Child Sacrifice and the Darker Cult

"They have built the high places of Topheth, which is in the valley of the son of Hinnom, to burn their sons and their daughters in the fire, which I didn't command, nor did it come into my mind." — Jeremiah 7:31¹

"There is, however, no historical evidence outside biblical polemic for child sacrifice at the high places." — Mark S. Smith, historian of Israelite religion, 1990²

Summary

Critics say the Hebrew Bible hides an older religion in which Israelites sacrificed children to Yahweh, and that some of its texts still command it.  This article tests that claim text by text.

1.  It happened.  The Bible itself says that Ahaz and Manasseh, kings of Judah, made their sons "pass through the fire," that the northern kingdom did the same, and that families burned children at the Topheth in the Hinnom valley outside Jerusalem until the exile.  The prophets say "burn," "slain" and "shed innocent blood," so the children died.  The rabbinic and modern readings of the rite as a harmless "dedication" cannot explain those words.

2.  Molech.  Scholars split on whether Molech was a god (John Day, George Heider) or the name of a kind of sacrifice, as the Punic word mlk is at Carthage (Otto Eissfeldt, Heath Dewrell).  Both sides agree that children were killed.  The choice matters only for whether the offerings could have been made to Yahweh.

3.  Carthage.  The infant urns of Carthage's tophet, its vow stones, its lamb substitutes and three Greek and Roman writers show that a Phoenician people sacrificed infants for centuries.  A 2010 bone study argued that most were stillborn or newborn;  the excavators' team found most had lived one to one and a half months.  Even the skeptical study allows some sacrifice.  But no tophet has been found in Phoenicia, Israel or Judah, and the old "proofs" from Canaanite Gezer and Egyptian reliefs have collapsed.

4.  Did anyone think Yahweh wanted it?  Yes, some did.  Jeremiah denies three times that God commanded it, which shows someone said he had.  Exodus 22:29, "You shall give the firstborn of your sons to me," lacks the redemption clause that Exodus 13 and 34 supply.  Ezekiel 20:25-26 has God give "statutes that were not good" and defile Israel through their firstborn.  The Hebrew there does not say "fire," and every reading makes the laws a punishment, but no reading makes the verse comfortable.

5.  Three stories.  Jephthah probably killed his daughter, as every ancient reader thought;  the reading that she was only dedicated to God is medieval, found in David Kimhi about AD 1200.  But God never asks for her, and Judges presents the age as one when "everyone did that which was right in his own eyes."  In Genesis 22 God tests Abraham and stops the knife.  Mesha of Moab sacrificed his heir, a foreign king's act that Kings reports without approval.

6.  The ban.  Ḥērem, devoting enemies to destruction, has a vow-and-offering strand (Numbers 21:2, the Mesha Stone), but Susan Niditch, who argues for that strand, calls justice the Bible's "dominant voice," and Philip Stern denies it is sacrifice at all.  It is not parents offering their children.

7.  Divination.  Necromancy, teraphim and sorcery appear among Israelites too, alongside the approved Urim and prophets, and the law condemns them in the same breath as child sacrifice.

Verdict.  Some Israelites, including kings, sacrificed children, and some believed Yahweh approved.  The Hebrew Bible records this in order to condemn it, and no law, prophet or story in it presents child sacrifice as something God wants.  An honest defender must concede the practice, the belief, the missing clause in Exodus 22, Jephthah's daughter and the hardness of Ezekiel 20:25.  The evidence does not show that the Bible approves child sacrifice, that it was normal Israelite religion, or that archaeology has found it in Israel.

Part 1: What the Bible itself says happened

The Hebrew Bible does not hide child sacrifice in Israel.  It reports it in three kinds of text:  the history of the kings, the prophets, and the law.  It never reports it with approval.

The kings

Kings names two kings of Judah who did it.  Of Ahaz, in the late eighth century BC:  "he walked in the way of the kings of Israel, and even made his son to pass through the fire, according to the abominations of the nations whom Yahweh cast out from before the children of Israel."¹  Chronicles tells the same reign in plainer words:  he "burned his children in the fire."¹  Of Manasseh, in the early seventh century:  "He made his son to pass through the fire, practiced sorcery, used enchantments, and dealt with those who had familiar spirits and with wizards."¹  Kings also charges the whole northern kingdom before its fall in 722 BC:  "They caused their sons and their daughters to pass through the fire."¹

Josiah's reform, about 622 BC, wrecked the place where it was done:  "He defiled Topheth, which is in the valley of the children of Hinnom, that no man might make his son or his daughter to pass through the fire to Molech."¹  So Kings knows a fixed site, a named recipient, and both sons and daughters as victims.

The prophets

Jeremiah and Ezekiel, writing in the decades around Jerusalem's fall in 586 BC, describe the practice as the people's, not only the kings'.  Jeremiah says three times that it continued at the Topheth:

1. "They have built the high places of Topheth, which is in the valley of the son of Hinnom, to burn their sons and their daughters in the fire, which I didn't command, nor did it come into my mind."¹
2. They "have built the high places of Baal to burn their children in the fire for burnt offerings to Baal, which I didn't command, nor speak, which didn't even enter into my mind."¹
3. "They built the high places of Baal, which are in the valley of the son of Hinnom, to cause their sons and their daughters to pass through fire to Molech, which I didn't command them."¹

Ezekiel, a priest in exile, uses the words of slaughter.  "Moreover you have taken your sons and your daughters, whom you have borne to me, and you have sacrificed these to them to be devoured."  He asks whether it was a small matter "that you have slain my children and delivered them up, in causing them to pass through the fire to them?"¹  He adds that "when they had slain their children to their idols, then they came the same day into my sanctuary to profane it."¹  In 591 BC he tells the elders of the exile that they still "make your sons pass through the fire ... to this day," though the words about sons and fire are missing from the Old Greek translation and may be a later scribe's gloss.³¹  Isaiah 57:5 speaks of those "who kill the children in the valleys,"¹ and Psalm 106 looks back on a people who "sacrificed their sons and their daughters to demons" and "shed innocent blood."¹

The law

The law forbids it in the strongest terms it has.  "You shall not give any of your children as a sacrifice to Molech.  You shall not profane the name of your God."¹  Leviticus 20 orders the offender stoned and has Yahweh say he will cut him off, "because he has given of his offspring to Molech, to defile my sanctuary, and to profane my holy name."¹  That clause pictures people who offered to Molech and also came to Yahweh's sanctuary, the same picture Ezekiel paints.  Deuteronomy calls it the worst of the nations' practices:  "they even burn their sons and their daughters in the fire to their gods."¹  It bans anyone "who makes his son or his daughter to pass through the fire," in the same list as diviners and necromancers.¹

What "pass through the fire" means

The usual phrase is "make pass through the fire" (Hebrew he'ĕbîr bā'ēš).  By itself it could describe a rite of walking through or between flames, and the rabbis read the Molech law that way (Part 2).  But the Bible describes the same acts with plainer words:  "burn" (Jeremiah 7:31, 19:5;  2 Chronicles 28:3), "burnt offerings" (Jeremiah 19:5), "slain" (Ezekiel 16:21, 23:39), "sacrificed ... to be devoured" (Ezekiel 16:20), "shed innocent blood" (Psalm 106:38).  Exodus 13:12-15 uses the same verb for handing over the firstborn of animals to Yahweh, and those animals were sacrificed.  John Day of Oxford, the author of the standard monograph on Molech, concludes that "it is clear from various Old Testament allusions that this phrase denoted actual sacrifice and not something more harmless like running through the fire."⁴

What this part settles

On the Bible's own testimony, at least two kings of Judah offered children by fire;  the practice had a fixed site in the Hinnom valley just outside Jerusalem;  and the prophets describe ordinary families doing it in the last century before the exile.  A defender who accepts the Bible's history has already conceded all of this.  The open questions are whom the children were offered to, whether any Israelite thought Yahweh asked for it, whether archaeology adds anything, and whether the Bible itself ever commands or approves it.

Part 2: Molech: a god, a kind of sacrifice, or a dedication?

The Hebrew word is molek, always with the article or the preposition "to":  lammolek, "to the Molech."  Scholars read it three ways.  The choice matters, because it decides whether the Molech laws describe offerings to another god or a kind of offering that could have been made to Yahweh.

Reading 1: a god named Molech

Until 1935 every scholar took Molech as a god's name.⁴  John Day of Oxford, in his 1989 monograph and again in 2000, and George Heider, in his 1985 The Cult of Molek, still do.⁵  Day gives three reasons.

1. Leviticus 20:5 condemns those who "play the prostitute after" Molech. The Bible often speaks of Israel prostituting itself after other gods, but, Day notes, "we never hear of people playing the harlot after a sacrifice."⁴ The phrase is not reserved for gods, though. The same chapter uses it of mediums, "to play the prostitute after them,"¹ and Judges 8:27 says Israel "played the prostitute" with Gideon's ephod, a cult object.¹ So the verse needs Molech to be something people worshipped, not necessarily a person.
2. The verbs used with Molech, "give," "burn" and "offer up," take "to" plus a god's name elsewhere in the Bible, never "to" plus a kind of sacrifice.⁴
3. A god mlk appears outside the Bible. Two serpent charms from Ugarit name him, tied to Ashtaroth in Transjordan, and two Mesopotamian god lists equate Malik with Nergal, the underworld god. Day concludes that Molech was a Canaanite god of the underworld.⁴ What Ugarit's texts show about the Bible more generally is the subject of Ugaritic and Ancient Near Eastern Parallels.

Reading 2: a kind of sacrifice

In 1935 the German scholar Otto Eissfeldt pointed out that Punic inscriptions from Carthage and its colonies use mlk (molk) for a kind of sacrifice.  On his reading, lammolek means "as a molk-offering," not "to Molech."⁴  The Punic phrases are telling.  Mlk 'dm is "a molk of a human," parallel to mlk 'mr, "a molk of a sheep," and some of the stelae show a priest holding a child.³

Heath Dewrell's 2017 study, the fullest recent one, revives Eissfeldt.  He grants that a god Malik is attested in some Near Eastern texts, but says "there is no evidence whatsoever that this deity was ever associated with child sacrifice," and that Leviticus 20:5, the one verse that seems to require a god, is textually doubtful.³  Mark Smith, another leading critical historian, also separates the two:  Ugaritic mlk "appears to be unrelated to either child sacrifice or the Phoenician sacrificial term mlk."²

Reading 3: a dedication by fire, not a killing

The oldest Jewish legal reading is that the Molech rite did not kill the child.  The Mishnah makes a parent liable only if he "hands over his child to the priests of Molekh and passes the child through the fire."⁶  In the Talmud, Abaye describes the child walked "between the fire on this side and the fire on that side," and Rava likens it to "the leaps of children on Purim" over bonfires.⁶  Rashi, the great medieval commentator, repeats that the priests "lit two large pyres one opposite the other and made the child to pass on foot between the two pyres."⁶

Moshe Weinfeld gave this a modern form in 1972.  He called the biblical evidence for killing ambiguous or "tendentious poetical fancy," and argued that "the children born of cultic prostitution associated with Molech were presumably delivered to the idolatrous priests, even as the offspring of a regular marriage may have been handed over to Molech."³  His main outside evidence was a set of Assyrian contracts that threaten to "burn" an offender's child to a god he read as Adad-milki.  In 1978 Stephen Kaufman showed that the god's name in those texts is Sin, the moon god, which, as Dewrell puts it, "eliminates the key piece of evidence for Weinfeld's thesis."³

Which reading the evidence favors

The dedication reading is the weakest.  It can explain the phrase "pass through the fire" by itself, but not the prophets' "burn," "burnt offerings," "slain" and "shed innocent blood" (Part 1).  Day adds Isaiah 30:33, which borrows the Topheth as a picture of the pyre prepared for Assyria's king:  "This makes sense only on the sacrificial interpretation of the Molech cult but does not agree with the dedicatory view."⁴  Dewrell, who disagrees with Day about almost everything else, agrees here:  "the one thing that we can say with certainty about biblical [molek] is that it involved the sacrifice of children, just like the Punic rite with the same name."³  The rabbinic reading is best taken as later law shaping how an extinct rite was to be prosecuted, not as a memory of what was done.

Between a god and a sacrifice term, the evidence is closer to even.  Day's argument from Leviticus 20:5 is strong if the verse's wording is original, and Dewrell's reply rests on a textual judgment that not all scholars share.  The Punic mlk is a real sacrifice term, but it is attested only in the western colonies, centuries after the Bible's texts and a sea away from Judah (Part 3).

What turns on the choice is the question of Part 4.  If Molech was a god, the Molech laws describe offerings to a rival god, which fits Jeremiah's charge.  If molek was a kind of sacrifice, the laws do not say to whom it was offered, and two verses suggest some offered it to Yahweh:  Leviticus 20:3, where the offender defiles Yahweh's own sanctuary, and Deuteronomy 12:31, "You shall not do so to Yahweh your God."  On either reading, children died.  Whether Israel's God was once worshipped alongside other gods, as the Molech texts might suggest, is a larger question argued in Did Israelite Monotheism Evolve?.

Part 3: What Carthage shows, and what it does not

The best evidence anywhere for child sacrifice in this cultural family comes from Carthage, the Phoenician colony in modern Tunisia, and its sister colonies in Sicily, Sardinia and Malta.  Scholars call their open-air precincts of infant burials "tophets," borrowing the biblical name.

The finds

From about 750 to 146 BC, Carthaginians buried urns of cremated infants and young animals, mostly lambs and kids, in a walled precinct at Salammbô.  Stone markers above the urns record vows to the gods Baal Hammon and Tanit.  Lawrence Stager of Harvard, who dug the site from 1976 to 1979, estimated about 20,000 urns for the fourth and third centuries BC alone.²  That is about 100 urns a year, or two a week, steady use rather than a rare crisis rite.⁷

Three ancient writers describe what the urns held:

1. Diodorus of Sicily, writing about 30 BC, says that when Agathocles of Syracuse besieged Carthage in 310 BC, the Carthaginians blamed their god Cronus (the Greek name they used for Baal Hammon) because they had taken to sacrificing bought children instead of their own. "They selected two hundred of the noblest children and sacrificed them publicly." A bronze statue of Cronus held out its hands "so that each of the children when placed thereon rolled down and fell into a sort of gaping pit filled with fire."⁸
2. Plutarch, about AD 100, says the Carthaginians "offered up their own children, and those who had no children would buy little ones from poor people and cut their throats as if they were so many lambs or young birds," while flutes and drums drowned the cries.⁹
3. A note on Plato's Republic credits a similar account, with the bronze statue and the brazier, to Kleitarchos, a historian of about 300 BC.¹⁰

All three are Greek or Roman writers hostile to Carthage, and their details overlap enough that they may draw on one tradition.  The stones are the Carthaginians' own words, but they speak of vows, not of killing.  Stager and his colleague Joseph Greene concede that of the thousands of inscriptions, "None refers explicitly to child sacrifice, only to vows made to Tanit and Ba'al Hammon."¹¹

The fight over the bones

Since 2010 the case has turned on the infants' ages.  A naturally caused infant cemetery should be weighted toward the stillborn and newborn, since that is when most babies die.  A precinct for children killed to fulfil a vow should show babies who had lived long enough to be offered.

1. Jeffrey Schwartz of the University of Pittsburgh and three colleagues studied 348 urns holding 540 individuals. They found that most "fell within the period prenatal to 5-to-6 postnatal months, with a significant presence of prenates," a pattern "consistent with modern-day data on perinatal mortality." In 50 milk teeth, 24 showed the neonatal line that forms when a baby survives birth by about two weeks, and 26 did not. They concluded that tophets "were cemeteries for those who died shortly before or after birth, regardless of the cause."¹⁰
2. Patricia Smith of the Hebrew University, with Stager and others, measured the same remains with other methods. In 2011 they reported that "the majority of the infants died between one and one and a half months. Sacrifice was thus very probable."¹² Stager summarized the two results in 2014: Schwartz's team has "24% fetuses, 32% died at birth," while Smith's team has "67% in the 1-2 month cohort, and a questionable 0.7% cremated fetuses."¹³

Both teams agree on more than the headlines suggest.  Schwartz's paper accepts that "at least some, if not all, of the cremated animal remains represent sacrificial offerings," and that "the Carthaginians may occasionally have practiced human sacrifice."¹⁰  Stager's side points to what an infant cemetery cannot explain:  lambs in their own urns under the same vow markers, and infants and lambs cremated together in about a quarter of the burnt offerings.¹³  A lamb buried under a vow stone makes sense as a substitute offering.  It makes little sense as a burial.

Weighed together, the classical writers, the vow inscriptions, the animal substitutes and at least one of the two age studies favor real child sacrifice at Carthage.  Even the strongest critics allow that some took place.  What remains open is how often, and the bone evidence alone cannot settle that, because the two teams read the same teeth to different ages.

What Carthage shows about Judah

1. It shows the rite was real in this cultural family. A Phoenician people, speaking a language close to Hebrew, offered infants by fire to a god in a dedicated precinct for six centuries. The Bible's picture of a Topheth outside Jerusalem is not an implausible slander; it matches a known type.
2. It supplies a word. The Punic mlk for a child offering is the main support for Eissfeldt's and Dewrell's reading of molek (Part 2).

It does not show as much as is often claimed.

1. There is no link in between. No tophet has been found in the Phoenician homeland. A supposed tophet at Tyre, found in 1990 by looters, held no infant bones and stelae with only personal names; in Dewrell's words, "the evidence for a tophet at Tyre has ultimately not held up to scholarly scrutiny."³ Carthage lies about 2,350 km from Jerusalem in a straight line, and "there is no reason to believe that Iron Age Judah and Carthage were in regular contact."³
2. Nothing has been dug up in Judah itself. No urns have been excavated from the Hinnom valley or anywhere in Israel or Judah. Dewrell, who believes Judah and Carthage shared the rite, concedes that "there is as yet no unambiguous material evidence from Syria-Palestine itself that corroborates the claims of the Hebrew Bible."³ The case for child sacrifice in Judah rests on the Bible's own testimony (Part 1), which is strong, and not on archaeology.
3. The older "proofs" from Canaan have failed. The infant jar burials that R. A. S. Macalister found near the "high place" at Gezer before World War I were long cited as Canaanite sacrifices. William Dever's re-excavation showed that the burials are older than the standing stones, so they were never offerings at that shrine.³ The Egyptian reliefs of children held over the walls of besieged Canaanite towns are better read, with Othmar Keel, as children handed over to Pharaoh in surrender.³
4. It cannot say to whom Judahites offered, or how often.

Part 4: Did any Israelite think Yahweh wanted it?

This is the hard question.  The Bible condemns child sacrifice offered to other gods.  But three groups of texts have led scholars to argue that some Israelites offered children to Yahweh, believing he asked for them:  the law of the firstborn, Ezekiel 20:25-26, and Jeremiah's repeated denial.

The law of the firstborn

Every version of the law says the firstborn belong to Yahweh.  Most say how to hand over a firstborn son:  redeem him.  "Sanctify to me all the firstborn, whatever opens the womb ... It is mine," says Exodus 13:2, and Exodus 13:13 adds, "you shall redeem all the firstborn of man among your sons."¹  Exodus 34:20 repeats it,¹ Numbers 18:15 orders the priests to "surely redeem the firstborn of man,"¹ and Numbers 3 takes the tribe of Levi "instead of all the firstborn" for the sanctuary's service.¹

One version has no redemption clause.  Exodus 22:29-30:  "You shall give the firstborn of your sons to me.  You shall do likewise with your cattle and with your sheep.  It shall be with its mother seven days, then on the eighth day you shall give it to me."¹

Critical scholars read this as the remains of an older practice.  Dewrell argues that "likewise" ties the sons to the animals, which were sacrificed, and that a child taken from its mother on the eighth day cannot be going to temple service:  "the literal wording of this ideal unambiguously demands child sacrifice."³  Jon Levenson of Harvard, the most influential writer on the theme, reads the verse as "a theological ideal about the special place of the first-born son, an ideal whose realization could range from literal to non-literal implementation, that is, from sacrifice to redemption, or even to mere intellectual assent without any cultic act whatsoever."¹⁴

The Jewish commentator Moshe Greenberg, in the Anchor Bible, answers that the redemption laws state how the claim was met:  "Outside of our passage no evidence for such an interpretation of these laws, or for such a practice, exists; indeed, it is intrinsically improbable."¹⁵  A reader of the Torah as it stands has the same answer:  Exodus 13:13 comes nine chapters before 22:29 in the same book and has already said how the firstborn son is "given."  That is a harmonizing reading, but not a forced one.  A law code that states a claim in one place and its procedure in another is ordinary.

Two arguments are often made against the literal reading.  One fails and one holds.

1. Demography does not rule it out. Roland de Vaux called a general law requiring the firstborn's death "absurd," and William Propp called it "Darwinian suicide." Dewrell answers with numbers: where women bore six or seven children and half died in infancy, the firstborn was one child in four to seven, and the firstborn of young mothers were already the likeliest to die.³ The arithmetic agrees. Losing the first child whenever it is a son removes about one birth in twelve to fourteen, which slows a growing population but does not end it.¹⁶
2. The silence of the narratives holds. Esau, Reuben, Manasseh, Aaron's son Nadab and David's son Amnon are all called firstborn, and none was sacrificed. Dewrell states the objection himself: "If there ever were an Israelite custom that all firstborn children were to be sacrificed on the eighth day, then it is remarkable that not a single biblical narrative alludes to the practice."³ Levenson agrees that a general practice would have left "ample testimony, in both law code and narrative."¹⁴ So Dewrell concludes only that "a small 'ultra-pious' or 'literalist' group of Yahwists" ever did it, and he admits that "it is impossible to say how extensive either practice was."³

Ezekiel 20:25-26

Ezekiel 20 retells Israel's history as rebellion.  God first "gave them my statutes ... which if a man does, he will live in them" (20:11).¹  When they rejected those, and "their eyes were after their fathers' idols" (20:24), God says:  "I gave them statutes that were not good, and ordinances in which they couldn't live.  I polluted them in their own gifts, in that they caused all that opens the womb to pass through the fire, that I might make them desolate, to the end that they might know that I am Yahweh."¹

Levenson notes that the passage "has over the centuries had most exegetes running for cover."¹⁴  There are four main readings.

1. God really gave a law that demanded the firstborn, as punishment. Francesca Stavrakopoulou: "in 20:25-26 it is claimed that YHWH deliberately demanded that his people sacrifice their firstborn children to him in order to punish them."¹⁷ Levenson: "the ultimate cause of the 'laws that were not good' was Israel's rebellion; the proximate cause was divine revelation."¹⁴ Dewrell: Ezekiel "acknowledged that Yahweh commanded that firstborn children be sacrificed as the Covenant Code literally says but avoided the conclusion that this was something that Yahweh actually wanted by transforming the command into a form of divine punishment."³
2. Ezekiel reports a popular belief, and the charge is exaggerated. Greenberg: "our vs. 25 was not spun out of thin air. The polemic against child sacrifice (to YHWH) in Deut 12:29ff.; Jer 7:31; 19:5; 32:35 indicates that at least from the time of the last kings of Judah it was popularly believed that YHWH accepted, perhaps even commanded, it." He calls the claim that Israelites "regularly offered up every firstborn as a sacrifice" "a manifest exaggeration."¹⁵
3. God handed a rebellious people over to their own misreading. This is the oldest reading. The Aramaic Targum already made God punish them "by allowing them to follow their own base instincts," and Levenson grants that it "comes as close to being standard as any position in the history of interpretation ever does."¹⁴ Ronald Hals put it as God hardening Israel's heart, as he hardened Pharaoh's: the people read "All that opens the womb is mine" literally and "ignored the subsequent clarification, 'All the first-born of your sons you shall redeem.'"¹⁴ Ezekiel himself speaks this way elsewhere: "If the prophet is deceived and speaks a word, I, Yahweh, have deceived that prophet."¹ Popular conservative writers take the same line. Paul Copan calls the statutes "sarcastic 'commands' that aren't 'good' and by which Israel can't 'live.'"¹⁸ Greg Koukl writes: "God gave them over to these statutes of death so He could judge them for their evil through their own actions."¹⁹
4. The "not good" laws are Deuteronomy's. Scott Hahn and John Bergsma argue that a priest like Ezekiel saw Deuteronomy's relaxed rules on firstling animals as defiling, and that verse 26 is "an Ezekielian polemic against the Deuteronomic code," not a reference to child sacrifice at all.²⁰

How much Ezekiel 20:25 resists a comfortable reading

Some of it can be answered.  The Hebrew of verse 26 does not say "through the fire."  It says, in Dewrell's translation, "in causing every breach of a womb to pass over," the wording of the firstborn law in Exodus 13:12, where the same verb covers sons who are then redeemed.³  "Fire" is added by translators from verse 31, and in verse 31 the words about sons and fire are missing from the Old Greek (Part 1).  Greenberg calls the merging of the two phrases "unique to our passage."¹⁵  So the verse does not plainly say that God ordered children burned.

The rest cannot be explained away.  God is the speaker, and he says "I gave them."  Copan's "sarcasm" and Koukl's "idols' statutes" do not fit those words;  Levenson's objection to the hardening reading stands:  "Where else are God's laws ever seen in such a light?"¹⁴  Every reading must accept that Ezekiel attributes to God laws that "were not good," given as punishment, which led his people into a defiling practice connected with their firstborn.  The hardening reading is the best a defender has, and it is a real reading with roots in Ezekiel 14:9, 1 Kings 22 and Exodus.  But it is an inference from parallels, not what verse 25 says.  An honest defender should say that this verse is hard, and that the comfortable readings are the weakest ones.

"Which I did not command"

Jeremiah says three times that Yahweh "didn't command, nor did it come into my mind" (Part 1).  Read one way, this is plain condemnation.  Day points out that the same phrase in Deuteronomy 17:3, about worshipping the sun and moon, means "which I forbade," and most translations render it so there.⁴

Read another way, the denial is evidence that someone claimed the opposite.  Levenson:  "The threefold denial of the origin of the practice in YHWH's will ... suggests that the prophet doth protest too much."¹⁴  Smith:  "these denials may suggest that offering this sacrifice in Yahweh's name occurred."²  Dewrell adds that Jeremiah's words contradict Ezekiel's:  "Although Ezekiel says that Yahweh did give such a law (albeit a 'not good' one), Jeremiah adamantly denies that he did."³

Both readings can be true at once.  A prophet who insists three times that God never wanted something is likely answering people who said he did.  Greenberg, who rejects the critical reconstruction, accepts exactly this.  The fair conclusion is that by the seventh century some Judahites believed Yahweh accepted or asked for child sacrifice, and that every biblical writer who mentions the belief rejects it.

Micah 6:7

"Shall I give my firstborn for my disobedience?  The fruit of my body for the sin of my soul?"¹  Micah's worshipper asks whether ever greater gifts will please God:  calves, rams, oil, and finally a child.  The answer follows at once:  "He has shown you, O man, what is good.  What does Yahweh require of you, but to act justly, to love mercy, and to walk humbly with your God?"¹  The question shows that offering a firstborn in desperation was thinkable to Micah's hearers.  The answer rejects it.

Part 5: Three stories: Jephthah, Isaac and Mesha

Three narratives put a child on an altar.  None of them is a law, and in none does God command the killing of a child and let it stand.  But each is used, by critics and defenders, to show what Israelites thought God might accept.

Jephthah's daughter

"Then Yahweh's Spirit came on Jephthah."  Marching against the Ammonites, he vowed that "whatever comes out of the doors of my house to meet me ... shall be Yahweh's, and I will offer it up for a burnt offering."¹  His daughter, "his only child," came out dancing.  He tore his clothes and said, "I have opened my mouth to Yahweh, and I can't go back."  She asked for two months to "bewail my virginity" on the mountains.  "At the end of two months, she returned to her father, who did with her according to his vow which he had vowed.  She was a virgin.  It became a custom in Israel that the daughters of Israel went yearly to celebrate the daughter of Jephthah the Gileadite four days in a year."¹

The ancient readers all say she died.  Josephus, about AD 93:  "when that time was over, he sacrificed his daughter as a burnt-offering, offering such an oblation as was neither conformable to the law nor acceptable to God."²¹  In Pseudo-Philo's Biblical Antiquities, from the first century AD, God is angry at the careless vow:  "if a dog should meet Jephthah first, will the dog be offered to me?"  The daughter, whom the book names Seila, goes to her death, and "he did everything that he had vowed and offered the holocausts."²²  The rabbis agree.  Genesis Rabbah says "that young woman was lost" between Jephthah's pride and the high priest Phinehas's, and "both of them were punished for the blood of the young woman."²³  The Talmud takes the words "And I did not speak" in Jeremiah 19:5 as "referring to Jephthah, who sacrificed his daughter as an offering."²⁴  The rabbis argued over whether the vow had ever bound him, since a person cannot be a burnt offering and the law allowed a vowed person to be redeemed with money (Leviticus 27:2-8).  None of them said she lived.

The dedication reading is medieval.  David Kimhi, about AD 1200, credited his father with reading the "and" in the vow as "or":  whatever came out would either belong to Yahweh or be offered up, so a human would be dedicated and an animal burned.  He argued that she wept for her virginity, not her life, and that verse 39 says "he did to her as he had vowed," not "he offered her as a burnt offering":  "This shows us that she was celibate/separated, and this is what he had vowed -- that she should be for God."²⁵  Gersonides followed him.²⁶  Modern defenders of the reading include Rabbi Jonathan Magonet, former principal of Leo Baeck College in London, who adds that the yearly custom may mean the women went "to" speak with her rather than to lament "about" her.²⁷

Most scholars, critical and conservative, read a sacrifice.  Dewrell:  "nearly all agree that the most natural reading of the narrative is that Jephthah in fact vowed to sacrifice the first thing coming out to meet him, that he knew it was possible or even probable that this would be a human being, and that Jephthah literally sacrificed his daughter as a result."³  Copan, an evangelical apologist, grants:  "Some Old Testament scholars argue that Jephthah didn't literally sacrifice his daughter.  Most, however, are convinced that the text asserts this."¹⁸  At Answers in Genesis, Tim Chaffey argues the same:  "I argued that I believe Jephthah did actually sacrifice his daughter to fulfill his rash vow."²⁸  The weight of the evidence is with them.  "Burnt offering" means burnt offering, the vow was made before anything came out of the door, and the earliest readers all took the story as a death.

But the story does not approve it.  The Spirit came on Jephthah before the vow, not as its reward;  the vow was his own idea.  God never asks for the girl and never accepts her.  The book of Judges presents its heroes as a declining series, and closes with the refrain:  "In those days there was no king in Israel.  Everyone did that which was right in his own eyes."¹  Copan answers the obvious objection:  the Spirit's coming on Jephthah is not "a wholesale divine endorsement of all Jephthah did—no more so than the Spirit's coming on Gideon (Judg. 6:34) was a seal of approval on his dabbling with idolatry."¹⁸  The law Jephthah ignored already offered a way out:  a vowed person could be redeemed at a set price.¹  Josephus and the rabbis drew the same lesson:  he did what God did not want.

The binding of Isaac

"God tested Abraham, and said ... 'Now take your son, your only son, Isaac, whom you love, ... Offer him there as a burnt offering.'"  At the last moment the angel stops him:  "Don't lay your hand on the boy or do anything to him.  For now I know that you fear God."  A ram is offered "instead of his son."¹

This is the one place where God himself commands a child sacrifice.  The narrator frames it as a test from the first verse, and God stops it.  Many readers, ancient and modern, take the story as written against child sacrifice:  the God of Israel, unlike the gods of Canaan, provides a substitute.  Levenson disagrees sharply:  "The cumulative evidence against the ubiquitous idea that the aqedah opposes child sacrifice and substitutes an animal cult is overwhelming:  nothing in Gen 22:1-19 suggests that God's command to immolate Isaac was improper."  The most he will grant is that "the story reflects a situation in which the sacrifice of a sheep as a substitute for the favored son can meet with God's approval."¹⁴

Nahum Sarna, in the Jewish Publication Society commentary, also rejects the polemic reading, but for the opposite reason.  He sees no sign that child sacrifice was ever normal in Genesis, where animals are the ordinary offering from Cain and Abel on:  "Such an understanding of the narrative cannot be supported either by history or by biblical tradition."²⁹  The evangelical Gordon Wenham reads the test as working only because the act was unthinkable:  "to offer one's child was quite out of the question for devout orthodox worshipers," though "it was done occasionally in the biblical world, especially in times of dire crisis."³⁰

Levenson is right that the text never calls the command wrong.  But the story's outcome is plain:  God stops the knife, and the only thing burned is an animal.  A story in which God demands a son and then refuses him cannot be evidence that Israel's God wanted children killed.  It shows that the idea of giving up one's dearest was in Israel's religious world, and that this story resolves it by substitution.  The New Testament reads it as faith in resurrection:  Abraham was "concluding that God is able to raise up even from the dead."¹

Mesha's son

When an Israelite, Judahite and Edomite army had Mesha, king of Moab, trapped in his last city, "he took his oldest son who would have reigned in his place, and offered him for a burnt offering on the wall.  There was great wrath against Israel; and they departed from him, and returned to their own land."¹

Every ancient reader took this as a real sacrifice.  Josephus says Mesha "offered him as a whole burnt-offering to God," and that the allied kings, moved by pity, lifted the siege.²¹  The Talmud records Rav and Shmuel disputing only whether he offered his son "for the sake of Heaven" or "for the sake of idol worship."³¹  Whose "wrath" fell on Israel is unclear.  Dewrell notes that elsewhere in the Bible "a great wrath came upon" always means divine anger, so the narrator seems to say the sacrifice worked.³  Greenberg, from the other side of the debate, agrees that offering a firstborn "as a token of extraordinary devotion in emergencies appears to have been widely held (II Kings 3:27; Micah 6:7)."¹⁵

Two things follow, and one does not.  The story shows that a neighbouring king sacrificed his heir in a crisis, and that an Israelite writer could report it without surprise.  It fits Kings' charge that Ahaz and Manasseh followed "the abominations of the nations."  It does not show that Israel's God accepted such an offering;  the victim was Moab's, offered by Moab's king, presumably to Moab's god Chemosh.

Mesha's own inscription, the Moabite Stone from his capital at Dhiban, says nothing about his son.  It does describe the ban:  "And Kemosh said to me, Go take Nebo against Israel.  And I went by night and fought against it from the break of dawn till the noontide, and I took it and slew all ... seven thousand m[en] and ... and women and ... and damsels, for I had devoted it to 'Ashtar-kemosh."³²  The verb is ḥrm, the same as Hebrew ḥērem.  That brings the article to the ban.

Part 6: Is the ban (ḥērem) a human sacrifice?

The ban is the Bible's order to "utterly destroy" certain enemies, "devoting" them to Yahweh.  Of the cities of Canaan, Deuteronomy says, "you shall save alive nothing that breathes ... that they not teach you to follow all their abominations, which they have done for their gods."¹  Jericho was "devoted ... to Yahweh," and Israel destroyed "both man and woman, both young and old."¹  Saul was told to kill the Amalekites, "infant and nursing baby."¹  Leviticus puts devoted people in the same class as devoted property:  "No one devoted to destruction, who shall be devoted from among men, shall be ransomed.  He shall surely be put to death."¹  And Numbers tells of a ban vowed in exchange for victory:  "If you will indeed deliver this people into my hand, then I will utterly destroy their cities."¹

Whether the ban is right is the subject of The Morality of the Conquest of Canaan, which argues that much of the ban language is stock war rhetoric, and the archaeology of the conquest is in Jericho, Ai, and the Conquest Narratives.  The question here is narrower:  is the ban a kind of human sacrifice, so that the Bible commands sacrificing people after all?

The case that it is

Susan Niditch of Amherst College makes it in War in the Hebrew Bible (1993).  She reads one strand of the ban, in Numbers 21 and the Mesha Stone, as a vow:  "Israel is promising a sacrifice to God, the cities and their content.  So the Moabite king had promised his Israelite enemies to his deity."³³  Of soldiers who vow to kill an enemy in exchange for victory, she writes, "They are offering human sacrifices to the deity," though the enemy was "usually imagined to be slaughtered by sword in the denouement of battle and not prepared Aztec-style for a separate sacrifical ritual."³³  Leviticus 27:29 supports her:  a devoted person cannot be ransomed, just as a devoted field or animal cannot.³³

But Niditch also finds that this is not the Bible's main view:  "In fact, the dominant voice in the Hebrew Bible treats the ban not as sacrifice in exchange for victory but as just and deserved punishment for idolators, sinners, and those who lead Israel astray or commit direct injustice against Israel."  On the same page:  "The dominant voice in the Hebrew Bible condemns child sacrifice as the epitome of anti-Yahwist and anti-social behavior."³³

The case that it is not

Philip Stern, in The Biblical Herem (1991), reads the ban as war against chaos:  "the struggle of a people to overcome the forces of chaos and to establish a world order."³⁴  He denies it is a sacrifice on a point of definition:  "a sacrifice is a free-will offering.  The categories may be fixed as to what is acceptable (cow, sheep, bird), but within the categories, the sacrificer may choose which animal or bird to offer, an element lacking in the [herem]."  He grants that the vow in Numbers 21 has the structure of an exchange, a ban vowed "in order to receive in return the aid of YHWH against the enemy," but concludes that "the [herem] is clearly not one of them," meaning the sacrifices.³⁴  On Samuel's killing of King Agag "before Yahweh," which some read as a sacrifice, he answers that "1 Sam 15:20 explicitly delimits the sacrifice to the sheep and cattle."³⁴  Norbert Lohfink, in the Theological Dictionary of the Old Testament, grants that "A few texts appear to associate hrm with sacrificial terminology," and reads the story of Saul and Agag as "deliberate polemic against any attempt to make hérem a sacrificial term."³⁵

What the evidence supports

The ban shares some features with sacrifice.  It gives something to God, it can be vowed for help, and what is devoted cannot be bought back.  It lacks others.  There is no altar, no parent offering his own, and no burning of people as an offering;  the enemy dies in battle or by execution, and the Bible's stated reasons are judgment and keeping Israel from the nations' worship (Deuteronomy 9:4, 20:18).¹  Niditch herself says the sacrifice reading is a minority strand, overlaid by the justice reading.

So the ban is not child sacrifice in the sense this article is testing:  parents killing their own children to win a god's favor.  It does kill children, and that is a real moral problem, but it is the problem of the conquest, not of the Molech cult.  A skeptic who calls the ban "human sacrifice" is borrowing Niditch's minority strand and dropping her own qualification.  A defender who says the ban has nothing to do with sacrifice is ignoring Numbers 21:2, Leviticus 27:29 and the Mesha Stone.

Part 7: Divination and the darker cult

Child sacrifice heads a longer list.  Deuteronomy 18:10-11 bans, in one breath, "anyone who makes his son or his daughter to pass through the fire, one who uses divination, one who tells fortunes, or an enchanter, or a sorcerer, or a charmer, or someone who consults with a familiar spirit, or a wizard, or a necromancer."¹  Kings charges Manasseh with the same cluster.¹  The Bible shows Israelites doing all of these, and it sorts them into allowed and forbidden ways of asking God.

1. Allowed: the Urim and Thummim, prophets and dreams. The high priest carried the Urim and Thummim in his breastpiece "when he goes in before Yahweh," and Joshua was to "inquire for him by the judgment of the Urim."¹ When God would not answer Saul, he did not answer "by dreams, by Urim, or by prophets."¹ These were the sanctioned channels.
2. Forbidden: the dead. Saul himself "had sent away those who had familiar spirits and the wizards out of the land." Then, desperate, he went to the medium at Endor and asked her to "bring Samuel up." She cried out, "I see a god coming up out of the earth," and Samuel spoke.¹ The story does not deny that the dead could be reached. It condemns reaching them: Chronicles says Saul died "because he asked counsel of one who had a familiar spirit, to inquire, and didn't inquire of Yahweh."¹ Leviticus makes the medium liable to stoning,¹ and Isaiah asks, "Should they consult the dead on behalf of the living?"¹ Mark Smith argues that the ban came late: "Condemnation of necromancy is not recorded for any prophet before Isaiah (8:19; cf. 19:3; 29:4; cf. 57:6) or any legal code before the Holiness Code," and the note in 1 Samuel 28:3 that Saul had banned mediums "may have been an editorial addition."² That argument depends on dating Leviticus and Deuteronomy late, which Deuteronomy and the Pious-Fraud Charge examines. What Israel believed about the dead is the subject of The Late Development of Afterlife and Satan.
3. Disputed: the teraphim. Teraphim were household images. Rachel stole her father's and hid them in a camel saddle (Genesis 31:19, 34).¹ David's wife Michal put "the teraphim" in his bed to fool Saul's men (1 Samuel 19:13).¹ Micah of Ephraim kept them with an ephod in his "house of gods" (Judges 17:5).¹ Hosea lists them, with the ephod, among the things Israel will "live many days without."¹ Yet Samuel ranks teraphim with witchcraft,¹ Josiah removed them with the mediums,¹ and Zechariah says "the teraphim have spoken vanity."¹ The texts agree that teraphim were consulted for guidance, as the king of Babylon did at the crossroads.¹ They show a household practice that some Israelites kept for centuries and that the prophets and reformers rejected.
4. Archaeology. Household shrines, figurines, standing stones and the Yahweh temple at Arad belong to Israelite Religion on the Ground, which finds that the excavated religion of Iron Age Israel and Judah matches the Bible's own picture of a people who often did not keep the religion its prophets demanded.

The pattern is the same as with child sacrifice.  The Bible reports these practices among Israelites, including kings, and the law and the prophets condemn them.  It never presents necromancy, sorcery or child sacrifice as approved worship of Yahweh.

What must be conceded, and what the evidence does not show

What an honest defender must concede

1. Child sacrifice was practiced in Judah and Israel. The Bible says so of at least two kings of Judah, of the northern kingdom, and of ordinary people at the Topheth in the generation before the exile. The writers had every reason to blame foreigners and instead blamed their own kings and people.
2. The children died. The prophets say "burn," "slain" and "shed innocent blood." The dedication reading of the rabbis and of Weinfeld cannot account for those words.
3. Some Judahites believed Yahweh accepted or wanted it. Jeremiah's threefold "which I did not command," Leviticus 20:3's offerer who defiles Yahweh's sanctuary, and Deuteronomy 12:31's "You shall not do so to Yahweh your God" make sense only if someone claimed otherwise. Greenberg, no friend of the critical reconstruction, grants that "it was popularly believed that YHWH accepted, perhaps even commanded, it."¹⁵
4. Exodus 22:29 has no redemption clause. The canonical answer, that Exodus 13 and 34 supply it, is reasonable, but it is an answer, not something the verse says.
5. Ezekiel 20:25-26 is hard. God says he gave laws "that were not good" and defiled his people through their offering of the firstborn. The verse does not say God ordered children burned, but no reading makes it comfortable.
6. Jephthah probably killed his daughter. The medieval dedication reading is possible but is not the natural one, and every ancient reader took it as a death.
7. The ban sometimes looks like a vowed offering. Numbers 21:2, Leviticus 27:29 and the Mesha Stone share a vow-and-devotion pattern, even if the ban is not a sacrifice in the strict sense.
8. Some popular apologetics are wrong. Halley's handbook said that at Gezer "Macalister found great numbers of jars containing the remains of children who had been sacrificed to Baal," and that "The whole area proved to be a cemetery for new-born babes."³⁶ Dever's re-excavation removed that evidence (Part 3). Apologists who cite the Schwartz study's press release as if it were the paper misquote it.³⁷

What the evidence does not show

1. That the Bible commands or approves child sacrifice. Every law that mentions it forbids it. Every narrative that reports it, by Israelite or foreigner, either condemns it or reports it without approval. The one command, to Abraham, is framed as a test and stopped. Even Stavrakopoulou's reading of Ezekiel 20 makes the "not good" laws a punishment, which is not approval.
2. That child sacrifice was normal Israelite religion. The critics' own scholars limit their claims. Smith: "There is, however, no historical evidence outside biblical polemic for child sacrifice at the high places," and "child sacrifice may not have been part of traditional religion; the biblical and inscriptional records do not indicate how widespread the practice was."² Dewrell: "probably far from most" sacrificed their firstborn, and "we have almost no information concerning how widespread any of these rites were."³
3. That archaeology confirms it in Israel. No tophet or burned-infant precinct has been excavated in Israel, Judah or Phoenicia. The evidence is textual.
4. That the prophets invented it. The argument that the biblical charges are "unreliable polemic" has no evidence behind it. Day: "no concrete evidence has been cited to support this supposition."⁴ Carthage shows the rite existed in the Phoenician world.

The strongest case each way, in its advocates' words

For Yahwistic child sacrifice:  Levenson argues that the gift of the firstborn son was an old ideal of Israel's own religion, that Jeremiah "doth protest too much," and that Ezekiel 20 shows the prophet who knew such laws had to make them a punishment.  Stavrakopoulou:  the Bible "offers, both implicitly and explicitly, a vivid portrayal of YHWH as a god of child sacrifice."¹⁷  Dewrell:  "there were at least three different rites that involved the sacrifice of children to Yahweh in Iron Age Israel."³

Against it:  Greenberg:  outside Ezekiel 20 there is "no evidence for such an interpretation of these laws, or for such a practice ... indeed, it is intrinsically improbable."  Day:  the Molech rites were held in the Hinnom valley, not the temple, and took daughters as well as sons, while Yahweh's claim was on firstborn sons only, so "These considerations militate against those who would equate the offerings to Molech and Yahweh."⁴  And the narratives name many firstborn sons, none of them sacrificed.

The balance:  some Judahites, including kings, sacrificed children, and some of them believed Yahweh approved.  The Hebrew Bible records them in order to condemn them.  Its laws, prophets and narratives agree that the God of Israel did not want it.  The one text that resists, Ezekiel 20:25-26, makes even the bad laws a judgment, not a desire.

Objections and responses

1. "The Bible commands child sacrifice:  'You shall give the firstborn of your sons to me' (Exodus 22:29)."  The same book has already said how a firstborn son is given:  "you shall redeem all the firstborn of man among your sons" (Exodus 13:13), and repeats it at 34:20.  Critics read Exodus 22 as an older law from a different source, and on their dating that is possible.  But even Dewrell, who reads Exodus 22 literally, concludes that only a small "literalist" minority ever acted on it, and the narratives name many firstborn sons who lived (Part 4).

2. "Ezekiel 20:25 admits that God gave Israel child-sacrifice laws."  It says God gave "statutes that were not good" as punishment, and that he defiled them through their offering of the firstborn.  The Hebrew of verse 26 does not say "through the fire";  it uses the firstborn law's own phrase.  The verse is hard, and this article does not pretend otherwise (Part 4).  But on every reading, including Stavrakopoulou's, the laws are a judgment, not something God wanted.  Jeremiah, Ezekiel's contemporary, says flatly that God never commanded it.

3. "Jephthah is praised in Hebrews 11, so the Bible approves his sacrifice."  Hebrews 11:32 lists him with Gideon, Barak and Samson, men whose faith is praised and whose failures Judges records at length.¹  Judges never says God accepted the vow, and Josephus and the rabbis called the sacrifice unlawful.

4. "Abraham's willingness shows that Israel's God wanted children."  The story calls it a test from its first verse, God stops the knife, and an animal is burned instead.  Levenson is right that the text never calls the command wrong.  A story in which God refuses the child cannot show that God wanted the child.

5. "Molech was just Yahweh under another name."  The god-or-sacrifice question is open (Part 2), and if molek is a sacrifice term some Judahites may have offered it to Yahweh.  That would show what some worshippers did, which the Bible already admits.  It would not show what the Bible approves, since every Molech text condemns the rite.

6. "Archaeologists found jars of sacrificed babies under Canaanite high places, proving the Canaanites deserved judgment."  Not at Gezer, Megiddo or Taanach;  the Gezer burials predate the high place (Part 3).  Defenders should drop this argument.  The Bible's testimony to Canaanite child sacrifice (Deuteronomy 12:31, Psalm 106:38) stands on its own, Carthage shows the rite in a Phoenician people, and Day sees "no reason to doubt the biblical testimony to Canaanite child sacrifice."⁴  But the archaeological proof from Canaan itself is gone.

7. "Schwartz's 2010 study proved the Carthaginians never sacrificed children."  Schwartz and his colleagues said only that their data "do not support the contention that all humans interred in the Tophet had been sacrificed," and allowed that "the Carthaginians may occasionally have practiced human sacrifice."¹⁰  The animal urns, vow stelae and ancient writers remain (Part 3).

8. "The prophets made up child sacrifice to smear their opponents."  No evidence supports this, and Carthage shows the rite was real in the same cultural world.  The prophets also blame their own kings and people, which is not how a smear against foreigners works.

9. "If the Bible condemns child sacrifice, why does the ban kill children?"  Because the ban is a different act with a different stated purpose:  judgment on the nations and keeping Israel from their worship, not parents offering their own children for favor.  Whether that judgment was just, and how much of its language is war rhetoric, is argued in The Morality of the Conquest of Canaan.

10. "Child sacrifice was a normal part of Judahite Yahwism, which the Bible's editors later disowned."  That is Stavrakopoulou's thesis.  The scholars who have studied the evidence most closely do not go that far.  Smith says there is no evidence outside biblical polemic and that the scale is unknown;  Dewrell thinks the firstborn rite was a minority practice and the Molech rite a late import under Ahaz.  "Some" is supported.  "Normal" is not.

11. From the other side:  "'Pass through the fire' meant only a dedication rite, as the rabbis said, so no children died."  The prophets say "burn," "slain" and "shed innocent blood."  Weinfeld's outside evidence turned out to name the moon god Sin (Part 2).  This defence fails.

12. From the other side:  "Jephthah only dedicated his daughter to a life of virginity."  Possible, but not the natural reading.  The vow says "burnt offering," the earliest readers all took it as a death, and the dedication reading is medieval:  Kimhi, about AD 1200, credits it to his father, and a modern writer credits Ibn Ezra a generation earlier.  An evangelical such as Copan grants the sacrifice and argues instead that the story does not approve it, which is the stronger ground.

Issues and unanswered questions

1. The Topheth has never been excavated. Nothing is known from archaeology about the scale of child sacrifice in Jerusalem. If urns were found in the Hinnom valley, they would settle how often and how young.
2. Ezekiel 20:25-26 has no fully satisfying reading. The hardening reading is the best a defender has, and it rests on parallels, not on the words of verse 25.
3. Leviticus 20:5 decides the god-or-sacrifice question. Day's case for a god rests on its wording, and Dewrell's case against rests on its textual history. Neither has persuaded the other side.
4. The relative dates of the firstborn laws are disputed. The critical reading of Exodus 22 as the oldest law depends on source criticism. The canonical reading, that Exodus 13 and 34 explain it, is a choice the text allows but does not force.
5. The Carthage age data have not been independently replicated. A blind count of neonatal lines by a third team on the same teeth would show which age profile is right. The 2011 and 2013 papers by Smith and colleagues were read here only in abstract, and Smith's own numbers are reported through Stager.
6. Some sources could not be read. Daniel Block's evangelical commentary on Ezekiel 1-24 (NICOT, 1997), Eissfeldt's 1935 monograph and Weinfeld's 1972 article (both read only through Day and Dewrell), John Day's 1989 Molech, and Ibn Ezra on Judges. Kimhi says the dedication reading agrees with "the opinion of our rabbis," but none of the rabbinic texts checked here says the daughter lived; which rabbis he meant is unknown.
7. The Kleitarchos account was not checked against a critical edition. It is cited here only through Schwartz and colleagues.

Verdict

Ancient Israelites, including kings of Judah, sacrificed children by fire, and some of them believed Yahweh accepted or even commanded it.  The Bible itself is the main witness to this, and it reports the practice in order to condemn it.  The Molech laws forbid it on pain of death, Deuteronomy calls it the worst of the nations' abominations, Jeremiah denies three times that God ever commanded it, and the stories of Isaac, Jephthah and Mesha either stop the sacrifice, report it without approval, or attribute it to a foreign king.

What an honest defender must concede is real:  the practice, the deaths, the belief that Yahweh wanted it, the missing redemption clause in Exodus 22:29, the probable death of Jephthah's daughter, and above all Ezekiel 20:25-26, where God gives laws "that were not good."  What the evidence does not show is that the Hebrew Bible commands or approves child sacrifice, that it was normal Israelite religion, or that archaeology has found it in Israel.  Even Ezekiel's hardest verse makes the bad laws a punishment.  The critics' strongest scholars agree that the evidence is the Bible's own polemic and that the scale is unknown.

Sources