Chapter 13
Chapter 13
Did God Command the Genocide of the Canaanites?
Matthew Flannagan
One of the most perplexing issues facing Christian believers is a series of jarring texts in the Old Testament. After their liberation from slavery in Egypt, the Israelites arrived on the edge of the Promised Land. In Deuteronomy, God commanded Israel to "destroy totally" the people occupying these regions (the Canaanites); the Israelites were to "not leave alive anything that breathes" (20:16 NIV). The book of Joshua records the carrying out of this command at Jericho: "they devoted the city to the LORD and destroyed with the sword every living thing in it—men and women, young and old, cattle, sheep and donkeys" (6:21 NIV). Several chapters later, we read that throughout the region Joshua "left no survivors. He totally destroyed all who breathed, just as the LORD, the God of Israel, had commanded" (10:40; cf. 11:14). The text mentions city after city where Joshua, at God's command, puts every inhabitant "to the sword," "totally destroyed everyone," and "left no survivors" (10:28,30,33,37,39–40; 11:8).
If one takes these passages literally, one must conclude this is a record of divinely authorized commission of genocide. Critics of Christian theism often ask a rhetorical question: How could a good and loving God command the extermination of the Canaanites?1
One response, which goes back to the patristic era, is to suggest that the strict, literal reading on which this rhetorical question is based is mistaken. Recently, several Protestant scholars have suggested a hyperbolic reading of the relevant passages.2 They suggest language such as "destroy totally," "do not leave alive anything that breathes," destroy "men and women, young and old," and so on, should be understood more like how we understand what a person means who yells, in the context of watching Lennox Lewis in a boxing match: "Knock his block off! Hand him his head! Take him out!" or hopes that the All Blacks will "annihilate the Springboks" or "totally slaughter the Wallabies." Now, the sports fan does not actually want Lennox Lewis to decapitate his opponent or for the All Blacks to become mass murderers. The same could be true here: understood in a nonliteral sense, the phrases probably were hyperbolic ways of telling the Israelites to defeat and drive out the Canaanites and "were not used and are not intended literally."3
Nicholas Wolterstorff suggests:
The Book of Joshua has to be read as a theologically oriented narration, stylized and hyperbolic at important points, of Israel's early skirmishes in the Promised Land, with the story of these battles being framed by descriptions of two great ritualized events. The story as a whole celebrates Joshua as the great leader of his people, faithful to Yahweh, worthy successor of Moses. If we strip the word "hagiography" of its negative connotations we can call it a hagiographic account of Joshua's events.4
In this essay I will defend Wolterstorff's position. In Part 1, I will sketch, adapt, and defend Wolterstorff's argument. In Part 2, I will argue that external evidence from comparative studies in ancient Near Eastern conquest accounts gives considerable support to Wolterstorff's position. Finally, in part 3, I will look at two implications of this position.
Wolterstorff's Argument
Wolterstorff's contention is that "a careful reading of the text in its literary context makes it implausible to interpret it as claiming that Yahweh ordered extermination." It is important to understand what he means by context. Here, it is clear that Wolterstorff is advocating a canonical approach. He notes that
Joshua as we have it today was intended as a component in the larger sequence consisting of Deuteronomy, Joshua, Judges, I and II Samuel, and I and II Kings. . . . I propose that we interpret the Book of Joshua as a component within this larger sequence—in particular, that we interpret it as preceded by Deuteronomy and succeeded by Judges.5
Joshua comes after Deuteronomy and before Judges. These books should be read as a single narrative—a connected literary unit. For example, Josh 24:28–31 mentions Joshua's death and burial place at Timnathserah "in the hill country of Ephraim" (NIV). This is deliberately connected to Judg 2:6–9, which mentions Joshua's death and burial place at Timnath-heres "in the hill country of Ephraim" (NIV). The first and last consonants are switched (from serah to heres)—a Hebrew literary device. Not only this, Judges connects to 1 (and 2) Samuel (the next book in the Hebrew canon) with its reference to "the hill country of Ephraim."6
When one does read these books as a connected literary unit, several issues are apparent. Joshua 6–11 summarizes several battles and concludes, "So Joshua took the entire land, just as the LORD had directed Moses, and he gave it as an inheritance to Israel according to their tribal divisions. Then the land had rest from war" (Josh 11:23 NIV).7 Judges, however, opens with a battle that occurs after Joshua's death; it states,
After the death of Joshua, the Israelites inquired of the Lord, "Who shall go up first for us against the Canaanites, to fight against them?" The Lord said, "Judah shall go up. I hereby give the land into his hand." Judah said to his brother Simeon, "Come up with me into the territory allotted to me, that we may fight against the Canaanites; then I too will go with you into the territory allotted to you." So Simeon went with him. Then Judah went up and the Lord gave the Canaanites and the Perizzites into their hand; and they defeated ten thousand of them at Bezek. (Judg 1:1–4)
Taken literally, the book of Joshua states that Joshua conquered the whole land and yet Judges states that much of the land was unconquered.
Similarly, Joshua affirms that he exterminated all the Canaanites in this region; repeatedly the text states that Joshua left "no survivors" and "destroyed everything that breathed" in "the entire land" and "put all the inhabitants to the sword." Alongside these general claims, the text identifies several specific places and cities where Joshua exterminated everyone and left no survivors. These include Hebron, Debir, the hill country, the Negev, and the western foothills (Josh 10:38,40). Yet in the first chapter of Judges we are told that the Canaanites lived in the Negev, the hill country, the western foothills, Hebron, and Debir (1:9–11). Moreover, they did so in such numbers and strength that they had to be driven out by force. These are the same cities that Joshua 10 tells us Joshua had annihilated, leaving no survivors.8
Likewise, Josh 11:23 states that "Joshua took the entire land" and then "gave it as an inheritance to Israel according to their tribal divisions" (NIV). Note: the conquered region is the same land that is later divided among the Israelite tribes. Only a chapter later, when the text turns to giving an account of these tribal divisions, the allotments begin with God telling Joshua, "You are very old, and there are still very large areas of land to be taken over" (13:1 NIV). Moreover, when one examines the allotment given to Judah, we see Caleb asking permission to drive the Anakites from the hill countries (14:11), and we also hear how Caleb has to defeat Anakites living in Hebron and, after this, marches against the people "living in Debir" (15:13–19). Similarly, it is evident with several of the other allotments that the people were yet to drive out Canaanites entrenched in the area and that the Israelites were not always successful in doing so.
We read, for example, that the Ephraimites and Manassites "did not dislodge the Canaanites living in Gezer; to this day the Canaanites live among the people of Ephraim" (Josh 16:10 NIV). Similarly, in chap. 17 it states, "Yet the Manassites were not able to occupy these towns, for the Canaanites were determined to live in that region. However, when the Israelites grew stronger, they subjected the Canaanites to forced labor but did not drive them out completely" (17:12–13 NIV). We hear that "Danites had difficulty taking possession of their territory, so they went up and attacked Leshem, took it, put it to the sword and occupied it. They settled in Leshem and named it Dan after their forefather" (19:47). Here, we see the same land said to be subdued and conquered by Joshua in battles where he exterminated and left alive nothing that breathed. This land was yet to be occupied by the tribes of Israel and is occupied by Canaanites, often heavily armed and deeply entrenched (17:17–18).9
Finally, the account of what God commanded differs in the two narratives. Joshua asserts: "He left no survivors. He totally destroyed all who breathed, just as the LORD, the God of Israel, had commanded" (10:40, my italics) and "exterminating them without mercy, as the LORD had commanded Moses" (11:20b, my italics). However, when this command is retroactively referred to in Judg 2:1, there is no mention of genocide or annihilation. Instead we hear of how God had promised to drive them out as well as His command to the Israelites not to make treaties with the Canaanites and to destroy their shrines. This silence is significant in the context. If God had commanded genocide, then it is odd that only instructions concerning treaties and shrines were mentioned.
Therefore, taken as a single narrative and taken literally, Joshua 1–11 gives a seemingly different account of events to that narrated by Judges and also to that narrated by the later chapters of Joshua itself.10
Wolterstorff makes a further point,
Those whose occupation it is to try to determine the origins of these writings will suggest that the editors had contradictory records, oral traditions, and so forth to work with. No doubt this is correct. But those who edited the final version of these writings into one sequence were not mindless; they could see, as well as you and I can see, the tensions and contradictions—surface or real—that I have pointed to. So what is going on?11
Wolterstorff's point is that, regardless of what sources or strata of tradition are alleged behind the final form of Joshua, the editors who put these books into a single narrative would have been well aware of the obvious tensions in the passages mentioned above. Moreover, they were not mindless or stupid. They would not want to affirm that both accounts were a literal description of what occurred. Yet they chose to put in the canon, next to Joshua, a book that began with a narration at odds with a literal reading of the early chapters of Joshua. Moreover, the author of Joshua chose to juxtapose the picture of Joshua 1–11 with the later chapters I mentioned above. They cannot, therefore, be asserting that both accounts are literally true.
I think Wolterstorff is right on this. However, his position could be strengthened on theological grounds. In Divine Discourse: Philosophical Reflections on the Claim that God Speaks, Wolterstorff provides an interesting and rigorous analysis of the notion that Scripture is the Word of God. Central to his analysis is that "an eminently plausible construal of the process whereby these books found their way into a single canonical text, would be that by way of that process of canonization, God was authorizing these books as together constituting a single volume of divine discourse."12
This understanding of Scripture provides the theological justification for reading the text as a single series. By examining what was affirmed in Judges and in later passages of Joshua, one can determine what the author of the early chapters of Joshua intended.13 Moreover, if the primary author of Scripture is God, then obviously the primary author of the final canonical text is unlikely to have deliberately (or accidentally) authored an obviously contradictory narrative. Hence, even if the contradictions were not obvious to the authors (and I think Wolterstorff correctly infers that these apparent contradictions would have been),14 they would be evident to God. Given that the process whereby these books were incorporated into a "single canonical text" constitutes God authorizing them, this process cannot have involved the authorization of a text that affirms, as literally true, two contradictory accounts.
It may be contended that an appeal to divine authorship in this way begs the question, but I think this is mistaken. As I understand the objection, the skeptic who claims that God commanded genocide is offering a reductio ad absurdum; he starts by assuming that whatever God commands is right and that Scripture is the Word of God and then derives from these assumptions the absurd conclusion that genocide is not wrong. The question then is whether, granting these assumptions, such a conclusion does, in fact, follow. If Scripture is a unified divine discourse, the skeptic's conclusion need not follow, because another assumption of the skeptic—namely, that all accounts were intended to be taken as literal—is not evidently true.
At this point Wolterstorff raises a further issue about the type of literature Joshua appears to be. He notes that the early chapters of Judges, by and large, read like "down-to-earth history." However, he continues, anyone carefully reading the book of Joshua will recognize in it certain stylistic renderings—"formulaic phrasings" and "formulaic convention[s]"15 as well as "the highly ritualized character of some of the major events described."16
The book is framed by its opening narration of the ritualized crossing of the Jordan and by its closing narration of the equally-ritualized ceremony of blessing and cursing that took place at Shechem; and the conquest narrative begins with the ritualized destruction of Jericho.17
A related ritualistic feature is "the mysterious sacral category of being devoted to destruction."18 Most significant is the use of formulaic language:
Anyone who reads the Book of Joshua in one sitting cannot fail to be struck by the prominent employment of formulaic phrasings. . . . Far more important is the formulaic clause, "struck down all the inhabitants with the edge of the sword."
The first time one reads that Joshua struck down all the inhabitants of a city with the edge of the sword, namely, in the story of the conquest of Jericho (6:21), one makes nothing of it. But the phrasing—or close variants thereon—gets repeated, seven times in close succession in chapter 10, two more times in chapter 11, and several times in other chapters. The repetition makes it unmistakable that we are dealing here with a formulaic literary convention.19
Thus Joshua itself appears to be full of ritualistic, stylized, formulaic language. It therefore looks like something other than a mere literal description of what occurred. In light of these facts Wolterstorff argues that Judges should be taken literally whereas Joshua is hagiographic history, a highly stylized, exaggerated account of the events designed to teach theological and moral points rather than to describe in detail what literally happened. Wolterstorff provides the example of North American morality tales of the noble Puritan or Washington crossing the Delaware. These are idealized, exaggerated accounts of the past designed to teach a moral lesson; they are not intended to be taken as accurate accounts of what actually occurred.
Ancient Near Eastern Conquest Accounts
Wolterstorff's argument has, I think, considerable force. Judges and Joshua cannot both be taken literally as their accounts are at odds; given the internal evidence Wolterstorff cites, it is reasonable to contend that Joshua is the one that is nonliteral. Wolterstorff, however, limits his case to what I call internal evidence, evidence from within the text itself. I think there is some interesting external evidence, evidence of how particular terms and language were used in other ancient Near Eastern histories of conquests and battles, which could be added to Wolterstorff's argument. Here I will cite three such lines of evidence.
The first is that comparisons between the book of Joshua and other ancient Near Eastern conquest accounts from the same period demonstrate some important stylistic parallels. Various studies have documented these similarities. Commenting on the structure of the campaigns mentioned in Joshua 9–12, Kenneth Kitchen reminds us:
This kind of report profile is familiar to readers of ancient Near Eastern military reports, not least in the second millennium. Most striking is the example of the campaign annals of Tuthmosis III of Egypt in his years 22–42 (ca. 1458–1438). . . . [T]he pharaoh there gives a very full account of his initial victory at Megiddo, by contrast with the far more summary and stylized reports of the ensuing sixteen subsequent campaigns. Just like Joshua against up to seven kings in south Canaan and four-plus up north.20
Kitchen adds,
The Ten Year Annals of the Hittite king Mursil II (later fourteenth century) are also instructive. Exactly like the "prefaces" in the two Joshua war reports (10:1–4; 11:1–5), detailing hostility by a number of foreign rulers against Joshua and Israel as the reason for the wars, so in his annals Mursil II gives us a long "preface" on the hostility of neighbouring rulers and people groups that lead to his campaigns.21
Kitchen offers other examples. He observes that the same formulaic style found in Joshua is also used in the Amarna letters EA 185 and EA 186.22 Similarly, before his major campaigns,
Joshua is commissioned by YHWH not to fear (cf. 5:13–15; 10:8; 11:6). So also by Ptah and Amun were Merenptah in Egypt, and Tuthmosis IV long before him; and likewise Mursil II of the Hittites by his gods (Ten-Year Annals, etc.), all in the second millennium, besides such kings as Assurbanipal of Assyria down to the seventh century.23
Similar studies have been done by John Van Seters24 and James Hoffmeier.25 However, the most comprehensive is that done by K. Lawson Younger. Younger notes similarities in the preface, structure, and even the way the treaty with the Gibeonites is recorded between Joshua and various ancient Near Eastern accounts.26 Joshua follows this convention in describing numerous battles occurring in a single day or within a single campaign.27 Ancient Near Eastern accounts also, like Joshua, repeatedly make reference to the enemy "melting with fear."28 Even the way postbattle pursuits are set out and described shows similarities with pursuits in ancient Near Eastern literature.29 I could mention more examples. The point is that "when the composition and rhetoric of the Joshua narratives in chaps. 9–12 are compared to the conventions of writing about conquests in Egyptian, Hittite, Akkadian, Moabite, and Aramaic texts, they are revealed to be very similar."30
Second, Younger notes such accounts are "highly figurative"31 and narrate military events via a common transmission code. The literary motif of divine intervention is an example. Both The 10 Year Annals of Mursilli and Sargon's Letter to the God record a divine intervention where the god sends hailstones on the enemy.32 Tuthmosis III has a similar story regarding a meteor.33 Younger observes that these accounts are very similar to parallel accounts in Joshua 10. Similarly, Younger points out that in many ancient Near Eastern texts "one can discern a literary technique in which a deity is implored to maintain daylight long enough for there to be a victory,"34 which has obvious parallels to Josh 10:13–14. Similarly, Richard Hess points to Hittite conquest accounts that describe the gods knocking down the walls of an enemy city in a manner similar to that described in the battle of Jericho.35 The fact that similar events are narrated in multiple different accounts suggests they are "a notable ingredient of the transmission code for conquest accounts"36—that is, part of the common hyperbolic rhetoric of warfare rather than descriptions of what actually occurred.
Third, part of this "transmission code" is that victories are narrated in an exaggerated hyperbolic fashion in terms of total conquest, complete annihilation and destruction of the enemy, killing everyone, leaving no survivors, and so on. Kitchen offers illuminating examples:
The type of rhetoric in question was a regular feature of military reports in the second and first millennia, as others have made very clear. . . . In the later fifteenth century Tuthmosis III could boast "the numerous army of Mitanni, was overthrown within the hour, annihilated totally, like those (now) non-existent"—whereas, in fact, the forces of Mitanni lived to fight many another day, in the fifteenth and fourteenth centuries. Some centuries later, about 840/830, Mesha king of Moab could boast that "Israel has utterly perished for always"—a rather premature judgment at that date, by over a century! And so on, ad libitum. It is in this frame of reference that the Joshua rhetoric must also be understood.37
Younger offers numerous other examples. Merenptah's Stele describes a skirmish with Israel as follows, "Yanoam is nonexistent; Israel is wasted, his seed is not."38 Here a skirmish in which Egypt prevailed is described hyperbolically in terms of the total annihilation of Israel. Sennacherib uses similar hyperbole, "The soldiers of Hirimme, dangerous enemies, I cut down with the sword; and not one escaped."39 Mursilli II records making "Mt. Asharpaya empty (of humanity)" and the "mountains of Tarikarimu empty (of humanity)."40 Mesha (whom Kitchen cited as stating "Israel has utterly perished for always") describes victories in terms of him fighting against a town, taking it, and then killing all the inhabitants of the town.41 Similarly, The Bulletin of Ramses II, a historical narrative of Egyptian military campaigns into Syria, narrates Egypt's considerably less-than-decisive victory at the battle of Kadesh with the following rhetoric:
His majesty slew the entire force of the wretched foe from Hatti, together with his great chiefs and all his brothers, as well as all the chiefs of all the countries that had come with him, their infantry and their chariotry falling on their faces one upon the other. His majesty slaughtered and slew them in their places. . . . He took no note of the millions of foreigners; he regarded them as chaff. 42
Numerous other examples could be provided. The hyperbolic use of language similar to that in Joshua is strikingly evident.43
It is equally evident that histories of this sort are highly stylized and often use this exaggeration for what could be called hagiographic purposes to commend the kings as faithful servants of the gods, rather than for giving literal descriptions of what occurred.44 They constitute "monumental hyperbole."45
Three things become evident about the book of Joshua as we study the evidence. First, taken as a single narrative and taken literally, Joshua 1–11 gives an account of events at odds with that narrated by Judges and also with that narrated by the later chapters of Joshua itself. Second is that, as Wolterstorff commented, "those who edited the final version of these writings into one sequence were not mindless," particularly if God speaks through them. Third, while Judges reads as "down to earth history," a careful reading of Joshua reveals it to be full of ritualistic, stylized, and formulaic language. This third point is supported by research into ancient Near Eastern conquest accounts. Such studies show (1) such accounts are highly hyperbolic, hagiographic, and figurative, and follow a common transmission code; (2) comparisons between these accounts and the early chapters of Joshua suggest Joshua is written according to the same literary conventions and transmission code; and (3) part of this transmission code is to portray a victory hyperbolically in absolute terms of totally destroying the enemy or in terms of miraculous divine intervention: "such statements are rhetoric indicative of military victory,"46 not literal descriptions of what occurred.
I think these three points, taken together, provide compelling reasons for thinking that one should interpret the text as a highly figurative and hyperbolic account of what occurred. In light of these factors, it seems sensible to conclude that the accounts of battles in Joshua 6–11 are not meant to be taken literally.
Implications
We can draw two important implications in light of what we have seen so far. First, as Wolterstorff asserts, on the assumption that Deuteronomy and Joshua are both literarily and linguistically connected and canonically sequenced,
this interpretation of Joshua forces a back-interpretation of Deuteronomy. If "struck down all the inhabitants with the edge of the sword" is a literary convention when used to describe Joshua's exploits, then it is likewise a literary convention when similar words are used by Moses in his instructions to Israel in general and to Joshua in particular.47
I think Wolterstorff is correct here: this interpretation of Joshua does force a back-interpretation of Deuteronomy. Deuteronomy 7:2 states "when the LORD your God has given them up before you and you have struck them, you shall utterly destroy them." Similarly, Deut 20:16–17 commands: "do not leave alive anything that breathes. Completely destroy them" (NIV). In Joshua 10 one sees the formulaic language of "and the LORD gave [the city]" and he/they "struck it and its king with the edge of the sword" until "there was no one remaining" (NRSV). The chapter is summarized with the phrase, "So Joshua defeated the whole land . . . ; he left no one remaining, but utterly destroyed all that breathed" (v. 40 NRSV). The similar phraseology is evident.
Moreover, the book of Joshua clearly, explicitly, and repeatedly identifies what Joshua did in these chapters with the command that Moses had given regarding the Canaanites in Deuteronomy: "He left no survivors. He totally destroyed all who breathed, just as the LORD, the God of Israel, had commanded" (Josh 10:40 NIV, emphasis added). So, if the language of "striking all the people by the sword," "leaving no survivors," "totally destroying," "striking all the inhabitants with the edge of the sword," and so on is hyperbolic (as the evidence suggests it is), then the command cannot have been intended to be taken literally.
This understanding of Deuteronomy's commands also solves some other interpretative problems. Here I will mention briefly three.
(1) What God commanded regarding the Canaanites differs in various canonical books. As noted above, in Judges 2 when the Angel of the Lord refers back to the original command, it is stated in terms of not making treaties with them, destroying their shrines, and driving them out; it is not stated in terms of literally exterminating them. Similarly, in the earlier book of Exodus the command is given in terms of not allowing the Canaanites to live in the land, again, not in terms of extermination. This is significant. "Deuteronomy" in Greek means "second law"; and throughout Deuteronomy, Moses repeats laws already laid down in the book of Exodus, sometimes expanding on them. The Decalogue, for example, which was delivered on Sinai in Exodus 20, is repeated in Deuteronomy 5. The laws about releasing an ebed (an indentured servant) in Exod 21:1 are repeated and expanded in Deut 15:12–18. Similarly, Deut 22:28–29 is a repetition of a law spelled out in Exod 22:15.48 The same occurs with the law under discussion. Deuteronomy 7 repeats the same promises and commands laid down in Exod 23:20–32; however, in Deuteronomy, the language of "destroy them" replaces the "do not let them live in your land" in Exodus. Wolterstorff's interpretation explains this.
(2) The word herem, which is translated "destroy" in Deut 7:2, has the primary meaning of the irrevocable giving over or devotion of something to Yahweh and hence implies renunciation. The term has also developed a secondary secular meaning of "to destroy,"49 but, a literal reading of "destroy" here does not fit the context well. The command to "destroy" the Canaanites occurs alongside several other commands, "Make no treaty with them, and show them no mercy. Do not intermarry with them. Do not give your daughters to their sons or take their daughters for your sons" (Deut 7:2–3 NIV). However, this seems odd: killing is not an obvious antithesis to marrying or making a covenant; moreover, the text goes on to elaborate the command in terms of smashing idols and driving them out—in a similar vein as Judges. For this reason Christopher Wright argues herem should be translated as "renounce," and is a command to shun the idolatrous Canaanites.50 This reading clashes with the parallel verse in Deut 20:17, where "herem is used epexegetically to verse 16, 'you shall not leave alive anything that breathes.'"51 However, taking the word as "destroy" and understanding it hyperbolically makes sense of this.52
(3) The hyperbolic reading addresses another apparent contradiction in the text noted by many readers of the Pentateuch. While Deut 7:2 and 20:16–17 command Israel to "utterly destroy" the Canaanites and to "not leave alive anything that breathes," numerous other texts claim the Canaanites are to be "driven out," "dispossessed," "thrust out," and so on. In fact, often the "drive out" language is juxtaposed with the language of "destroy." Taken literally, these pictures are inconsistent. If I stated that I had driven an intruder from my house, one would not assume the intruder was dead in my lounge. Similarly, if I said I had killed an intruder, one would not normally think this meant the intruder had fled. The Hebrew confirms this; the language of driving out and casting out is used elsewhere to refer to Adam and Eve being driven from Eden (Gen 3:24), Cain being driven into the wilderness (Gen 4:14), and David being driven out by Saul (1 Sam 26:19). All are cases where the meaning precludes something being literally destroyed.53 Moreover, when the "drive out" language is used of Canaan, it often is used in a context where it does not literally mean destroyed, but rather dispossessed. In Lev 18:26–28, in the same way the Canaanites would be driven out, so Israel would be for violating the covenant.54 This ultimately happened during the Babylonian exile—which was obviously not an annihilation of ethnic Jews. Furthermore, the language of destroying whole nations in several places in Deuteronomy is used in a rhetorical or hyperbolic sense, alongside references to "driving out" the nation in question or dispossessing them; it does not mean exterminating them. Hence, Wolterstorff's suggestion has ample precedent from within the text itself.55
A second implication of Wolterstorff's position is that Joshua does not assert that Israel engaged in divinely authorized genocide:
. . . when a high-school basketball player says his team slaughtered the other team last night, what is he asserting? Not easy to tell. That they scored a decisive victory? Maybe. But suppose they just barely eked out a win? Was he lying? Maybe not. Maybe he was speaking with a wink-of-the-eye hyperbole.56
In the same way, when one realizes that Joshua is hagiographic and highly hyperbolic in its narration of what occurred, the best one can conclude from the accounts of killing everyone that breathed is this:
Israel scored a decisive victory and once you recognize the presence of hyperbole it is not even clear how decisive the victories were. Joshua did not conquer all the cities in the land nor did he slaughter all the inhabitants in the cities he did conquer. The book of Joshua does not say that he did.57
Canonical factors force the same conclusion. I noted above that in Judges and Exodus the command is expressed in terms of avoiding treaties and driving the Canaanites out. In Joshua and Deuteronomy the command is expressed in the language of "utterly destroying them." The conclusion we have reached is that the latter is figurative language and the former is literal. If this is the case, then the command was to drive them out and it was not to literally exterminate them.
Conclusion
I contend that the widely held view that the book of Joshua teaches that God commanded the genocide of the Canaanites is questionable. Joshua is accepted as part of a canon. Read in this context, taking the account of total annihilation of the Canaanite populations as a literal description of what occurred contradicts what is affirmed to have literally occurred in Judges. Moreover, it conflicts with how the command is described elsewhere in Judges and Exodus. The writers would have known this and, not being mindless, could not have meant both accounts to be taken literally. This means one must be nonliteral. The literary conventions Joshua uses are highly stylized, figurative, and contain hyperbolic, hagiographic accounts of what occurred. The conventions in Judges are less so. Consequently, the so-called genocide in Joshua and the command to "utterly destroy" the Canaanites should not be taken literally.
Appendix: The Case of the Midianites
Wolterstorff's approach may also shed light on some other troubling texts such as the apparent genocide of the Midianites in Numbers 31. After the Israelites "fought against Midian, as the LORD commanded Moses, and killed every man" (v. 7 NIV), Moses commanded them to "kill all the boys. And kill every woman who has slept with a man, but save for yourselves every girl who has never slept with a man" (vv. 17–18 NIV). Taken in isolation, this text affirms that every Midianite was killed and only female virgins survived so they could be assimilated into the Israelite community. However, read in its literary context, as part of a single narrative—a connected literary unit—similar issues are apparent.
First, Numbers 31 is one part of a broader context; it is both part of the Pentateuch and also part of a larger canonical sequence. The Pentateuch contains the Torah or Law. Normally in the Torah when Moses utters a command on God's behalf the passage begins with "The LORD commanded Moses"; this preface is absent from the commands in Numbers 31. The passages merely state that God commanded them to make war on Midian; v. 7 states, "They fought against Midian, as the LORD commanded Moses, and killed every man." This suggests the Israelites fulfilled this command. Moses' command to kill women and children occurs after this and appears to be on his own authority.58 If one reads the laws of war that are elaborated in the book of Deuteronomy, which follows Numbers, God commanded Israel not to kill noncombatants, such as women and children. He condemns the kind of conduct Moses commands here.59
If one looks at the larger canonical sequence, the subsequent narrative states quite emphatically that the Midianites were not, in fact, literally wiped out. In Judges 6 and 7, the Midianites invade Israel in numbers said to be "like swarms of locusts. It was impossible to count the men and their camels" (Judg 6:5 NIV). Israel was so overrun with Midianites that they fled to "mountain clefts, caves and strongholds" (Judg 6:2 NIV). Unable to win in open battle, Gideon was forced to use deception to defeat them. This is not congruous with the Midianites having been "utterly destroyed."
Second, these tensions in the text are fairly obvious. As with Joshua, whoever "edited the final version of these writings into one sequence"60 juxtaposed several accounts, which, if taken literally, describe Israel, at Moses' command, annihilating the entire population of Midian (including noncombatants); yet several other accounts affirm that God prohibits the killing of noncombatants and that the Midianites were not annihilated. Assuming this person was not "mindless," he cannot have intended to affirm both as literally true.
Third, the genre and style of the accounts suggests Numbers 31 is the nonliteral account. Numbers 31 appears highly hyperbolic; it contains obvious rhetorical exaggeration.61 The Israelite army is said to have killed every Midianite man in battle without a single Israelite fatality (Num 31:50).62 Moreover, the spoil from the battle is said to be 32,000 maidens and 675,000 sheep and goats—this is astronomically and absurdly large.63 Daniel Fouts notes that exaggerated numbers are common forms of hyperbole in ancient Near Eastern battle accounts.64 Wolterstorff suggests, "These are all hyperbolic descriptions of battles that took place."65
So, if we read the text in the literary context of the broader canon, we again see the author juxtaposing two accounts. One claims that God prohibits killing noncombatants and that the Midianites continued to live in the land as a serious military threat; another account, using rhetoric known to be used hyperbolically in military contexts, states that Israel, at Moses' command, wiped them all out. Assuming the author was an intelligent person,66 we are at least owed an argument as to why one should read these texts as literally claiming that God commanded genocide.67
A similar phenomenon occurs with the case of the Amalekites,68 the Babylonian invasions,69 and the sacking of the Jebusite city of Jerusalem.70 In each case a battle is narrated in totalistic terms of complete destruction of all the people, yet later narration goes on to assume matter-of-factly that it did not literally occur. The fact that this happens on multiple occasions in different books rapidly diminishes the probability that these features are coincidental or careless errors. Why is it that almost every time a narration of "genocide" occurs, it is followed by an account that presupposes it did not happen? These facts significantly raise the probability that this is a deliberate literary construction by the authors.
1 Raymond Bradley develops this objection in "A Moral Argument for Atheism," in The Impossibility of God, ed. Michael Martin and Ricki Monnier (Amherst, NY: Prometheus Books, 2003), 144–45. Similar arguments to Bradley's have been made by Wes Moriston, "Did God Command Genocide? A Challenge to the Biblical Inerrantist," Philosophia Christi 11/1 (2009): 8–26; Randal Rauser, "Let Nothing that Breathes Remain Alive: On the Problem of Divinely Commanded Genocide," Philosophia Christi 11/1 (2009): 27–41; Michael Tooley, "Does God Exist?" in Michael Tooley and Alvin Plantinga, The Knowledge of God (Malden, MA: Blackwell Publishers, 2008), 73–77; Edwin Curley, "The God of Abraham, Isaac, and Jacob," and Evan Fales, "Satanic Verses: Moral Chaos in Holy Writ," in Divine Evil? The Moral Character of the God of Abraham, ed. Michael Bergmann, Michael J. Murray, and Michael C. Rea (New York: Oxford University Press, 2011), 58–78, 91–108; Louise Antony, "Atheism as Perfect Piety," and Walter Sinnott-Armstrong, "Why Traditional Theism Cannot Provide an Adequate Foundation for Morality," in Is Goodness without God Good Enough? A Debate on Faith, Secularism, and Ethics, ed. Robert K. Garcia and Nathan L. King (Lanham, MD: Rowman & Littlefield Publishers, 2009), 67–84, 101–16.
2 Alvin Plantinga, "Comments on 'Satanic Verses: Moral Chaos in Holy Writ,'" in Bergmann, Murray, and Rea, Divine Evil? 110–11; Paul Copan, "Yahweh Wars and the Canaanites: Divinely Mandated Genocide or Corporate Capital Punishment," Philosophia Christi n.s. 11/1 (2009): 73–90; idem, Is God a Moral Monster? Making Sense of the Old Testament (Grand Rapids: Baker Books, 2011), 158–97; Christopher Wright, The God I Don't Understand: Reflections on Tough Questions of Faith (Grand Rapids: Zondervan, 2008), 87–88. In his latest discussion on the issue, William Lane Craig states, "I've come to appreciate that the object of God's command to the Israelis was not the slaughter of the Canaanites, as is often imagined. The command rather was primarily to drive them out of the land. The judgment upon these Canaanite kingdoms was to dispossess them of their land and thus destroy them as kingdoms." See "Question 147: Divine Command Morality and Voluntarism," accessed October 29, 2010, http://www.reasonablefaith.org/site/News2?page=NewsArticle&id=7911.
3 Plantinga, "Comments," 111.
4 Nicholas Wolterstorff, "Reading Joshua," in Divine Evil? The Moral Character of the God of Abraham, ed. Michael Bergmann, Michael J. Murray, and Michael C. Rea (New York: Oxford University Press, 2010), 252–53.
5 Ibid., 249.
6 Note this phrase "the hill country of Ephraim" in Joshua (17:15; 19:50; 20:7; 24:30,33), then in Judges (2:9; 3:27; 4:5; 7:24; 10:1; 17:1,8; 18:2,13; 19:1,16,18), and then in 1 Samuel (1:1; 9:4). The linkage of Judg 17:1 and 19:1 with 1 Sam 1:1 is connected by the general threefold pattern of (1) "There was a (certain) man . . . "; (2) "from the hill country of Ephraim"; (3) "and his name was _______."
As mentioned in the text, Josh 24:28–31 mentions Joshua's death and burial place at Timnath-serah "in the hill country of Ephraim," and then Judg 2:6–9 refers to Joshua's death and burial place at Timnath-heres "in the hill country of Ephraim." That is, Joshua and Judges are literarily connected in their mention of Joshua's (1) death and (2) burial. (1) Both books state that "Joshua son of Nun, the servant of the LORD, died" at the age of "a hundred and ten" (Josh 24:29; Judg 2:8). (2) Then a deliberate literary connection is made on the burial place of Joshua, using the Hebrew letter-substitution cipher known as atbash, in this case reversing the first and last consonants of the burial place from "-serah" to "-heres." Compare "they buried him in the territory of his inheritance in Timnath-serah, which is in the hill country of Ephraim, on the north of Mount Gaash" (Josh 24:30, NASB, my emphasis) with "they buried him in the territory of his inheritance in Timnath-heres, in the hill country of Ephraim, north of Mount Gaash (Judg 2:9, NASB, my emphasis). I owe these examples to Paul Copan.
7 All Scripture quotations are from the NIV unless otherwise stated.
8 In addition to these general claims about exterminating populations, Josh 11:21–22 states, "Joshua went and destroyed the Anakites from the hill country: from Hebron, Debir and Anab, from all the hill country of Judah" (NIV). This happened after Joshua is already said to have killed the inhabitants in these areas in Josh 10:30–40. Joshua 11:21 states that no Anakites were left living in Israeli territory after this campaign. In Judg 1:21 the text explicitly states that Anakites are in Hebron.
9 Kenneth A. Kitchen, On the Reliability of the Old Testament (Grand Rapids: Eerdmans, 2003). Kitchen notes that a careful reading of the earlier chapters makes it clear that Israel did not actually conquer the areas mentioned at all. Kitchen notes that, according to Josh 4:19, after crossing the Jordan the Israelites set up camp in Gilgal on the eastern border of Jericho (161). Rather, after every battle in the next six chapters the text explicitly states that they returned to Gilgal:
The conflict with Canaanite city-state rulers in the south part of Canaan is worth close observation. After the battle for Gibeon, we see the Hebrews advance upon six towns in order, attacking and capturing them, killing their local kings and such of the inhabitants as had not gotten clear, and moving on, not holding on to these places. Twice over (10:15, 43), it is clearly stated that their strike force returned to base camp at Gilgal. So there was no sweeping takeover and occupation of this region at this point. And no total destruction of the towns attacked (162).
Kitchen goes on to comment:
What happened in the south was repeated up north. Hazor was both leader and famed center for the north Canaanite kinglets. Thus as in the south, the Hebrew strike force defeated the opposition; they captured their towns, killed rulers and less mobile inhabitants, and symbolically burned Hazor, and Hazor only, to emphasise the end of its local supremacy. Again Israel did not attempt to immediately hold on to Galilee; they remained based at Gilgal (cf. 14:6) (ibid.).
Kitchen points out that "the first indication of a real move in occupation outward beyond Gilgal comes in 18:4" (162). This is after the first allotment of "lands-to-be-occupied had been made," and as we saw above, the Israelites did not find occupying these allotments easy. He concludes: "these campaigns were essentially disabling raids: they were not territorial conquests with instant Hebrew occupation. The text is very clear about this" (ibid.).
10 In this section I am arguing that the record of Joshua fulfilling the command is hyperbolic. Even many evangelical scholars who believe the command was literal accept this. For example, David M. Howard Jr. argues that, on the one hand, statements claiming Joshua subdued "the whole region" (10:40) are hyperbolic assertions emphasizing that "there was indeed a sweeping victory" and that "no significant opposition remained." Such was intended "to reiterate the theological point made many times in the book that God was indeed giving Israel the entire land." On the other hand, "the author acknowledged elsewhere that the conquest was indeed not complete" and that "there was still much work to do" (Joshua, New American Commentary [Nashville: B&H, 1998], 259). See also Richard Hess, Joshua: An Introduction and Commentary, Tyndale Old Testament Commentary (Downers Grove, IL: InterVarsity, 1996), 284–86. Where I differ from them is over the implications of this conclusion. I argue below that if fulfillment is hyperbolic, then a plausible implication is that the command itself was hyperbolic.
11 Wolterstorff, "Reading Joshua," in Bergmann et al., Divine Evil? 251.
12 Nicholas Wolterstorff, Divine Discourse: Philosophical Reflections on the Claim that God Speaks (Cambridge: Cambridge University Press, 1995), 295. See also "Unity behind the Canon," in One Scripture or Many? The Canon from Biblical, Theological and Philosophical Perspectives, ed. Christine Helmer and Christof Landmesser (New York: Oxford University Press, 2004), 217–32.
13 Alvin Plantinga argues in Warranted Christian Belief (New York: Oxford University Press, 2000), 385:
[A]n assumption of the enterprise is that the principal author of the Bible—the entire Bible—is God himself (according to Calvin, God the Holy Spirit). Of course each of the books of the Bible has a human author or authors as well; still, the principal author is God. This impels us to treat the whole more like a unified communication than a miscellany of ancient books. Scripture isn't so much a library of independent books as itself a book with many subdivisions but a central theme: the message of the gospel. By virtue of this unity, furthermore (by virtue of the fact that there is just one principal author), it is possible to "interpret Scripture with Scripture." If a given passage from one of Paul's epistles is puzzling, it is perfectly proper to try to come to clarity as to what God's teaching is in this passage by appealing not only to what Paul himself says elsewhere in other epistles but also to what is taught elsewhere in Scripture (for example, the Gospel of John).
14 Nicholas Wolterstorff in the question-and-answers session following presentation of his paper "Reading Joshua" (presented at the "My Ways Are Not Your Ways: The Character of the God of the Hebrew Bible" conference at the Center for Philosophy of Religion, University of Notre Dame, 12 September 2009, accessed Dec. 19, 2009, <http://www.nd.edu/~cprelig/conferences/documents/HBprogram_006.pdf>, stated that the phrase "he killed all the inhabitants with the edge of the sword" occurs at least 15 times in Joshua 6–11 in close succession. This is "hammered home with emphasis." This is then followed in the next chapter by the claim that Joshua had not conquered the whole land. In the next five chapters it is stressed repeatedly that the land is not yet conquered. This is followed by the opening chapters of Judges, which affirms eight times in a single chapter that the Israelites had failed to conquer the land or the cities, finishing with the Angel of the Lord at Bokim rebuking them for failing to do so (Judg 2:1–5). These are not subtle contrasts. They are, in Wolterstorff's words, "flamboyant"; it is unlikely that an intelligent author or editor would have missed this.
15 Wolterstorff, "Reading Joshua," Bergmann et al. in Divine Evil? 251.
16 Ibid., 252.
17 Ibid. The ritualized nature of the narration is also stressed by Duane L Christensen, Deuteronomy 1:1–21:9 (Nashville: Thomas Nelson, 2001).
18 Wolterstorff, "Reading Joshua," in Bergmann et al., Divine Evil? 252. That the word herem, translated as "devoted to destruction," often serves a figurative or rhetorical function is also noted by Christopher Wright in The God I Don't Understand: Reflections on Tough Questions of Faith (Grand Rapids: Zondervan, 2008), 87–88:
Now we need to know that Israel's practice of herem was not in itself unique. Texts from other nations at the time show that such total destruction in war was practiced, or at any rate proudly claimed, elsewhere. But we must also recognize that the language of warfare had a conventional rhetoric that liked to make absolute and universal claims about total victory and completely wiping out the enemy. Such rhetoric often exceeded reality on the ground.
At the other end of the spectrum, minimalist scholar Thomas L. Thompson, writing on the use of herem in the Mesha Stele, notes that the "use of the ban at both Ataroth and Nebo are clearly part of the totalitarian rhetoric of holy war rather than historical considerations" ("Mesha and Questions of Historicity," Scandinavian Journal of the Old Testament 22/2 [2007]: 249).
19 Wolterstorff, "Reading Joshua," in Bergmann et al., Divine Evil? 251.
20 Kitchen, Historical Reliability, 170 (emphasis added).
21 Ibid. (emphasis added).
22 Ibid., 172.
23 Ibid., 174–75.
24 John Van Seters, "Joshua's Campaign of Canaan and Near Eastern Historiography," Scandinavian Journal of the Old Testament 2 (1990): 1–12.
25 James K. Hoffmeier, "The Structure of Joshua 1–11 and the Annals of Thutmose III," in Faith, Tradition, and History: Old Testament Historiography and Its Ancient Near Eastern Context, ed. A. R. Millard, James K. Hoffmeier, and David W. Baker (Winona Lake, IN: Eisenbrauns, 1994), 165–81.
26 K. Lawson Younger Jr., Ancient Conquest Accounts: A Study in Ancient Near Eastern and Biblical History Writing (Sheffield: Sheffield Academic Press, 1990), 200–204.
27 Ibid., 216.
28 Ibid., 258–60.
29 Ibid., 220–25.
30 Ziony Zevit, The Religions of Ancient Israel: A Synthesis of Parallactic Approaches (London and New York: Continuum, 2001), 114.
31 Younger states: "As the ancient historian (whether Near Eastern, biblical, or otherwise) reconstructed "historical" referents into a coherent description, he produced a figurative account, a 're-presenting representation.'" K. Lawson Younger Jr., "Judges 1 in Its Near Eastern Literary Context," in Millard et al., Faith, Tradition, and History, 207. He also suggests such a historian functions as "a literary artist." He adds: "Ancient Near Eastern conquest accounts are figurative in three ways: (1) the structural and ideological codes that are the apparatus for the text's production; (2) the themes or motifs that the text utilizes; and (3) the usage of rhetorical figures in the accounts" (ibid.).
32 Younger, Ancient Conquest Accounts, 208–11.
33 Ibid., 217.
34 Ibid., 219. For further discussion of the relationship between Joshua's long day and other ancient Near Eastern texts, see John Walton, "Joshua 10:12–15 and Mesopotamian Celestial Omen Texts," in Millard et al., Faith, Tradition, and History, 181–90.
35 Richard Hess, "West Semitic Texts and the Book of Joshua," Bulletin for Biblical Research 7 (1997): 68.
36 Younger, Ancient Conquest Accounts, 211.
37 Kitchen, Reliability of the Old Testament, 174.
38 Younger, Ancient Conquest Accounts, 227.
39 Ibid., 228.
40 Ibid.
41 Ibid., 227.
42 Ibid., 245 (emphasis added).
43 In addition, both Kitchen and Younger observe that such hyperbolic language is used in several places within the book of Joshua itself. Joshua 10:20, for example, states that Joshua and the sons of Israel had "finished destroying" and "completely destroyed" their enemies. Immediately, however, the text affirms that the "survivors went to fortified cities." In this context, the language of total destruction is clearly hyperbolic. Similarly, the account of the battle of Ai is clearly hyperbolic. After Joshua's troops feign a retreat, the text states that "all the men of Ai" are pressed to chase them. "Not a man remained in Ai or Bethel who did not go after Israel. They left the city open and went in pursuit of Israel" (8:17 NIV). Joshua lures the pursuers into a trap "so that they were caught in the middle, with Israelites on both sides. Israel cut them down, leaving them neither survivors nor fugitives" (8:22 NIV). Then the text immediately following states: "When Israel had finished killing all the men of Ai in the fields and in the desert where they had chased them, and when every one of them had been put to the sword" (8:24 NIV), they went to the city of Ai and killed all the men in it.
44 Thomas Thompson examines several different ancient Near Eastern conquest accounts of this type and notes they have a hagiographic function. See his "A Testimony of the Good King: Reading the Mesha Stele," in Ahab Agonistes: The Rise and Fall of the Omri Dynasty, ed. Lester L. Grabbe (Edinburgh: T&T Clark, 2007), 236–92.
45 John Goldingay goes on to give yet another example from within the Bible itself: "While Joshua does speak of Israel's utterly destroying the Canaanites, even these accounts can give a misleading impression: peoples that have been annihilated have no trouble reappearing later in the story; after Judah puts Jerusalem to the sword, its occupants are still living there 'to this day' (Judg. 1:8, 21)." See his "City and Nation," in Old Testament Theology, vol. 3 (Downers Grove, IL: InterVarsity, 2009), 570.
46 K. Lawson Younger Jr., "Joshua," in The IVP Bible Background Commentary: Old Testament, ed. John H. Walton, Victor H. Matthews, and Mark W. Chavalas (Downers Grove, IL: InterVarsity Press, 2000), 227.
47 Wolterstorff, "Reading Joshua," in Bergmann et al., Divine Evil? 252.
48 Gordon Wenham, "Bethulah: A Girl of Marriageable Age," Vetus Testamentum 22 (1972): 326–48.
49 J. P. U. Lilley, "Understanding the Herem," Tyndale Bulletin 44 (1993): 170–73.
50 Christopher J. H. Wright, Deuteronomy, New International Biblical Commentary (Peabody, MA: Hendrickson, 1996), 109.
51 Lilley, "Understanding the Herem," 174.
52 Other commentators such as Duane L. Christensen and J. Gordon McConville suggest "destroy" is being used in a figurative sense. McConville, for example, states "the concept of complete annihilation of the nations is always a kind of ideal, symbolizing the need for radical loyalty to Yahweh on the part of Israel." J. Gordon McConville, Deuteronomy, Apollos Old Testament Commentary (Downers Grove, IL: InterVarsity, 2002), 161.
Some object that a hyperbolic interpretation does not fit the context, which draws a contrast between sparing "the women, the children, the livestock" (v. 14) and totally destroying them (v. 16: "do not leave alive anything that breathes"). This is mistaken. First, the emphasis in v. 14 is not on sparing noncombatants but rather on the permissibility of marrying the women of conquered enemies, adopting their children, and using their cattle. Second, the contrast is not between vv. 14 and 16, but between v. 16 and the whole set of instructions regarding nations that are far away (vv. 10–15). These verses command Israel to seek to make peace treaties first and if they go to war and kill combatants they can marry the women, adopt children, and keep the livestock. In other words, as much as possible they are to seek peaceful coexistence with these nations. A command to go to war and drive them out expressed hyperbolically as "totally destroy them" and "leave nothing alive that breathes" would stand in contrast with this. A final point on this is that the crucial issue is whether the hyperbolic interpretation is more plausible than a literal one, even if a literal interpretation fits Deuteronomy 20 better. Above, I have argued that a literal interpretation puts Joshua 6–11 at odds with Judges 1–2 and the later chapters of Joshua. It would be odd to reject a hyperbolic interpretation because one passage in Deuteronomy 20 does not cohere with it and instead embrace a literal interpretation, which creates an even greater incoherence in the text.
53 I owe this point to conversations with Paul Copan. See his comments on this in Is God a Moral Monster? 181–82.
54 "But you must keep my decrees and my laws. The native-born and the aliens living among you must not do any of these detestable things, for all these things were done by the people who lived in the land before you, and the land became defiled. And if you defile the land, it will vomit you out as it vomited out the nations that were before you" (NIV).
55 Deut 2:10–12,20–22; 4:26–30; 28:63.
56 Nicholas Wolterstorff, "Reply to Antony," in Bergmann et al., Divine Evil? 263.
57 Nicholas Wolterstorff, "Reading Joshua," presented at the "My Ways Are Not Your Ways: The Character of the God of the Hebrew Bible" conference; this paragraph was in the paper presented at the conference but was omitted from the published version in Divine Evil?
58 Alvin Plantinga, "Comments on Satanic Verses in Holy Writ," in Bergmann et al., Divine Evil? 110. Plantinga is responding to Evan Fales, who based his reading on the mistaken reading of the verse "kill all the women but save for yourselves every girl who has never slept with a man" as commanding rape—a reading he also ascribes to Deuteronomy 21. In fact, a glance at the immediate context shows that the reference to a woman not sleeping with a man in the former passage was mentioned to distinguish the women in question from those who had seduced Israelite men into idolatry; it is not there to emphasize their availability for sex. Deuteronomy 21, in fact, protects female captives from being raped or sold as concubines. See Copan, Is God a Moral Monster? 118–21, 180.
59 Some might object that God did command the killing of men, women, and children in 1 Samuel 15, where God commanded Saul to "utterly destroy [haram]" and "not spare" the Amalekites. God used the sweeping language of "put to death both man and woman, child and infant, ox and sheep, camel and donkey" (15:3 NASB). It is important to note that Saul carried this command out, see 1 Sam 15:7–9:
Saul attacked the Amalekites all the way from Havilah to Shur, to the east of Egypt. He took Agag king of the Amalekites alive, and all his people he totally destroyed with the sword. But Saul and the army spared Agag and the best of the sheep and cattle, the fat calves and lambs—everything that was good. These they were unwilling to destroy completely, but everything that was despised and weak they totally destroyed. (NIV)
The language of "utterly destroy" (haram, and its noun cognate herem), the reference to cattle and sheep, and the reference to "sparing" all parallel the language of the command. So it would be difficult to justify interpreting the command as literal and the fulfilment as hyperbolic; the command and fulfilment need to be read in the same sense. Agag, the sole survivor, was executed a few verses later (15:33). So taken in isolation and read literally, this passage records a divinely commanded genocide that was carried out.
However, if when one reads this passage as part of a single narrative—a connected literary unit—then several issues are apparent. First, the narrative states quite emphatically that the Amalekites were not, in fact, literally wiped out. In 1 Sam 27:8–9 David invaded a territory full of Amalekites:
Now David and his men went up and raided the Geshurites, the Girzites and the Amalekites. (From ancient times these peoples had lived in the land extending to Shur and Egypt.) Whenever David attacked an area, he did not leave a man or woman alive, but took sheep and cattle, donkeys and camels, and clothes. Then he returned to Achish.
Not only does this text affirm the Amalekites still exist, but the reference to Egypt and Shur states they exist in the same area that Saul attacked in the previous passages. Moreover, David killed Amalekite men and women living in his area and took sheep and cattle as plunder. However, these are the very things Saul was said in 1 Samuel 15 to have eradicated.
As we showed above, Saul was said to have killed all the Amalekites, men and women from "Havilah to Shur, to the east border of Egypt" and all the livestock had either been destroyed or taken as plunder. Further, in 1 Samuel 30, a sizeable Amalekite army attacked Ziklag (1 Sam 30:1). David pursued this army and fought a long battle with them and 400 Amalekites fled on horseback (1 Sam 30:7–17). In 2 Sam 1:8, an Amalekite took credit for killing Saul, and 1 Chron 4:43 tells us Amalekites were still around during the reign of Hezekiah. So, if you view 1 Samuel and its canonical context as a single literary unit the text cannot be sensibly claiming that 1 Samuel 15 and 27 are both literally true accounts of battles with the Amalekites. Third, the genre and style of the accounts suggest 1 Samuel 15 is the nonliteral; it appears highly hyperbolic and contains obvious rhetorical exaggeration. Saul's army was said to be 210,000 men, and thus larger than any army known at this time in antiquity. Moreover, Saul struck the Amalekites from Havila to Shur. Shur is on the edge of Egypt; Havila is in Saudi Arabia. This is an absurdly large battlefield. The text used standard hyperbolic rhetoric of exaggerated numbers and geography. Similarly, 1 Samuel 15's use of the language of "utterly destroying [haram]" populations "with the sword," is the same as that used hyperbolically in Joshua. Further, the language of "do not spare" the Amalekites but "put to death both man and woman, child and infant, ox and sheep, camel and donkey" appears to be stereotypical language; it is very similar to the hyperbolic description of military defeat in 1 Chron 36:16–1, where Jewish defeat and exile to Babylon were described in similar language. There are also obvious parallels between 1 Samuel 15 and the hyperbolic rhetoric of the Mesha Stele. I discuss the issue of the Amalekites in more detail in Paul Copan and Matthew Flannagan, "Some Reflections on the Ethics of Yahweh Wars," in Old Testament 'Holy War' and Christian Morality: Perspectives and Prospects, ed. Jeremy Evans, Heath Thomas, and Paul Copan (Downers Grove, IL: InterVarsity Press Academic, 2012).
60 Wolterstorff, "Reading Joshua," 251.
61 Jacob Milgrom, Numbers: The Traditional Hebrew Text with the New JPS Translation (Philadelphia: Jewish Publication Society of America, 1990), 490–91.
62 Ibid., 490.
63 Ibid.
64 Daniel M. Fouts, "A Defense of the Hyperbolic Interpretation of Numbers in the Old Testament," Journal of the Evangelical Theological Society 40/3 (1997): 377–87.
65 Personal correspondence between Nicholas Wolterstorff and Paul Copan (May 2, 2011).
66 One again needs to remember the dialectical context here. The skeptic who claims that God commanded genocide is offering a reductio ad absurdum; she starts by assuming that whatever God commands is right and that Scripture is the Word of God, she then derives from these assumptions the absurd conclusion that genocide is not wrong. The question then is whether, granting these assumptions, such a conclusion does, in fact, follow; hence, one is quite entitled, in this context, to assume the author was an intelligent person.
6767 In Divine Discourse: Philosophical Reflections on the Claim That God Speaks (Cambridge: Cambridge University Press, 1995), Wolterstorff defends an account of how the Bible constitutes God's Word in terms of appropriation. On page 54 he writes, "All that is necessary for the whole [Bible] to be God's book is that the human discourse it contains have been appropriated by God, as one single book, for God's discourse." An implication of Wolterstorff's position is that when one reads the Bible as the Word of God one must read the canon as appropriated discourse. Citing as an example an imaginary case where he appropriates the writings of another person, Wolterstorff elucidates the principles of interpreting appropriated discourse on page 205:
The fundamental principle, I submit, is this: the interpreter takes the stance and content of my appropriating discourse to be that of your appropriated discourse, unless there is good reason to do otherwise—such "good reason to do otherwise" consisting, at bottom, of its being improbable, on the evidence available, that by my appropriation in this situation, I would have wanted to say that and only that. At those points where the interpreter does have good reason to do otherwise, he proceeds by selecting the illocutionary stance and content which have the highest probability of being what I intended to say in this way. If the most probable of those is nonetheless improbable, then he adopts some such fall-back option as that I didn't really appropriate the discourse but only appeared to do so, that in appropriating it I said something I never intended to say, that I misunderstood the discourse I appropriated—or that he has misunderstood the appropriated discourse.
Wolterstorff suggests the same principle applies when God appropriates the "discourse by inscription" of another human being except that some of the fall-back options are excluded. "God does not unwittingly say things God never intended to say, nor does God misunderstand the discourse God appropriates!" The upshot of this is that one cannot just offer a literal reading of the relevant passages and conclude from this that the Bible is not the Word of a just and loving God. To conclude God did not appropriate the text the critic needs to contend that, given the evidence available, the most probable of all alternative interpretations is that God is literally affirming that He commanded genocide and all other alternatives are improbable so that the only other alternative is the fall-back option that God did not really appropriate the text. Hence the critic needs to argue that the literal reading is the most plausible of all available readings and also that it is more plausible than claiming that the critic has misunderstood the appropriated discourse.
68 See n60.
69 Compare 2 Chron 36:17 with 36:20 and also 36:18 with 36:19.
70 See n45.