Deep Bible
Back to Questions
Deep dive · 13 min read
Moral problem

The Amalek command: divine genocide?

1 Samuel 15:3 orders Saul to kill 'both man and woman, infant and suckling, ox and sheep, camel and ass.' The verse names the categories. Four positions have circulated since the patristic period. The chapter also has its own seams (the Amalekites who survive in 1 Sam 30, the Agagite who shows up in Esther) that every position has to handle.

What's at stake

Samuel speaks the command at 1 Samuel 15:3. 'Now go and smite Amalek, and utterly destroy all that they have, and spare them not; but slay both man and woman, infant and suckling, ox and sheep, camel and ass.' The chapter offers no internal softening. Saul wins the battle, spares Agag the king and the best of the livestock, and is rejected by the LORD for the partial obedience. The chapter's verdict is on Saul, not on the order. Four families of reading have been on the table since at least Origen, and each one has to handle three things the chapter itself preserves: the infant clause, the Amalekites who appear alive ten chapters later in 1 Samuel 30, and the Agagite line in the book of Esther.

What the chapter says, and what the chapter does not say

Samuel arrives at Saul with a charge anchored in older texts. The reason given at 15:2 is what Amalek did to Israel coming up out of Egypt, the attack at Rephidim recorded in Exodus 17:8-16. Deuteronomy 25:17-19 takes that earlier event and turns it into a standing obligation: when the LORD gives Israel rest from her enemies, the memory of Amalek is to be blotted out. 1 Samuel 15 frames the campaign as the execution of that obligation, centuries delayed.

Saul musters 200,000 foot soldiers and 10,000 from Judah. He warns the Kenites to leave the area, since they had shown kindness to Israel during the exodus. Then he strikes Amalek from Havilah to Shur. He brings back Agag alive and the best of the sheep, the cattle, the fatlings, the lambs, and everything good. Everything despised and worthless he destroys. Samuel hears about it, comes out to confront Saul, and renders the verdict: 'to obey is better than sacrifice' (15:22). Samuel himself then hacks Agag in pieces before the LORD at Gilgal.

What the chapter does not say is also part of the data. It does not give a casualty count. It does not say the Amalekites were extinct after the campaign. It does not name any survivors except Agag, and it does not address the relationship of this campaign to the much later Amalekite raid on Ziklag in 1 Samuel 30. Whatever the reader makes of the order, the canon then preserves Amalekite raiders ten chapters later, still operating in the Negev, still capable of mounting an attack. Esther 3:1 then names Haman 'the Agagite' more than half a millennium after Agag's death at Gilgal. The chapter's stated outcome and the canon's later record do not match cleanly, and each position has to say why.

The four positions

How the chapter has been read

Four families of reading. Each takes the order at 15:3 differently, and each has to make sense of the infant clause, the surviving Amalekites in 1 Samuel 30, and the Agagite line in Esther.

The command at 15:3 is a real divine order, and the warrant is the centuries of Amalekite hostility codified in Exodus 17 and Deuteronomy 25. The execution is delayed (from the Sinai oath at Exod 17:16 to Saul's reign) but the sentence is already on record. Saul's incomplete obedience is the chapter's point, not its problem.
Held by
  • Origen, Homilies on Joshua 11 and 12 (c. 240s CE) (typological frame, literal command granted)
  • Augustine, Quaestiones in Heptateuchum VI (c. 419 CE) on herem and divine sovereignty over life
  • Thomas Aquinas, Summa Theologiae I-II q. 105 a. 3 ad 2 (c. 1270s)
  • John Calvin, Commentary on 1 Samuel (1563), at 1 Sam 15
  • Matthew Henry, Exposition (1710), at 1 Sam 15
  • C. F. Keil and Franz Delitzsch, Commentary on the Books of Samuel (1875)
  • Gordon J. Wenham, The Book of Leviticus (NICOT; Eerdmans, 1979)
  • Eugene H. Merrill, in Show Them No Mercy (Zondervan, 2003)
Evidence
  • • Exodus 17:14-16 records the LORD's oath after the Rephidim attack: 'I will utterly blot out the remembrance of Amalek from under heaven.' The Sinai oath is the warrant the chapter says it is executing
  • • Deuteronomy 25:17-19 reactivates the obligation: when Israel has rest from her enemies, the memory of Amalek is to be blotted out. Saul's reign is the first sustained period of post-conquest rest, which the reading takes as the trigger
  • • 15:2 explicitly cites the older history: 'I will punish what Amalek did to Israel in opposing them on the way when they came up out of Egypt.' The chapter understands itself as the long-delayed sentence rather than as a fresh judgment
  • • The chapter records that Saul warns the Kenites to leave (15:6), because the Kenites had shown kindness to Israel during the exodus. The narrative distinguishes between groups based on historical relationship to Israel. The Amalekites are not being struck for ethnic reasons but for what their ancestors did and what their nation continued to do
  • • 1 Samuel 14:48 records Saul's earlier victory over Amalek before chapter 15. The Amalekites were ongoing raiders against Israel through the Saul era, not a distant memory
  • • Within the canonical frame, herem is restricted to specific covenantal moments (the conquest generation, the Amalekite case, the moments at Jericho and Ai). It is never extended past the early monarchy and is treated by later prophets as belonging to a closed era
Challenges
  • • The judicial-warrant argument does not by itself answer the infant clause. Infants and sucklings have no responsibility for what Amalek did at Rephidim, and the delayed-execution frame does not explain why ritual non-involvement is not a category here
  • • The chapter records the order as Samuel's relay of the LORD's command. Modern critics question the prophetic mediation: how does the narrative establish that the relayed order matches the original divine intent. The reading depends on accepting prophetic mediation as reliable
  • • If herem is a closed category restricted to a particular era, the question of why the LORD instituted it at all remains open. The judicial frame names the warrant but does not address the moral logic
  • • The reading has to handle 1 Samuel 30 (Amalekite raiders surviving) and Esther (Haman 'the Agagite'). On the literal-execution reading, the line should have ended at Gilgal; the canon preserves it continuing

How each position handles the three problems

The chapter does not stand alone. The infant clause is inside 15:3. The Amalekite survivors are five chapters and ten chapters later, in 1 Samuel 27 and 1 Samuel 30. Haman the Agagite is centuries later in Esther 3:1 and 9:24. Each position has to make sense of all three. The columns below set out how each reading handles them.

Three problems, four readings

The infant clause, the Amalekite survivors in 1 Samuel 30, and Haman the Agagite. Each position has to handle all three; the differences are where the readings come apart.

The infant clause (15:3)
Delayed-execution reading
The infants are inside the judicial scope as members of the corporate national entity under judgment. The category does not depend on individual ritual involvement; it depends on covenantal status. The reading grants the moral difficulty and locates it within divine sovereignty over life.
ANE-hyperbole reading
The 'man, woman, infant, suckling, ox, sheep, camel, ass' formula is a totality marker, the conventional shape of a holy-war oath. The list is rhetorical scope, not a kill-list. The reading lowers the literal body count by reading the language against the ANE comparators.
Accommodation reading
The infant clause reflects the moral horizon of the early-monarchic community, not the divine intent. The cross is the criterion. The clause sits in the canon as witness to the community's understanding rather than as a divine command.
Unresolved reading
The infant clause is the residue no smoothing strategy reaches. The chapter says what it says, and the modern reader has to hold the difficulty without resolution.
Survivors in 1 Sam 30
Delayed-execution reading
1 Samuel 30 records a separate Amalekite group not encompassed in Saul's campaign, or the campaign succeeded against the populations it reached without exterminating the wider Amalekite presence in the Negev. The original order was carried out as far as Saul's reach extended.
ANE-hyperbole reading
Direct in-canon evidence that the rhetoric of chapter 15 was not literal extermination. The four hundred camel-mounted escapees of 30:17 are exactly what the rhetorical reading predicts: the campaign was real, the language was conventional, and Amalek continued.
Accommodation reading
The chapter's preservation of Amalekite survivors confirms that the order at 15:3 was not the divine intent. The canon itself records the gap between what the text says happened and what actually happened in the community's later life.
Unresolved reading
The survivors complicate the chapter's own framing. The narrator presses Saul against the literal arithmetic, then preserves Amalekites raiding the Negev fifteen chapters later. The internal tension is part of what the reader has to weigh.
Haman the Agagite (Esther 3:1)
Delayed-execution reading
Haman is a literary descendant of Agag, marking the unfinished business of the chapter. The book of Esther is the canonical sequel: Saul failed to complete the herem, Mordecai (a Benjaminite like Saul) and the Jewish community finally finish what Saul did not. The Agagite line shows the canon treating Amalek as continuous.
ANE-hyperbole reading
Haman the Agagite is further evidence that Amalek continued for centuries after Gilgal. The canon is comfortable preserving the line because the chapter never claimed actual extermination. The literary echo is doing thematic work within the rhetorical convention.
Accommodation reading
The Esther echo is part of the canon's complex witness to Israel's continuing experience of opposition. The Agagite line is theological language for inveterate enmity rather than literal genealogy from Agag.
Unresolved reading
Esther's use of 'Agagite' shows the canon itself developing the Amalek category beyond the literal genealogy. By Esther, Amalek has become a moral and literary category for the implacable enemy. The development is part of how Jewish tradition has handled the chapter's residue.

The Exodus 17 backstory

1 Samuel 15:2 grounds the chapter in something Israel did long ago. The reference is Exodus 17:8-16. Israel has crossed the sea and is in the wilderness at Rephidim. Water has just come from the rock. Amalek attacks. Joshua leads the battle in the valley; Moses lifts his hands on the hilltop; Israel prevails. The chapter closes with the LORD swearing 'war with Amalek from generation to generation' (Exod 17:16) and Moses writing the oath in a book.

Deuteronomy 25:17-19 reactivates the same memory in legal form. The text is set as Moses's final address to the next generation. 'Remember what Amalek did unto thee by the way, when ye were come forth out of Egypt; how he met thee by the way, and smote the hindmost of thee, even all that were feeble behind thee, when thou wast faint and weary; and he feared not God. Therefore it shall be, when the LORD thy God hath given thee rest from all thine enemies round about, in the land which the LORD thy God giveth thee to inherit it for an inheritance to possess it, that thou shalt blot out the remembrance of Amalek from under heaven; thou shalt not forget it.' The framing is precise: the obligation activates when Israel has rest from her enemies.

1 Samuel 14:47-48 records Saul as the king who delivered Israel from her enemies (Moab, Ammon, Edom, the kings of Zobah, the Philistines, the Amalekites). The chapter sets up 1 Samuel 15 as the moment the Deuteronomy 25 condition is met. Whatever the modern reader makes of the order, the chapter is reading itself inside a canon-internal trajectory that the prior books established.

Where the arguments actually disagree

Stepping back from the four positions, the disagreements cluster around three questions the chapter itself does not settle. First, what kind of statement is 15:3. The delayed-execution reading treats it as a judicial sentence with a long backstory. The hyperbole reading treats it as a rhetorical totality marker shaped by ANE convention. The accommodation reading treats it as the community's understanding of what divine warfare looked like. The unresolved reading treats it as a hard limit the canon preserves without softening.

Second, what is the relationship between 15:3 and the canon's later record. All four positions agree that Amalekites appear alive in 1 Samuel 27, 1 Samuel 30, 1 Chronicles 4:42-43, and (as 'the Agagite') in Esther. The delayed-execution and hyperbole readings handle this directly. The accommodation and unresolved readings treat the canonical preservation of the line as part of what makes the chapter hard.

Third, what is the chapter's relationship to present-day ethics. All four positions agree the chapter does not authorize any current military action. The category that allowed the order (a singular covenantal context, a particular historical grievance, an ANE convention, a particular priestly-prophetic frame) is not a category any modern community sits inside. The disagreement is about why the chapter does not authorize present-day action, not whether it does.

Reading the chapter with the question open

1 Samuel 15 is one of the chapters readers come back to. The four positions above do not collapse into each other, and none of them dissolves the discomfort the chapter produces. What they do is name what each reader is trading off. The delayed-execution reading preserves the chapter's plain force and accepts the cost of a divine command that runs against later moral instincts. The hyperbole reading preserves divine character at the cost of changing what the text was reporting. The accommodation reading preserves a Christ-shaped account of God at the cost of locating the violence inside the community's perception rather than the divine will. The unresolved reading preserves the chapter's hard surface at the cost of leaving the canonical authority question open.

Most readers who stay with the chapter end up holding a position with borrowed pieces. Few delayed-execution defenders deny that the chapter draws on ANE conventions. Few accommodation readers reject the historical grievance the chapter cites. The labels above are the families. They are not airtight cells. What the chapter requires is that the reader pick a position knowing what each one costs.

Sources

Primary sources
  • 1 Samuel 15:1-35 (Hebrew Masoretic Text; Leningrad Codex B19a; KJV cited above for familiarity)
  • Exodus 17:8-16 (Amalek at Rephidim and the LORD's oath)
  • Deuteronomy 25:17-19 (the standing Amalek obligation)
  • 1 Samuel 14:47-48 (Saul's earlier campaign against Amalek)
  • 1 Samuel 27:8 (David raiding Amalekites during his Philistine period)
  • 1 Samuel 30:1-20 (Amalekite raid on Ziklag; four hundred camel-mounted escapees)
  • 1 Chronicles 4:42-43 (Simeonites striking the Amalekite remnant under Hezekiah)
  • Esther 2:5; 3:1; 9:24 (Mordecai's Benjaminite genealogy; Haman the Agagite)
  • Mesha Stele (KAI 181, c. 840 BCE; British Museum) (ḥrm used for towns continuing in occupation)
  • Origen, Homilies on Joshua (c. 240s CE), Sources Chrétiennes 71
  • Origen, On First Principles IV.2.9 (c. 220s CE), SC 268
  • John Chrysostom, Homilies on Genesis 17.1 (c. 380s CE), PG 53
  • Augustine, Quaestiones in Heptateuchum VI (c. 419 CE), CCSL 33
  • Thomas Aquinas, Summa Theologiae I-II q. 105 a. 3 ad 2 (c. 1270s)
  • Babylonian Talmud, Yoma 22b (Saul's reasoning and the rabbinic response)
  • Maimonides, Mishneh Torah, Hilchot Melakhim 5:4-5; 6:1-5 (Amalek as category and the procedural constraints)
  • Pseudo-Philo, Liber Antiquitatum Biblicarum 58 (Saul and Agag tradition)
Modern scholarship cited
  • John Calvin, Commentary on 1 Samuel (1563)
  • John Calvin, Institutes of the Christian Religion (1559), I.13
  • Matthew Henry, Exposition of the Old and New Testament (1710)
  • C. F. Keil and Franz Delitzsch, Commentary on the Books of Samuel (T&T Clark, 1875)
  • C. S. Lewis, Reflections on the Psalms (Geoffrey Bles, 1958)
  • Gordon J. Wenham, The Book of Leviticus (NICOT; Eerdmans, 1979)
  • K. Lawson Younger Jr., Ancient Conquest Accounts (JSOT Press, 1990)
  • Lawson G. Stone, 'Ethical and Apologetic Tendencies in the Redaction of Joshua,' CBQ 53 (1991)
  • Susan Niditch, War in the Hebrew Bible (Oxford, 1993)
  • Eugene H. Merrill and Daniel L. Gard, in Show Them No Mercy: Four Views on God and Canaanite Genocide (Zondervan, 2003)
  • Hector Avalos, Fighting Words: The Origins of Religious Violence (Prometheus, 2005)
  • Christopher J. H. Wright, The God I Don't Understand (Zondervan, 2008)
  • Richard S. Hess, in War in the Bible and Terrorism in the Twenty-First Century (Eisenbrauns, 2008)
  • Kenton L. Sparks, God's Word in Human Words (Baker Academic, 2008)
  • Eric A. Seibert, Disturbing Divine Behavior (Fortress, 2009)
  • Walter Brueggemann, Divine Presence amid Violence (Cascade, 2009)
  • Wesley Morriston, 'Did God Command Genocide?' Philosophia Christi 11 (2009)
  • Nicholas Wolterstorff, 'Reading Joshua,' in Divine Evil? The Moral Character of the God of Abraham (Oxford, 2010)
  • Carolyn J. Sharp, Wrestling the Word (Westminster John Knox, 2010)
  • Paul Copan, Is God a Moral Monster? (Baker, 2011)
  • Philip Jenkins, Laying Down the Sword (HarperOne, 2011)
  • Paul Copan and Matthew Flannagan, Did God Really Command Genocide? (Baker, 2014)
  • Peter Enns, The Bible Tells Me So (HarperOne, 2014)
  • Gregory A. Boyd, The Crucifixion of the Warrior God, 2 vols. (Fortress, 2017)
Read 1samuel 15 →