Edom Arguing with Edom: The Intra-Protestant Interior
A Review Essay on O. Palmer Robertson, The Israel of God: Yesterday, Today, and Tomorrow (Phillipsburg, NJ: P&R Publishing, 2000)
O. Palmer Robertson’s The Israel of God is the best book ever written inside the wrong debate. For fifty years the argument has run on two rails. The dispensationalist says God has two peoples — Israel and the church — running on separate tracks toward separate destinies. The covenant theologian answers: no, one people, and the church has inherited the name. Robertson takes the second rail with skill and conviction. He refuses the prophecy charts, refuses the two-peoples scheme, and lands where his tradition always lands: the church is the Israel of God, and the story is complete.
But both rails were laid on the same severed ground. Both assume that “Israel” in Paul’s mouth meant what it means in ours — a single ethnic bloc, either kept or replaced. Neither asks what the word meant to the people who bore it.
Here is what they knew, and we forgot. “Israel” and “the Jews” were never the same word. The Jews were the surviving remnant of Judah. Israel was the whole house — twelve tribes, ten of them swallowed by the nations since Assyria, scattered like seed no one could find. Every prayer of the period aches with this. Ezekiel’s two sticks. Jeremiah’s new covenant, cut — read it slowly — with the house of Israel and the house of Judah. The tenth blessing of the synagogue liturgy, still prayed today: gather our exiles. The question of the age was never will God keep Israel or replace her? It was: how does God raise a nation dissolved among the gentiles from the dead?
Paul answers that question. “A hardening has come upon part of Israel until the fullness of the nations comes in — and so all Israel will be saved.” The fullness of the nations. Jacob spoke those words over Ephraim, the lost son whose seed would become melo ha-goyim — the fullness of the gentiles. The nations streaming into Messiah are not Israel’s replacement. They are Israel’s resurrection. The wild branches are how the tree becomes whole.
Robertson never sees this, because his conversation partners are all Christians arguing with Christians. His bibliography has no Second Temple sources, no rabbis, no memory of the ten tribes. Edom arguing with Edom — the estate divided among the heirs while the missing brothers go unmentioned at the table.
So the challenge, for those of us raised on either rail: the Israel of God is not the church renamed. It is not an ethnic bloc awaiting a separate program. It is the whole family being gathered — Judah’s remnant and Israel’s fullness, one olive tree, one root. And the wild branch has exactly one instruction: do not boast against the branches. We didn’t inherit the estate. We came home to it.
What follows is the argument in full.
I. The Specimen and Its Placement
O. Palmer Robertson’s The Israel of God is the most disciplined covenant-theological entry in that quarrel, and for that reason the most instructive. Robertson is no polemical hack. His earlier work on the covenants remains a standard of Reformed biblical theology, and this book displays the same virtues: exegetical patience, structural clarity, a principled refusal of apocalyptic speculation about the modern Middle East. He dismantles the dispensational two-peoples scheme with genuine skill. And yet the book fails — not at the level of argument but at the level of question. Robertson answers, definitively, a question the first century never asked, and his very success in answering it has helped seal an entire tradition inside the wrong debate.
The purpose of this essay is to document that failure as a specimen: to show that the intra-Protestant argument over Israel is not, in the strict sense, a debate at all, because it lacks a shared court of appeal; that the historical evidence capable of adjudicating it has been excluded by both parties in advance; and that rapprochement between the two positions is neither possible nor desirable, since both are wrong in the same direction.
II. The Architecture of the Book
Robertson organizes the work around five topics — land, people, worship, lifestyle, and the coming of the kingdom — with a sixth chapter devoted to Romans 11. The controlling method is announced early and applied consistently: the Old Testament is to be read through the fulfillment declared in the New, so that every promise made to Israel finds its terminus in the Messiah and, through him, in the church. The land promise expands to the cosmos on the warrant of Romans 4:13. The people of God are redefined around faith in the Messiah, so that “Israel” becomes, in the new covenant era, a designation the church rightly bears. Worship is deterritorialized; the lifestyle of the people of God is universalized; and the kingdom comes not through a restored national program but through the present reign of the risen Messiah.
The Romans 11 chapter, anticipated by Robertson’s earlier essay on the same text, executes the architecture’s most consequential move.¹ Against the broad exegetical consensus that Paul anticipates a future turning of Israel, Robertson reads the houtōs of 11:26 as purely modal — “in this manner all Israel will be saved” — and takes “all Israel” as the aggregate of elect Jews gathered piecemeal alongside the gentiles throughout the present age. Nothing remains to be awaited; the text describes a process, not a promise. The olive tree of 11:17–24 becomes an illustration of covenantal continuity and individual engrafting, and the declaration of 11:28–29 — beloved on account of the fathers, for the gifts and the calling of God are irrevocable — is absorbed into the general faithfulness of God to his one elect people.
The internal coherence of this construction should be conceded at once. Granted its premises, the book is nearly seamless. The question is whether the premises could survive contact with the semantic world in which Paul’s sentences were first intelligible.
III. The Semantic Evidence: Israel and Judah in Second Temple Usage
The premise that cannot survive is the assumption, shared by Robertson and his dispensational opponents alike, that “Israel” in the first century named a single ethnic-national bloc — the Jewish people — whose covenantal standing was either perpetual (the dispensational horn) or transferred (the covenantal horn). Jason Staples has now demonstrated at length what a careful reading of the sources always suggested: “Israel” and “Jews” (Ioudaioi) were not coterminous terms in the Second Temple period.² “Judah” and its derivatives named the surviving remnant polity descended from the southern kingdom; “Israel” named the twelve-tribe covenantal totality, the greater part of which had been dissolved among the nations since the Assyrian deportations and never returned. To speak of Israel in the first century was therefore to speak, unavoidably, of a body that did not presently exist in gathered form — a nation whose restoration was the object of prayer, not the description of a constituency.
The restoration hope saturates the literature. Ezekiel’s two sticks, Judah and Joseph, joined into one in the prophet’s hand (Ezek 37:15–28); Jeremiah’s new covenant, cut explicitly “with the house of Israel and with the house of Judah” (Jer 31:31), a double formula that Christian exegesis has habitually read past; Tobit’s confidence that the scattered will be gathered from among the nations; Ben Sira’s plea that God gather all the tribes of Jacob and give them their inheritance as at the beginning (Sir 36:13); the Psalms of Solomon’s expectation that the Messiah will assemble a holy people and distribute them upon the land according to their tribes (Pss. Sol. 17:26–28); the tenth benediction of the Amidah, prayed daily to this hour, blessing the God who gathers the dispersed of his people Israel; the Qumran community’s self-organization as a proleptic twelve-tribe camp awaiting the reconstitution of the whole.³ The eschatological question of the age was not whether God would keep or replace a presently existing Israel. It was how the God of Israel would raise from the dead a nation swallowed by the gentiles — how seed sown among the nations and grown indistinguishable from them could ever be recovered.
Into this semantic field Paul speaks, and within it his most contested sentences become lucid. The “fullness of the nations” (plērōma tōn ethnōn) whose incoming precedes the salvation of all Israel (Rom 11:25–26) echoes Jacob’s blessing over Ephraim, the lost son whose seed would become melo ha-goyim, the fullness of the nations (Gen 48:19) — a connection Staples presses to its full weight.⁴ The gentile ingathering is not Israel’s replacement but the very mechanism of Israel’s restoration: the wild branches are how the tree becomes whole. And the phrase “all Israel” (kol yisrael) belongs to a recognizable technical register. The Mishnah’s most famous sentence — “All Israel has a share in the world to come” (m. Sanh. 10:1) — deploys it for the covenantal totality, and the tractate’s continuation immediately debates whether the ten tribes are included in that totality (m. Sanh. 10:3), which establishes beyond reasonable doubt the semantic field in which the phrase lived.⁵ The rabbis arguing over the ten tribes’ share in the age to come and Paul announcing the salvation of all Israel through the incoming fullness of the nations are working the same question from opposite sides of a single inheritance.
Robertson engages none of this. His bibliography is confessional and canonical; the Pseudepigrapha, the liturgy, Qumran, and the rabbinic corpus are functionally absent. The omission is not incidental but structural: the book’s either/or — Israel kept as ethnic nation or Israel redefined as church — is only available to a reader for whom the third referent, the twelve-tribe qahal awaiting reunification, has dropped out of memory. Robertson inherited the amnesia and built a system on it.
IV. The Two Load-Bearing Texts
The construction rests principally on two texts, and in both cases the restoration reading is at least as natural grammatically and considerably more coherent historically.
Galatians 6:16 — “peace and mercy upon all who walk by this rule, and upon the Israel of God” — supplies Robertson his title and his thesis. He takes the final kai as epexegetical: the Israel of God simply is the community of those who walk by the rule, the church under its inherited name. But the epexegetical reading was resisted by careful exegetes long before the Paul-within-Judaism school made resistance fashionable. Burton’s classic commentary took the kai as genuinely conjunctive; Richardson demonstrated that no Christian writer applies “Israel” to the church before Justin Martyr in the mid-second century — a silence devastating to the claim that Paul had already made the transfer; Eastman has argued the connective force afresh on contextual grounds.⁶ And the context is decisive. Galatians is the letter in which Paul refuses, with maximal vehemence, to let gentile believers become proselytes to Judah. That such a letter should close by quietly awarding gentiles Israel’s own name is exegetically perverse. On the restoration reading the benediction is doubled and coherent: peace upon those who walk by this canon — the gentile assemblies holding the line against proselytization — and mercy upon the Israel of God, the whole covenant house now being reconstituted, natural branches and wild together. The verse is not a deed of transfer; it is a blessing over a family in the process of reunion.
II. The Architecture of the Book
Robertson organizes the work around five topics — land, people, worship, lifestyle, and the coming of the kingdom — with a sixth chapter devoted to Romans 11. The controlling method is announced early and applied consistently: the Old Testament is to be read through the fulfillment declared in the New, so that every promise made to Israel finds its terminus in the Messiah and, through him, in the church. The land promise expands to the cosmos on the warrant of Romans 4:13. The people of God are redefined around faith in the Messiah, so that “Israel” becomes, in the new covenant era, a designation the church rightly bears. Worship is deterritorialized; the lifestyle of the people of God is universalized; and the kingdom comes not through a restored national program but through the present reign of the risen Messiah.
The Romans 11 chapter, anticipated by Robertson’s earlier essay on the same text, executes the architecture’s most consequential move.¹ Against the broad exegetical consensus that Paul anticipates a future turning of Israel, Robertson reads the houtōs of 11:26 as purely modal — “in this manner all Israel will be saved” — and takes “all Israel” as the aggregate of elect Jews gathered piecemeal alongside the gentiles throughout the present age. Nothing remains to be awaited; the text describes a process, not a promise. The olive tree of 11:17–24 becomes an illustration of covenantal continuity and individual engrafting, and the declaration of 11:28–29 — beloved on account of the fathers, for the gifts and the calling of God are irrevocable — is absorbed into the general faithfulness of God to his one elect people.
The internal coherence of this construction should be conceded at once. Granted its premises, the book is nearly seamless. The question is whether the premises could survive contact with the semantic world in which Paul’s sentences were first intelligible.
III. The Semantic Evidence: Israel and Judah in Second Temple Usage
The premise that cannot survive is the assumption, shared by Robertson and his dispensational opponents alike, that “Israel” in the first century named a single ethnic-national bloc — the Jewish people — whose covenantal standing was either perpetual (the dispensational horn) or transferred (the covenantal horn). Jason Staples has now demonstrated at length what a careful reading of the sources always suggested: “Israel” and “Jews” (Ioudaioi) were not coterminous terms in the Second Temple period.² “Judah” and its derivatives named the surviving remnant polity descended from the southern kingdom; “Israel” named the twelve-tribe covenantal totality, the greater part of which had been dissolved among the nations since the Assyrian deportations and never returned. To speak of Israel in the first century was therefore to speak, unavoidably, of a body that did not presently exist in gathered form — a nation whose restoration was the object of prayer, not the description of a constituency.
The restoration hope saturates the literature. Ezekiel’s two sticks, Judah and Joseph, joined into one in the prophet’s hand (Ezek 37:15–28); Jeremiah’s new covenant, cut explicitly “with the house of Israel and with the house of Judah” (Jer 31:31), a double formula that Christian exegesis has habitually read past; Tobit’s confidence that the scattered will be gathered from among the nations; Ben Sira’s plea that God gather all the tribes of Jacob and give them their inheritance as at the beginning (Sir 36:13); the Psalms of Solomon’s expectation that the Messiah will assemble a holy people and distribute them upon the land according to their tribes (Pss. Sol. 17:26–28); the tenth benediction of the Amidah, prayed daily to this hour, blessing the God who gathers the dispersed of his people Israel; the Qumran community’s self-organization as a proleptic twelve-tribe camp awaiting the reconstitution of the whole.³ The eschatological question of the age was not whether God would keep or replace a presently existing Israel. It was how the God of Israel would raise from the dead a nation swallowed by the gentiles — how seed sown among the nations and grown indistinguishable from them could ever be recovered.
Into this semantic field Paul speaks, and within it his most contested sentences become lucid. The “fullness of the nations” (plērōma tōn ethnōn) whose incoming precedes the salvation of all Israel (Rom 11:25–26) echoes Jacob’s blessing over Ephraim, the lost son whose seed would become melo ha-goyim, the fullness of the nations (Gen 48:19) — a connection Staples presses to its full weight.⁴ The gentile ingathering is not Israel’s replacement but the very mechanism of Israel’s restoration: the wild branches are how the tree becomes whole. And the phrase “all Israel” (kol yisrael) belongs to a recognizable technical register. The Mishnah’s most famous sentence — “All Israel has a share in the world to come” (m. Sanh. 10:1) — deploys it for the covenantal totality, and the tractate’s continuation immediately debates whether the ten tribes are included in that totality (m. Sanh. 10:3), which establishes beyond reasonable doubt the semantic field in which the phrase lived.⁵ The rabbis arguing over the ten tribes’ share in the age to come and Paul announcing the salvation of all Israel through the incoming fullness of the nations are working the same question from opposite sides of a single inheritance.
Robertson engages none of this. His bibliography is confessional and canonical; the Pseudepigrapha, the liturgy, Qumran, and the rabbinic corpus are functionally absent. The omission is not incidental but structural: the book’s either/or — Israel kept as ethnic nation or Israel redefined as church — is only available to a reader for whom the third referent, the twelve-tribe qahal awaiting reunification, has dropped out of memory. Robertson inherited the amnesia and built a system on it.
IV. The Two Load-Bearing Texts
The construction rests principally on two texts, and in both cases the restoration reading is at least as natural grammatically and considerably more coherent historically.
Galatians 6:16 — “peace and mercy upon all who walk by this rule, and upon the Israel of God” — supplies Robertson his title and his thesis. He takes the final kai as epexegetical: the Israel of God simply is the community of those who walk by the rule, the church under its inherited name. But the epexegetical reading was resisted by careful exegetes long before the Paul-within-Judaism school made resistance fashionable. Burton’s classic commentary took the kai as genuinely conjunctive; Richardson demonstrated that no Christian writer applies “Israel” to the church before Justin Martyr in the mid-second century — a silence devastating to the claim that Paul had already made the transfer; Eastman has argued the connective force afresh on contextual grounds.⁶ And the context is decisive. Galatians is the letter in which Paul refuses, with maximal vehemence, to let gentile believers become proselytes to Judah. That such a letter should close by quietly awarding gentiles Israel’s own name is exegetically perverse. On the restoration reading the benediction is doubled and coherent: peace upon those who walk by this canon — the gentile assemblies holding the line against proselytization — and mercy upon the Israel of God, the whole covenant house now being reconstituted, natural branches and wild together. The verse is not a deed of transfer; it is a blessing over a family in the process of reunion.
Romans 11:26 receives parallel treatment. Robertson’s modal reading of houtōs is grammatically possible but proves nothing by itself, since the manner in view — hardening in part, fullness of the nations incoming, and so all Israel saved — is precisely the sequence of restoration eschatology. His real work is done by the redefinition of “all Israel” as the trickle of elect Jews across the age, a reading that renders Paul’s climactic mystērion — his revealed truth, in the proper Pauline sense of disclosure rather than riddle — a solemn announcement of the obvious. That Jews were being saved alongside gentiles was empirically evident in every Pauline assembly; no unveiling was required to assert it. The mystery discloses something structurally unexpected: that the gentile mission itself is the instrument of Israel’s regathering, Ephraim’s fullness returning through the nations into which he was sown. Against this, Robertson’s reading empties the chapter of its own crescendo and leaves 11:28–29 — beloved on account of the fathers; the gifts and calling irrevocable — without a referent his system can name, since the only Israel his system retains has already been absorbed into the church that is supposed to be doing the loving.
V. Land, Typology, and the Erasure of Particularity
The land chapter exhibits the same pattern in a different key, and here Robertson’s partial rightness must be honestly weighed. Romans 4:13 does declare Abraham heir of the kosmos; the promise does expand; and Robertson’s refusal to read the modern state of Israel off prophecy charts is sound instinct, the same instinct that makes his anti-dispensational polemic effective. But expansion from a particular center is not abolition of the center. Isaiah’s eschatology moves the nations toward Zion, not Zion into abstraction; Zechariah sees ten men from the nations grasping the garment of a Jew because they have heard that God is with him (Zech 8:23). The biblical universal is always mediated through the particular — this is the architecture Benamozegh grasped more clearly than any modern Christian theologian: the nations blessed alongside Israel’s abiding vocation, a dignified universal path that does not require the dissolution of the particular covenant to make room for it.⁷ Robertson’s typology, lacking that two-track architecture, can conceive universalization only as erasure. The cosmos swallows the land, the church swallows the name, and the result is presented as fulfillment. It is, rather, the severance completing itself with a clear conscience — the cutting of the shoots performed as though it were pruning.
VI. No Shared Court of Appeal: The Structure of the Impasse
It remains to name why this quarrel, of which Robertson’s book is the ablest covenant-theological monument, has produced sixty years of polemics and no movement. A genuine debate requires a shared court of appeal: some body of evidence both parties acknowledge as capable, in principle, of overturning their position. The dispensational–covenant argument has none. Each side’s hermeneutic is load-bearing for its entire system. The dispensationalist cannot permit “Israel” to mean anything other than the perpetual ethnic nation without the architecture of distinct programs collapsing; the covenant theologian cannot permit Romans 11 to carry restoration eschatology without conceding that the confessional tradition read Paul through a frame he never occupied. Neither side can afford to lose, and therefore neither side is listening. What presents itself as debate is two closed systems performing self-maintenance in each other’s presence.
This explains the otherwise puzzling non-engagement with the historical sources — an avoidance too consistent to be laziness. The Second Temple and rabbinic evidence is dangerous to both parties for the same reason: it enters from outside the loop, unauthorized by either confessional tradition, and cannot be metabolized by either without structural renovation. It is institutionally easier to keep m. Sanhedrin 10 out of the bibliography than to explain why the rabbis debate the ten tribes’ inclusion using the very phrase Paul deploys at his climax. The pattern replicates in the succeeding generation with documentary precision: Schreiner’s Romans converses with Käsemann, Cranfield, Moo, and Dunn — the Protestant commentary succession — while Nanos, whose Mystery of Romans placed the restoration question on the table three decades ago, and the scholarship that followed him, are footnoted as curiosities or passed over entirely.⁸ McDermott’s 2013 survey glimpsed the divide and even named the Hebraic horizon visible from the Global South, then prescribed as remedy a deeper submission to the patristic-medieval consensus — the physician recommending the disease.⁹ Polemics, in the end, is what remains when a tradition has decided in advance which witnesses may testify.
From this the conclusion about rapprochement follows with some rigor: it is not possible on the current terms, and its impossibility is a mercy. The two rails cannot be reconciled with each other because they are wrong in the same direction — both are answers to a question the first century never asked, both conducted entirely within the severed frame, both innocent of the whole house of Israel. A treaty between them would be Edom consolidating the estate more efficiently: the heirs settling the inheritance while the missing brothers remain unmentioned at the table. What is required is not a treaty but the reopening of the inheritance question itself, and that reopening, where it is occurring, is occurring at the margins the guild does not police — in the Paul-within-Judaism scholarship, in Global South churches whose Hebraic instincts have not yet been trained out of them, in pastors who stumble upon the two-house material and find their categories dissolving without their permission. Grafting is not negotiated between wild branches. It happens at the root.
Nor is the impasse victimless, and the dossier should record the pastoral cost. Two generations of laypeople have been catechized into believing their only options are a prophecy-chart Zionism perpetually embarrassed by history or a church that has quietly absorbed Israel’s name and closed the story. When both collapse under scrutiny — and both do — the casualties do not upgrade their theology; they abandon the narrative altogether. The recovery of the story we left behind is not an academic adjustment. It is the difference between an estate under litigation and a family table with places still set.
VII. Conclusion: The Question Reopened
Robertson’s book deserves the tribute of serious refutation rather than dismissal, and the tribute is this: he is right against his opponents and wrong against the sources. The dispensational bifurcation he demolishes deserved demolition. The Israel-Christology he offers — the Messiah recapitulating Israel’s story, called out of Egypt as the true son — contains real truth, but truth of a representative rather than eliminative kind: in the logic of corporate solidarity the king embodies the kingdom in order to restore it, as David embodies Israel without abolishing the tribes. The king is not the qahal. Robertson converts recapitulation into replacement and hands the estate to the church, and the entire transaction is conducted in a courtroom from which the decisive witnesses — Ezekiel’s two sticks, Jeremiah’s double formula, Jacob’s word over Ephraim, the Mishnah’s ten tribes, the daily prayer for the ingathering — were excluded before the proceedings began.
Once those witnesses are readmitted, the either/or that structures The Israel of God simply evaporates. The Israel of God is neither the church renamed nor an ethnic bloc awaiting a separate program. It is the whole house being rebuilt: Judah’s remnant and Israel’s fullness among the nations, one olive tree, one root, and the wild branch under a single standing instruction it has spent most of its history disobeying — do not boast against the branches.
We didn’t inherit the estate. We came home to it.
End Notes
1. O. Palmer Robertson, “Is There a Distinctive Future for Ethnic Israel in Romans 11?” in Perspectives on Evangelical Theology, ed. Kenneth S. Kantzer and Stanley N. Gundry (Grand Rapids: Baker, 1979), 209–27; developed in The Israel of God: Yesterday, Today, and Tomorrow (Phillipsburg, NJ: P&R Publishing, 2000), chap. 6.
2. Jason A. Staples, The Idea of Israel in Second Temple Judaism: A New Theory of People, Exile, and Israelite Identity (Cambridge: Cambridge University Press, 2021); idem, Paul and the Resurrection of Israel: Jews, Former Gentiles, Israelites (Cambridge: Cambridge University Press, 2024).
3. Ezek 37:15–28; Jer 31:31; Tob 13:5–6; Sir 36:13; Pss. Sol. 17:26–28; the tenth benediction of the Amidah; on Qumran’s twelve-tribe self-understanding, see e.g. 1QS 8:1–2; 1QM 2:2–3.
4. Staples, Paul and the Resurrection of Israel, on Rom 11:25–26 with Gen 48:19; the LXX’s plēthos ethnōn and Paul’s plērōma tōn ethnōn rendering melo ha-goyim.
5. m. Sanhedrin 10:1, 3.
6. Ernest de Witt Burton, A Critical and Exegetical Commentary on the Epistle to the Galatians, ICC (Edinburgh: T&T Clark, 1921), 357–58; Peter Richardson, Israel in the Apostolic Church (Cambridge: Cambridge University Press, 1969), 74–84; Susan Grove Eastman, “Israel and the Mercy of God: A Re-reading of Galatians 6.16 and Romans 9–11,” New Testament Studies 56, no. 3 (2010): 367–95.
7. Elijah Benamozegh, Israel and Humanity, trans. Maxwell Luria (New York: Paulist Press, 1995).
8. Mark D. Nanos, The Mystery of Romans: The Jewish Context of Paul’s Letter (Minneapolis: Fortress, 1996); cf. Paula Fredriksen, Paul: The Pagans’ Apostle (New Haven: Yale University Press, 2017); Matthew Thiessen, Paul and the Gentile Problem (Oxford: Oxford University Press, 2016).
9. Gerald R. McDermott, “The Emerging Divide in Evangelical Theology,” Journal of the Evangelical Theological Society 56, no. 2 (2013): 355–77.
