rabbisaul.com home

Covenantal nomism?
A comparative review of Sanders and Carson et al

by Tim Gallant


The field of biblical studies has perhaps seen no more dramatic revolution in recent years than in the scholarly approach to first century Judaism, particularly with how it intersects with Paul's writings. The publication of Paul and Palestinian Judaism by E. P. Sanders in 1977 triggered what has come to be known as the "New Perspective on Paul" (NPP), and has cut a wide swath across New Testament studies. In can well be said that Sanders' work has been significant enough that even those biblical scholars who disagree with his thesis must interact with it in order to be taken seriously.

Ironically, the Sanders thesis, which calls into question long-standing opinions regarding Judaism as a "merit theology," was not genuinely revolutionary. Many Judaic specialists had quarrelled with the accepted view for quite some time, at least as early as G. F. Moore, and numerous Jewish scholars had argued the same point in one fashion or another (Montefiore, Schoeps, Sandmel).

The particular contribution of Sanders, however, was to (1) investigate a very large body of primary sources with this singular question in mind; and (2) write a book that would be widely read by those in the community of biblical scholarship. The result was a widespread paradigm shift of significant proportions. Many top New Testament scholars in the ensuing years found themselves agreeing with Sanders, and their approach to passages in Paul's writings, previously thought to be a polemic against "merit theology," was radically affected.

This shift has not gone unchallenged. Many scholars have resisted the Sanders thesis. Many conservative Protestants, who are understandably nervous about any attempt to reinterpret Paul (given the theological weight that Protestants have always placed upon a generally Lutheran reading of Paul's letters), have responded with a great deal of fire. Much of the latter response has attempted very little in terms of refuting Sanders' work in the primary sources, however.

This, however, is changing, as more biblical scholars recognize that Sanders must be answered on his own terms if the momentum of the NPP is to be slowed or stopped.

In light of these developments, I have chosen here to reflect on (1) the seminal work itself, Sanders's Paul and Palestinian Judaism; and (2) a recent compilation aimed at reevaluating Second Temple Judaism, under the leadership of the respected conservative Protestant biblical scholar D. A. Carson. Along with Peter O'Brien and Mark Seifrid, Carson edited Justification and Variegated Nomism: Volume 1 - The Complexities of Second Temple Judaism (the second volume is aimed at a fresh interpretation of Paul himself).

The Challenge of Sanders

The fundamental point which E. P. Sanders sought to establish in Paul and Palestinian Judaism was that, rather than being a ladder-climbing religion in which Jews attempted to earn God's favour through meritorious works, Judaism was a religion whose pattern could be best described as "covenantal nomism." Sanders summarizes this way:

Briefly put, 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 transgression. (Sanders, 75)

In Sanders's view, Judaim employs righteous language (the Greek dik- word group) differently from Paul. In Paul, it is a verb (usually translated justify) employed as a "transfer term" to describe the movement into salvation; for Judaism it refers as an adjective to refer to the correct behaviour of those who are already part of the covenant (544-545). (The exception is Qumran, which did "righteous" language as a transfer term, although, unlike Paul, Qumran employed it with reference to those who were righteous, not with reference to undeserved justification.)Paul and Palestinian Judaism book cover

Sanders's treatment handles a great deal of material: Tannaitic literature, the Dead Sea Scrolls, the Apocrypha, and the Old Testament Pseudepigrapha. His overarching argument is that, with the exception of the apocalyptic IV Ezra, the literature as a whole presupposes election as the foundation of the covenant. Israel is God's people due to His own choice long ago; they need not earn entrance. The case of IV Ezra is indeed otherwise: here, "the human inability to avoid sin is considered to lead to damnation. It is this pessimistic view of the human plight which distinguishes the author from the rest of Judaism as it is revealed in the surviving literature" (418). "Otherwise, however, in all the literature surveyed, obedience maintains one's position in the covenant, but it does not earn God's grace as such. It simply keeps an individual in the group which is the recipient of God's grace" (420). Obedience to Torah (which would include employing the law's means of atonement for transgression), according to Sanders, indicated an intention to continue in the covenant. Those who repudiated obedience were repudiating the covenant, and thus disqualified themselves from it.

Judgment according to works

Sanders acknowledges that the literature envisions a future judgment according to works. It must be understood, however, that (with the exception of IV Ezra) the theme of strict retribution is interwoven with the theme of God's mercy.

There are two different formulations concerning mercy and justice. One is that of Rabbinic literature: God's mercy is greater than his justice. In the other literature, the usual formulation is that God punishes the wicked for their deeds, while bestowing mercy on the righteous.... The themes of mercy and retribution or justice are not actually in competition, but serve different functions. Statements to the effect that God pays each man his just due serve to assert the justness of God and to assure both sinners and the righteous that what they do matters. God is not capricious. He will neither punish for obedience nor reward transgression. The theme of mercy - whether put in terms of God's mercy in electing Israel, God's mercy in accepting repentant sinners (repentance does not earn a reward, but is responded to by God in mercy), or God's 'rewarding' the righteous because of his mercy - serves to assure that election and ultimately salvation cannot be earned, but depend on God's grace. One can never be righteous enough to be worthy in God's sight of the ultimate gifts, which depend only on his mercy. The theme of God's mercy as being the final reliance even of the righteous appears in all the literature surveyed except IV Ezra.... (421-422)


The question that arises is: what about the Rabbinic language that those who do a certain thing merit God's favour? If the literature explicitly employs the language of merit (Heb. zekut), how can Sanders' scheme stand?

Unsurprisingly, Sanders challenges the common view of the meaning of zekut. Zekut does not refer to works of "supererogation," or "merit," for that matter. George Foot Moore had already argued that zekut "cannot be so construed" (45).

Sanders writes,

We should note here that zekut, especially when prefaced by the preposition bet (by, in), does not necessarily bear the full meaning of the English word 'merit'. That is, one should not necessarily suppose that the appearance of zekut always implies the full doctrine of stored-up merits which some scholars have found in Rabbinic literature (and compared to the Roman Catholic 'treasury of merits'). Zekut is closer to the English word 'virtue' in one way: both can bear a full or a weak meaning. Thus to say that someone is a person of virtue is a significant use of the word; but in the phrase 'by virtue of' the meaning is weakened to little more than 'because of'. One should not, therefore, grow too excited over every appearance of the expression bizekut - 'by virtue of' or 'by the merit of'. (90; for a good example of zekut being used in this sense, see the discussion of Aboth 5.18 on p. 187.)

Moore in fact argued that (at least frequently) bizekut should be translated "for the sake of" (183). Phrases commonly taken to mean "because of the merit of the fathers" should be understood rather as "for the sake of the fathers" - a phrase that could mean no more than the prevalent biblical notion that because of God's promises to the patriarchs, He faithfully treats their descendants with His favour. (It should be pointed out, however, that even Sanders does not claim that this necessarily means that the phrase always refers to pure grace, without reference to the righteousness of the fathers. While he doesn't see merit being at issue, he does seem to agree with Sjoberg that "the actual righteousness of the fathers" seems to have been in view in most of the Rabbinic references [184]. Nor does Sanders appear to deny that later, medieval Judaism may have developed something resembling the merit theology of the Roman Church. One tends to wonder if these developed in parallel.)

Beyond the usual problematic understanding of zekut, Sanders argues that the object has usually been ignored. When there are "rewards for merit" these are "almost without exception quite concrete historical rewards and are not soteriological" (189). Consequently, the idea that Judaism was a religion of salvation by merit is simply wrong-headed all round.

Central issues

Some have characterized Sanders's view as suggesting that Judaism and Paul had essentially the same "pattern of religion." This is incorrect. Sanders' view on Paul (which are very problematic on a number of fronts) is that Paul's soteriology is essentially "participationist" rather than covenantal. It is not so much that Paul disagrees with covenantal nomism as that the category is beside the point. Moreover, unlike Paul, Sanders notes that Judaism had no doctrine of original sin (114ff.). Sanders believes, however, that Paul's main objection to Judaism was simply that it was not Christianity, and he therefore developed all sorts of arguments to oppose Judaism.

Fundamentally, Sanders's criticism of the received tradition of Judaism as meritorious is that it was built on a false methodology from the beginning. The Reformers saw Paul refuting merit theology; thus, let's go to the primary sources and find the merit theology there. "We have here the retrojection of the Protestant-Catholic debate into ancient history, with Judaism taking the role of Catholicism and Christianity the role of Lutheranism" (57). Consequently, the monumental works on Judaism done by biblical scholars (Weber, Schurer, Bousset) presupposed merit theology before the texts had even been examined. This was corrected by Moore, but because he carried on little polemic, his own work simply was read in the light of the others (33-34, 58-59). Sanders thus represents an attempt to read Judaism on its own terms, try to detect a pattern of religion, and only thereafter compare it with Christianity.

How should we evaluate Sanders's work? This remains an open question, and as someone who is not an expert in the Jewish literature, it would folly to pretend competence simply to accept or reject Sanders' thesis on any basis other than what is evident in his own writings. In other words, I'm not prepared to go behind his work to the original sources and claim that he misunderstood context, for example.

A book, however, can be judged by the evidence that it does contain. When the writer quotes a passage, do his comments follow from the apparent meaning of that passage? Do the conclusions follow from the cumulative argument? It is possible for a non-specialist to speak to such questions without getting in too far over his head. And of course, there are related issues which have little to do with the primary sources of Judaism at all. Here, then, are some critical observations.

1. Sanders's point regarding approach is well-taken. When one's primary reason for investigating Jewish literature is to prove what you think you already know, there is a tremendous pressure to demonstrate what you have already presupposed. (It is telling, however, that Sanders shows little awareness that this could work two ways. If your primary reason for investigating the literature is to exonerate Judaism from prejudicial Protestant assumptions, there is surely a tremendous pressure to validate that purpose as well.) In such a context, it is difficult to ask the right questions of the literature. Sanders has an interesting discussion of David Flusser's work, which went in the opposite direction: Flusser examined Qumran, and then searched for links in Christianity. Not surprisingly, the resultant view of Christianity was greatly distorted (15-16).

2. In connection with (1), we also need to acknowledge that the most extreme views regarding Judaism were filtered through two sets of (potentially distorting) lenses: i. the Reformers' struggles with the legalism of medieval scholastic Roman Catholic theology; and ii. the scholarship of heavily anti-Semitic 19th century German scholars. If this sort of pressure is applied, it is difficult to believe that no distortion follows.

3. Sanders himself, however, distorts the body of evidence by refusing to deal with the New Testament witness to first century Judaism. The New Testament is not to be admitted as witness, because it is "polemical" and thus an unsafe guide (cf. 152 n. 23). This is where we must face up squarely to the problems of dealing with unbelieving critical scholars. (To get a clear picture of Sanders's "faith," look up his cynical entry "Truth, ultimate" in the subject index.)

4. Having said that, it is important that Protestant scholars recognize that the problem of retrojection does not begin with the treatment of the Judaic texts, but with the Scriptures themselves. If it is possible to impose the polemical concerns of the Reformation era onto the Jewish literature, it is equally possible to impose them upon Scripture itself. In a religious context (Roman Catholic scholastic theology) where "law" had been dehistoricized into a "merit principle," it was very easy to presuppose that meaning when one read of "nomos" in the New Testament. Likewise, when one's polemical foes are promoting salvation by merit, it is easy to unselfconsciously see merit-mongers in the polemical foes of Jesus and Paul. We need to ask ourselves whether, quite apart from the extrabiblical evidence, we can reread all of those texts in any different light; whether the critique of Judaizers and non-Christian Jews might make better sense without presupposing an omnipresent merit theology. (On this point, cf. my article on the shape of Paul's argument in Galatians, "What Saint Paul Should Have Said.")

5. As he himself foresaw, Sanders's attempt to treat the whole of Second Temple Judaism as one, and determine a single "pattern of religion" is itself a questionable enterprise. As has been repeatedly stressed, it is more accurate to speak of Judaisms (plural) than of some monolithic entity that never existed. (To be fair to Sanders, he was more aware of this difficulty than some of his critics seem to credit him for, and he was generally careful not to build too large of a "pattern of religion" that could not be shared between disparate groups.) The significance of this disparity, however, should not be exaggerated (more on this point below).

6. Sanders's reading of Paul himself has been (rightly) questioned by other so-called NPP biblical scholars. In my view, Sanders's handling of Paul reveals somewhat limited theological competence (see e.g. his apparent assumption that substitutionary and participationist themes must be mutually incompatible [463-468]). He also makes unqualified assertions which are demonstrably untrue. According to Sanders, Paul "never appeals to the fact that the Messiah has come as a reason for holding the law invalid" (479-480; italics original). That, however, is precisely what Paul argues in Galatians 3:15-25. In view of these limitations, one may well question Sanders's theological capability to evaluate the literature in question and reliably determine a "pattern of religion."

7. Finally, it should not be necessary to stress that the so-called New Perspective on Paul, broadly considered, should not be identified with Sanders in a way that what he says is ascribed to other scholars. Few biblical scholars have followed Sanders far in his exegesis of Paul; most of them are at fundamental odds with him. In that regard, there really is no such thing as a New Perspective on Paul - rather, there are a great many biblical scholars attempting to make new sense of Paul in light of a new perspective on Judaism, and there is a wide variety in these efforts. Even in connection with Judaism itself, many so-called NPP adherents have provided correctives and modifications to Sanders along the way.

Justification and Variegated Nomism

One way for a non-expert to test the validity of a thesis is by studying the responses by other experts to that thesis. This, of course, is not fail-safe. Scholars do not always employ the best arguments and cite the most damaging material when they seek to refute a thesis. But generally, we do expect that when a field of top scholars each work in narrower areas than the writer of the original thesis did, we would expect a more careful body of work that is likely to reflect reasonably wise choices being made with regard to the available material.

Justification and Variegated Nomism book coverHere then is the value of the first volume of Justification and Variegated Nomism (JVN). Rather than employing one scholar to cover the whole range of literature in response to Sanders, this work specializes. Each scholar examines a genre, sub-genre or a single author (such as Josephus or Philo).

The book cover identifies JVN as "a fresh appraisal of Paul and Second Temple Judaism." This volume constitutes the fresh appraisal of Judaism; the second volume will be devoted to Paul.

Carson's summary

To "get at" what this volume actually accomplishes, I will begin with Carson's summary evaluations of the findings.

1) Carson acknowledges that the contributors come to a variety of positions (543).

2) Several scholars find at least partial evidence of covenantal nomism (543).

3) There is "strong agreement that covenantal nomism is at best a reductionist category" (543).

4) The notion of covenantal nomism is misleading, because it "engenders an assumption that there is more uniformity in the literature than there really is.... Sanders' formula is rather difficult to falsify, precisely because it is so plastic that it hides more than it reveals." Even "more importantly, covenantal nomism is a false alternative to merit theology." Although earlier treatments were not accurate readings of the material, Sanders' reading is likewise inaccurate. By contrasting merit theology to covenant theology, Sanders has constructed a system which "reserves grace in the 'getting in' while preserving works (and frequently some form or other of merit theology) in the 'staying in.'" Because the covenantal nomism category is so elastic, it "includes and baptizes a great deal of merit theology" (544). In connection with this, the whole idea of "staying in" is itself too flexible, since "obedience may mean 'faithful conformity to God's gracious revelation... enabled and empowered by God's help' or 'the human contribution to the entire scheme'" (545).

5. In some of the literature, even Sanders' flexible category (covenantal nomism) seems simply mistaken. Carson here cites Josephus and Jubilees, where election "is mere initiation; final salvation finally depends on obedience." Because one is born "in," all the focus ends up on "staying in." Thus the covenantal election is merely preliminary, and salvation really is "accomplished by the 'staying in'" (545-546).

6. Seifrid argues that "righteousness" language is creational, rather than covenantal; thus "the fundamental assumptions behind covenantal nomism begin to crumble" (547).

7. Even "getting in" is not always tied to grace, because there is a prospect of human sinlessness, which does not require grace to begin with (547).

8. Carson notes Bockmuehl's developmental thesis. Bockmuehl "argues that its direction is precisely toward the sort of 'works of the law' that the apostle Paul would have opposed" (547).

9. Even if the "constitutional documents" of Judaism were not merit-mongering, in practice ordinary people may not think in the same line (547-548).

Carson concludes,

Examination of Sanders's covenantal nomism leads one to the conclusion that the New Testament documents, not least Paul, must not be read against this reconstructed background - or, at least, must not be read exclusively against this background. It is too doctrinaire, too unsupported by the sources themselves, too reductionistic, too monopolistic. The danger is that of the "parallelomania" about which Sandmel warned us, by which texts are domesticated as they are held hostage to the ostensible background called forth by appealing to certain other antecedent texts.

My own evaluation of the accomplishment of this work is, frankly, somewhat different from Carson's. This is for several reasons. 1) Carson's concluding summaries do not always accurately reflect the chapters in question. 2) Some of the chapters themselves do not demonstrate their own conclusions. 3) Many of the contributors evince a failure to wrestle profoundly enough with the issues before them.

It is also to be noted that, while several of Carson's points are indeed well-taken and poignant, we need to be well aware that even Carson's summary - while itself a stretch - does not dare claim that the traditional viewpoint was accurate. This, I think, is telling.

Assertion rather than demonstration

Non sequiturs

Aside from Carson's misleading summarization (on which, see below), some of the authors themselves are guilty of attempting to undercut the force of Sanders's thesis by way of unsupported speculation rather than by way of actual demonstration. An example of this is Daniel Falk's suggestion:

It cannot be ruled out that in some instances the penitential prayer itself may become a pious activity that merits God's favor. This of course is virtually impossible to track [!], but certainly in the sectarian covenant ceremony the penitential prayer is transformed into a formulaic part of a ritual to reinforce the boundaries between those under God's favor and those excluded in the last days. (16; exclamation mark mine.)

In response, I note: (1) It is altogether too convenient to make a prejudicial but inherently undemonstrable claim (i.e. assume that the prayer functions in a manner that contradicts its actual contents, which is what the case in question would amount to); (2) Falk's second sentence above does not support the first: just because a community has clear boundaries does not necessitate a theology or system of merit.

Falk is repeatedly guilty of overexegeting. He claims that in the Hodayot there "are statements that God will purify individuals of sin because of their obedience, and in relation to special divine knowledge granted to them" (32). But the three quotations he provides do not suggest that this purification is on the basis of obedience.

Philip Davies commits almost the ultimate faux pas when he first admits, "The narrative itself gives little insight into the manner in which Jews were seen to be justified by God," and yet somehow immediately "concludes" thus: "One is justified, whether Jew or Gentile, by obedience to the Law, but also by obedience to reason, which is the same thing" (121). Go figure. The narrative itself provides little insight into the question, but I suppose we can tell what it really means, anyway.

Prejudicial questions

Philip Davies simply poses the question prejudicially from the start:

...we should be aware of the different possibilities for understanding the nature of the divine-human relationship that the genres entail, including the ethical system that guarantees a person divine favor and, indeed, what it is that earns our "salvation." (105)

So now we start with the notion that it is the ethical system that guarantees favor and earns salvation. Kind of like: "Have you stopped beating your wife - yes or no?"

Fundamental confusion

There are some arguments that reveal further problems. Falk, for example, claims that Qumran cannot be construed as holding to covenantal nomism, because in Qumran, membership was conditioned upon taking the law upon oneself (32). Falk, however, misunderstands the point. Qumran regarded Judaism as a whole as apostate, so it became necessary to re-enter the covenant by joining the sect. This re-entry, however, was attributed to election. Contra Falk, this does not demonstrate that Qumran was not under a covenantal nomism pattern; it still falls under the rubric Sanders has provided. (Falk is not the only writer who seems to presuppose that covenantal nomism can only exist if the covenant is in force with the whole nation.)

Evans suggests that in Joseph and Aseneth,

...one is left with the impression that the change of lifestyle, not least the change of diet, plays a vital role in the redemption of Aseneth. God's grace is the presupposition, to be sure, but apart from wholesale adoption of Jewish food and purity laws, the conversion of Aseneth could not have taken place. (65-66)

The problems here are at least twofold: (1) Aseneth is a convert, and thus has to enter somehow. It would be unthinkable for a convert to Judaism to say no to the Jewish law, as much as it would be unthinkable for a Christian "convert" to decline baptism and participation in the Lord's Supper. (2) The food and drink in the passage which Evans cites is quite clearly figurative; it is questionable to refer them to the dietary laws, in any case.

Martin McNamara questions the value of "covenantal nomism" as "an apt description of any form of Jewish religion." Why not?

The term "nomism" tends to denote a static position, conformity to a defined set of rules. Covenant, on the other hand, has reference to a living God, who was and who is, and will be, a God active in the past, the present and in the future. The covenant is not a term that describes a static religion. (355)

These observations, however, seem to me to miss the point. For first, McNamara himself has shown that the Targums do not present a "consistent" viewpoint, such that Law-religion would be seen as incompatible with covenant, as too "static." In addition, McNamara has also shown that the Targums do not treat the Law as merely objectified and static; indeed they use Law terminology to refer to God Himself.

Mark Seifrid seems unclear on what would be entailed in disproving Sanders's thesis. He spends considerable time showing that berith (covenant) had various uses, including the promises given to Abraham, but also to Torah and its individual demands. Consequently, Seifrid asserts, "Sanders's use of the phrase 'being in the covenant' to convey the idea of 'participating in salvation,' does not fit the nature of rabbinic usage, since here again the idea of obedience is often attached to berith, as in the Scriptures" (438). This line of reasoning appears flawed on several counts. First, Sanders never claimed that his view derived from rabbinic usage of the term berith as such. Second, Sanders never claimed that the covenant did not entail obedience, but precisely the opposite. Third, Seifrid's entire argument is carried by an unestablished presupposition that the (biblical) covenant is not soteriological in nature.

Is the Old Testament okay? How about the New?

Some of the discussions appear to have some implicit charge against the literature for material that sounds a great deal like the Old Testament itself.

For example, Falk writes concerning the Psalms of Solomon: "the motifs of God's righteousness and mercy, Israel and covenant are redefined in terms of the opposing categories pious/sinners rather than Israel/Gentiles. Consequently, individual choice and behavior are determinative for one's destiny" (37). It is difficult to see how the first sentence undermines covenantal nomism; presumably the pious are those who affirm the law, sinners those who reject it - precisely Sanders's categories. But apart from that, one may wish to reflect upon the biblical language that makes the same sorts of distinctions throughout without ever inferring a works-righteousness.

Peter Enns, whose overall treatment is very responsible and evenhanded, asks in connection with Jubilees, "How can there be sins unto death when election is the basis of salvation?" That would seem to be a good question to ask of 1 John 5:16-17....

Markus Bockmuehl evaluates Qumran:

The texts manifest an uneasy co-existence of the belief that atonement for sins remains emphatically an act of God with other statements that the faith and life of the community, and of its priestly leaders in particular, is in some sense instrumental to that atonement.

Aside from "in some sense instrumental" vagueness, presumably Bockmuehl would also see the Old Testament pattern as made up of an "uneasy co-existence" between divine atonement and the sacrificial cult.

Bauckham on 2 Baruch: "the notion of salvation as the reward for the righteous does not imply a nice calculation of merit and reward, but it does make salvation dependent on adherence to God and his Law" (182). Hm. Sounds something like David:

But the lovingkindness of the LORD is from everlasting to everlasting on those who fear Him, and His righteousness to children's children, to those who keep His covenant, and who remember His precepts to do them. (Ps. 103:17-18)

Some of the contributors do recognize the analogy, however. Philip Alexander, for example, says that the Tannaitic literature is a faithful mirror of Scripture itself (268-269, 273). (Unfortunately, what he means by that is that both the Old Testament and the Tannaitic literature are rife with not only tensions, but contradictions. However, he does concede, "The implication may be, if we really want a rationalization of such a paradoxical viewpoint, that the contradiction exists only at the level of human understanding, but presumably does not exist at the level of the mind of God" [273].)

Alexander later points out, "The charge of legalism often ignores the fact that the rabbis inherited from the Bible a comprehensive system of law which was meant to serve as the constitution of a people" (279-280).


When a point cannot be proven decisively, one can always resort to language that could be understood in radically differing ways.

Evans concludes that "there is no indication... that [the writers] believed that people could gain God's acceptance apart from obedience to the Law" (72). Evans does not define what he means either by "apart from" or by "obedience." One may wonder if he would have a problem with the assertion in Hebrews that "without holiness no man shall see the Lord" (Heb. 12:14).

Bauckham similarly vacillates between terms of adherence and terms of performance (see e.g. 182: "It is in his grace that God makes the covenantal promises and lays down the requirement of obedience to the Law as the condition for receiving them"). What is meant by "requirement of obedience"? What is meant by "condition"? We are left unsure what Bauckham has proven, or hoped to prove.

Philip Alexander, too, seems to have some impulse to place the discussion in terms of "works-righteousness," without being clear on what that must mean. He writes, "Tannaitic Judaism can be seen as fundamentally a religion of works-righteousness"; there is little indication "that God can simply forgive the sinner without any action whatsoever on the sinner's part" (300). This vagueness, however, leaves us with the impression that a requirement of repentance would constitute "works-righteousness." Here, as elsewhere, the issues have not been grappled with at anywhere near a profound enough level.

Not so legalistic after all?

Lest anyone think that I am simply taking shots at the contributors in order to avoid the force of their own conclusions, I wish to make clear that many of them admit openly that they have not established merit theology from their sources.

...throughout the Psalms of Solomon the righteous display confidence in the face of their sin and God's discipline, but it is not expressed as confidence in their behavior. Rather, the righteous appeal to God's mercy on the basis of the election of Israel and God's covenant with their forefathers. (Falk, 47)

The Apocalypse of Zephaniah teaches "that even the righteous can be saved only by the mercy of God in forgiving their sins" (Bauckham, 158).

Robert Kugler's findings in connection with the study of the Testament of Moses is interesting. He does not think that this work fits with covenantal nomism. Why not? Too legalistic? Actually, no. This work embraces the concept of election but rejects the idea that "obedience [to the laws]... or atonement and repentance for transgression" are ultimately necessary "for remaining in the covenant community." And likewise, "while the Testament of Job admits a God-fearer into the community of the elect, it only requires of him trust in God to maintain that relationship" (189-190). All I can say is that if this analysis is correct, it may not be covenantal nomism, but looks even less like the traditional characterization of first century Judaism.

Donald Gowan on Sirach: "what he says of the relationship between God and the Israelite differs little from that of the book of Proverbs" (216). "People are judged according to their works (16:12b), but that is no more a strict works-salvation scheme in Sirach than in the Old Testament" (219). Similarly, Gowan coments on Baruch: "God's justice has always been tempered with mercy, however (2:27), and... [He is] ready to forgive the undeserving (4:28-29). Israel has no merit which might enable them to claim a continuing relationship to God" (222). As a whole, Gowan sees the material thus: "There is no suggestion that works of the law are expected to prove anything to God, or that God counts up merits.... The books thus begin with the assumption that the readers are 'in,' and that there is another group, the enemies, who are 'out'" (239).

On the matter of the meaning of zekut, so often translated "merit," McNamara admits: "It is doubtful that in the matter [of the phrase, zekut] of the Fathers, the Targums are far removed from what Paul writes to the Romans concerning Israel" (332; citing Rom. 9:4-5; 11:28; McNamara notes that the latter text's phrase dia tous pateras "corresponds well" to the Targumaic use of "the zekut of the Fathers").

Bockmuehl on salvation and justification in Qumran: "God's righteousness and God's righteous acts, therefore, constitute the salvation and justification of the individual" (399).

Final evaluation

In the light of what we have seen above, let's return to Carson's concluding summaries and observations.

1) It is clear enough that the contributors come to a variety of conclusions.

(Thus no further comment is necessary.)

2) It is certain that several of the scholars found evidence for covenantal nomism.


3) There is "strong agreement that covenantal nomism is at best a reductionist category" (543).

This observation is true as far as it goes, but one must also be aware of Sanders's original purpose which was, after all, quite limited. Those who have been largely persuaded by Sanders's thesis are not under any illusion that they now have some comprehensive understanding of Judaism. That would be beside the point.

4a) The notion of covenantal nomism is misleading, because it "engenders an assumption that there is more uniformity in the literature than there really is.... Sanders' formula is rather difficult to falsify, precisely because it is so plastic that it hides more than it reveals."

The diversity of Second Temple Judaism has been stressed repeatedly, and Sanders himself acknowledged it. Yet, in connection with our interests in New Testament studies, this stress on diversity is actually somewhat misleading.

Interestingly, an essay in this volume illustrates this very well. In his monumental and refreshingly clear-headed contribution, Roland Deines argues that Sanders undervalues Pharisaic influence (444ff.), and states "when the historical course of events is presented, the Pharisees appear as the most clearly recognizable and socially active group over the entire span of time" (447).

While we do well to take note of the diversity in Judaism, the notion that first century Judaism was made of dozens or more of tiny groups competing with one another has little to say for itself. Recent scholarship has made it quite clear that Pharisaism was the dominant voice, and was largely viewed by the common people as the mainstream faith, while the Sadducees held most of the temporal power. Deines argues that the Pharisees were not a sect - exclusive and closed; "The purpose of the associations of Pharisees was not to fence themselves off, but to mark themselves as a reform group within the society (449-450; emphasis mine).

Deines writes:

Conversely, everyone who recognizes the Torah and its tradition as handed down by the Pharisaic teachers as normative can call himself a Pharisee (which, however... was not customary). A lot of what Sanders calls "Common Judaism" can thus be labeled in a wider sense as a pharisaic influenced Judaism. (494)

Deines concurs with Moore that the Am ha-Aretz ("people of the land") constituted "the majority of the Jewish population, which, as far as its values were concerned, belonged to Pharisaic Judaism, even though they lived these ideas in reality only to a very limited extent." It is thus possible to speak of Pharisaism as "normative Judaism, not because all lived according to Pharisaic halakah, but because Pharisaism was by the majority acknowledged as legitimate and authentic interpretation of the divine will for the chosen nation" (501).

This surely is a point that we could have derived from the Gospels: the ostensible spiritual authority which we find the Pharisees exercising was not strictly "political" (they were not strong in civil or official ecclesiastical positions); Pharisaism surely would have needed a considerable accepted public base for such actions to make sense.

It is all fine, then, to recognize literature representing sectarian strands of various stripes, but in biblical studies, it is not so strange to think that the overwhelming majority of cases of confrontation between Christianity and Judaism would have involved either the Pharisees or the Sadducees. Nor is this mere assumption; in many cases, this is explicit in the biblical text. Elsewhere, it is not all that difficult to deduce. It is frivolous for New Testament scholars to waste time arguing over what some tiny sect may have believed, when the chance of intersection with the biblical text is frankly minimal.

Deines, however, goes even further than this: Sanders is "to be thanked for placing the concept of what was 'common' to Judaism so prominently in the center of his works. What Sanders has persuasively shown is that Judaism is to be understood as one religion, and was perceived to be such by outsiders." Non-Jews had very specific conceptions of the Jewish religion, involving "Sabbath, circumcision, food laws, and the temple cult" on the practical level, and "monotheism, exclusivity, and a particular ethic" on the belief level, as well as common roots, particularly in the Old Testament (453-454). The party divisions between Essene, Pharisee, and Sadducee were primarily halakhic, not "theological" (499-500). (Although cf. the Sadducee's denial of resurrection, etc.) "All three Jewish movements oriented themselves basically around the Torah as the center of individual and national Jewish existence" (500).

4b) Even "more importantly, covenantal nomism is a false alternative to merit theology." Although earlier treatments were not accurate readings of the material, Sanders' reading is likewise inaccurate. By contrasting merit theology to covenant theology, Sanders has constructed a system which "reserves grace in the 'getting in' while preserving works (and frequently some form or other of merit theology) in the 'staying in.'" Because the covenantal nomism category is so elastic, it "includes and baptizes a great deal of merit theology." In connection with this, the whole idea of "staying in" is itself too flexible, since "obedience may mean 'faithful conformity to God's gracious revelation... enabled and empowered by God's help' or 'the human contribution to the entire scheme.'"

Here we grant the problem with Sanders's category, although it is interesting that only Bauckham, in connection with IV Ezra and 2 Baruch can clearly establish "staying in" by way of merit.

Beyond this, these observations must also be placed in the context of our wrestling with Scripture itself. When Paul excommunicates a man for committing fornication with his father's wife (1 Cor. 5), for example, is he or is he not implying something about "staying in" in connection with works? Being able to point to such phenomena in the Judaic texts cannot prove merit theology in itself; that would be proving too much. Then we would have Paul blaming Judaism for what may appear to be his own position. If there is a fundamental differentiation between what is going on in 1 Corinthians 5 (and numerous other biblical passages) and what is going on in the Second Temple literature, that difference needs to be articulated, not just assumed. Part of the problem throughout this book, as well as other discussions of the subject, has been a profound failure to grapple with the biblical data which could easily be paralleled to a great deal of the supposedly problematic Judaic texts.

The point of this observation, of course, is not to imply that there really is no difference between Paul and Second Temple Judaism. The point, rather, is to challenge us to think a little more clearly about just what that difference is.

5. In some of the literature, even Sanders' flexible category (covenantal nomism) seems simply mistaken. Carson here cites Josephus and Jubilees, where election "is mere initiation; final salvation finally depends on obedience." Because one is born "in," all the focus ends up on "staying in." Thus the covenantal election is merely preliminary, and salvation really is "accomplished by the 'staying in.'"

The citation of the Josephus article is somewhat misleading here. It is widely recognized that Josephus, as an apologetic writer, deliberately cast the various Jewish movements as "philosophical schools" in order for them to be comprehensible to his Greco-Roman audience. Given this, placing a strong accent on the covenant would have been mystifying on his part. It is debatable that we can detect a "pattern of religion" in Josephus, for the simple reason that he is deliberately posing Judaism(s) as philosophy, not religion.

As for the Jubilees discussion, I find it somewhat curious that Carson bothers to appeal to Enns to support his point. Out of all the contributions, that of Enns is one of the least amenable to subverting the notion of graciousness in Judaism.

Enns does point out some unclarity in Sanders's own comment on salvation in Jubilees: "It is repeatedly emphasized that the basis of salvation is membership in the covenant and loyalty to it" (Sanders 367). Here, it is to be doubted whether Sanders is attempting to speak with anything approaching theological precision (which he rarely does), and Enns capitalizes upon it.

Nonetheless, even here, the difficulties which Enns raises are difficulties which could be raised against Scripture itself. "How is it that Israel is God's eternal, elect, covenantal people, whom he will never forsake, while at the same time there are sins for which there is no possibility of atonement, and this results in this person's elect status?" (95) Frankly, I do not think this question is particularly telling against Sanders's thesis, since Sanders himself hardly believes in individual predestination; election refers to God's choice of the nation, and staying in that chosen nation requires cleaving to God and His commandments. But apart from that, as I noted above, such a question is every bit as much a challenge to 1 John 5:16-17 as it is to Sanders. And this reminds us again that we need to think more profoundly about the biblical issues before we will be able to provide a genuinely biblical critique of the "pattern of religion" we happen to discover in the Judaic literature.

The substance of Enns's critique of Sanders, however, seems to be a result of his own misunderstanding: "It seems that the whole thrust of Sanders's argument is that the 'basis of salvation' is only 'membership in the covenant' (i.e., election) and not 'loyalty to it,' which is obedience" (95).

Here, Enns manifests the confusion between adherence and performance which Bauckham and a number of other contributors likewise stumbled over. While Sanders's choice of the word basis may not be particularly helpful, Enns is taking the problem to an entirely different key by understanding loyalty as obedience. Loyalty could simply mean (and clearly does for Sanders) an intention not to forsake God's commandments (cf. Ps. 119:176), without any connotations regarding the "success" of carrying out the commands. There is, after all, forgiveness provided for in the means of atonement. Sanders is clear throughout his work that such adherence is required for "staying in," but it seems strange to think that a requirement of wanting to stay in the covenant (which by the nature of the case means acknowledging and affirming the covenant law) is somehow necessarily veering us toward works-righteousness. If it were, such a charge could be levelled against Scripture itself (compare e.g. John 15:2, 6; Heb. 10:38 etc.).

6. Seifrid argues that "righteousness" language is creational, rather than covenantal; thus "the fundamental assumptions behind covenantal nomism begin to crumble."

Seifrid's own creation vs. covenant dichotomy is what is biblically problematic. Only an assumption that "covenant" only has reference to Jews and Jews alone could make the division which Seifrid does. Seifrid says that "righteousness" fits more with "creational theology" than with covenant. But what is "creational theology"? The "Noahic covenant" is clearly creational. If Seifrid should claim that that is not the "covenant" in view in "covenantal nomism," he is resting his viewpoint on the supposed absolute particularity of God's covenant with Israel. Yet the Abrahamic covenant in Genesis 12 has reference to all the families of the earth.

Seifrid tries to get around this by suggesting, "In Genesis, righteousness is attributed to Noah and to Abraham prior to God's establishing a covenant with either of them" (425). This, however, is not tenable. As N. T. Wright has noted, Paul in Romans 4:11 paraphrases the "sign of the covenant" (Gen. 17:11) as "seal of righteousness" - a paraphrase which by itself overthrows Seifrid's entire argument. Quite apart from that, Seifrid looks too hard for the berith terminology, and wrongly assumes that if "covenant" language is not present, covenant itself is not present. Frankly, a great many biblical scholars do not share those sorts of presuppositions.

7. Even "getting in" is not always tied to grace, because there is a prospect of human sinlessness, which does not require grace to begin with.

The notion of sinlessness to which Carson refers has to do with the patriarchs, who had no need of repentance. This, however, apparently refers specifically to the sin of idolatry; Carson fails to qualify the point as Falk does (14-15). Consequently, it seems that Carson's summary is fundamentally misleading.

8. Carson notes Bockmuehl's developmental thesis. Bockmuehl "argues that its direction is precisely toward the sort of 'works of the law' that the apostle Paul would have opposed."

It should be noted that Bockmuehl was dealing with the materials from the Qumran sect, with whom Paul has no apparent confrontation in Scripture.

Moreover, even Bockmuehl does not seem to recognize the same problematics with Qumran as Carson apparently derives from his essay. Bockmuehl essentially defines the Qumran view, not as a strict replacement theology, but as a remnant theology (390-394). Further, Bockmuehl wrote: "Salvation, on this view, could never be a matter of human merit. The covenanters do not know themselves elect by their works but, on the contrary, their works bear witness to their election" (397). Even more striking, as noted above: "God's righteousness and God's righteous acts, therefore, constitute the salvation and justification of the individual" (399).

What of the "developmental thesis" which Carson mentions? Bockmuehl argues that the redactional history of 1QS indicates a "tightening religious practice in which atonement and forgiveness were increasingly limited to the sect itself" (411). This development highlights the sectarian nature of Qumran. But it is not at all clear that these works became the basis of justification even at Qumran; nor is it clear on what basis Carson can simply assume (apparently) direct relevance to what apostle Paul "would have opposed."

The difficulty here remains one of fundamental unclarity. For surely the point of the NPP is not to develop some sort of "Judaism" that Paul would not have opposed. The point in question is precisely what it was that Paul did oppose. And presumably the exploration of the Second Temple texts is with a view to helping us understand that. But simply asserting that Qumran was going in the direction of something that Paul would have opposed is no aid whatsoever to our understanding.

9. Even if the "constitutional documents" of Judaism were not merit-mongering, in practice ordinary people may not think in the same line.

Speaking of suppositions which cannot be falsified.... That which cannot be demonstrated cannot be disproved. This observation is virtually worthless, and the Christian faith is no less subject to this criticism (even among hard-line sola fide Protestants) than is Judaism.

Carson concludes,

Examination of Sanders's covenantal nomism leads one to the conclusion that the New Testament documents, not least Paul, must not be read against this reconstructed background - or, at least, must not be read exclusively against this background. It is too doctrinaire, too unsupported by the sources themselves, too reductionistic, too monopolistic. The danger is that of the "parallelomania" about which Sandmel warned us, by which texts are domesticated as they are held hostage to the ostensible background called forth by appealing to certain other antecedent texts.

It strikes me that Carson's concluding remarks are very ironic for those who wish to see Justification and Variegated Nomism as the tome that vindicates the now-traditional view of Judaism. For here, Carson admits that well, covenantal nomism should not be the exclusive background against which the New Testament is read. Even he cannot say that employing Sanders's paradigm will get things all wrong. Moreover, Carson takes up Sandmel's warning of domesticating the texts... a warning that had to be issued due to the traditional reading in the first place. We say nothing of how "doctrinaire" it is for a number of the contributors to arrive at conclusions they have not proven in their argumentation, or for Carson to "summarize" the contributions of those who never quite said what he makes them say.

This critique has been searching, but it should not be thought that this volume is a wasted effort. In my judgment, Justification and Variegated Nomism is a significant contribution to the discussion. By carefully studying what the contributors actually demonstrate (as opposed to what Carson claims for them, as well as what some of them claim for themselves), we are provided with necessary qualifications to a too-naive and simplistic adoption of the Sanders' thesis. And we are also given hints about the strength of Sanders's view, precisely where falsification has not proven possible.


In conclusion, I would like to note a few areas which our developing understanding of Second Temple Judaism, and our corresponding approach to the New Testament, must keep in mind.

Judaism is a religion of Torah

There is no getting around the fact that Torah holds pride of place in Second Temple Judaism. This does not yet address precisely how it functions (i.e. as a system of merit, as an extended and precise exposition of the content of covenantal gratitude, etc.). But the fundamental centrality of Torah may not be overlooked.

Getting a true picture of first century Judaism involves recognizing our own presuppositions and concerns. This is as true for liberal "Protestants" (agnostics?) like Sanders as it is for hard-core Lutherans. Philip Alexander aptly warns what sort of risk we run by having our view of Judaism heavily driven by critical Protestants. While the traditional articulation of Judaism's "legalism" was "profoundly mistaken," yet it is likewise to be observed that "the grounds on which the rabbis have sometimes been defended may be hardly less mistaken and may run the risk of denying the absolute centrality of law to rabbinic Judaism and of turning rabbinic Judaism into a rather pale reflection of liberal Protestantism" (279). One should not be blind to the fact that liberals have a habit of projecting their own weak-kneed "theology" onto others (witness the attempt to paint mainstream Islam as peace-loving).

Deines echoes this point in his analysis:

What is at stake in the proper observance of the Law is the nation's standing with God and thus its future. It is the Law's soteriological relevance (which is inseparably linked with its traditional application) that requires its 'precise' interpretation and observance. This Sanders does not adequately take into account. (493)

Also involved are the hints which the New Testament itself provides. Alexander notes that a great deal of "rabbinic jurisprudence" was aimed at "moderating the rigors of biblical law," not at making it heavier (281). This must be balanced with Christ's charges in e.g. Matthew 23:4 (even if the primary fault there is not so much the binding of heavy burdens as the refusal to relieve God's people, Alexander's point would appear to be relativized). Yet it resonates well with what Christ said elsewhere: the Pharisees made the absolute law of honour for father and mother of no effect through the Corban tradition (Mt. 15:3ff.). Similar things may be said about the rabbinic attitude toward divorce, which Jesus considered too loose.

Merit theology probably not pervasive

My provisional conclusion, having read this work by scholars, most of whom were looking for material to counter Sanders's thesis, is that merit theology is probably not very pervasive in the literature. With the exception of IV Ezra, and 2 Baruch (which itself is apparently based on IV Ezra), which are in any case widely recognized as late material, these scholars were very hard-pressed to demonstrate merit theology here. As seen above, many of the cases where the contributor made the suggestion, the passage cited did not bear the conclusion. Not being a Judaic scholar myself, I cannot be triumphalistic on Sanders's behalf, but I will say that if this is what top scholars can unearth in opposition, the traditional characterization of Second Temple Judaism as a system of merit-based works-righteousness looks rather tenuous.

Remnant doctrine

One weakness of Sanders's exposition is that he pays little attention to the "remnant" theme in Second Temple Judaism. (On this, see now Mark Adam Elliott's massive study, The Survivors of Israel, published in 2000.) This was certainly a fundamental dimension at Qumran; whether it is integrated into and pervasive in Pharisaism is a key question. (Carson is not entirely correct in his apparent assumption, however, that remnant theology must be a negation or contradiction of "covenantal nomism." He apparently presupposes that the "covenant" must be national in scope in order for covenantal nomism to function, which appears to me to be a gratuitous assumption [see 540-541].)

Things we don't understand aren't necessarily the point of conflict with Paul

In connection with the broader issues of intersection and collision between Second Temple Judaism and the gospel of the New Testament, many questions remain which only a few of these contributors have acknowledged. Simply bringing up passages in the literature which speak of a "judgment according to works," as if here is irrefutable evidence of "legalism," fails to account for the same theme that appears, not only in the Old Testament (Ecc. 12:14), but in Paul himself (Rom. 2:5ff.) - including explicitly with reference to Christians (see e.g. 1 Cor. 3:12-15; 2 Cor. 5:10-11).

We need a major, careful study of the Biblical texts

Appealing to extrabiblical texts in order to illuminate our understanding of Scripture, I wish to stress, is not an illegitimate exercise, as some claim. It is true that there is always the danger of "parallelomania" (Sandmel again), of trying to "find" our extrabiblical knowledge all over the pages of the New Testament. But that hardly means there is little or no value in extrabiblical knowledge. Our knowledge of the biblical languages themselves is based to a great deal upon extrabiblical knowledge. And it should be obvious (although it appears that it is not) that it is at least a lot more legitimate to read the Bible in the light of extrabiblical first century texts, that actually belong to the Bible's milieu, than to read the Bible in light of extrabiblical late medieval Roman Catholic theology texts. I frankly am puzzled with the radical upside-down methodological standard being employed by many conservative Protestants. It is simply wrong-headed to presuppose that first century Judaism must have looked a lot like 16th century Romanism.

The challenge and task remaining is a systematic, careful review of the biblical data. This review will not presuppose that there really wasn't anything wrong with Judaism, after all. But neither is it safe to presuppose that the problems in Roman Catholic theology were the precise issues in the Christian gospel's confrontation with the Judaizers and unbelieving Israel. Even texts which seem so "obviously" to manifest a merit theology, when you presuppose it, do not necessarily stand up so well upon closer prodding.

I think, for example, of the parable of the publican and the Pharisee (Lk. 18:9-14). The Pharisee prayed, "I thank you that I am not like other men...." There is clearly something smelly and self-righteous in the Pharisees' prayer (which is Luke's own assessment of the Pharisees in verse 9), but that is not automatic evidence that he thinks he is "earning his way to heaven." Note that while Luke says that some pepoithotas eph' heautois that they were righteous - usually translated "trusted in themselves" - we need to be aware of the possibilties with this construction. Cf. 2 Cor. 2:3, where it apparently means no more than having confidence in someone. The point in Luke seems to be a contrast in the Pharisees' minds: they were confident that they were righteous, but they despised others. Again, the issue is not whether the Pharisees were "self-righteous" or not, but what that really means.

The NPP is a step forward, but not the final word

Finally, Sanders's notion of "covenantal nomism" may or may not be an adequate category. But it appears that it is a substantially more accurate characterization of Second Temple literature than the merit-monger view it displaced. While a great deal of questions remain, the New Perspective has changed the course of responsible scholarship, as is clear enough by Justification and Variegated Nomism. It is very telling that even this response aimed at subverting the Sanders viewpoint does not claim to vindicate the traditional characterization of Judaism. And even in its chastened claims, it often proves less than it claims for itself. This indicates, I think, that the advent of the New Perspective on Judaism is a welcome one. But the task of interpreting Scripture remains.

tim gallant creative © 2006