by works of law is faith in Jesus as Messiah. The only restriction, that is, to covenantal nomism is faith in Christ. But, in this first clause, covenantal nomism itself is not challenged or called in question – restricted, qualified, more precisely defined in relation to Jesus as Messiah, but not denied. Given that in Jewish self-understanding covenantal nomism is not antithetical to faith,35 then at this point the only change which the new movement calls for is that the traditional Jewish faith be more precisely defined as faith in Jesus Messiah. This is evidently the accepted view of Jewish Christians to which Paul appeals.


The point, then, is that the common ground from which Paul’s argument moves out need not be understood as setting covenantal nomism and faith in Christ in antithesis. As Peter’s conduct and the conduct of the rest of the Jewish believers at Antioch made abundantly clear, so far as the Jewish Christian was concerned, belief in Jesus as Messiah did not require him to abandon his Jewishness, to give up the badges of his natonal religion, to call in question works of the law as the still necessary response of the Jew to God’s covenant grace. And why not? Why should a Jewish belief in a Jewish Messiah make any difference to these long-established Jewish distinctives?


But Paul followed a different logic – the logic of justification by faith: what is of grace through faith cannot depend in any sense, in any degree, on a particular ritual response. If God’s verdict in favour of an individual comes to effect through his faith, then it is dependent on nothing more than that. So, in repeating the contrast between justification by works of law and justification through faith in Jesus Christ, Paul alters it significantly: what were initially juxtaposed as complementary, are now posed as straight alternatives – ‘…knowing that a man is not justified from works of law except through faith in Jesus Christ, we have believed in Christ Jesus in order that we might be justified from faith in Christ, and not from works of law …’ Moreover, in describing justification by faith in Christ, Paul varies the formula slightly: we are justified not only through faith in Christ but also from faith in Christ – the implication quite probably being that in Paul’s view faith in Christ is the only necessary and sufficient response that God looks for in justifying anyone.


In other words, in verse 16 Paul pushes what began as a qualification of covenantal nomism into an outright antithesis. If we have been accepted by God on the basis of faith, then it is on the basis of faith that we are acceptable, and not on the basis of works. Perhaps, then, for the first time, in this verse faith in Jesus Messiah begins to emerge not simply as a narrower definition of the elect of God, but as an alternative definition of the elect of God. From being one identity marker for the Jewish Christian alongside the other identity markers (circumcision, food laws, sabbath), faith in Jesus as Christ becomes the primary identity marker which renders the others superfluous.


This line of exposition can be re-expressed in a slightly different way, with more emphasis on the salvation-history significance of Christ. The question Paul was in effect grappling with at this point is this: How do we Jewish believers relate our covenantal nomism, our works of law, our obligations under the covenant, to our new faith in Jesus as the Christ? Or, in slightly broader terms: What difference does the coming of Jesus the Messiah make to our traditional understanding of the covenant? The answer of many Jerusalem believers seems to have been, None; no difference; it is still God’s covenant with Israel into which Gentiles can be received on the recognized and well-established conditions. Others, including the leading apostles, were willing to dispense Gentile believers from the need to be circumcised as an entry requirement, but when it came to the ‘crunch’ they still in effect expected the Gentile believers to live as those within the covenant in traditional terms, to maintain covenant status by, in particular, conforming with the food and purity regulations which governed the meal table – even Peter and Barnabas (2.12-14). Their answer to the question was in effect: Christ’s coming has made some difference, but in the day-to-day event not much; the people of God are still to be defined in essentially and distinctively Jewish terms. But at precisely this point Paul begins to develop a different answer.


In brief, Paul’s new answer is that the advent of Christ had introduced the time of fulfillment, including the fulfillment of his purpose regarding the covenant. From the beginning, God’s eschatological purpose in making the covenant had been the blessing of the nations: the gospel was already proclaimed when God promised Abraham, ‘In you shall all the nations be blessed’ (Gal. 3.8; Gen. 12.3; 18.18). So, now that the time of fulfillment had come, the covenant should no longer be conceived in nationalistic or racial terms. No longer is it an exclusively Jewish qua Jewish privilege. The covenant is not thereby abandoned. Rather it is broadened out as God had originally intended – with the grace of God which it expressed separated from its national restriction and freely bestowed without respect to race or work, as it had been bestowed in the beginning. This is roughly the argument of Galatians 3-4, as also developed later in Romans 3-4.


The decisive corollary which Paul saw, and which he did not hesitate to draw, was that the covenant is no longer to be identified or characterized by such distinctively Jewish observances as circumcision, food laws and sabbath. Covenant works had become too closely identified as Jewish observances, covenant righteousness as national righteousness.36 But to maintain such identifications was to ignore both the way the covenant began and the purpose it had been intended to fulfil in the end. To continue to insist on such works of the law was to ignore the central fact for Christians, that with Christ’s coming God’s covenant purpose had reached its intended final stage in which the more fundamental identity marker (Abraham’s faith) reasserts its primacy over against the too narrowly nationalistic identity markers of circumcision, food laws and sabbath.


If this understanding of Galatians 2.16 is correct, then we are in fact being given the unique privilege in this verse of witnessing a very crucial development for the history of Christianity taking place, before our very eyes, as it were. For in this verse we are seeing the transition from a basically Jewish self-understanding of Christ’s significance to a distinctively different understanding, the transition indeed from a form of Jewish Messianism to a faith which sooner or later must break away from Judaism to exist in its own terms.


Once again two clarificatory corollaries.


1. We should not let our grasp of Paul’s reasoning slip back into the old distinction between faith and works in general, between faith and ‘good works’. Paul is not arguing here for a concept of faith which is totally passive because it fears to become a ‘work’. It is the demand for a particular work as the necessary expression of faith which he denies. As he puts it later in the same letter, ‘In Christ Jesus neither circumcision nor uncircumcision is of any avail, but faith working through love’ (5.6).


2. Nor should we press Paul’s distinction between faith and works into a dichotomy between faith and ritual, simply because the works of the law which he has in mind belong to what has often been called the ritual or ceremonial law. There is a distinction between outward and inward, between ritual and spiritual, but no necessary antithesis. Paul has no intention here of denying a ritual expression of faith, as in baptism or the Lord’s Supper. Here again we should keep the precise limitations of Paul’s distinction between faith in Christ and works of law before us. What he is concerned to exclude is the racial not the ritual expression of faith; it is nationalism which he denies not activism. Whatever their basis in the Scriptures, these works of the law had become identified as indices of Jewishness, as badges betokening race and nation – inevitably so when race and religion are so inextricably intertwined as they were, and are, in Judaism. What Jesus has done by his death and resurrection, in Paul’s understanding, is to free the grace of God in justifying from its nationalistically restrictive clamps for a broader experience (beyond the circumcised Jew) and a fuller expression (beyond concern for ritual purity).


(d) Finally, we should take note of the last clause of our verse, where Paul probably alludes to Psalm 143.2.37 Our thesis also helps explain why Paul should use the Psalm in the way he does, why he both modifies and adds to the Psalmist’s words. In Psalm 143.2 we read the plea:


Enter not into judgement with your servant;

For no man living is righteous before you.


Paul does two things to the second half of the Psalm verse: he adds ‘from works of law’, and he substitutes ‘all flesh’ for ‘all living’. Where the Psalmist said


No living (being) will be justified before you,


Paul rephrases thus,


By works of law no flesh will be justified.38


How can he justify restricting the more general statement by adding ‘from works of law’? The simplest answer is probably given in the substitution of ‘all flesh’ for ‘all living’. But it has the merit, for Paul, of focusing the unacceptability of man in his fleshliness. By that, of course, Paul will not intend a dualism between spirit and matter, however dualistic his antithesis between spirit and flesh may seem later on in Galatians 5. He certainly has in mind man’s weakness, his corruptibility, his dependence on the satisfaction of merely human appetites (4.13-14; 5.16-17; 6.8). But the word ‘flesh’ also embraces the thought of a merely human relationship, of a heritage determined by physical descent, as in the allegory of Galatians 4 (4.23, 29).39 That is to say, in speaking of ‘all flesh’ Paul has in view primarily and precisely those who think their acceptability to God and standing before God does depend on their physical descent from Abraham, their national identity as Jews. It is precisely this attitude, which puts too much stress on fleshly relationships and fleshly rites, precisely this attitude which Paul excoriates in his parting shot in 6.12-13 – ‘they want to make a good showing in the flesh … they want to glory in your flesh’.


With the Psalm reference thus more sharply defined in terms of physical and national identity, the addition of ‘from works of law’ becomes merely clarificatory. It does not narrow the Psalmist’s assertion any further; rather it ties into and emphasizes more clearly the ‘all flesh’. For works of the law, epitomized in this letter by circumcision, are precisely acts of the flesh. To insist on circumcision is to give a primacy to the physical level of relationship which Paul can no longer accept. ‘Works of the law’, because they put such an emphasis on such marks of racial identity, are, ironically, no different from ‘works of the flesh’ (5.19), so far as acceptability before God is concerned – precisely because these works of the law in effect imprison God’s righteousness within a racial and national, that is, fleshly framework. Whereas those who belong to Christ, from Paul’s perspective, have passed through a different starting-point (the gift of the Spirit – 3.3), have crucified the flesh (5.24), and the life they now lead in the flesh they live not in terms of fleshly rites or fleshly relationships but by faith in the Son of God (2.20). God’s purposes and God’s people have now expanded beyond Israel according to the flesh, and so God’s righteousness can no longer be restricted in terms of works of the law which emphasize kinship at the level of the flesh.

Two final corollaries by way of clarification.


1. Yet once more we must note that it is works of the law that Paul disparages, not the law itself or law-keeping in general. In his latest contribution to the discussion Sanders recognizes the nationalistic significance of circumcision, food laws and sabbath,40 but he keeps taking the phrase ‘works of the law’ as though it was simply a fuller synonym for ‘law’. So far as Sanders is concerned, ‘no man shall be justified by works of law’ is just the same as saying, ‘no man shall be justified by the law’.41 But Paul is as little opposed to the law per se as he is to good works per se. It is the law understood in terms of works, as a Jewish prerogative and national monopoly, to which he takes exception. The law understood in terms of the command to ‘love your neighbor as yourself’ is another matter (Gal. 5.14).


2. So, too, lest the point still be confused, I repeat, Paul here is not disparaging works in general or pressing a dichotomy between outward ritual done in the flesh and inward grace operative in the spirit. Once again we must observe the limited target he has in his sights. It is works which betoken racial prerogative to which he objects, acts done in the flesh because faith in Christ is reckoned insufficient as the badge of covenant membership which he denounces. Over against Peter and the other Jewish Christians Paul insists that God’s verdict in favour of believers comes to realization through faith, from start to finish, and in no way depends on observing the works of law which hitherto had characterized and distinguished the Jews as God’s people.




So much for Galatians 2.16. Time does not permit me to follow the development of the same line of argument through the rest of the letter, though I believe that it helps resolve more than one crux in subsequent chapters. Likewise, Paul’s later letter to the Roman Christians gains considerably in coherence when viewed from the same perspective. For example, when in Romans 3.27 Paul affirms that boasting is excluded, he is not thinking of boasting in self-achievement or boasting at one’s good deeds.42 It is the boasting of the Jew which he has in mind – the boasting in Israel’s special relationship with God through election, the boasting in the law as the mark of God’s favour, in circumcision as the badge of belonging to God (Rom. 2.17-29). Among other things, this means that there is no significant development in Paul’s thought on this particular point, at least, between Galatians and Romans. However, further exposition will have to await the commentary on Romans which I mentioned at the beginning, and which, as you may appreciate, I am now a good deal more enthusiastic about writing than I was when first asked.


It would also be premature, of course, to build extensive conclusions on the basis of just one verse. Nevertheless, there is some obligation at the end of a lecture like this to attempt some summing up, and at least to sketch out the preliminary results which seem to follow so far from this new perspective on Paul, but which must naturally be subjected to further testing.


(a) In Galatians 2.16 Paul actually addresses Judaism as we know it to have been in the first century – a system of religion conscious of its special relationship with God and sensitive to its peculiar obligations within that relationship. The criticisms of Paul for his misunderstanding of Judaism therefore involve a double failure of perspective. What Jewish scholars rejected as Paul’s misunderstanding of Judaism is itself a misunderstanding of Paul, based on the standard Protestant (mis)reading of Paul through Reformation spectacles. When we take these Reformation spectacles off, Paul does not appear to be so out of touch with his first-century context as even Sanders thinks. Sanders in effect freed Pauline exegesis from its sixteenth-century blinkers, but he has still left us with a Paul who could have made little sense to his fellow Jews and whose stated willingness to observe the law elsewhere (1 Cor. 9.19-23) must have sounded like the most blatant self-contradiction.


(b) The major exegetical flaw of Sanders’ reconstruction of Paul’s view of the law (and of course not only his)43 is his failure to perceive the significance of the little phrase ‘works of the law’. He recognizes rightly that in disparaging ‘works of the law’ Paul is not disparaging good works in general, far less is he thinking of good works as earning merit. But by taking ‘works of law’ as equivalent to ‘doing the law’ in general (the normal exegesis), he is led to the false conclusion that in disparaging ‘works of the law’ Paul is disparaging law as such, has broken with Judaism as a whole. To be fair, the mistake is a natural one, since Judaism had itself invested so much significance in these particular works, so that the test of loyalty to covenant and law was precisely the observance of circumcision, food laws and sabbath.44 But it is these works in particular which Paul has in mind, and he has them in mind precisely because they had become the expression of a too narrowly nationalistic and racial conception of the covenant, because they had become a badge not of Abraham’s faith but of Israel’s boast.45 Although Sanders has seen this point quite clearly,46 he does not follow it through, and his failure to distinguish ‘works of the law’ from ‘doing the law’ prevents him from developing the insight properly.47


This failure has had serious consequences for Sanders’ larger thesis. For had he delimited more precisely the force of Paul’s negative thrust against works of the law, he would have been able to give a more adequate account of Paul’s more positive attitude to the law elsewhere. In particular, he would not have had to press so hard the distinction between ‘getting in’ (not by doing the law) and ‘staying in’ (by keeping the law), a distinction which seems very odd precisely at Galatians 2.16, where the issue at Antioch was the day-to-day conduct of those who had already believed (2.14), and where Paul’s concern regarding the Galatians is over their ending rather than their beginning (3.3).48 In consequence also he would not have had to argue for such an arbitrary and abrupt discontinuity between Paul’s gospel and his Jewish past, according to which Sanders’ Paul hardly seems to be addressing Sanders’ Judaism. Whereas, if Paul was really speaking against the too narrow understanding of God’s covenant promise and of the law in nationalist and racial terms, as I have argued, a much more coherent and consistent reconstruction of the continuities and discontinuities between Paul and Palestinian Judaism becomes possible.


(c) All this confirms the earlier important thesis of Stendahl, that Paul’s doctrine of justification by faith should not be understood primarily as an exposition of the individual’s relation to God, but primarily in the context of Paul the Jew wrestling with the question of how Jews and Gentiles stand in relation to each other within the covenant purpose of God now that it has reached its climax in Jesus Christ.49 It is precisely the degree to which Israel had come to regard the covenant and the law as coterminous with Israel, as Israel’s special prerogative, wherein the problem lay. Paul’s solution does not require him to deny the covenant, or indeed the law as God’s law, but only the covenant and the law as ‘taken over’ by Israel. The models of the man of faith are for Paul the founding fathers, Abraham, Isaac and Jacob, where covenant membership was neither determined by physical descent (racial consanguinity) nor dependent on works of law (Rom. 4; 9.6-13). This certainly involved something of an arbitrary hermeneutical procedure, whereby the example of Abraham in particular was treated not only as typical and normative, but also as relativizing those subsequent Scriptures which emphasize Israel’s special place within God’s affections. But it is a procedure which Paul is more than willing to argue for and defend rather than simply to state in a take-it-or-leave-it, black-and-white way.


Once again, however, we are beginning to push too far beyond the proper limits of the present essay, and I must desist. But hopefully I have said enough to show how valuable the new perspective on Paul may be in giving us a clearer insight into and appreciation of him and his theology.




  1. F.F. Bruce, Paul: Apostle of the Free Spirit (Exeter, 1977); O. Kuss, Paulus: die Rolle des Apostels in der theologischen Entwicklung der Urkirche (Regensburg, 1971); H. Schlier, Grundzüge einer paulinischen Theologie (Freiburg/Basel/Wien, 1978).
  2. A. Suhl, Paulus und seine Briefe: ein Beitrag zur paulinischen Chronologie (Gütersloh, 1975); R. Jewett, Dating Paul’s Life (London, 1979); G. Lüdemann, Paulus, der Heidenapostel Band I: Studien zur Chronologie (Göttingen, 1980).
  3. S. Kim, The Origin of Paul’s Gospel (Tübingen, 1981); J. C. Beker, Paul the Apostle: the Triumph of God in Life and Thought (Philadelphia, 1980).
  4. A. Lindemann, Paulus im ältesten Christentum (Tübingen, 1979).
  5. J. W. Drane, Paul: Libertine or Legalist? (London, 1975); H. Hübner, Das Gesetz bei Paulus (Göttingen, 1978, 21980).
  6. See particularly the work of G. Theissen, Studien zur Soziologie des Urchristentums (Tübingen, 1979), partial ET, The Social Setting of Pauline Christianity (Edinburgh, 1982).
  7. See, e.g., K. P. Donfried, ed., The Romans Debate (Minneapolis, 1977); R. McL. Wilson, ‘Gnosis in Corinth’, Paul and Paulinism: Essays in Honour of C. K. Barret, ed. M. D. Hooker and S. G. Wilson (London, 1982), pp. 102-14; G. Howard, Paul: Crisis in Galatia (Cambridge, 1979).
  8. See particularly H. D. Betz, ‘The Literary Composition and Function of Paul’s Letter to the Galatians’, NTS, 21 (1974-5), pp. 353-79; also Galatians (Hermeneia: Philadelphia, 1979); W. Wuellner, ‘Paul’s Rhetoric of Argumentation in Romans’, CBQ, 38 (1976), pp. 330-51, reprinted in The Romans Debate (n. 7 above), pp. 152-74; also ‘Greek Rhetoric and Pauline Argumentation’, Early Christian Literature and the Classical Intellectual Tradition: in honorem R. M. Grant, ed. W. R. Schoedel and R. L. Wilken (Paris, 1979), pp. 177-88; R. Jewett, ‘Romans as an Ambassadorial Letter’, Interpretation, 36 (1982), pp. 5-20.
  9. I refer particularly to Jesus and the Spirit (London, 1975) and Christology in the Making (London, 1980).
  10. E. P. Sanders, Paul and Palestinian Judaism: a Comparison of Patterns of Religion (London, 1977). Cf. the estimate of W. D. Davies in the Preface to the fourth edition of his Paul and Rabbinic Judaism (Philadelphia, 1981): ‘a work of immense learning and penetration, a major milestone in Pauline scholarship … of potentially immense significance for the interpretation of Paul’ (pp. xxix-xxx).
  11. Sanders, Paul, p. 6. See the fuller survey ‘Paul and Judaism in New Testament scholarship’ on pp. 1-12.
  12. Sanders traces the dominance of this very negative evaluation of the Judaism of Paul’s time back to F. Weber, System der altsynagogalen palästinischen Theologie aus Targum, Midrasch und Talmud (1880), revised as Jüdische Theologie auf Grund des Talmud und verwandter Schriften (Leipzig, 1897). For the following paragraph see also Sanders on ‘The persistence of the view of Rabbinic religion as one of legalistic works-righteousness’ (Paul, pp. 33-59).
  13. K. Stendahl, ‘The Apostle Paul and the Introspective Conscience of the West’, HTR, 56 (1963), pp. 199-215, reprinted in his Paul Among Jews and Gentiles (London, 1977), pp. 78-96. See also the several recent contributions by W. D. Davies in this area – ‘Paul and the People in Israel’, NTS, 24 (1977-8), pp. 4-39; also Paul and Rabbinic Judaism4, pp. xxvii f.; also ‘Paul and the Law: Reflection on Pitfalls in Interpretation’, Paul and Paulinism (n. 7 above), pp. 4-16.
  14. e.g. R. Bultmann, Jesus Christ and Mythology (London, 1960); ‘Demythologizing is the radical application of the doctrine of justification by faith to the sphere of knowledge and thought’ (p. 84); E. Käsemann, Das Neue Testament als Kanon (Göttingen, 1970): ‘Die Rechtfertigung des Gottlosen … muss als Kanon im Kanon betrachtet werden …’ (p. 405).
  15. Hübner (n. 5 above).
  16. Sanders, Paul, pp. 75, 420, 544. Worth noting is the fact that J. Neusner, though fiercely critical of Sanders’ methodology, nevertheless accepts Sanders’ understanding of Judaism in terms of ‘covenantal nomism’ as valid. That rabbinic discussions presupposed the covenant and ‘were largely directed toward the question of how to fulfil the covenantal obligations’ is to Neusner a ‘wholly sound and … self-evident proposition’. ‘So far as Sanders proposes to demonstrate the importance to all the kinds of ancient Judaism of covenantal nomism, election, atonement, and the like, his work must be pronounced a complete success’ – ‘Comparing Judaisms’, History of Religions, 18 (1978-9), pp. 177-91 (here pp. 177, 180).
  17. See particularly Sanders, Paul, pp. 550-2.
  18. H. Hübner, ‘Pauli Theologiae Proprium’, NTS, 26 (1979-80), pp. 445-73.
  19. H. Räisänen, ‘Legalism and Salvation by the Law’, in Die Paulinische Literatur und Theologie, hrsg. S. Pedersen (Göttingen, 1980), pp. 63-83.
  20. M. Hooker, ‘Paul and Covenantal Nomism’, Paul and Paulinism (n. 7 above), pp. 47-56.
  21. Sanders, Paul, the Law, and the Jewish People, p. 47.
  22. cf. H. Räisänen, ‘Paul’s Theological Difficulties with the Law’, Studia Biblica 1978, vol. III, ed. E. A. Livingstone (JSNTSupp. 3: Sheffield, 1980), pp. 301-20.
  23. See J. D. G. Dunn, ‘The Incident at Antioch (Gal. 2.11-18)’, reprinted above, ch. 6.
  24. ‘Incident at Antioch’, ch. 6 above, p. 172, n. 117.
  25. It is unlikely that Paul wrote είδότες δέ. (1) δέ is omitted by P46 as well as by other important manuscripts, and was probably introduced by a scribe who misread the flow of Paul’s thought and assumed that an adversative particle should be added. (2)  Had Paul wished to give adversative force he would have more probably written ήμεîς έσμέν φύσεί Ιουδαĩοι … οĭδαμεν δέ … (contrast Rom. 6.9 and 2 Cor. 4.14 with Rom. 8.28). In fact what he wrote is ‘We Jews by nature … knowing that …’ (cf. H. Schlier, Galater [Göttingen, 41965], p. 89). If he did not follow the construction through consistently, that is hardly untypical of Paul. (3) The έάν μή confirms that v. 16a is intended to express the Jewish (Christian) understanding of justification through faith (see p. 195 above).
  26. Clem. Hom. 11.16 – ‘The Jew believes God and keeps the law … But he who keeps not the law is manifestly a deserter through not believing God; and thus is no Jew, but a sinner …’ Cf. K. Kertelge, ‘Zur Deutung des Rechtfertigungsbegriffs im Galaterbrief’, BZ, 12 (1968), p. 213; U. Wilckens, ‘Was heisst bei Paulus: “Aus Werken des Gezetzes wird kein Mensch gerecht”?’, (1969), Rechtfertigung als Freiheit: Paulusstudien (Neukirchen, 1974), pp. 87-8; F. Mussner, Galaterbrief (Freiburg/Basel/Wien, 31977), pp. 167-9.
  27. See Dunn, ‘Incident at Antioch’, ch. 6 above, §4.1c; also J. D. G. Dunn, Romans (Dallas, 1988), pp. lxix-lxx.
  28. See particularly S. K. Williams, ‘The “Righteousness of God” in Romans’, JBL, 99 (1980), pp. 260f.; Dunn, Romans, pp. 40-2. For references to the Dead Sea scrolls see also Mussner, Galaterbrief, pp. 168f.
  29. Sanders repeatedly emphasizes that ‘to be righteoused’ (sic) in Paul is ‘transfer terminology’. See further, Dunn, Romans, p. 97.
  30. Full details in M. Stern, ed., Greek and Latin Authors on Jews and Judaism (Israel Academy of Sciences and Humanities: Jerusalem, Vol I, 1976, Vol. II, 1980), §§ 194, 195, 258, 281, 301.
  31. See Philo, Migr., 89-93; cf. Qu. Ex., II, 2.
  32. See particularly J. Neusner, From Politics to Piety (Englewood Cliffs, 1973), pp. 80, 83-90.
  33. Kertelge (n. 26 above): ‘Die erga nomou in v. 16 sind also der Ausdruck des jüdischen Selbstbewusstseins von v. 15’ (p. 215).
  34. R. Bultmann, Theology of the New Testament, I (ET, London, 1952), p. 254. Cf., e.g., H. Ridderbos, Paul: an Outline of his Theology (1966; ET, London, 1977), p. 139; E. Käsemann, Romans (HNT, 1973; ET, London, 1980), pp. 93, 102, 284; Hübner, Gesetz, p. 102; Becker, Paul, p. 247.
  35. Mussner, Galaterbrief: ‘Der Jude lässt die pln. Antithetik “Glaube” – “Werke des Gesetzes” – nicht gelten, ja sie ist ihm unverständlich’ (p. 170).
  36. A phrase I owe to N. T. Wright; see his Oxford D. Phil. thesis: The Messiah and the People of God: a study in Pauline Theology with particular reference to the argument of the Epistle to the Romans (1980), pp. 89f.
  37. Despite Mussner’s misgivings (Galaterbrief, pp. 174f.), Paul probably did intend an allusion to the Psalm, as the parallel with Rom. 3.20 confirms, since the allusion is clearer there.
  38. The omission of ‘before you’ from Ps. 143.2 in Gal. 2.16 has no significance, as the retention of the phrase in the Rom. 3.20 allusion to the same passage makes clear.
  39. See J. D. G. Dunn, ‘Jesus – Flesh and Spirit: an Exposition of Romans 1.3-4’, JTS, 24 (1973), pp. 43-9.
  40. See n. 46 below.
  41. See also E. P. Sanders, ‘On the Question of Fulfilling the Law in Paul and Rabbinic Judaism’, Donum Gentilicum: New Testament Studies in Honour of David Daube, ed. C. K. Barrett, E. Bammel and W. D. Davies (Oxford, 1978), pp. 103-26.
  42. Contrast those cited in n. 34 above.
  43. It is unfair to pick out Sanders, since this is the common view of the matter, usually the result of basing exegesis primarily on Gal. 3.11 and of reading 3.10 in the light of it without sufficient reference to the initial emphatic statement of 2.16. See, e.g., N. A. Dahl, Studies in Paul (Minneapolis, 1977), pp. 106, 170; U. Wilckens, ‘Zur Entwicklung des paulinischen Gesetzverständnis’, NTS, 28 (1982), pp. 166-9; Mussner, Galaterbrief: ‘Nur eine naive Exegese könnte … “die Werke des Gesetzes” auf die rituellen Vorschriften des Judentums beschranken’ (p. 170). But see n. 45 below.
  44. We may compare the way in which in fundamentalist circles of substitutionary atonement and inerrancy of Scripture have been regarded as touchstones of orthodoxy, even when several other doctrines are acknowledged to be of equal or greater importance.
  45. The same point applies to the distinction between the ritual and the moral law frequently attributed to Paul. The point is that Paul does not presuppose or develop that distinction as such. His more negative attitude to the ritual prescriptions of the law arises from the fact that it is precisely in and by these rituals as such that his Jewish kinsmen had most clearly marked themselves out as God’s people the Jews – and been identified by others as ‘that peculiar people’ (see n. 30 above).
  46. See, e.g., his Paul, the Law, and the Jewish People, p. 33 – ‘Boasting’ in Rom. 3.27 refers to ‘the assumption of special status on the part of the Jews’ (also p. 35); his recognition of the significance of circumcision, sabbath and food laws (pp. 101-2) – ‘the most obvious common denominator to these laws is the fact that they distinguish Jews from Gentiles’ (p. 114); and his quotation from Gaston (‘Israel as a whole interpreted the righteousness of God as establishing the status of righteousness for Israel alone, excluding the Gentiles’) and Howard (‘Their own righteousness’ is their ‘collective righteousness to the exclusion of the Gentiles’) in his notes (p. 61, n. 107). The earlier article by J. B. Tyson, ‘”Works of Law” in Galatians’, JBL, 92 (1973), pp. 423-31, shares similar strengths and weaknesses.
  47. e.g. ‘The explanation of “not by faith but by works”, then, is “they did not believe in Christ” … Israel’s failure is not that they do not obey the law in the correct way, but that they do not have faith in Christ’ (p. 37) – where I would rather say, ‘they relied on their covenant status, as attested by the works of the law, rather than on Christ’; ‘His criticism of his own former life is not that he was guilty of the attitudinal sin of self-righteousness, but that he put confidence in something other than faith in Jesus Christ’ (p. 44) – Tertium datur!, … guilty of putting his confidence in his being a Jew and in his zeal as a devout Jew; ‘The only thing that is wrong with the old righteousness seems to be that it is not the new one’ (p. 140) – No! that it was too narrowly and nationalistically ‘Jewish’; ‘In Pauline theory, Jews who enter the Christian movement renounce nothing’ (p. 176) – except their claim to a Jewish monopoly of divine righteousness.
  48. Sanders tries to grapple with this point in his first main section of Paul, the Law, and the Jewish People (p. 52, n. 20), and in effect acknowledges that the issue is ‘being in’ (what covenant membership involves) rather than a distinction between getting in and staying in as such. The Jewish Christians and Judaizers wanted not simply a one-off action from the Gentile believers, but a continuing lifestyle in accordance with the Torah.
  49. cf. Stendahl, Paul among Jews and Gentiles, passim – e.g. ‘… a doctrine of faith was hammered out by Paul for the very specific and limited purpose of defending the rights of Gentile converts to be full and genuine heirs to the promises of God to Israel’ (p. 2).


Additional Note


‘The New Perspective’ brought a number of responses – particularly rejoinders from H. Räisänen and H. Hübner, the former delivered at the Basel SNTS Seminar at which I also delivered the first version of my ‘Works of the Law’ paper (ch. 8 below), and in which Hübner also participated. Räisänen’s paper was published as ‘Galatians 2.16 and Paul’s Break with Judaism’, NTS 31 (1985) 543-53, reprinted in The Torah and Christ, Helsinki: Finnish Exegetical Society 45 (1986) 168-84. And in the same year Hübner’s response also appeared, ‘Was heisst bei Paulus “Werke des Gesetzes”?’, Glaube und Eschatologie, W. G. Kümmel Festschrift, ed. E. Grasser and O. Merk, Tübingen: Mohr (1985) 123-33.


Räisänen’s criticism is the most sustained. He makes four principal points.


1. The noun phrase δικαιοσύνη θεοû and the verb διακαιοûσθαι must be carefully distinguished. Galatians 2.16 does have in view entry into the community; δικαιοûσθαι there is ‘transfer terminology’ and is different from pre-Christian Jewish usage.1


2. The continuity between Judaism and Paul is overstressed. The faith of which Galatians 2.16 speaks is not faith in the sense of (Jewish) recognition of God’s covenant grace. It is faith in Jesus Christ, and this is something novel in Judaism.2 And even if ‘justification by faith’ could be described as a Jewish theologoumenon, Dunn does acknowledge that Paul’s further emphasis on justification by faith in Christ as an antithetical alternative to covenantal nomism was bound to result in a break away from Judaism.3


3. The suggestion of a movement in thought within Galatians 2.16 does not square with the text. ‘The flow of thought in verse 16 is quite smooth. There is no formal indication of a contrast between the beginning and the end … Justification by works of the law is denied throughout verse 16, as it is in the rest of the letter.’4 Stuhlmacher (see below) also made the same point: can Galatians 2.16a and 16b be read in the sense of a ‘radicalizing gradation’?


4. The attack Paul mounts is not simply on a particular attitude to the law,5 but on the law itself. Jews as well as Gentiles must enter the new community. Paul’s positive comments on the law are not to be explained away: they reflect Paul’s attempt to maintain and emphasize continuity with Judaism (‘Dunn comes close to describing Paul’s position as Paul himself wished it to be understood’). But his actual teaching on the law was inconsistent with that objective; his position amounted to a de facto break with Judaism (as Sanders has so clearly seen).6 In short, ‘Paul’s critique of the law is much more radical than Dunn allows and we should not shrink from speaking of his “break” with Judaism.’7


In response:


1. I am not impressed by attempts to pull δικαιοσύνη θεοû and δικαιοûσθαι too far apart. Of course the terms δικαιοσύνη, δικαιοσύνη θεοû and δικαιοûσθαι have different functions and different ranges of reference. But these functions and ranges also overlap. In particular, it is ‘the righteousness of God’ which ‘makes righteous’, as is surely clear in Romans 3.21-4, 26; and ‘righteousness’ is the effect of God ‘making righteous’, as is clear in Romans 4.2-3, 5, Galatians 3.6, 8 and 5.4-5. My point is that this degree of overlap prevents a neat dichotomy of reference, whereby the verb can be limited to ‘transfer terminology’.


(a) ‘Righteousness’ is not only a status granted at conversion (by ‘transfer’), but can also be used in reference to an ongoing status, or living relationship (as in Rom. 5.21), and to describe the end-point of the whole process (as in Rom. 6.16 and Gal. 5.5). ‘The righteousness of God’ is to be seen, therefore, as the outgoing power of grace which grants, sustains and finally secures that ‘righteousness’, not just a once-for-all act of transfer.


(b) The verb certainly does denote ‘transfer’ on several occasions; nearly half the relevant Pauline uses are aorist and perfect tenses. But more than half are present and future tenses. To be sure, the present tenses could be taken as ‘timeless’ presents, but most of the future tenses are best taken as referring to future (=final) justification (on the day of judgement) (Rom. 2.1; 3.20; Gal. 2.16; 5.4). Moreover, the aorists in Romans 8.30 appear to cover (in retrospect) the whole salvation process that lies between ‘being called’ and ‘being glorified’; that is to say, they probably embrace the whole process from initial acceptance by God (‘transfer’) to the final vindication at the seat of judgement.8 So too in Galatians 2.17 the aorist includes the seeking of justification as an ongoing goal. When these two sets of considerations are taken together (a and b), it becomes clearer that by ‘the righteousness of God’ Paul means the power of divine grace which is effective ‘for salvation’ (Rom. 1.16-17), from first to last.9


(c) The crucial fact remains that in the Antioch incident, and in Galatians, Paul was confronted by a view which insisted that covenant status could not be sustained without ‘works of the law’. In Jewish covenant theology, that also meant final vindication could not be assured without ‘works of the law’. And in the Jewish-Christian adaptation of that, covenant status and final vindication depended on justification by faith completed by ‘works of the law’ (the clear implication of Gal. 3.2-5; cf. Jas. 2.22-4). Paul’s point is to insist precisely that the ongoing process of salvation is wholly of a piece with its beginning; that as their initial acceptance by God was through faith, so is their continuation (Gal. 3.2-5) and their final acceptance (Gal. 5.5).10 Consequently the range of tenses in Galatians 2.16 probably denotes a richer theology of justification than Räisänen allows. To paraphrase the verse: ‘Since man is justified through faith in Jesus Christ (the present tense can cover the whole process), we have believed in Christ Jesus (aorist = ‘transfer’) in order that we might be justified from faith in Christ and not from works of law (the aorist tense can refer to the goal of the whole process, as in 2.17 – the point being that justification is by faith from start to finish) because (as will become apparent at the last judgement) “no flesh will be justified by works of the law”.’ This seems a superior solution to Räisänen’s, who can only maintain his attempt to limit the verb to ‘transfer terminology’ by allowing that ‘in effect one has to “enter” twice: first here and then at the final judgment’.11 With this admission my point has been largely conceded: Galatians 2.16 has in view not only the initial act of acceptance, but the question of what then is necessary to ensure final acceptance.


2. Of course Paul has in mind not just justification by faith, but justification by faith in Christ. Justification by faith in Christ is, if you like, the Jewish-Christian refinement of Jewish election theology, which I characterized as ‘justification by faith’ to underscore the presupposition of divine grace which is central to that theology. It is that Jewish-Christian understanding which provides Paul with sufficient common ground for his dialogue with his fellow Jewish believers in Christ, and out of that Paul develops his own more characteristic emphasis (Gal. 2.15-16). I do not dispute that the end result of this development was a breach between (rabbinic) Judaism and Christianity. I do dispute that this was ever Paul’s intention or that it was inevitable within the context of the much broader stream of pre-70 Judaism. Within that broader stream Paul’s interpretation of covenant and promise was a legitimate option for Jews (and Judaism) within a wider range of options.


3. I recognize that the lack of a clear adversative at the start of the second clause of Galatians 2.16 makes for a distinct weakness in my exegesis. However, there is an equal or more serious weakness in the more normal exegesis, which can be posed thus. What is it that ‘we Jews know’ (Gal. 2.15-16)? That ‘justification is by faith in Jesus Christ’ – Yes! So clearly Paul has Christian Jews in mind, and almost certainly his rebuke to Peter at Antioch, or at least that situation.12 But the problem is that these same Christian Jews (at Antioch) had also evidenced their belief that Christian Jews must still practise ‘the works of the law’ (in this case the Jewish food laws). So they evidently did not accept, in any practice at any rate, that justification is by faith in Christ and not by works of law. In Galatians 2.16 Paul therefore attempts to press this antithesis as an ‘either-or’ (against the Jewish-Christian ‘both-and’). As initial justification was by faith in Christ, and not by works of the law, so the outworking of that initial ‘transfer’ should be consistent with it – justification by faith in Christ and not by works of the law, for that will be the basis of the final judgement (Ps. 143.2). In short, Räisänen has grounds for his objection when Galatians 2.16 is read on its own, but when the context is taken into consideration the logic of the verse seems to favour the exegesis in chapter 7 above.13


4. I am not particularly happy with Räisänen’s characterization of my understanding of Paul as attacking ‘a particular attitude to the law’, though I do use the phrase as one of a number of possible ways of distinguishing my view of the matter from that of Räisänen. What I am more clear on is that Räisänen’s own characterization of Paul as critiquing the law itself is much too blunt. Paul does seem to critique the law in a passage like Galatians 2.19. But he also defends the law in Romans 7.7-25 and affirms the law in passages like Romans 3.31, 8.4 and Galatians 5.14. And his critique is regularly more nuanced – law as ‘letter’ (Rom. 2.27; 7.6; 2 Cor. 3.6), ‘works of the law’ (Gal. 2.16; etc.), law as used by sin (Rom. 7.7-14), law understood as defining the scope of God’s righteousness (Rom. 10.3-4), law as keeping Israel in a state of immaturity (Gal. 3.23 – 4.5), and so on. Whether this is better described as an ‘attitude to the law’ or the law perceived as operating in a certain way, defining the people of God, making specific demands on that people (‘works of the law’), is perhaps a merely semantic quibble. What I am more certain of is that Räisänen has missed (as have many writing in this area) both such nuances and what I go on to describe in the next chapter as ‘the social function of the law’.14 Consequently I find Paul’s treatment of the law a good deal more consistent than either Sanders or Räisänen. And talk of a ‘break’ with the law or with Judaism becomes less justified. To assume that baptism in the name of Jesus meant not just entry into the community (sect?) of believers in Jesus, but also in effect an exit from Judaism is a serious misperception of the self-understanding of both Jewish Christians and Paul. The ‘new covenant’ did not mean a new religion, but the eschatological fulfillment of the old in Messiah Jesus, not a third olive tree!,15 but the full flourishing of the one olive tree cultured by God’s electing grace (Rom. 11.16-24).


5. Besides Räisänen’s sharp critique of ‘The New Perspective’, Hübner seems to have been content largely to describe ‘The New Perspective’ and to echo the classic treatments of such as Bultmann and Käsemann that by ‘works of the law’ Paul has in mind the human attempt to lay claim on God, the thought of self-boasting before God, the law perverted as a means to self-justification.16 He has therefore not really recognized or addressed the challenge posed to that view by ‘the new perspective’. And since I have continued to elaborate that perspective and to show how deep are its foundations in the Pauline texts and the historical context, it may be sufficient to refer to the following chapter and to my Romans.17


In contrast to both the above, one may simply note the much more positive acceptance of the basic thesis of Chapter 7, from the Catholic side, by K. Kertelge, ‘Gesetz und Freiheit im Galaterbrief’, NTS 30 (1984) 382-94, particularly 391 and n. 24.


‘The New Perspective’ also drew some private correspondence with P. Stuhlmacher (24.9.83, my reply 3.10.83), who was then presumably working on his own response to Sanders and Räisänen;18 and with E. P. Sanders (12.9.94, my reply 1.10.84), the principal debate partner in ‘The New Perspective’.


The correspondence with Stuhlmacher clarified some points, and he lodged two further objections. (6) ‘Works of the law’ cannot be reduced to circumcision, food commandments and sabbath observance, as Romans 2.17ff. and 3.9-20 show. (7) Philippians 3.4-11 shows that the justification issue is as old as his calling; it is from here that his ‘teaching’ is to be determined, and not just from the Antioch controversy.


6. On the first point, I fully agree that Paul’s critique does not reduce to questions of circumcision, food laws and sabbath, and I acknowledge that my formulation of the issue in ‘The New Perspective’ could be read too easily as though that is all that Paul had in mind when he spoke of ‘works of the law’. This was one of the reasons why I returned to the subject in the following essay (ch. 8). There I make the point that the law had become too closely identified with Israel, as marking out Israel and marking off Israel from other nations, so that Gentiles by definition were ‘outside the law’, ‘sinners’ = outlaws. For those ‘within the law’, ‘works of the law’ are what the law requires; the two phrases are co-ordinate; ‘works of the law’ are what Sanders refers to in his phrase ‘covenantal nomism’. But Israel’s history had reinforced the reality of the law as a boundary dividing Israel from the (other) nations, and the Maccabean crisis in particular had focused that boundary function on two or three key ‘make or break’ issues – especially circumcision and food laws. They remained prominent at the time of Paul, for the same reason. In short, that is why it is precisely circumcision and food laws which are so much to the fore when Paul speaks of ‘works of the law’ in Galatians – not because they are the only ‘works’ which the law requires, but because they had become the crucial test cases for covenant loyalty and for maintaining Jewish identity as the people chosen by God for himself alone.


7. The second charge had in effect already been made by Houlden in his response to ‘The Incident at Antioch’ (ch. 6), and I may refer simply to the Additional Note 1 there. Paul’s insight does go back to his call on the Damascus road, but I do think it took the circumcision issue at Jerusalem and the food laws controversy at Antioch to bring out for Paul the full significance of his initiatory revelation and of the agreement that justification is by faith (alone!).


8. The letters exchanged with Sanders were lengthy and served to clarify the areas of our agreement and disagreement. I hope it is sufficiently clear from the above chapter how great is my indebtedness to Sanders. ‘The New Perspective’ is the one which he has made possible. I did attempt to document that indebtedness,19 and could have noted further examples of agreement. Moreover, I was probably unfair in the original text at n. 46, in which I speak of Sanders only ‘glimpsing’ the point, and which I therefore have modified for this reprint. In view of the extent of our agreement it became important to understand the ‘why’ and to clarify the ‘what’ of our disagreement. In my letter of 1.10.84 I suggested two main reasons.


(a) In his earlier book Sanders posed Paul’s Christianity and Judaism as two different ‘systems’.20 It is probably this which heightens the impression for Sanders that Paul’s ‘conversion’ was a ‘transfer’, involving a complete ‘break’ with the law and rejection of the law.21 In my opinion, however, the break was not as radical as Sanders maintains and Paul was not so out of tune with the rest of first-century Judaism as Sanders believes. So, for example, when Sanders says, ‘Israel’s failure is not that they do not obey the law in the correct way, but that they do not have faith in Christ’,22 I disagree. Paul does criticize his fellow Jews for ‘not obeying the law in the correct way’ (Rom. 3.27, 31; 9.30-2; Gal. 3.10-11). Nor do I agree that Paul ‘cast off, denied … the election and the law’,23 only an abuse and misunderstanding in particularist (Sanders’ term) and nationalist terms (my word).24 In all these cases I would argue that my distinction between ‘works of the law’ and ‘the law’ is important, because ‘works of law’ denotes that way of regarding the law (as an identity and boundary marker defining the people of God) which Paul rejects – but not the law itself, the law seen from a different perspective and in different terms. But I have already said enough on this in response to Räisänen (point 4 above).


(b) The mistake of posing too radical an antithesis between Paul and Judaism/the law is compounded by the distinction between ‘getting in’ and ‘staying in’. The debate in Galatians cannot be subsumed under the head of ‘entry’, not least under ‘entry’ in sharp distinction from ‘staying in’.25 The disagreement is pointed up by Sanders’ assertion that ‘Paul’s statement “not by works of law” has to do with entry into the body of Christ’.26 ‘Works of law’ are what good Jews do to enact and demonstrate their loyalty to the covenant as given to Israel as Israel’s special prerogative; entry requirements are only part of it, as again the Antioch incident shows. It is because Sanders seems to focus so much of Paul’s critique of Judaism on questions of entry that Sanders’ Paul does not seem to be addressing Sanders’ Palestinian Judaism.


(c) It is this double, and in this context doubly false, distinction (Judaism and Christianity are two quite different ‘systems’; ‘getting in’ and ‘staying in’) which, I believe, causes Sanders to criticize Paul’s view as not making a logical whole, as inconsistent, as not harmonizable, etc.27 The awkwardness in Paul results from Sanders’ distinctions. In particular, he tries to resolve the tension between Paul’s negative and positive statements regarding the law by confining the negative to the ‘getting in’ question and the positive to the issue of ‘staying in’.28 In other words, the illogicality Sanders finds in Paul is as much or more of his own making, the result of his pushing through this latter distinction. Paul seems inconsistent because his statements do not in fact fit into that distinction. The negativeness in his attitude to the law embraces both aspects (getting in and staying in), and the positive statements cannot be confined to the latter either. Whereas my thesis regarding ‘works of the law’ makes it clear that Paul’s criticism was primarily directed at a way of regarding the law and allows the positive thrust of Romans 2 in particular to stand precisely as a counter to that wrong attitude.29


Since Hübner and Stuhlmacher have both chided me for taking up from Sanders too uncritically, I hope this clarification serves also for them.


F. F. Bruce, ‘Paul and the law in recent research’, Law and Religion, ed. B. Lindars, Cambridge: James Clarke (1988) 124-5, criticizes me at two points. (9) My exegesis of Galatians 2.16 is defective. ‘To translate έάν μή by “except” in this construction seems to me to run counter to Greek idiom’ (p. 125). (10) ‘Paul’s argument throughout Galatians makes it plain that it was not just identity markers that were at issue. Identity markers, for Paul, were optional … But if Paul’s Gentile converts took over Jewish identity markers … it was their giving any place to law or law-keeping in the matter of salvation that Paul saw as a subversion of the gospel of free grace’ (p. 125).


9. On the force of έάν μή I think it would be generally agreed among grammarians that the phrase is properly exceptive and not adversative.30 It is true that Moulton and Howard cite a number of examples to demonstrate that εί μή or έάν μή can function as equivalent to άλλά.31 But Hort had already pointed out that the force in such cases is not simply ‘but’, but ‘but only’.32 The point, then, is that there is some degree of ambiguity as to the precise force of the exception – as also, significantly, with the εί μή of 1.19. Clearly the exception qualifies the main verb – ού δικαιοûται. This is what is agreed by Paul and his fellow Jewish Christians – that no acquittal is possible except through faith in Jesus Christ. What is ambiguous is the relation of the έξ έργων νόμου to the έάν μή clause – ‘knowing that a person is not justified from works of law but only (that he is justified) through faith in Jesus Christ’. The Jewish Christians evidently believed (as the Antioch incident demonstrated) that ‘works of law’ continued to be the appropriate if not essential expression of that faith in Jesus Christ, for Jewish Christians at least. But Paul (now) found it necessary to clarify the ambiguity into outright antithesis.33


10. Here we are back with the issues already dealt with in the response to Räisänen (point 4 above). As with Räisänen, Bruce seems to ignore the positive role Paul gives to the law ‘in the matter of salvation’; how else can one read passages like Romans 2.13-16 and 8.4? Bruce is of course right when he says that Paul seeks to defend and promote his gospel of free grace. But what was threatening that gospel, in Paul’s judgement, was not simply law-keeping as such, but law-keeping which made it impossible for Gentiles to experience salvation as Gentiles, that is, a Jewish insistence that only those participate in the promises of the covenant who practise ‘the deeds of the law’ laid down for the people of the covenant. For Paul, ‘the obedience of faith’ could not be defined so restrictively.


11. Most recently T. R. Schreiner, ‘The Abolition and Fulfilment of the Law in Paul’, JSNT 35 (1989) 47-74, echoes the same criticism as Stuhlmacher. ‘Dunn’s attempt to limit “works of law” to these identity markers (circumcision, food laws and observance of certain days) is not successful.’34 With regard to ‘works of law’ I refer to my response to Stuhlmacher above, and to the fuller treatment in Chapter 8 below.35 What is somewhat surprising is that Schreiner’s criticism (like that of Westerholm, see pp. 237-40) is in effect directed primarily at the preliminary formulation of the above chapter, and that he has failed to give due acknowledgement to the more careful formulation of Chapter 8 below, even though he cites it in his n. 46 (again like Westerholm). This is all the more surprising since he (unlike Westerholm) has taken the point about the social function of the law as that to which Paul’s critique of the law is directed. The laws which Paul specifically excludes are precisely those practices which separated Jews from Gentiles in the Greco-Roman world, which uniquely characterized the Mosaic covenant and the Jews. ‘The nature of that covenant was such that it divided Jews from Gentiles, and thus the covenant was intrinsically nationalistic’ (pp. 56-8). But then he rather spoils it by reducing the issue back to one of a distinction between moral and ritual law (pp. 59ff). In so doing he has fallen into the confusion against which I attempted to give warning in Chapter 7 above (pp. 198, 200 = BJRL, pp. 115-16, 118). The fact that Paul’s critique of the law produces such a distinction in practice is not reason enough for reducing the critique to that distinction. It only works out in these terms because, as I have observed in Chapter 8 (pp. 216-17 = NTS, p. 524), social identity is defined and expressed to such a large extent in and through ritual. Paul’s objection is not to ritual law, but to exclusivist or particularist attitudes which came to expression in and are reinforced by certain rituals. Not the rituals as such, but the attitude behind them, expressed typically as a ‘boasting’ in works of the law (Rom. 2.17-23; 3.27ff). Paul does not in fact object to circumcision except as it constitutes a breach of the law by limiting the grace of God to the outwardly circumcised Jew (Rom. 2.25-9), except as it becomes an occasion for ‘boasting’ (Gal. 6.13).36




  1. NTS 31 (1985) = NTS, p. 545; Torah and Christ = TC, pp. 172f.
  2. NTS, p. 546; TC, pp. 174f.
  3. NTS, pp. 546f.; TC, p. 176.
  4. NTS, p. 547; TC, pp. 177-8. He is followed here by R. Y. K. Fung, Galatians, NICNT, Grand Rapids: Eerdmans (1988) 116n.
  5. Here Räisänen lumps me with Cranfield and refers to his critique of the latter in Paul and the Law, WUNT 29, Tübingen: Mohr (1983) 42-50.
  6. NTS, pp. 548-50; TC, pp. 179-84.
  7. NTS, p. 544; TC, p. 171. On two small points Räisänen is in error. (1) NTS, p. 550 n. 11 = TC, p. 169 n. 6. I do not ascribe to Sanders the view that Paul misunderstood Judaism. My point refers to earlier Jewish critiques of Paul – critiques themselves based on Protestant misreading of Paul and his situation. (2) Despite Räisänen’s charge to the contrary (NTS, p. 544 = TC, p. 170), I do note Sanders’ recognition of the significance of circumcision, sabbath and food laws (my n. 46).
  8. See also J. D. G. Dunn, Romans, WBC 38, Dallas: Word (1988) 485.
  9. See further Dunn, Romans 39-42.
  10. See further ch. 9 below.
  11. NTS, p. 551 n. 31; TC, p. 173 n. 2.
  12. See above ch. 6 n. 117.
  13. cf. J. M. G. Barclay, Obeying the Truth: A Study of Paul’s Ethics in Galatians, Edinburgh: T. & T. Clark (1988). Since the textual evidence relating to δέ at the beginning of 2.16 is so evenly balanced ‘it would be unwise to build much on either reading … But we can still tell from the content of 2.16 (Jew and Gentile alike cannot be justified by works of the law) that it modifies the typically Jewish perspective of 2.15 (Jews are distinct from and superior to “Gentile sinners”)’ (78 n. 9). See also point 9 below.
  14. See further ch. 8, including the Additional Note on Westerholm.
  15. As Räisänen claims (NTS, p. 549; TC, p. 182).
  16. ‘Werke’ 131-2. So also R. Yates, ‘Saint Paul and the Law in Galatians’, ITQ 51 (1985) 105-24 (here 107).
  17. See Dunn, Romans index: ‘Boasting’, ‘Works of law’.
  18. Published as ‘Paul’s Understanding of the Law in the Letter to the Romans’, SEA 50 (1985) 87-104.
  19. BJRL 97-100, 102 and n. 46 = ch. 7 above pp. 184-6, 187-8 and n. 46.
  20. Paul and Palestinian Judaism, London: SCM (1977) 550-2.
  21. Law 4, 25.
  22. Law 37.
  23. Law 78.
  24. See further n. 47 above.
  25. The criticism is well made by R. H. Gundry, ‘Grace, Works, and Staying Saved in Paul’, Bib 66 (1985) 8-12.
  26. Law 105.
  27. Law 4, 77, 86, 122.
  28. Law 84.
  29. See again my Romans on Rom. 2, and Introduction pp. 7-8.
  30. See E. D. Burton, Galatians, ICC, Edinburgh: T. & T. Clark (1921) 121; BAGD, έάν 3.b.
  31. J. H. Moulton and W. F. Howard, A Grammar of New Testament Greek, Vol. 2, Edinburgh: T. & T. Clark (1929) 468.
  32. F. J. A. Hort, The Epistle of St James, London: Macmillan (1909) xvi.
  33. See also points 2 and 3 above, in response to Räisänen.
  34. 71 n. 46; referring particularly to D. J. Moo, ‘”Law”, “Works of the Law”, and Legalism in Paul’, WTJ 45 (1983) 90-99, and Räisänen’s article discussed above. F. Watson, Paul, Judaism and the Gentiles. A Sociological Approach, SNTSMS 56, Cambridge University (1986) 198 n. 79 makes a similar criticism.
  35. Including the Additional Note on Westerholm; also my Romans 153-5.
  36. Despite my general agreement with Barclay’s thesis, at this point he misses the social dimension of Paul’s critical comment – ‘they want you circumcised in order that they might boast in your flesh’ (6.13, my emphasis). In other words, he criticizes them not for self-glorification (Barclay, Obeying 65), but for national aggrandizement – that is, because circumcision will mark the Galatians’ full and physical identification with Israel as a people. Cf. my Romans 124-5 (on Rom. 2.28-9).