N.T. Wright discusses Romans 4:1

NT Wright wrote the commentary on Romans for the New Interpreter’s Bible. Here is his response, with citations below to Hays, at length (reproduced here by permission of Abingdon Press):
4:1. “What then shall we say? Have we found Abraham to be our forefather according to the flesh?” This is not, of course, what any of the commentaries or translations say, but it has a strong claim to represent Paul’s mind.142 Three reasons stand out. First, it introduces the chapter Paul is writing, as opposed to the one that many think he should have written; in other words, a chapter about the scope and nature of Abraham’s family, rather than a chapter about “justification by faith” as a doctrine about how people become Christians. This, as we shall see, results in the straightforward solution of at least one major exegetical problem. Second, it recognizes that when Paul introduces an argument with ti”v ou\n ejrou’men ti oun eroumen, “what then shall we say?” this phrase is frequently complete in itself, requiring a question mark at once. (There is, of course, no punctuation in the earliest mss.)
Obvious examples are 6:1; 7:7; see also tiv ou\n (ti oun) in 3:9. Third, it avoids at a stroke the awkwardness of sense, and hence of translation, in the usual readings (of which the NRSV and the NIV are typical) in which Abraham is the subject of euJrhke”nai heure kenai, “to have found,” rather than the object as in the reading proposed; since it is not clear what “to have found” could possibly mean in this context, the sense of the verb has to be stretched as in the NRSV (“was gained by”) and NIV (“discovered”), neither of which lead in to what Paul is actually going to say.143 The proposal, then, is that Paul raises in v. 1 a possible conclusion that could be drawn from what has been said so far, in order to argue against it.
At this point, however, I diverge from the meaning Hays gives to his own proposed reading. He suggests that Paul wants to say “Have we Jews normally considered Abraham to be our forefather only according to the flesh?” I suggest, rather, that the whole of Romans 4 hinges on the question, whether 3:21-31 means that we Christians, Jews and Gentiles alike, now find that we are to be members of the fleshly family of Abraham (note how the word “find” suddenly makes perfect sense).144 In other words (Paul is proposing this as a hypothetical question), if in Christ God has been true to the covenant with Abraham, might that not mean, as the Galatians had been led to believe, that members of the Christ-family in fact belong to Abraham’s fleshly family? When we read Romans 4 as the answer to this question, it gains in coherence and force.145
4:2. Verses 2-8 are regularly appealed to by those who still argue that Paul was after all attacking a theology of self-help legalism, in which “righteousness” is earned by moral effort. By themselves these verses might indeed bear that sense. But within the present argument they are much better understood as a further metaphorical expansion, rather than the inner substance, of Paul’s point.
Paul’s main argument is that “works” (i.e., of Torah) were not the reason for Abraham’s justification; and the idea of “working” is then expanded metaphorically in vv. 4-5 into the idea of doing a job for which one earns wages. The critical connection is established with “for” at the start of v. 2 (“in fact” in the NIV is a loose way of making the same point) and depends on the link between “works of Torah” and “Jews only” that Paul had established in the immediately preceding verses. It is ethnic Jews who possess Torah; so if Abraham were the forefather of an ethnic family only, this family would be defined by Torah, and hence defined visibly by Torah’s “works.” Thus (v. 2a) if Abraham was reckoned to be in covenant with God (i.e., was justified) on the basis of works of Torah, he and his family would be able to “boast,” in the way that Paul described in 2:17-20 and then firmly excluded in 3:27-30. Verse 2a, in other words, explains the question of v. 1, as follows: If Abraham’s covenant membership was indeed defined in terms of “works of Torah” (v. 2a), then he and his family would be able to sustain an ethnic boast, and so (v. 1) any Gentiles wishing to belong to this family would then have to consider themselves ethnic Jews–would, in other words, need to become proselytes, with the males among them becoming circumcised.
Paul’s response, to be filled out as usual in what follows, is brusque: “but not before (lit., “toward”) God,” i.e., “not as far as God is concerned.” Verse 2b is thus Paul’s initial reaction to the suggestion, rather than part of the “if . . . then” clause of the earlier part of the verse.146 This now forces Paul into saying what is true “before God,” to cut the ground from under any potential ethnic boast, and to establish once and for all the non-ethnic nature of Abraham’s true family, on the basis of the original covenant itself.

Notes
142. See esp. Richard B. Hays, “ ‘Have We Found Abraham to Be Our Forefather According to the Flesh?’ A Reconsideration of Rom 4:1,” NovT 27 (1985) 76-98. See also Richard B. Hays, Echoes of Scripture in the Letters of Paul (New Haven: Yale University Press, 1989) 54-55. The suggestion goes back at least to J. A. Bain, “Romans iv. 1,” Expository Times 5 (1893—94) 430. It is not clear that subsequent commentators (e.g. Byrne, Fitzmyer, Moo, Bryan) have recognized the force or point of Hays’s proposal; the same is true of S. K. Stowers, A Rereading of Romans: Justice, Jews, and Gentiles (New Haven: Yale University Press, 1994) 234, 242, who unlike them claims to follow it.
143. Abraam is indeclinable in Greek, and hence its grammatical role in the sentence is unclear until defined from elsewhere.
144. R. B. Hays, “Adam, Israel, Christ,” in Pauline Theology Volume III: Romans, ed. D. M. Hay and E. E. Johnson (Minneapolis: Fortress, 1995) 81, has graciously accepted my amendment of his proposal.