Paul and the Patriarch: The Role of Abraham in Romans 4

1. Introduction

Abraham features in several passages in Romans and Galatians, but sooner or later all discussions of Paul and the patriarch come back to Rom. 4, the focus of the present article. A full discussion of the place of Abraham in Paul’s theology would include, and would indeed be strengthened by consideration of, Rom. 9–11 and 15.8-9, as well as Gal. 3.1-29 and 4.21-31. I shall allude to these, but my present argument relates principally to the most obviously central chapter. ¹

Almost every section of Romans is load bearing, holding up some of the weight of Paul’s overall argument. But with many sections there is no agreement as to what the load in question might be, or how the passage in question might be bearing it. In the case of Rom. 4, this problem is acute. The question of Abraham brings into sharp focus several key questions about the shape of Paul’s thought. We could construct a kind of spectrum of possibilities (which would work for Gal. 3 as well): (a) Paul only refers to Abraham because his opponents have done so and he must defeat them on their own ground, but left to himself he would not have mentioned the patriarch; ² (b) Paul is happy to introduce reference to Abraham, but only because this provides him with a convenient but random scriptural proof-text for a doctrine, in this case ‘justification by faith’, whose real ground is elsewhere; ³ (c) Abraham is a kind of ‘test case’ for Paul’s doctrine, not just a proof-text; Paul needs to be able to show some continuity with Israel’s founding fathers; ⁴ (d) Abraham is the ‘narrative prototype’ whose faith prefigures the faithfulness of the Messiah (Hays 2005: xii-xiii); (e) Paul is expounding the covenant-making chapter (Gen. 15) in order to show that the revelation of God’s righteousness in the gospel is (however shocking and paradoxical it may be) the fulfilment of this ancient promise. I can imagine a sixth possibility, though I know of nobody who holds it: (f) that Paul envisages a smooth, continuous, salvation-history in a crescendo all the way from Abraham to Jesus. Since I hold (and shall here expound) (e), and since this is often mistaken for (f) and criticized on those grounds, it may be worth making this clear from the start.

There are at least three problems here, each affecting the others. First, what is Rom. 4 really saying? Second, what place does it have in the developing arguments of the letter? (How does Rom. 4 relate, that is, to the flow of 1–4 as a whole, and in particular to the dense argument of 3.21-31? And how does it help 1–4 prepare the way for 5–8, 9–11 and 12–16?) Third, how is Paul using scripture here? ⁵

Somewhere in the middle of this triangle, we meet some sharp exegetical problems. How, for a start, are we to understand and translate the opening verse (4.1)? What has happened to the argument at vv. 16 and 17, where many translations bracket off a passage which, to other interpreters, looks central rather than peripheral? And, as a particular focus, why does Paul speak of a ‘reward’ in v. 4?

Finally, hovering over the whole thing, but whether as vultures or doves remains to be seen, there are larger questions of Pauline interpretation: the question of ‘old’ and ‘new perspectives’; of ‘justification’ and ‘incorporation’ (‘juridical’ and ‘participationist’ patterns of thought); of ‘apocalyptic’, ‘covenant’ and ‘salvation history’; of soteriology and ecclesiology. ⁶ My hope is that, by exploring some usually ignored exegetical possibilities, we might gain a new angle on all these matters, though of course the present article will not have space to follow them up.

My proposal, then, is a variation on (e) above. I shall propose and argue (1) that in Rom. 4 Paul expounds the story of Abraham, especially in Gen. 15, the whole of which he has in mind, in terms of God’s establishment of the covenant which always envisaged a single worldwide family; (2) that in Romans this forms part of Paul’s demonstration that God has been faithful to the covenant, in the gospel and its forming of a multi-ethnic family (while in Galatians it forms part of his argument that the people of Abraham is a single family and cannot therefore be demarcated by the ethnically divisive Torah-works; these two arguments are of course complementary); (3) that Paul’s use of δικαιοσύνη language is closely related to the idea of the ‘covenant’ to which God has been faithful and into which believers are incorporated; (4) that Paul speaks of ‘reward’ (μισθός) in Rom. 4.4, not because he is after all thinking in ‘old perspective’ terms about people earning (or not earning) their ‘justification’, but because he is thinking of the whole of Gen. 15, which is all about the μισθός of a huge family promised by God to Abraham; ⁷ and (5) that this opens up a better way of reading the difficult but related passages 4.1 and 4.16-17. Were there more space, we could go on to show how this reading of Rom. 4, and a parallel one of Gal. 3, points to a fresh and satisfying reading of the other patriarchal references in Gal. 4, Rom. 9 and 11, and Rom. 15. One might then conclude, again granted more space, that these readings of the relevant passages provide a better account of the flow of thought in both letters than is normally envisaged. ⁸

2. How Paul Read Genesis 15

We begin with Gen. 15 itself and the question of how Paul was using this chapter as a whole. ⁹ I take it he was using the chapter as a whole, since as well as quoting 15.6 in Rom. 4.3 and 4.22-23, and referring back to it in 4.5, 6, 9, 10, 11 and 13, he refers to God’s promise in Gen. 15.5 in 4.18, and throughout the discussion gives every appearance of wrestling with the message of the whole chapter, looking at it from one angle after another. In vv. 10-11 he relates Gen. 15 sequentially to Gen. 17 (the command to circumcise), making the historical sequence between them a key point in his argument, and then quoting 17.5 at 4.17 and alluding to 17.17 at 4.19. There are other echoes as well, notably of Gen. 18 and 22 in v. 13. It looks, then, as though Paul is not plucking out a proof-text (Gen. 15.6) without regard for its context. Nor is he appealing to a sense of ‘scriptural authority’ as a mere rhetorical move (‘I can appeal to some ancient texts!’). He is working with the actual context of Genesis, and particularly ch. 15.

So what is that context? The context is of God’s promise to, and covenant with, Abraham. ¹⁰ God had made promises to Abraham before, at the start of Gen. 12; Paul quotes 12.3 in Gal. 3.8, and indeed seems in Rom. 4 to presuppose that promise, that God will bless all the nations in Abraham. But in Gen. 15.1 the promise is repeated, though in a particular form: God will be Abraham’s shield, and his reward shall be very great (ὁ μισθός σου πολὺς ἔσται σφόδρα). But what is this ‘reward’? What is it that God has promised to Abraham? ¹¹

To answer this we must take a step back. At the end of Gen. 14 Abraham has just returned from defeating the pagan kings and rescuing his nephew Lot. That is when (15.1) God promises him a ‘very great reward’, which puzzles Abraham because he has no heir. He clearly (15.2) understands ‘reward’ in terms of the inheritance, both human and geographical, which he has been expecting on the basis of God’s earlier promises. ‘O Lord God’, he says, ‘what will you give me, for I continue childless, and the heir of my house is Eliezer of Damascus?’ In other words, he assumes that his ‘reward’ ought to consist of his family and their inheritance. Without such a physical family to inherit, the inheritance will go to ‘a slave born in my house’, an οἰκογενής. ¹² The prominence of the theme of slavery later on in both Romans and Galatians should make us prick up our ears, but that is not for now.

Abraham then expands his puzzled question in 15.3, though now it is not so much a question as a statement of the apparently impossible situation: ‘You have given me no offspring (σπέρμα), and so a slave born in my house is to be my heir (κληρονομήσει με).’ Abraham, in other words, assumes in a quite uncomplicated fashion that the promise of ‘reward’ means that he is to have a family, heirs to inherit from him. This way of reading the passage is confirmed by God’s answer in vv. 4 and 5. ‘This man [i.e. Eliezer of Damascus] shall not be your heir; no one but your very own issue (ὃς ἐξελεύσεται ἐκ σοῦ) shall be your heir.’ That is what the conversation is about. Abraham’s own physical offspring will indeed inherit from him, κληρονομήσει σε; the phrase is repeated.

Then, taking him outside, God invites Abraham to look up at the stars and to try to count them (15.5). Then comes the amplified promise, which Paul quotes in 4.18: οὕτως ἔσται τὸ σπέρμα σου, ‘so shall your descendants be’; in other words, ‘that is what your “seed” will be like’. This promise of an enormous, uncountable family is the promise which Abraham believes; and this in turn is the belief, the faith, of which the text says that God ‘reckoned it to him as righteousness’, ἐλογίσθη αὐτῷ εἰς δικαιοσύνην.

The theme of κληρονομία is repeated yet again in 15.7 (though many translations fail to bring it out), this time in relation not to the enormous family but to the promised land: ‘I am YHWH, ¹³ who brought you out of Ur of the Chaldeans to give you this land to inherit’ (κληρονομῆσαι). This precipitates Abraham’s request for an assurance that he will indeed possess the land (ὅτι κληρονομήσω αὐτήν), to which the divine response is to make a covenant, with Abraham cutting animals in two and the fiery symbol of the divine presence passing between the pieces. The meaning of this covenant, giving assurance to Abraham that the promises will hold good, has to do with the means by which his σπέρμα will indeed inherit (κληρονομήσει) the promise: they will themselves be slaves (δουλώσουσιν) in a foreign land for four hundred years, ¹⁴ but God will judge those who enslave them and will bring them out with great possessions. This, it seems, is the means by which Abraham’s σπέρμα is to gain the promised κληρονομία. ¹⁵ The point is repeated one more time in Gen. 15.18: God made a covenant with Abram, διέθετο κύριος τῷ Αβραμ διαθήκην, saying, ‘to your descendants’, τῷ σπέρματί σου, ‘I give this land’. ¹⁶ Insofar as we can gauge the ways in which Gen. 15 was being read by other Jews in the second-Temple period, it looks as though this is the emphasis that they would have brought out: that the Abraham narrative as a whole, and Gen. 15 within it, forms the launching-point for a long narrative about the patriarch’s family and their promised land, which the later writers are expounding in order to emphasize the fulfilment of that in their own day. ¹⁷

Paul appears not only familiar with all this but actually to be presupposing it, in both Rom. 4 and Gal. 3. ¹⁸ He is tracking, and engaging with, exactly this sequence of thought about the σπέρμα and the κληρονομία. But at this point he appears to make a jump. True, Gen. 12.3 had indicated that through Abraham God would bless all nations. But, despite the insistence of Gen. 15.4 (echoed in Heb. 11.12) that Abraham’s true heir would be his own physical offspring, Paul in Rom. 4.18 quotes ‘so shall your seed be’ as a reference to Abraham being ‘the father of many nations’. It is as though he advances the change of name, from Abram to Abraham, from Gen. 17 to Gen. 15. This, I think, is what lies behind what Paul says in Rom. 4.13, that the promise to Abraham and his seed was that they should ‘inherit the world’, τὸ κληρονόμον αὐτὸν εἶναι κόσμου. This apparent widening of the promise of the land to a promise about the whole world is made elsewhere in second-Temple Judaism. ¹⁹ It is true, as well, that Gen. 15 ends up with the covenant promise of the geographical land (15.18-21, demarcated both geographically and ethnically), and that in Rom. 8 Paul uses language relating to the ancient ‘promised land’ to refer now to the entire cosmos. But I think that at the moment, here in ch. 4, this reference to the worldwide inheritance seems to be not so much about geography but about descendants: about the uncountable family whom Paul is taking to be not simply a very large family of Abraham’s own physical descendants, but a ‘family’ composed of many nations, indeed of the whole world.

Romans 4 as a whole comes back again and again to this same point. In 4.11-12 Paul speaks of Abraham as the father of uncircumcised as well as circumcised. In 4.16-17 he speaks of the promise being valid for ‘the entire seed’, παντὶ τῷ σπέρματι, ‘not only the [seed] “out of the law”’, ἐκ τοῦ νόμου, but ‘the [seed] that is out of the faith of Abraham’ (τῷ ἐκ πίστεως’Αβραάμ), because he is ‘father of us all’, quoting Gen. 17.5. ²⁰

How then is Paul putting together, in Rom. 4 in particular, the specific point about the promise referring to many nations (including, therefore, many who are not Abraham’s physical descendants) with the equally specific point, to which he returns emphatically at the end of the chapter, that Abraham’s faith had to do with believing that God would give to him and Sarah, despite their advanced age, an actual physical son of their own? It looks as though Paul is holding the two in a closely linked parallel. In 4.18 it is clear that he refers the promise of Gen. 15.5 (‘so shall your seed be’, looking up at the uncountable stars) to the many nations that God would give him, but in 4.19-21 he clearly describes Abraham’s faith, in believing precisely this promise, in terms of the specific son to be born to Abraham and Sarah in their old age, and says that this was the faith because of which ‘it was reckoned to him as righteousness’. This leads him to the chapter’s triumphant conclusion, that this same faith in the God who raises the dead is the faith of the Christian (4.24-25; compare 10.9-10).

This in turn sends us back to the verse which introduces the final section of the chapter. In 4.17 Paul describes Abraham’s God in two ways, corresponding exactly to this parallel. Abraham, he says, believed in the God who

I suggest that Paul, in reading Gen. 15, sees these two reflected in Abraham’s request and God’s promise. Abraham asked God about an actual physical offspring; this is answered by God ‘raising the dead’, giving life to his and Sarah’s ‘dead’ bodies by giving them a son of their own. God promised, in addition, something far more abundant than Abraham’s specific request: a family consisting of many nations, like the stars of heaven. ²¹ This has to do with God ‘calling non-existent things into existence’, giving Abraham ‘offspring’ from many nations.

This double statement then corresponds closely, of course, to the repeated double statement about Jewish and Gentile members in Abraham’s faith-family, as in 4.11-12. There the order is reversed, because of the flow of thought of the chapter at that point. Abraham is described as

The same happens, with the original order again, in the decisive 4.16, where the promise is validated for ‘all the seed’, not only

This leads straight into the double statement of Abraham believing in God as both the giver of life in place of death and the one who creates out of nothing.

It seems clear from all this that Paul understands the crucial promise of Gen. 15.5, and hence the nature of Abraham’s faith in Gen. 15.6, in terms of the promise of the ‘seed’, starting with Isaac but broadening out to include that much larger family.

Paul has, as so often, compressed many complex ideas here into a tight space. But it seems clear that in Rom. 4 at least he is reading the promise of Gen. 15.5 as a promise about a family, a ‘seed’. This consists of ‘heirs’, ‘inheritors’, starting with Abraham’s own son Isaac but then going well beyond the boundaries of his subsequent physical family into the ‘many nations’ of 4.17 (quoting Gen. 17.5). ²² This was, of course, according to both Genesis and Paul, the promise which Abraham believed, with a ‘faith’ that was then ‘reckoned as righteousness’. ²³

My proposal, then, is that when Paul spoke of Abraham’s μισθός in Rom. 4.4, he intended to refer to this promise and to this worldwide family, starting with the life-out-of-death Isaac and moving on to the created-out-of-nothing ‘many nations’. ²⁴ The verse is then much more than ‘an example taken from everyday life’ or a ‘universally valid rule’. ²⁵ So far from the reference to a μισθός introducing a general point, only loosely related to the Abraham story, the link between that word and Gen. 15.1 indicates (as does the sequel in the rest of the chapter) that Paul is still talking about the patriarch himself. The promise about the worldwide family (Abraham’s ‘reward’ in Genesis) is after all the subject of much of the rest of the chapter. It is also, arguably, the subject of the key passage which leads in to ch. 4, namely 3.27-31. A person is justified by faith apart from works of the law (3.28), because otherwise—that is the significance of the little word ἤ at the start of 3.29—God would be the God of the Jews only, whereas in fact he is God of Gentiles as well, since God is one. Paul here invokes the Shema itself against any suggestion that his own Jewish people might be the sole people of God. No, he says: God will justify the circumcision ἐκ πίστεως and the uncircumcision διὰ τῆς πίστεως. ²⁶ And this, so far from nullifying Torah, makes Torah more firmly grounded (3.31). It is exactly this point, rather than a generalized or abstract scheme of salvation, which Paul is then anchoring in the story of Abraham, part of the foundational narrative of the five-book Torah itself.

The underlying thrust of this proposal will, I hope, be clear. It strikes exactly against a position which has become one of the last strongholds of the ‘old perspective’ on Paul.

There is already a tradition among the opponents of the so-called ‘new perspective on Paul’ in which Rom. 4.4-8 becomes the key evidence that Paul, despite the arguments of Sanders, Dunn, myself and others, was after all mainly if not exclusively concerned with the contrast between (a) the human performance of good works as the means of justification and (b) the simple trust in God’s grace. ²⁷ I and others have argued in the past that, on the contrary, the metaphor of ‘working for a reward’, and the idea of earning wages, is here simply a secondary metaphor which Paul never employs in similar contexts elsewhere, and which cannot be allowed to become a sort of Old Perspective tail wagging the otherwise large and furry New Perspective dog. ²⁸ But once we make the connection of μισθός in v. 4 with the same word in Gen. 15.1, ²⁹ and once we follow through the implications of Paul’s virtually certain reference to μισθός, not as a ‘reward’ for ‘obedience to Torah’ or some equivalent thereof, but as the ‘reward’ which consists of a family (and indeed a land) promised by God to Abraham, not only does the entire passage make much more sense, as I shall shortly show, but this last refuge of the ‘old perspective’ is dismantled, leaving the occupants nowhere to hide. Yes, Paul does then develop a very brief book-keeping metaphor in v. 4. But the reason for the metaphor itself (‘working’ for a ‘reward’ which one is then ‘owed’) emerges not from an underlying implicit second-Temple Jewish soteriology of ‘doing good works’ to earn God’s favour, an idea for which there is scant evidence, but from Gen. 15 itself, which is innocent of all such notions, and which speaks instead, as Paul does, of covenant and family. Verse 4 embroiders this with a particular colour, but this embroidery carries no weight in the passage as a whole.

The same is true for the language of ‘works’ in 4.2 and 4.6. These belong closely, of course, with the discussion of ‘works of the law’ in 3.20, 27-28. Those verses are themselves the subject of current controversy, and one cannot therefore appeal to 4.2, 6 as though they constituted an unambiguous reference to ‘works’ in the older reformational sense. In fact, the often ignored opening word of 3.29—the ἤ which means ‘or’—indicates clearly enough, as we just noted, that Paul is here talking about the ‘works of the law’ which would separate Jew from Gentile (see Wright 2002, ad loc.). Though we cannot pursue this further at the moment, this is perhaps the point to emphasize that this reference to the abolition of barriers that would exclude Gentiles from the one people of God does not reduce Paul’s soteriology to sociology, as some have suggested. ³⁰ As I have stressed elsewhere, it is a mainstream Jewish understanding of God’s purpose in choosing Abraham that he was to be the one through whom the problem of human sin and its effects would be dealt with. The question of ‘who are the children of Abraham’ is thus precisely the question of ‘who are the people whose sins have been/are being/will be dealt with?’, and, more widely, ‘how is the creator God going to deal with the problem of sin in humankind as a whole?’ The presence of the dealing-with-sin theme—for instance, in the discussion of David and the quotation from Ps. 32 in vv. 6-8—cannot be adduced as though it proved that Paul was ‘really’ or ‘only’ speaking of salvation from sin, and not primarily mounting an argument about the worldwide meaning of the covenant made in Gen. 15.

3. The Justification of the Ungodly

How then should we read ‘the justification of the ungodly’ in Rom. 4.5? It has long been customary to understand the passage something like this:

One might summarize this viewpoint by saying that Abraham is justified by faith because he believes in justification by faith (as opposed to the justification of the godly).

I suggest, instead, that we should read the passage like this:

What then does ‘justification’ itself consist of? There is no sign here, I suggest, of the elaborate mediaeval apparatus concerning iustitia or of the radical adjustment of that apparatus that was achieved by the sixteenth-century reformers. That is not what Rom. 4 is about. Rather:

1. Genesis 15, the chapter with which Paul is here working, is the chapter in which God makes the covenant with Abraham, promising him both a huge family and a particular geographical homeland—and promising him that, in order to get the former into the latter (the family into the land), there will be a long period of slavery followed by a dramatic rescue: in other words, the Exodus. ³⁴

2. I suggest that one should read (and that one should understand Paul as reading) Gen. 15.7-21 as epexegetic of 15.6. In other words, one should not understand the extended promise of the land and how it will be attained, focused on the covenant-making ceremony with the divided animals and the fiery apparition that goes between them, as something other than what is said in 15.6, that Abraham believed God and ‘he reckoned it to him as righteousness’. Genesis 15, in other words, is not saying two different things, (a) ‘God reckoned it to Abraham as righteousness’ and (b) ‘God made a covenant with Abraham in which he promised him a family and a land, to be attained through the Exodus’. Genesis 15.6 is sandwiched between the two halves of the chapter: 15.1-5 is all about God’s promise to Abraham concerning his family and his inheritance, and 15.7-20 is all about the making of the covenant which consists of the divine promise about land, family and inheritance. When, in the middle of all this, we read in v. 6 that Abraham believed a promise, about those very things, and that God ‘reckoned it to him as righteousness’, the context alone strongly suggests that ‘reckoned it to him as righteousness’ means, more or less, ‘God reckoned this in terms of covenant membership’, or perhaps ‘God made a covenant with him on this basis’. Abraham’s faith in God, and in God’s promise, was the sign and badge of the covenant which God then proceeded to make, the covenant which guaranteed precisely that promise. Nothing in the text of Genesis suggests that ‘he reckoned it to him as righteousness’ means anything other than this.

3. In case this seems too big a stretch from normal readings of the verse, consider a further point which comes into focus at Rom. 4.11. Describing Abraham’s circumcision, Paul writes that he received circumcision as ‘a sign and seal’ (σημεῖον, σφραγῖδα) of the ‘righteousness of faith which [he had] in uncircumcision’, τῆς δικαιοσύνης τῆς πίστεως τῆς ἐν τῇ ἀκροβυστίᾳ. But in the passage Paul is referring to (Gen. 17.10-11), God declares that circumcision will function as a sign of the covenant: καὶ ἔσται ἐν σημείῳ διαθήκης ἀνὰ μέσον ἐμοῦ καὶ ὑμῶν (17.11). Thus where Genesis has ‘sign of the covenant’, σημείον διαθήκης, Paul has ‘sign of δικαιοσύνη’, σημείον δικαιοσύνης, the δικαιοσύνη which is now characterized, and recognized, by πίστις. One might bring this out by rendering δικαιοσύνη here as ‘covenant membership’, or even ‘the status of covenant membership’. Circumcision was a sign of the covenant membership Abraham already had on the basis of believing God’s promise about his family. Paul seems to have made this equation, and so should we.

One possible objection to this move would be to say that Paul has deliberately avoided the use of ‘covenant’, and has substituted something quite different, namely ‘righteousness’, in its place. The best and clearest answer to this objection comes in the one and only parallel in the Hebrew Bible to the key phrase in Gen. 15, to which I now turn. It both answers the present point and makes a further strong case of its own.

4. The only parallel in Israel’s scriptures to the phrase ἐλογίσθη αὐτῷ εἰς δικαιοσύνην is found in Ps. 106 [lxx 105].30-31. Here the psalmist celebrates the action of Phinehas who, in Num. 25, intervenes as the plague is destroying the Israelites who have worshipped the Baal of Peor. Phinehas was one of the great biblical examples of ‘zeal’, and he and his story would have been well known to the young, self-confessedly zealous, Saul of Tarsus. ³⁵ The psalmist, commenting on Phinehas’s action, declares

And that has been reckoned to him as righteousness from generation to generation for ever.

καὶ ἐλογίσθη αὐτῷ εἰς δικαιοσύνην εἰς γενεὰν καὶ γενεὰν ἕως τοῦ αἰῶνος ³⁶

The psalmist had provided a brief and dense summary of what Phinehas had actually done. In Numbers, Phinehas took a spear and, in obedience to Moses, killed an Israelite and a Midianite woman with a single thrust. In the Psalm, he ‘stood up and intervened’. In exactly the same way, the psalmist has provided a brief and dense summary of the divine blessing which came upon Phinehas as a result.

In Numbers, this blessing is very specific:

YHWH spoke to Moses, saying: ‘Phinehas son of Eleazar, son of Aaron the priest, has turned back my wrath from the Israelites by manifesting such zeal among them on my behalf (ἐν τῷ ζηλῶσαί μου τὸν ζῆλον ἐν αὐτοῖς) that in my jealousy (ἐν τῷ ζήλῳ μου) I did not consume the Israelites. Therefore say, ‘I hereby grant him my covenant of peace (διαθήκην εἰρήνης). It shall be for him and for his descendants after him (καὶ τῷ σπέρματι αὐτοῦ) a covenant of perpetual priesthood, because he was zealous for his God (ἀνθ’ ὧν ἐζήλωσεν τῷ θεῷ αὐτοῦ), and made atonement for the Israelites.’ ³⁷

So too in Sirach, drawing back like the psalm from an explicit description of what Phinehas actually did but keeping the Pentateuchal reference to the covenant:

Phinehas son of Eleazar ranks third in glory for being zealous in the fear of the Lord, and standing firm, when the people turned away, in the noble courage of his soul; and he made atonement for Israel. Therefore a covenant of friendship (διαθήκη εἰρήνης) was established with him, that he should be leader of the sanctuary and of his people, that he and his descendants should have the dignity of the priesthood for ever. Just as a covenant was established with David … (Sir. 45.23-25).

How might one provide a brief and dense summary of God’s gift of ‘my covenant of peace’, ‘a covenant of perpetual priesthood’ and ‘a covenant of friendship’? ³⁸ The psalmist seems to think that one can perform this task by writing ‘it was reckoned to him as righteousness’. The psalmist has no reason to eliminate a reference to ‘covenant’ by substituting something quite different. ‘Reckoned as righteousness’ must therefore be understood as being broadly synonymous with ‘established a covenant’.

The phrase cannot, in any case, be a reference to Phinehas’s own personal ‘justification’ in some detached sense. ³⁹ The ‘reckoning’ continues to all his succeeding generations, all his σπέρμα. In the context both of Numbers and of the psalm, the point is clear: Phinehas’s descendants are priests for ever. The phrase ‘it was reckoned to him as righteousness’ thus means, fairly straightforwardly, ‘God made a covenant with him and with his family’.

Phinehas is linked with Abraham, interestingly, in the classic exhortation to ‘zeal’ issued by Mattathias of Modein. He himself had imitated Phinehas’s example by killing a Jew offering pagan sacrifice, killing the presiding official at the same time and tearing down the altar. ‘Thus’, says the writer, ‘he burned with zeal for the law, just as Phinehas did against Zimri son of Salu’. ⁴⁰ This leads to the formation of the anti-Syrian revolutionary movement. When Mattathias is about to die, he summarizes his position, and exhorts his sons, by telling them the story of Israel, highlighting the parts considered particularly relevant. ⁴¹ This is what it will mean, he explains, to ‘show zeal for the law, and give your lives for the covenant of our ancestors’. ⁴² The story then starts with Abraham and continues, through Joseph, to Phinehas:

Was not Abraham found faithful when tested, and it was reckoned to him as righteousness? Joseph in the time of his distress kept the commandment, and became lord of Egypt. Phinehas our ancestor, because he was deeply zealous, received the covenant of everlasting priesthood (ἐν τῷ ζηλῶσαι ζῆλον ἔλαβεν διαθήκην ἱερωσύ— νης αἰωνίας). ⁴³

One cannot, of course, argue directly from the parallel between these three sentences to an exact identification of ‘it was reckoned to him as righteousness’, spoken concerning Abraham, and ‘received the covenant of priesthood’, spoken of Phinehas. For a start, Joseph’s ‘becoming lord of Egypt’, the parallel in his case, is a different sort of thing entirely. But the correspondence is striking all the same, granted the way Ps. 106 uses the phrase ‘reckoned as righteousness’ to summarize what elsewhere is described as the establishing of a covenant.

As we just saw, a very similar point is made about Phinehas, in a very different list of heroes, in Sir. 45. The writer places the ‘covenant of peace’ established with Phinehas chronologically between the covenant made with Abraham and the covenant established with David, distinguishing the Davidic from the Aaronic covenant through Phinehas by commenting that the inheritance (κληρονομία) of the Davidic covenant passes from the king to his son but that the κληρονομία of Aaron passes to his descendants (καὶ τῷ σπέρματι αὐτοῦ). ⁴⁴

Four conclusions follow from this parallel with Phinehas. (1) There is good reason to suppose that the phrase καὶ ἐλογίσθη αὐτῷ εἰς δικαιοσύνην was taken to mean ‘and on this basis God established a covenant with him’. Δικαιοσύνη then means, more or less, ‘the status of covenant membership’. (2) There is every reason to suppose that Saul of Tarsus was extremely familiar with the tradition of ‘zeal’ exemplified in 1 Macc. 2, that he would have thought of Phinehas (and Elijah) as among his predecessors in this tradition, and that he would have known and cherished texts which pointed to this tradition of ‘showing zeal for the law’ (including quite possibly a reference to Abraham) as the appropriate way of supporting ‘the covenant of the ancestors’. Romans 4 provides strong evidence that, when Paul rethought his beliefs in the light of the revelation of the risen Jesus on the Damascus Road, he retained the covenantal theology and the idea of God ‘reckoning it as righteousness’, but filled it with fresh content culled from the story of Abraham itself. (3) When Paul speaks of God making promises to Abraham καὶ τῷ σπέρματι αὐτοῦ he is echoing, in an almost formulaic way, what was said about not only Abraham but also others with whom covenants were made. (4) It is therefore highly likely that when Paul quotes Gen. 15.6 (‘he reckoned it to him as righteousness’) he intends it to mean what it meant in its original context, that is, that ‘on this basis, God established his covenant with Abraham, reckoning Abraham as his covenant partner’, so that the promises of a worldwide family and ‘inheritance’ would be valid καὶ τῷ σπέρματι αὐτοῦ, ‘to his seed, also’.

From these conclusions there follows another, perhaps the most important. Paul the apostle had decisively rejected the model of ‘zeal’ he had formerly embraced, namely, that of the violence exemplified in Phinehas, Elijah and Mattathias. But he had not rejected its basis, namely God’s promise to Abraham and its further extension through David. He had, instead, radically reinterpreted it in the light of Jesus the Messiah and his death and resurrection. Now, instead of the ‘ungodly’ and all their ways being rejected, so that ‘zeal’ consisted in defending Israel against them, the Abrahamic promises were to be seen as the basis for their transformative inclusion within the covenant. This, Paul affirms, is what God had promised to Abraham as his ‘reward’: the ‘many nations’ that would be his extended σπέρμα, and the whole κόσμος that would now be the extended ‘promised land’. That is what Abraham had believed, and that was the basis of the covenant that was then established.

This, then (to confirm what was said above), is what Paul means by the phrase ‘the one who justifies the ungodly’ in v. 5. Romans 4 as a whole is about the bringing of Gentiles into the one family, a theme repeated again and again from different angles. ‘Justification’ would seem, then, to be Paul’s way of denoting either the bringing into the family of those outside or the recognition or demarcation of newcomers as being within that family. Either way, it has to do with covenant membership. The point is very close to the argument of Galatians, particularly 2.15-21 and 3.1-29. Indeed, when Paul speaks in Gal. 2.15 of ‘Gentile sinners’, ἐξ ἐθνῶν ἁμαρτωλοί, it appears that ἁμαρτωλοί there has a very similar meaning to ἀσεβής in Rom. 4.5. ⁴⁵

Romans 4 demands to be read, then, in terms of the fulfilment of the covenant of Gen. 15. This concerns Abraham’s enormous promised worldwide family, that is, his extended σπέρμα, and their extended κληρονομία, ‘inheritance’, which is the whole world. The ‘reward’ spoken of in 4.4 refers to the ‘reward’ in Gen. 15.1, which is precisely this huge family, the whole sperma. The ‘justifying of the ungodly’, spoken of in v. 5 as the characteristic action of the God in whom Abraham believed, is not a reference to Abraham’s own justification from within a supposed ‘ungodly’ state, but refers rather to ‘the inclusion within the σπέρμα’ (in other words, the ‘justification’) of non-Jews. Since that was the main subject of 3.27-31, we should not be surprised.

In and through it all, Paul is referring to Abraham, not because some hypothetical opponents had raised, or might raise, questions about the patriarch. He is doing so because the point he is making throughout Rom. 1–4, and particularly now in 3.21–4.25, is that the God of Israel has been faithful to the covenant promises he made to Abraham, and that this covenant fidelity, fully unveiled in the faithful death and resurrection of the Messiah, is the basis of the covenant status and membership (δικαιοσύνη) of all who believe the gospel, Gentile and Jew without distinction. Genesis 15 does not supply Paul with a detached ‘proof from scripture’. He is strongly affirming, even while he is radically reinterpreting, the ancient Jewish tradition in which he already stood, of envisaging the story of God’s people as going back to, and being characterized by, the origin of the covenant in God’s promises to Abraham.

This makes excellent sense of another key paragraph in Rom. 4, namely vv. 9-12. Many exegetes have treated this in terms of ‘circumcision’ being a ‘human work’ by which one might seek to gain God’s favour, but for Paul the question at stake is about the welcome now offered to Gentiles within Abraham’s extended family. The question of 4.9, whether the blessing is restricted to the circumcised or extended to the uncircumcised also, is in effect a repetition of the question with which the chapter opens (see below). Paul’s initial answer, in v. 10, is to point out that Gen. 15, already quoted, comes chronologically before Gen. 17, the giving of circumcision, so that ‘it was reckoned to Abraham as righteousness’ before he was circumcised. Circumcision cannot therefore be a condition of this ‘reckoning’. As we saw a moment ago, Paul explains that Abraham received circumcision as a sign or seal of the covenant that had already been made, of the ‘covenant membership’ (δικαιοσύνη) he already possessed on the basis of Gen. 15.

This, he now points out, has a double implication. Two conclusions are set in parallel, one about uncircumcised believers and the other about circumcised believers. The first one, relating to Gentile believers, itself divides into two: both parts are introduced by εἰς τό, but without any καί to link the two clauses. This indicates, I suggest, that the second εἰς τό clause is not an addition to the first, a separate topic as it were, but is rather an explanation of the first:

He received circumcision as a sign and seal of the δικαιοσύνη τῆς πίστεως which was ‘in uncircumcision’:

This parallelism suggests that ‘having righteousness reckoned to them’ is another way of saying the same thing, namely, ‘that Abraham might be the father of all who believe in uncircumcision’. As with Phinehas’s having ‘righteousness reckoned to him’, this is about the establishment of a covenant with the person concerned and with his σπέρμα after him. This provides further support for the larger argument I have been making, that the theme of the chapter is the fatherhood of Abraham over an extended family, and that this is denoted, not merely supplemented, by the language about the ‘reckoning of righteousness’, which always had to do with the covenant which God was making with the patriarch.

Paul’s second conclusion follows, now concerning the circumcision (4.12): Abraham is the father of the circumcised as well—provided that they do not rely on their circumcision by itself, but follow the steps of the faith-in-uncircumcision of ‘our father Abraham’. Abraham and his fatherhood remain the subject. Having argued that Abraham’s family is thrown open to the ‘ungodly’, Paul now indicates an important restriction: imitating Abraham’s faith is the necessary badge of membership. This reflects what Paul says elsewhere in Romans and Galatians. ⁴⁶

Reading the body of Rom. 4 in this way offers us a fresh point of view on several other issues in the chapter.

4. Romans 4.1: The Question which Structures the Chapter

First, the question which structures the chapter and the answer which it evokes. It has long been recognized that Rom. 4.1 is a puzzle, mainly because of the verb εὑρηκέναι, ‘to have found’. Commentators and translators have struggled to make sense of this with the normal punctuation of the verse, as apparently did early copyists, some of whom missed out εὑρηκέναι altogether, making ‘Abraham’ somewhat awkwardly the object of ἐροῦμεν, brought out in English by ‘what then are we to say about Abraham’. ⁴⁷ Reckoning this as a smoothing out of an obvious difficulty and hence almost certainly secondary, the verse as normally understood has to mean something like NRSV’s ‘What then are we to say was gained by Abraham?’ However, found and gained are not quite the same thing, and the difficulty of giving εὑρηκέναι anything like its proper (and regularly Pauline) sense ought to push us to look for alternative answers. ⁴⁸

This puzzle, I suggest, is exactly cognate with the puzzle of vv. 16 and 17. Many translators, all the way back to the King James Version, have experimented with brackets around part or all of 16b and 17. ⁴⁹ All these suggestions cause as many problems as they appear to solve, because the quote about Abraham being the father of many nations comes up again in the next verse, so the first time it occurs can hardly be an aside, as that bracketing suggests. But, much more important, we have seen good reason to regard the idea of Abraham as ‘the father of us all’ as the very heart of the chapter, and of Paul’s exposition of Gen. 15. It is not peripheral to the whole line of thought, as the introduction of a parenthesis would suggest, but is absolutely central. ⁵⁰ The promise Abraham believed was the promise about God giving him this worldwide family, about (in other words) him becoming the father of many nations. This cannot be shoved aside into a bracket without doing violence to the whole.

To recognize that this theme of Abraham’s universal fatherhood is the theme of the whole chapter enables us to embrace a quite different reading of v. 1. Richard Hays argued for the basic point nearly 30 years ago, and I have since then modified his proposal, with (I am glad to say) Hays’s approval. ⁵¹

Hays’s proposal began with an observation so obvious that one might wonder why nobody had said it in the recent past. Normally when Paul says ‘What then shall we say?’, τί οὖν ἐροῦμεν, there is an implicit question mark (there are, of course, no punctuation marks in the early MSS), followed by a proposal of something that one might go on to say at this point but which Paul will in fact deny. Examples abound, not least in Rom. 6.1: ‘What then shall we say? Shall we continue in sin, that grace may abound?’; 7.7: ‘What then shall we say? That the law is sin?’ ⁵² If we read Rom. 4.1 in this way, it opens the possibility of reading εὑρηκέναι with the subject not as ‘Abraham’ but as ‘we’ or ‘us’, implied from the first person plural of ἐροῦμεν and of the possessive ἡμῶν four words later. Hays therefore translated, ‘What then shall we say? Have we Jews considered Abraham to be our forefather only according to the flesh?’ expecting the answer ‘no’; and he read the chapter accordingly. Paul, from this angle, is wanting to stress that ‘we Jews’ have all along regarded Abraham as much more than simply ‘our fleshly forefather’. ⁵³

But this, as Hays himself has now admitted, is far less likely, both as a translation of the verse and as an introduction to what the chapter actually says, than the alternative which I proposed in my commentary and now have in my translation: ‘What shall we say, then? Have we found Abraham to be our ancestor in a human, fleshly sense?’ ⁵⁴ Or, in the recent Common English Bible, for which Hays provided the first draft of Romans: ‘So what are we going to say? Are we going to find that Abraham is our ancestor on the basis of genealogy?’ We might also compare the Italian translation La Bibbia, which, rendered into English, has ‘What then shall we say? That we have found Abraham to be our forefather according to the flesh?’ ⁵⁵

This proposal brings Rom. 4.1 closely into line with the overall question which Paul is addressing in Galatians. If someone comes to put their faith in Jesus the Messiah, ought they to regard Abraham as their father in the normal, fleshly sense, meaning that if the person concerned was a (male) Gentile they ought at once to get circumcised, to become a full proselyte, and so to become a member of the family κατὰ σάρκα? ⁵⁶ This question can equally be put the other way round: can Gentiles be considered to be members of Abraham’s family when he is obviously not their physical forefather? ⁵⁷

Most commentators have either ignored this proposal or simply pushed it aside. ⁵⁸ But the reasons they give for rejecting it show, time and again, that they have not in fact understood the way in which the proposal belonged closely with a fresh and highly plausible reading of the whole chapter. ⁵⁹

There are, for instance, the objections concerning grammar and wording. ⁶⁰ Can the question that results (εὑρηκέναι ’ Αβραὰμ τὸν προπάτορα ἡμῶν κατὰ σάρκα) really stand without expressing a subject (ἡμᾶς, presumably) for the infinitive? ⁶¹ I see no problem here: we already have ‘we’ understood from the opening τί οὖν ἐροῦμεν, and understood again through the ἡμῶν in the clause itself. For Paul to repeat ἡμᾶς, though pedantically correct, would seem otiose, and quite unlike his regular style, not least in the quick-fire argument of the central chapters of Romans. We may compare 4.16 (Διὰ τοῦτο ἑκ πίστεως, ’ίυα κατὰ χάριν, εἰς τὸ εἶναι βεβαίαν τὴν ἐπαγγελίαν παντὶ τῷ σπέρματι): no subject, no verb, but the sense is clear. Or, indeed, 5.12-21, where Paul almost rivals Aristotle in a kind of telegraphic style. Another problem has been raised in connection with the definite article before προπάτορα: if this noun is the predicate (‘have we found Abraham to be our forefather’) it should not (so it is claimed) have the article (Engberg-Pedersen 2000: 363-64 n. 3). Zahn already responded to this by pointing out that the assertion would not be that Abraham might be ‘a’ forefather, but that he was the forefather. This is not to be swept aside, since it corresponds exactly with the category outlined in Blass-Debrunner: despite the general rule that predicate nouns are anarthrous, ‘the article is inserted if the predicate noun is presented as something well known or as that which alone merits the designation (the only thing to be considered)’. ⁶² This is what we find here, especially I think in the light of the following ἡμῶν. It is possible, in any case, that there may be here, as often in Paul, an ellipsis: ‘Have we found Abraham, our forefather, [to be our forefather] according to the flesh?’ Avoiding the repetition of ‘forefather’ would be natural, and would then leave the article unchallenged. ⁶³

Other relevant objections have been raised in relation to κατὰ σάρκα if the question were to be read the way Hays has suggested. Watson, for instance, has drawn attention to the fact that the phrase does not occur again in the chapter, and Davies follows Käsemann in claiming that the phrase, used in the way Hays requires, would need to carry a negative connotation which it has not had so far in Romans. ⁶⁴ Wider reflection suggests answers to both points: in both Romans and Galatians Paul regularly links ‘flesh’ with ‘works’ and indeed with ‘circumcision’. ⁶⁵ The category of a family κατὰ σάρκα is precisely what Paul is discussing and setting aside in 4.13-17, though there he refers to them as οἱ ἐκ νόμου. Davies and Käsemann are wrong to suggest that σάρξ here would need to carry a negative connotation; Paul is not saying there was anything bad or wrong with Abraham’s ‘fleshly’ identity and family, just that the promise which God made, and which Abraham believed, was always about a larger unit.

A further objection has been raised: if Paul were following his normal pattern of beginning with τί οὖν ἐροῦμεν followed by a possible inference that might be drawn, but which he is going to reject, we might have expected him to follow up with μὴ γένοιτο. ⁶⁶ But this is unnecessary. Not only does Paul not always make that move in the other similar instances; in 3.9, for instance, he rejects the implication with οὐ πάντως. But in this case things are slightly more complicated. He does indeed reject the implication at the end of v. 2: ἀλλ’ οὐ πρὸς θεόν, ‘not in God’s presence’, looking forward to ‘in the presence of the God in whom he believed’ in v. 17. But before offering that functional equivalent of μὴ γένοιτο he first thinks it important to explain the force of the question he has raised. If it were to be the case that by finding oneself justified through the death of the Messiah, and brought thereby into the single family in whom the principle of the Shema is upheld, so that Torah itself is not abolished but established and God’s covenant faithfulness itself at last unveiled—if, in other words, we note the positive things that are said about the Jewish tradition, and not merely the negative ones, in 3.21-31—then does this mean that we have in fact discovered Abraham to be ‘our forefather’ in the sense of the one through whom a people of God κατὰ σάρκα has been established, which would then entail Gentile believers needing to be circumcised and to keep the whole Mosaic law?

That, I suggest, is how the question of v. 1 is to be understood, and v. 2 then amplifies it as follows. Paul has already, in 2.17-24, made it clear that the ‘boast’ of Israel—his own ‘boast’ from his former life as a zealous Pharisee—is ruled out. That boast, however, was not a matter of mere moral superiority (‘people like me are exempt from the charges you bring against the pagans’), but a matter of Israel’s vocation, the divine call of Israel to be ‘a light to the nations’: ‘a guide to the blind, a light to people in darkness, a teacher of the foolish, an instructor for children’. ⁶⁷ The boast in question was that (a) we are ethnically Jews, (b) we possess Torah, (c) the one true God is our God, so that (d) we can fulfil Israel’s ancient vocation to be the people through whom God will solve the problem of the world. This, not merely moral superiority, is the ‘boast’ which Paul is answering in 2.21-24, summing up the problem in 3.2-3, where Israel has been ‘entrusted’ (ἐπιστεύθησαν) with God’s oracles for the nations but has been faithless to that trust. The ‘boast’ which Abraham might have had, but does not have, here in 4.2 is the possibility that he might have become the father of the world-rescuing covenant family on the basis of something that was true in and of himself, a status ἐξ ἔργων which would then lead to a family that was purely κατὰ σάρκα. That was, of course, precisely how Abraham was seen in much if not all second-Temple literature (not as one who earned his salvation by doing good works, but as the one who kept Torah and so was able to be the forefather of Israel). ⁶⁸ That is the point which Paul is ruling out, and he will not do it simply by sweeping it aside with a μὴ γένοιτο. It is more serious than that. It has to do with the place of God himself in the picture: ἀλλ’ οὐ πρὸς θεόν. Just as the ‘righteousness of God’ is the real theme of the whole section, as, in a measure, of the whole epistle, so here the question of God, God’s grace and God’s life-giving power are all at stake. That is why, when Paul has dealt with the earlier arguments, he comes back to his real point in vv. 16 and 17: it must be ἐκ πίστεως, so that it can be κατὰ χάριν, because it all happens ‘in the presence of the God in whom he believed, the God who gives life to the dead and calls into existence things that do not exist’.

Paul, then, answers ‘no’ to the question of 4.1. He does this, in fact, throughout the chapter, whose climax is precisely vv. 16 and 17, the part frequently bracketed out or at least downplayed. ⁶⁹ But—in answer to another common objection to Hays’s proposal—it is the whole chapter that has this theme, not merely the section beginning with v. 9 or v. 13. ⁷⁰ That is the point of my earlier argument about μισθός in v. 4, and of the argument to which we will shortly turn about the psalm quotation in vv. 7-8. In short, if the opening question is not properly understood, the whole chapter is thrown out of balance; and that is what has happened in most readings. In vv. 16 and 17 Paul declares that ‘It is of faith, ἵνα κατὰ χάριν, so that [it might be] according to grace’. So what, here, is deemed to be ‘according to grace’? It is not, here, ‘justification’, let alone ‘salvation’, but Abraham’s fatherhood of the worldwide family. The κατὰ χάριν of v. 16 answers to the κατὰ σάρκα of v. 1. ‘We have’, in other words, ‘found Abraham to be our father according to grace, not “according to the flesh”, so that the promise might be valid for the whole seed, παντὶ τῷ σπέρματι, for all those who share Abraham’s faith’. This brings us, triumphantly, to the climax of the chapter, the complete answer to the question of the first verse: ‘And this Abraham is the father of us all, in accordance with the scriptural promise that “I have made you a father of many nations”.’ Abraham is not merely, in other words, the father of Jews, of the circumcised, of a κατὰ σάρκα family, into which Gentile converts would then have to come as circumcised proselytes, but of the whole, larger gathering which God had promised. This is what the chapter is all about. Verses 1, 16 and 17, read in this way, encapsulate and frame the theme perfectly. Commentators have either rejected or ignored this proposal because, in my view, they have continued to suppose that Rom. 4 was really a ‘scriptural proof’ of ‘justification by faith’ seen in something like the normal reformational sense. As so often, close attention to particular texts has taken second place to overarching schemes. There is nothing wrong with overarching schemes; indeed, they are vital. But they must justify themselves by showing that they make more sense of every line, every word that Paul actually wrote.

When, therefore, we read vv. 1 and 16-17 of Rom. 4 in the light of my proposal about μισθός and the justification of the ungodly Gentiles in vv. 4 and 5, these key verses gain clarity and force both in themselves and as the main frame for the argument of the chapter. What then about the obvious challenges that might be put at this point? Are there not other elements in the passage that might be taken as evidence for a more traditional view?

5. Works, Boasting and Forgiveness: Signs of an Older Perspective?

The most obvious objection that might be brought against my whole argument is the presence, in this same introductory paragraph of Rom. 4, of the theme of ‘works’ in vv. 2-6, and of the forgiveness of sins in the quotation from Ps. 32 [lxx 31] in vv. 7 and 8.

First, ‘works’. I hold to a version of the view made popular by James Dunn: that the ‘works’ which Paul says do not justify are the ‘works’ which, through their obedience to the distinctive marks of Israel’s Torah, mark out the Jews from their pagan neighbours. The long post-reformation tradition of seeing the ‘works’ in question as good moral works performed in the hope of justifying oneself by good behaviour simply does not fit the texts. In particular, it does not fit Rom. 3.28 and 29, where, as I said before, the ἤ at the start of 3.29 reminds us that when Paul says χωρὶς ἔργων νόμου at the end of v. 28 he is thereby ruling out, specifically, the ‘works’ which mark out Jewish people from pagans.

The argument for this position is long, and would involve a consideration not only of Gal. 2–4 but also texts such as 4QMMT, the only certain place in second-Temple Jewish literature where ‘works of the law’ is a key theme. ⁷¹ There, as I have argued at length elsewhere, the ‘works’ are specific points of Temple liturgy which demarcate one type of Jew from another, producing a form of inaugurated eschatology in which the person concerned may tell, in the present, that they are among the people whom God will vindicate in the future. This provides a fascinating parallel, in shape though not of course in content, with Paul’s own inaugurated eschatological framework for justification. The main difference of content is that the ‘works’ in MMT are those which differentiate one Jew from another, being post-biblical legal developments, whereas the ‘works’ in Paul are the biblical laws which differentiate Jew from Gentile. But the shape is the same: these ‘works’ will mark out those who perform them as the advance guard of God’s eschatologically vindicated people. That is ‘justification by works’. The boasting’ which Paul then excludes in Rom. 3.27 is the boast, not of the successful self-help moralist, but of the Jew whose marks of identity are a permanent barrier, keeping out Gentiles. We may of course compare 2.17-29. ⁷²

How then does this work out in Rom. 4? Verse 1 asks, as I have suggested, whether the inclusion of Gentiles in the family, and the affirmation of the law at the end of ch. 3, leaves us in a position where all Christians now have to regard Abraham as their forefather according to the flesh. As in Galatians, ‘flesh’ and ‘works’ go very closely together, not least (we may suppose) because two of the key works are circumcision and the food laws. Thus, in Rom. 4.2, Paul points out that Abraham would have a reason to ‘boast’ if he was ‘justified ἐξ ἔργων’. In the light of 3.27-31, with 2.17-29 behind it, this cannot mean that if Abraham had been a successful self-help moralist he would have a ground to boast, but that Abraham would have a ground for boasting if he and his family were marked out according to the ‘works’ which, as laid down in Torah, would keep them separate from the rest of the world and enable them to be, after all, ‘a light to the nations’. Paul answers this point with a brief negation: ἀλλ’ οὐ πρὸς θεόν. Once again we refer back to 2.17-29, where God’s comment on the Jewish ‘boasting’ is that it is empty, and worse than empty. ⁷³ In the present passage, of course, Paul answers with Gen. 15.6, invoking not only the idea of believing and being justified but, as I have suggested, the entire context of the Genesis chapter, not least the promise of the μισθός in the first verse.

It is that promise to which Paul then returns in vv. 4 and 5. Here I suggest that the word μισθός itself has indeed triggered in Paul’s mind the metaphor of someone doing a job in order to earn a ‘reward’ as of right. Paul has picked up μισθός from Genesis, which is firmly in the front of his mind, and allows an illustration to develop sideways out of it, which by coincidence happens to overlap with one way of expounding an ‘old perspective’ view of justification. But such an interpretation can safely be set aside in favour of Paul’s Genesis-based ‘covenantal’ reading. This, as we shall see, includes within it the notion of ‘forgiveness’, because the covenant was always there in the first place to deal with the sin of Adam. I have argued at length elsewhere, in fact, that the whole point of Abraham and his family in Jewish thought, going all the way back to the redaction of Genesis and all the way forwards to the post-Pauline rabbinic commentaries on the book, is that Abraham and his family are God’s chosen means of dealing with the problem entailed by Adam’s sin and its consequences. ⁷⁴ Thus if God is going to include ‘ungodly’ within Abraham’s family, this does not mean that God is condoning sin or ungodliness, ⁷⁵ but rather that God’s inclusion of them will contain within itself that dealing with their sinful state which is necessary if the inclusion is to be rescuing and restorative, which was the whole point in the first place. That, indeed, is how the larger section of 3.21–4.25 works: God unveils his ‘righteousness’ in the gospel of Jesus, in that the death of Jesus provides the way for people to be forgiven in order that God could thereby give to Abraham the worldwide family, including the ‘ungodly’, that he has always promised. Let me stress again: this is not at all to say that salvation from sin was unimportant to Paul. He presupposes it throughout. But the present argument, which refers back to God’s act of salvation in the death of Jesus, does so in order to make a different point, which is that through this saving death God has unveiled, ‘revealed-as-in-an-apocalypse’, his own covenant faithfulness as the active attribute of his character. The metaphor of ‘working for reward’, then, does not here need to go far into the finer abstract points of ‘earning’ as against ‘receiving something as a gift’. Those questions were indeed of interest to some Jews, and indeed Romans, in the period (and far more Europeans in the late Middle Ages), but they were certainly not Paul’s primary concern, here or elsewhere. ⁷⁶

Paul is here, in fact, making a double contrast: first, between ‘the one who works’ and ‘the one who believes’, and second, between κατὰ χάριν and κατὰ ὀφείλημα. The latter contrast—grace and debt—looks ahead once more to v. 16, where the emphatic ’ίυα κατὰ χάριν has nothing to do with the absence of self-help moral effort and everything to do with the promise being valid ‘for all the seed’, for Gentiles as well as Jews. ⁷⁷ We are right, then, to read these two verses not as the usual post-reformation (and post-romantic and post-existentialist!) contrast of human effort and divine grace, but as the contrast of (a) a human family κατὰ σάρκα, marked out by the ‘works’ which give them their distinctive ethnic identity, and (b) the worldwide family promised by God as Abraham’s ‘reward’, the family created by God’s gracious act in ‘justifying the ungodly’.

It is because Abraham believes in this God, ‘the God who justifies the ungodly’, that (according to Paul) he himself is ‘justified’, declared to be God’s covenant partner. As has often been pointed out, this definition of the God in whom Abraham believed (4.5) then looks forward to the further two definitions, (a) the God who raises the dead and calls the non-existent things into existence (v. 17, summing up once more the point about God’s making Abraham the father of many nations), and (b) the God who, more specifically, raised Jesus from the dead (vv. 24-25). ⁷⁸

It is that latter passage which then offers the way back to vv. 6-8, which are otherwise surprising, granted the direction of my argument so far, in their sudden introduction of the notion of forgiveness, of God not reckoning sin. I stress once more that this entire theme of ‘Abraham’s worldwide family’ cannot be played off against the more familiar themes of ‘sin and forgiveness’. Nothing that I have said here means that (as some have suggested) I have allowed ecclesiology (the single worldwide family) to elbow soteriology (how people are rescued from sin and its consequences) out of the picture. Anything but. Once we have Abraham’s family sorted out—but only once we have done that—we are able to address head on the question of how the underlying purpose of there being an Abrahamic family in the first place can now be fulfilled.

Without this perspective, it might be strange, even in reformational terms, to see Paul’s sudden switch from ‘Abraham and the covenant’ to ‘David and forgiveness of sins’. ⁷⁹ Paul has not mentioned sins, or forgiveness, elsewhere in this passage, and will not do so again until the formula in 4.25, where ‘Jesus our lord’ was ‘put to death for our trespasses and raised for our justification’. The obvious verbal link between Abraham and David is made by the idea of ‘reckoning’: the quote from the psalm speaks of God ‘not reckoning sin’, but Paul turns this round, speaking of the man to whom God ‘reckons righteousness’. This is made easier, perhaps, by the use of ‘the righteous’ at the end of the psalm (32.11). For the psalm, those whose sins are forgiven are ‘the righteous’ who are then to rejoice in YHWH.

But why, if I am right in my analysis of the chapter as a whole, does Paul want even to mention the forgiveness of sins? The answer lies in his strenuous emphasis on the inclusion of Gentiles. Obviously, the psalm might be taken to refer to Jews. ⁸⁰ But Paul, I think, makes it point to God’s determination to ‘justify the ungodly’, to bring pagans into his family. The psalm goes on to speak of the psalmist’s own experience of confession and forgiveness, but the opening two verses, cited here, appear to generalize from that into a much broader statement, as v. 6 also indicates (‘therefore let all who are faithful offer this prayer to you’). Paul seems to be picking up this much more general point, so that his ‘David’ here is not, any more than Abraham, spoken of as himself a sinner (though no doubt Paul could have said that too), but rather invoked as one who gives testimony to the blessing of forgiveness on anyone who has no ‘works’, no outward sign of belonging to God’s people. ⁸¹ In fact, as Paul says in vv. 9-12, ‘this blessing’—the blessing that David has pronounced—is intended not primarily for the circumcision but for the uncircumcision. ⁸² They have lived in a world where, by Jewish estimates, everything they are and do is by definition sinful. ⁸³ And Paul, stressing that God’s ‘reward’ to Abraham is the worldwide family for which the justification of the ungodly will be required, calls on David to emphasize too that God, after all, is the sin-forgiving God, and therefore the God who includes Gentile sinners in his family. ⁸⁴

When we check out Paul’s other references to David in Romans—which are normally screened out because they do not fit the standard profile of Paul’s theology—the result is striking. In 1.3-5, Jesus is the son of David, the risen son of God, through whom the Gentiles are summoned to faithful obedience. In 15.12, as the climax of the theological exposition of the whole letter, the ‘root of Jesse’ rises from the dead ‘to rule the nations; and in him the nations shall hope’. The vision of Ps. 2 (so important for Paul as for all early Christians), and of Ps. 72, is fulfilled, not by David smashing the pagans to pieces like a potter’s vessel (as in the Psalms of Solomon), but by him announcing God’s blessing, in line with the Abrahamic covenant, on the ungodly whose sins are forgiven as they are welcomed into the family. As Paul says in 15.8-9, the Messiah became a servant to the circumcised because of God’s truthfulness, to confirm the promises made to the patriarchs and in order that the Gentiles might glorify God for his mercy. That covenantal and eschatological scheme, rather than an atomized system of individual sin and salvation, is the point of the whole letter. Were there more space, it would be possible to follow this line of thought through chs. 9–11 in particular.

6. Conclusion

Romans 4 is, in the last analysis, about God and God’s unveiling of his worldwide saving purposes. Paul fully affirms the standard Jewish belief: when God wants to save the world and humankind from the plight of Gen. 3–11, he calls Abraham and makes promises to him, promises which are then kept, ultimately, in the Messiah. Jesus is hardly mentioned throughout the opening section of Romans, but the last two verses of ch. 4 show that he has been under the argument all along, and he will now emerge in the letter’s next section as the framework and energizing force for the exposition of what it means to be the renewed people of God. But the point Paul is making in these concluding verses of Rom. 4 is tightly integrated with the entire argument of 3.21–4.25 as a whole. The death and resurrection of Jesus is seen by Paul as the means by which God deals with, and so can forgive, the trespasses which would have kept both Jew and Gentile out of the category of ‘righteous’. Believing in the God who raised Jesus from the dead is seen by Paul as the crucial mark of the church (compare, of course, 10.1-13), the full expression of the faith which Abraham had when he believed in the God who raises the dead and creates out of nothing (4.17-22), which was itself the larger belief enclosing and including the specific belief in ‘the God who justifies the ungodly’ (4.5). People still try to make out that the ‘new perspective’ on Paul is a matter of sociology rather than soteriology, of the removal of minor inconveniences for Gentile converts rather than God’s victory over sin and death through Jesus the Messiah. There may be some who have taken it that way, though neither Ed Sanders himself nor Jimmy Dunn has been guilty of any such reductionism. Romans 4 shows how the whole picture hangs together.

It is nearly forty years since the late Bishop Stephen Neill remarked to me that the most important tool for understanding the New Testament was the concordance to the Septuagint. I paid attention, but not enough attention, at the time. The little word μισθός in Rom. 4.4, once we recognize it as an allusion to Gen. 15.1, enables us to clarify the entire argument of the chapter and, in so doing, to firm up a holistic and deeply Jewish reading of Romans in general and ch. 4 in particular. Paul has not introduced Abraham as an ‘example’ or ‘scriptural proof’ of a ‘doctrine’ (Käsemann 1980: 105), not even as ‘a decisive test case’ (Dunn 1988: 199). He is not quoting scripture merely to create a rhetorically powerful impression of his great learning; he has the whole context of Gen. 15 in mind, and his discussions of Abraham only make sense when we have it in mind as well. ⁸⁵ Paul’s actual use of Genesis dovetails perfectly with his theological perspective. Abraham, throughout Rom. 4, is the one with whom God made a covenant to rescue the whole world from the Adamic plight of sin and death, a promise now at last fulfilled in the Messiah. This contributes to the larger argument of the letter, which is a vindication of the covenant faithfulness of God, as seen in the powerfully saving gospel of Jesus the Messiah, and a plea, on that basis, for a radical ecclesial unity which will sustain the missionary theology Paul expounds and, as he explains in ch. 15, is still intent on implementing. Whereas Galatians, we might suggest, presents an ecclesiology in order to hold the ‘truth’ of the high ground we might characterize in terms of ‘apocalyptic’ and/or ‘eschatology’, ⁸⁶ Romans offers a theology (‘God’s righteousness’), revealed in the saving death and resurrection of Jesus, which issues in an ecclesiology. The soteriology of Romans and the ecclesiology of Galatians, and behind both of them the vision of God’s faithfulness and truth, hinges for Paul on his belief that in the crucified and risen Messiah God had done what he told Abraham he would do. Abraham is therefore the father of all believers, not ‘according to the flesh’, but according to grace. The ungodly have been justified. That, according to Paul, is the patriarch’s reward.