The Legacy of Child Sacrifice in Early Judaism and Christianity


William Blake, “The Flight of Moloch” (1809). WikiMedia


Tantum religio potuit suadere malorum…”

In this post I want explore how in the earliest strata of Israelite religion, it was understood that the Jewish god, YHWH/Elohim, commanded ritual sacrifice of the firstborn child from his followers—and that this was not originally considered a heterodox rite, but indeed what we’d call an “orthodox” or official part of the Jewish religion of the time.¹

This idea is sometimes thought to be so inherently implausible that it’s often assumed (erroneously) to be a fringe theory by non-academic audiences.² In this particular case, it also doesn’t help that the charge of the practice of child sacrifice in Jewish religion really has been part of the fabric of (genuine) legend: see the blood libel, a staple of medieval anti-Jewish polemic. Similarly, early Christians too were falsely accused of participating in infanticidal and cannibalistic rites.

A sort of blanket incredulity that human sacrifice was ever a part of “official” Jewish religion might even be detected in the Wikipedia entry for blood libel itself:

The supposed torture and human sacrifice alleged in the blood libels run contrary to the teachings of Judaism. According to the Bible, God commanded Abraham in the Binding of Isaac to sacrifice his son, but ultimately provided a ram as a substitute.

This reflects the common idea that the narrative of Genesis 22 is directed against the practice of child sacrifice, often associated with early Israel’s regional neighbors.

To be sure, there is polemic against child sacrifice in the Hebrew Bible.³ But the Bible is a collection of texts from many different eras of Israel’s history, and of different theological perspectives; and sometimes what’s condemned in one text is in fact condoned, or even demanded, in another.⁴

I would urge readers, however, to resist the impulse to automatically reject child sacrifice as a part of early Jewish religion. First, the kneejerk reaction that this is “inherently implausible” in this instance often comes merely from privileging one’s own religious heritage—and Jews and Christians obviously share the same heritage here—as being somehow radically different from other more “primitive” traditions: traditions in which it’s uncontroversial that human sacrifice was practiced in various forms.⁵

Second—and most important—the presence of child sacrifice (as religiously prescribed⁶ᵃ) in early Israelite religion should be considered because this is the consensus of academic scholars of early Judaism on the issue.⁶ᵇ Of course, as with most other consensuses, this doesn’t mean that there aren’t dissenters; but at the end of the day, we’ll find that the the evidence clearly weighs in favor of the consensus. 

The smoking gun for child sacrifice having been religiously prescriptive in at least one stratum of Israelite religion can be found in Ezekiel 20:25-26; and in some sense this can guide the interpretation of other “data,” too.

The wider context here is an oracle of God given to Ezekiel in the seventh year of the exile—or 591 BCE. The oracle proper begins in verse 5, prefaced with “Thus says Adonai LORD,” and continues with a sort of retrospective of the Israelites’ original enslavement in Egypt.

Beginning at v. 10, the exodus is mentioned, followed by the sojourn in the wilderness. Following this, verses 11-13 read

I gave them my statutes and showed them my ordinances, by whose observance everyone shall live. Moreover I gave them my sabbaths, as a sign between me and them, so that they might know that I the LORD sanctify them. But the house of Israel rebelled against me in the wilderness; they did not observe my statutes but rejected my ordinances, by whose observance everyone shall live; and my sabbaths they greatly profaned.

The next verses repeat, several times, the Israelites’ disobedience and subsequent punishment (for example being denied entrance to the “promised land” of Canaan), until we arrive at the very interesting lines of 20:25-26:

Moreover, I gave them statutes that were not good and ordinances by which they could not live. I defiled them through their very gifts — in their having caused all of those who “(first) open the womb” to “pass over/through” [בְּהַעֲבִיר כָּל־פֶּטֶר רָחַם] — in order that I might horrify/desolate them, so that they might know that I am the LORD.

The reference to “statutes” and “ordinances” throughout the chapter is clearly to Mosaic Law: the laws given to the Israelites via Moses on Mount Sinai, including the Ten Commandments, and enumerated throughout the books of Exodus-Deuteronomy. Yet in Ezekiel 20:25-26, we appear to have something like the negative counterparts to the positive laws mentioned earlier. These are similarly described as “statutes” and “ordinances”; and yet unlike the other ones—the ones “by whose observance everyone shall live” (20:11, 13, 21)—these are called laws “by which they could not live.”

As for the other language in Ezekiel 20:26 here, there are several hints that it is sacrificial, and particularly concerned with children. Those who “open the womb” clearly refers to firstborn—though that this specifically suggests human firstborn here needs to be supported. Further, the word translated as “pass over/through” is elsewhere associated with ritual child sacrifice. For example, just a few verses later in Ezek 20:31, it’s used in a condemnation of those who “make [their] children pass through the fire.” (That this word can itself suffice to denote actual sacrifice, separate from more explicit contextual clues, is uncertain; though really it’s a moot point, because these clues are in fact always provided, as they are in Ezek 20:25-26.⁷)

Here we have bizarre concatenation of ideas: that God gave the Israelites certain laws that, in their following, would itself end in a sort of punishment for them—a punishment that’s presumably meted out because of their disobedience in the wilderness, as it is elsewhere in ch. 20.

As mentioned above, the “statutes” and “ordinances” here almost certainly refer to the Mosaic Law—the Torah. Which laws from this, then, does Ezekiel refer to?⁷ᵇ

Two laws in particular are universally singled out here, both from the book of Exodus:

The LORD said to Moses: “Consecrate to me all the firstborn; whatever is the first to ‘open the womb’ among the Israelites, of human beings and animals, is mine.” (Exodus 13:1-2)


You shall not delay to make offerings from the fullness of your harvest and from the outflow of your presses. The firstborn of your sons you shall give to me. 30 You shall do the same with your oxen and with your sheep: seven days it shall remain with its mother; on the eighth day you shall give it to me. (Exod 22:29-30)

To this we might also add Exodus 13:12, “you shall set apart to the LORD all that ‘open the womb’, and all the firstborn of your livestock that are males shall be the LORD’s,” with a close parallel in Exodus 34:19.

With the exception of the final verse (Exod 13:12)—where the verses following this attempt to modify this law—here we see little distinction between humans and animals, both grouped together in what’s otherwise a sacrificial context. (For eighth day sacrifice, see also Leviticus 22:27.)

Before getting further into those, however, we make take a second to assess the academic landscape, and the dissent about early Israelite child sacrifice that I mentioned earlier—and particularly how this might relate to the Ezekiel passage under discussion.

One of the most forceful arguments against the existence of early Israelite child sacrifice has come by way of Jacob Milgrom’s essay “Were the Firstborn Sacrificed to YHWH? To Molek? Popular Practice or Divine Demand?” Discussing the Ezekiel passage, Milgrom notes that “[r]ather than denying that God ever sanctioned human sacrifice as does his older contemporary Jeremiah (Jer 7:31; 19:5; 32:35), Ezekiel uniquely takes the tack that God deliberately gave such a law in order to desolate them.”⁸

Milgrom continues, that

The only way to justify Ezekiel’s theodicy is that the people misinterpreted either Exod 22:28b (de Vaux 1964: 72) or Exod 13:1-2 . . . or that God deliberately misled them to punish them (Greenberg 1983, 368-70; Hals 1989: 141), on the analogy of God hardening Pharaoh’s heart or Israel’s heart (52)⁹

It should be noted here that while I haven’t offered the best corroborating evidence that the Ezekiel passage really does concern child sacrifice yet (something that, we’ll see, has been disputed, though not persuasively so), there’s little ambiguity—even for Milgrom, who, again, is otherwise highly critical of the idea of a positive attitude toward child sacrifice in official Israelite religion¹⁰—that Ezekiel 20:25-26 refers to some law in the Torah for which God had some direct agency in terms of influencing the Israelites to follow, with the result that they would practice child sacrifice (in order to punish them, or perhaps bring punishment on themselves).

The question of dispute for Milgrom, with regard to Ezekiel 20:25-26, is simply whether 1) this passage suggests that God misled the Israelites by (somehow) influencing them to misinterpret the Torah laws as condoning or demanding child sacrifice, whereas the laws were not originally formulated/given with this intention; or else that 2) these laws really did condone/demand child sacrifice originally, and the Israelites would be “punished” merely through their straightforward (and correct) interpretation and practice of them.

Milgrom’s discussion of Ezekiel 20:25-26 itself is very short, though even here it’s unusually shortsighted. First, it should be said that there’s no support for the interpretation that Ezekiel 20:25-26 suggests that the Israelites misinterpreted the Torah laws as condoning/demanding child sacrifice. Rather, the laws themselves are called “not good.”

Yet Milgrom puzzlingly suggests that

Ezekiel does not contradict Jeremiah’s view that the people were mistaken in believing that God demanded human sacrifice; he supports it by the example of the firstborn males, whom the people sacrifice because they erroneously assumed it was God’s will (or because they did not realize it was God’s condign punishment). (53)

I’ve already suggested that there’s no warrant for the interpretation that Ezekiel thought that adherents of the laws “erroneously assumed it was God’s will.” Further though, whether God instituted the practice as something that was intended to have a positive or negative outcome for the Israelites seems irrelevant for whether Ezekiel believed that God really did ordain it (and thus Ezekiel does contradict “Jeremiah’s view that the people were mistaken in believing that God demanded human sacrifice”). Indeed, there’s certain sense in which, no matter the case here, from its perspective the Israelites were “right” to follow it—if only in the sense that it would enact the punishment that God had ordained for them in so doing.

Milgrom’s arguments against texts like Exodus 13:1-2 and 22:29-30 as originally condoning child sacrifice are also surprisingly weak. He notes that attention has been paid “particularly to Exod 22:[29-30]”; yet in attempting to countering intimations of child sacrifice here, he hardly engages with the strongest arguments for this. Rather, he begins by critiquing what might be best characterized as idiosyncratic “supplemental” arguments that (erroneously) try to make more out of these verses than there is—arguments presumably made in order to make the case more comprehensively, but I suppose here backfiring in a way—but as such Milgrom attacks something of a straw man, neglecting the stronger and more cautious arguments.

A somewhat more sophisticated critique of the orthodoxy/orthopraxy of early Israelite child sacrifice can be found in Hahn and Bergsma’s article “What Laws were ‘Not Good’? A Canonical Approach to the Theological Problem of Ezekiel 20:25-26.”

Hahn and Bergsma propose a rather convoluted scenario in which Ezekiel 20:25-26 was never intended to refer to child sacrifice at all, but rather only to animal sacrifice. They do this by reframing these verses as responsive to certain legal innovations regarding animal sacrifice that were introduced by the Deuteronomistic law code (Deut 12-26), and which contradicted several earlier laws as were formulated in Leviticus (“Priestly” laws, of which Ezekiel was a staunch proponent). Specifically, these innovations related to firstlings—firstborn animals and how they were to be sacrificed.

There are several problematic ways, however, in which Hahn and Bergsma construe Ezekiel 20:25-26 as referring specifically to animal sacrifice—as well as problems with the specific connections they make between Ezekiel 20:25-26 and the Deuteronomistic laws that it purportedly opposed.

Recall, again, the original verses in Ezekiel under discussion here:

Moreover, I gave them statutes that were not good and ordinances by which they could not live. I defiled them through their very gifts — in their having caused all of those who “(first) open the womb” to “pass over/through” — in order that I might horrify/desolate them…

Hahn and Bergsma write that

many scholars recognize that the phrase [כל־פטר רחם, “all who (first) open the womb,” in Ezek 20:26] is a reference to Exod 13:12, since Ezek 20:26 uses virtually the same diction. Notably, Exod 13 goes on to refer specifically to “every first-born (בכר) of man” (v. 13 RSV), only to exclude them from the consecrated “firstlings” mentioned in the previous verse. In other words, Exod 13:13 distinguishes human firstborn from “every opener of the womb” in order to exclude them from being offered. Thus, in the closest biblical parallel to Ezek 20:26a, the context makes clear that human sacrifice is not the referent. (212)

Yet it’d be hasty to suggest that the phrase “all who (first) open the womb” must specifically be a “reference to Exod 13:12″—or, more still, that this clearly suggests that humans aren’t in view here. The same phrase, or a close variation, is used at a few different points in the Hebrew Bible, and in contexts where it clearly does suggest human children.¹¹ But, significantly, even if it did refer specifically to Exod 13:12, there’s nothing that says that this verse has to be read in light of the subsequent verses. (And in fact there’s some indication that texts like Exod 13:12 were originally formulated in isolation, with the subsequent verses being secondary additions.¹²ᵃ)

There are other things to critique in Hahn and Bergsma’s paragraph, quoted above. There’s no indication that ‘Exod 13:13 distinguishes human firstborn from “every opener of the womb” in order to exclude them from being offered’; and it’s just as likely, as mentioned above, that “opener of the womb” could refer toor at least includehuman children.¹²ᵇ

And one important thing that they neglect to discuss is the use of the word translated as “pass over/through” in Ezekiel 20:26, which I discussed earlier (see especially Note 7). Significantly, this a sort of “technical term” in sacrificial contexts, which can in fact be understood to refer exclusively to human sacrifice!¹³

As for Hahn and Bergsma’s proposal about the connection between Ezekiel 20:25-26 and (opposition to) the Deuteronomistic laws pertaining to firstlings: not all of the innovations relating to firstlings are relevant to discuss here. Most important among these is the first one that Hahn and Bergsma highlight:

Whereas under the Priestly legislation the people were required to visit the sanctuary or sanctuaries for the slaughter of any and all animals (Lev 17:1-8), the Deuteronomic code required the sanctuary visit only for the (annual) slaughter of the firstlings (Deut 12:6, 17; 15:19, 20) and voluntary sacrifices. (213-14)

Here, they suggest that “[a] priest like Ezekiel observing the crowds of Israelites coming to the Jerusalem temple to perform their annual sacrifice of firstlings would be struck by the mute testimony these visits bore to the absence of these same crowds the rest of the year-in” (215). Based on this and, again, other factors, in the end Hahn and Bergsma argue that

Ezekiel refers to the Deuteronomic code as “not good laws” and “rules by which they could not live,” because, on the one hand, they degraded the pristine Priestly standards and, on the other, they were interwoven with predictions of human disobedience and inevitable divine judgment.

and that

In this defective Deuteronomic sacrificial system (“I defiled them by their very gifts”), Ezekiel singles out for special censure the distinctively Deuteronomic practice of the annual pilgrimage to present tithes and firstlings (“when they offer [only] all the firstlings”), since the Deuteronomic regulations governing firstlings were so wholly deficient. (217)

(Earlier they had elaborated that “[d]uring the rest of the year, as Ezekiel knew, the Israelites were slaughtering clean animals promiscuously and pouring out the sacred blood upon the ground like water (contra Lev 17:1-9). In that sense, the annual sacrifice of firstlings was a painful reminder for a priest trained in the Holiness Code of the deficiency of sacrificial practice among the populace.”)

Yet even here, it is disingenuous to construe the Ezekiel verses as suggesting “when they offer [only] all the firstlings.” There is nothing in the text that suggests “only“; and indeed the much more natural interpretation here is that there’s something about this “firstborn” sacrifice itselfnot merely its infrequency—that is particularly disturbing. (And for that matter, I think its portrayal as being something intended to “horrify/desolate” those who follow this is much more suited for the horrifying practice of child sacrifice than any other perceived ritual violation.)

In the end, although Hahn and Bergsma’s is a valiant and creative attempt to reeinterpret these traditions, there are numerous places in which it falters, and can be firmly rejected.

Although at some points it’d be unwise to ascribe well-intentioned academic efforts to ideological/emotional bias, it’s also hard to avoid that the conclusion that the resistance to the sanction of child sacrifice in early Israelite religion is at least in part due to the fact that this religious heritage still survives—and that the very texts whose archaic strata indeed sanctioned this practice are still revered as holy by Jews and Christians.

It’s even more interesting in this regard how the acknowledgment of the historical practice of this, even when not “officially” sanctioned but as “heterodox,” has also been resisted—and this for centuries. For example, as mentioned earlier, the most well-known condemnation of heterodox child sacrifice in the Hebrew Bible was aimed at the pagan Canaanite burnt (child) offerings to a god Moloch/Molek. (Incidentally, it’s the association of this burnt sacrifice with Gehenna, or the Valley of Hinnom, which led to the characterization of Gehenna as the main Jewish “hell” realm, discussed by Jesus and others.)

Yet even though this practice is clearly condemned in the Hebrew Bible, the very intimation that it was ever practiced at all was sharply resisted, for example among the early Jewish rabbis. J.E. Hartley and T. Dwyer note that

Rabbinic tradition held that Molek worship involved the dedication of a child by pagan priests to Molek. In the second century C.E., for example, R. Judah ben Elai commented on Deut 18:10 that the prohibition involved commanding one not to go over to a pagan god. There is some rabbinic tradition that the prohibition in Lev 18:21 involved sexual intercourse with a cult prostitute. As a whole, rabbinic tradition denied that actual sacrifice of the children took place. (“An Investigation into the Location of the Laws on Offerings to Molek in the Book of Leviticus,” 84)¹⁴

Yet this rabbinic interpretation is not without even earlier precedent. In fact there seems to be a certain cross-cultural impulse where earlier violent rituals are replaced with “substitute” rites that preserve some of their essence, yet without the destruction of actual killing. For example, where a firstborn child (or a war captive) might have originally been “devoted” (=sacrificed) at the temple as an offering to a god, in a substitute ritual they are instead devoted to temple service. Similarly, in the Hebrew Bible, the Levites are themselves portrayed in terms of being a type of substitutionary “sacrificial offering.” (Cf. Numbers 8. I’ve elaborated on this in great detail here, as well as here—the latter link particularly focusing on circumcision itself as a substitute rite for child sacrifice.)

Further, the resisted acknowledgment of the practice of human sacrifice in ancient religion has not been limited to just Israelite religion, but in the ancient world more broadly. The reluctance of scholars to acknowledge the practice of Phoenician/Carthaginian child sacrifice has been detailed in Brien Garnand’s dissertation “The Use of Phoenician Human Sacrifice in the Formation of Ethnic Identities.”

Interestingly, on the converse, the 1st century CE Hellenistic Jewish philosopher/theologian Philo of Alexandria—famed for being the most influential figure in promoting allegorical interpretation of the Old Testament among Jews, and eventually Christians—seems to have had a rather cavalier attitude toward the prospect of child sacrifice, if truly to God. For example, in the context of Abraham’s near-sacrifice of Isaac, Philo suggests “[f]or a father to surrender one of a numerous family as a tithe to God is nothing extraordinary, since each of the survivors continues to give him pleasure, and this is no small solace and mitigation of his grief for the one who has been sacrificed.”¹⁵

I began this post with a quotation from the 1st century BCE Roman philosopher Lucretius’ De rerum natura: “Tantum religio potuit suadere malorum“: “To such depths of evil has religion been able to drive men.”

Lucretius delivers this line in reference to the famed Greek mythological figure Agamemnon‘s sacrifice of his daughter Iphigenia, in order to appease the goddess Artemis. Although it’s easy to dismiss this as an archaism or barbarism, there are several elements of Iphigenia’s sacrifice that closely resemble Jephthah’s sacrifice of his daughter, as recounted in the 11th chapter of the Biblical book of Judges.

An important perspective that we might take away from all this is that there was a certain shared understanding of human and child sacrifice that was felt in many different areas of the ancient Mediterranean and Near Eastern world; and echoes of this continued to be heard down through the ages. In the apostle Paul’s epistle to the Romans, he asks at one point “If God is for us, who is against us? He who did not withhold his own Son, but gave him up for all of us—will he not with him also give us everything else?” The language of God having not “withheld” Jesus is clearly indebted to language of Abraham’s attempted sacrifice of Isaac (Genesis 22:16), and indeed ultimately reflects the wider vocabulary of ancient child sacrifice.

(For more on hints of human sacrificial tradition in the Pauline epistles and beyond, see especially Lampe’s essay “Human Sacrifice and Pauline Christology”: “Picking up the category of human sacrifice as an interpretive tool and at the same time transcending and breaking it apart, the early Christians played with fire without getting burned” [208].)

Further, in the book of Acts, at one point Paul declares “Keep watch over yourselves and over all the flock, of which the Holy Spirit has made you overseers, to shepherd the church of God that he obtained with his own blood” (20:28). Although there’s often confusion about what this means,”his own blood” refers not to God‘s own blood, as if this means to suggest that Jesus = God here (and that Jesus spilled his blood). Rather, “blood” here is a metonym for one’s offspring in general; thus the most accurate interpretation would be “…that he ‘obtained’ with his own son.” The use of the phrase “his own blood” to denote “his own son” is particularly interesting, as this can be uniquely connected with the vocabulary of several sources that attest ancient child sacrifice, including native Phoenician sources themselves.

In particular, here we might cite a work by the Tyrian Neoplatonist philosopher Porphyry:

Originally, then, sacrifices to the gods were made with crops. In time we came to neglect holiness, and when crops were lacking and through the dearth of lawful food people took to eating each other’s flesh, then, imploring the divine power with many prayers, they first offered the gods sacrifice from among themselves, not only consecrating to the gods whatever was finest among them, but taking in addition others of the race who were not among the best . . . From then until now, it is not only in Arcadia at the Lykaia . . . and in Carthage for Kronos that everyone engages in public human sacrifice, but periodically, in remembrance of the custom, they stain altars with the blood of their own kind [ἐμφύλιον αἷμα]. (De abstinentia 2.27.1-3)

I highlight this for more than one reason. Although the language of sacrifice of a person’s “own blood/flesh” (=child) can, independent of this, be connected with the sacrificial vocabulary of Phoenicia, it’s uncertain whether Porphyry’s use of “blood of their own kind” means something more like “blood of their own ethnicity” instead of “blood of their own family” (=children).

However, elsewhere in Porphyry’s work, we read that “[i]n great disasters, such as wars and plagues and droughts, the Phoenicians used to choose by vote, for sacrifice to Kronos, one of those dearest to them” (2.56.1). (“Kronos” in these texts is understood variously as either the Phoenician deity Baal-Ḥamān, or as the common god El, according to Philo of Byblos.) But there’s ambiguity here, too. In Helen Dixon’s dissertation on Phoenician sacrifice, she notes

the translators clearly struggle to sort out whether (a) the sacrifice of τῶν φιλτάτων τινὰ “one of the most beloved” refers to a beloved child, or an adult who is valuable or “dear” to the community, (b) the sacrifice is to be made by every Phoenician family, or as a single sacrifice on behalf of all the Phoenicians. (“Phoenician Mortuary Practice in the Iron Age I – III (ca. 1200 – ca. 300 BCE) Levantine ‘Homeland’,” 348)

Yet most importantly, in the text of Philo of Byblos preserved by the 3rd/4th century Christian bishop Eusebius, “Kronos” himself is said to have offered his own son as a sacrifice during a plague. Even more, Kronos’ son is called here his μονογενῆ υἱὸν, “only-begotten son”—the exact title by which Jesus would come to be known in the New Testament and elsewhere. (See Hebrews 11:17 for a use of this term in connection with the Abraham/Isaac story.)

One other interesting Christian text that might display a curious connection with Near Eastern human sacrificial traditions is the 2nd century Epistle to Diognetus. In the relevant passage we read

But [God] was patient, he bore with us, and out of pity for us he took our sins upon himself. He gave up his own Son as a ransom for us, the holy one for the lawless, the innocent one for the wicked, the righteous one for the unrighteous, the imperishable one for the perishable, the immortal one for the mortal. . . . Oh, the sweet exchange! (Translation by Ehrman, The Apostolic Fathers, II, 151)

Here we seem to have a puzzling idea: God simultaneously “took our sins upon himself” as well as “gave up his own Son as a ransom for us.” While it would be tempting to understand this in line with some of the more classic tropes of binitarian/Trinitarian theology, that this simply refers to God the Son, there are several indications that this would be anachronistic or otherwise inappropriate here. Indeed, here we still seem to have an ontological distinction between God and Son; or at the most a sort of “proto-Trinitarianism.”

In light of this, although it’s somewhat speculative, perhaps this could be understood in light of ancient traditions of a king sacrificing his son in times of crisis (such as in 2 Kings 3, and also closely related to those about Kronos, described earlier). Cristiano Grottanelli suggests that, in these traditions,

Through [his first-born son], the king supplicates the angry gods and pays a great price to ransom his people; but through him the king also ransoms himself, as he covers his child with the insignia of his own rank and person. (“Cosmogonia e sacrificio II,” quoted in Levenson, Death and Resurrection of the Beloved Son, 27; emphasis mine)

Here, then, in a sense God can himself be said to have taken on sin—through the ancient understanding of a sort of “corporate” identity of father and son, almost in a sense of them being interchangeable, or the father having “extended” his person onto the son. (If anything, then, this seems to come closer to a sort of patripassianism than to any orthodox Christology.)

Finally, the connection between ancient traditions of child sacrifice and the death of Jesus may even have been felt as late as the 5th century (or beyond), in a text attributed to the Christian bishop and theologian Cyril of Alexandria, suggesting that the sacrifice of Jesus was a sort of reflection or fulfillment of the laws of firstborn sacrifice in the Torah.¹⁶

What could be made of all this, theologically speaking?

For one, this might push us to further acknowledgment a prominent role for child sacrifice in the religious heritage of Judaism and Christianity. (In this regard, no book has been quite so influential in having made room for this view as Jon Levenson’s The Death and Resurrection of the Beloved Son: The Transformation of Child Sacrifice in Judaism and Christianity has.)

Second: in regard to the clear support for child sacrifice in early strata of the Torah—laws ascribed to God himself—we can certainly see how this is the exact sort of thing that would be interpreted in line with a kind of truism that’s developed among certain types of progressive Christians, that God simply could not have actually said or done some of the things that were ascribed to him in the Old Testament. (And, in fact, the “bad laws” text of Ezekiel played into a wider early Christian tradition that certain Old Testament texts were inauthentic or otherwise fabricated. This motif has been explored at length in Vaccerella’s dissertation “Shaping Christian Identity: The False Scripture Argument in Early Christian Literature.”)

The alternative perspective—usually held by those who are more orthodox/conservative—is to appeal to what’s known as “divine command theory“: that since God is, by very definition, all-good and all-just, he’s incapable of immoral actions (no matter how seemingly horrific).

And there’s a sense in which the text of Ezekiel, with which we’ve been largely concerned, offers precisely divine command theory here—though with a twist. God purposely commanded things knowingly “not good.” Of course, we can surely conclude that, in the perspective of the author of Ezekiel, this does not make God “not good”; rather this was only to enact just retribution for human sin.

Divine command theory certainly solves several problems. One added benefit is that it avoids a problem that progressive Christians have, in their simply discarding certain problematic Old Testament texts (and in so doing calls into question how one might discern what’s truly “inspired” or “uninspired” in a non-arbitrary way). Yet, at least in this particular instance, divine command theory only “works” at the expense of (the Israelite) God’s uniqueness, vis-à-vis other deities of the ancient Near Eastern and Mediterranean world—making him ever more like Canaanite/Phoenician El or Baal-Ḥamān, “pagan” gods who indeed could be honored and propitiated by child sacrifice.

⁂       ⁂       ⁂

[Edit: in the wake of this post, most of the substantial responses to this came on Reddit. Here‘s an important conversation where, among other things, I clarified some stuff about Ezekiel; and here‘s a bibliography I added.]

I’ve now written a Part 2 and Part 3 to this post.


[1] I use “official” most uncontroversially to refer to things that were sanctioned in what’s otherwise universally considered authoritative (inspired) “Scripture.”

[2] This is also expressed in the academic literature: e.g. Greenberg speaks of the interpretation that certain laws of the Torah condoned child sacrifice as “intrinsically improbable.” Milgrom follows De Vaux in denying it on pragmatic grounds: “It would indeed be absurd to suppose that there could have been in Israel or among any other people, at any moment of their history, a constant general law, compelling the suppression of the firstborn, who are the hope of the race” (1964: 71).

Beyond the general ethical revulsion toward this idea, one major basis of criticism here is—as it’s expressed by others—the general strain on Israelite society that there would have been if they had indeed practiced regular child sacrifice (which is sometimes also mentioned in conjunction with the virtual absence of archaeological evidence for child sacrifice in Israel/Canaan).

Among other responses to these things, though, one could respond that this relies on an assumption that the Israelite law were realistic, or actually followed in practice, as opposed to simply idealistic. (This is a distinction that’s been increasingly recognized with respect to ancient Near Eastern and Israelite law.) In this regard, Levenson suggests the possibility “that Exod 22:28b articulates a theological ideal about the special place of the firstborn 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” (The Death and Resurrection of the Beloved Son, 9).

(That being said: also, more generally, the history of religion is full of things which might have otherwise seemed too outlandish to be true, and yet were/are. We might think of cargo cultsone of which is centered around reverence for Prince Philip as a deity—or the continuing foothold that Scientology still has among certain [thankfully fairly small] sectors of the populace; or that there are still several adherents of the Heaven’s Gate cult.)

[3] There are ambiguities about the original nature of the Binding of Isaac narrative, often cited as actually among the strongest polemic against child sacrifice.

Here I mainly refer to the idea that, in a (hypothesized) more primitive form of the narrative in Genesis 22, Abraham actually did go through with the sacrifice of Isaac, with no interference from the angel and no ram substitution. (Levenson suggests, however, that even with the story in current form, ‘[i]f the point . . . is “abolish human sacrifice, substitute animals instead,” then Abraham cannot be regarded as having passed the test to which Gen 22:1 tells us God is here subjecting him.’) There are in fact several reasons why the proposal of Abraham’s having actually gone through with the sacrifice in an original version of the narrative is not inherently implausible. Perhaps most important among these is that comparative data shows us that very similar stories can go through a similar process of evolution, where the original distributing element(s) is/are gradually excised in their retelling and redaction. (See my post here for more.)

That being said, I can’t say that I’ve ever spent too much time looking at Genesis 22 in this regard. From what I know though, I’m not as enthusiastic about a hypothetical more primitive form of this narrative as others are.

[4] Particularly in terms of sacrifice, one of the most striking examples of this is Jeremiah 7:22: “in the day that I brought your ancestors out of the land of Egypt, I did not speak to them or command them concerning burnt offerings and sacrifices.” On the surface, this seems to stand in stark contradiction to what we find in the Torah narratives / Mosaic Law (cf., for example, Exodus 20:24). While there have been several apologetic harmonizing explanations offered here to attempt to relieve this contradiction, I think the best evidence to explain the two traditions is simply that these are competing theological/historical perspectives. In general, that “sometimes what’s condemned in one text is in fact condoned, or even demanded, in another” is entirely uncontroversial among scholars.

(Interestingly, the “I did not speak to them or command them…” language of Jeremiah 7:22 can itself be connected with similar language in a context of child sacrifice—including just verses later, in 7:31!: “And they go on building the high place of Topheth, which is in the valley of the son of Hinnom, to burn their sons and their daughters in the fire—which I did not command, nor did it come into my mind.” On this verse, cf. Lange, “‘They Burn Their Sons and Daughters — That was No Command of Mine’ (Jer 7:31): Child Sacrifice in the Hebrew Bible and in the Deuteronomistic Jeremiah Redaction.” See, similarly, Jeremiah 32:35.)

[5] See especially the volume The Strange World of Human Sacrifice, edited by Jan N. Bremmer, and the volume Not Sparing the Child: Human Sacrifice in the Ancient World and Beyond. (For a more regional approach, see Tatlock’s dissertation “How in Ancient Times They Sacrificed People: Human Immolation in the Eastern Mediterranean Basin with Special Emphasis on Ancient Israel and the Near East.”)

[6a] Again see Note 2 for what I mean by “religiously prescribed.”

[6b] Frank Moore Cross had opined that “the biblical evidence for child sacrifice among the Israelites (and their neighbors), thanks to recent studies, has been put beyond doubt. Paul Mosca’s treatment of the evidence is definitive (1975:117-284; see Day 1989; Heider 1985)” (102); and in Susan Niditch’s 1993 War in the Hebrew Bible, she writes that “the consensus of scholars over the last decade concludes that child sacrifice was a part of ancient Israelite religion to large segments of Israelite communities of various periods” (47). (I’ve quoted Niditch more fully on this in my comment here.) 

Of course, many have questioned the extent to which child sacrifice was actually practiced in Israel; but that some Israelites practiced child sacrifice, in the name of YHWH—and that a positive (or at least neutral) attitude to this practice is indeed reflected in several Biblical texts—is a consensus that’s only been strengthened over the subsequent two decades since Niditch’s monograph.

Robert Knuse, in his 1997 monograph No Other Gods: Emergent Monotheism in Israel, writes “Previously scholars assumed that child sacrifice was a Canaanite custom, but increasingly they suspect that it, too, was a natural part of the Yahwistic religion practiced by kings in times of crisis,” citing

Green, Human Sacrifice, pp. 156-87; Heider, Molek, pp. 223-408; Day, Molech, pp. 29-71; Kennedy, ‘Isaiah 57’, pp. 50-51; Smith, Early History of God, pp. 137-38, 146; Ackerman, Green Tree, pp. 101-63; idem, ‘Child Sacrifice’, pp. 20-29, 56; Levenson, Death, pp. 3-52, 111-24. (188 n. 17)

He continues

Exod. 22.29-30 states, ‘The firstborn of your sons you shall give to me. You shall do the same with your oxen and your sheep’. The passage implies that sacrifice of the child should be undertaken as surely as the sacrifice of the animals. However, Exod. 13.13 and 34.19 provide for the replacement or ‘redemption’ of the chilid with an animal sacrifice. The omission of any reference to redemption in Exod. 22.29-30 leads scholars to suspect that some Yahweh devotees indeed sacrificed their children as burnt offerings to Yahweh, and this may be the earliest legislation on the custom. (188)

In the entry for “Molech” in Eerdmans Dictionary of the Bible (written by Brian Schmidt), we read

The Deuteronomistic history, although critical of the practice of human sacrifice, preserves the memory of former Israelites and their neighbors who indeed ritually killed their children — Jephthah (Judg. 11:34-40); Hiel (Josh. 6:26; 1 Kgs. 16:34); Ahaz (2 Kgs. 16:3); and Manasseh (2 Kgs. 21:16). It further acknowledges the efficacious power of such a ritual as in the sacrifice of the Moabite prince offered up by his father.

. . .

Human sacrifice as more generally referred to in the phrase, “the one who makes his son or daughter pass through the fire,” is frequently and exclusively attributed to Canaanite origins by some biblical writers (e.g. Deut. 12:31). Nonetheless, some form of human sacrifice was apparently part of the Yahwistic cult in pre-exilic (and perhaps exilic) times.

. . . 

As added confirmation of the endurance and pervasiveness of the practice, Ezekiel implies that Yahweh had commanded the Israelites to participate in the sacrifice of their firstborn (Ezek. 20:25-26), but qualifies this law as a form of punishment. Similarly, Exod. 22:29-30 (MT 28-29) comprises an unqualified demand to make the firstborn sacrifice to Yahweh; the option to redeem the firstborn is not offered here as in later Priestly texts. In the light of Jeremiah’s condemnation of the practice and Ezekiel’s recognition that Yahweh had once condoned the ritual killing of humans, it is self-evident that for many it was an acceptable form of Yahweh worship. (913)

(I’ve quoted this at greater length here.)

Paolo Xella, in his “‘Tophet’: An Overall Interpretation,” has particularly strong words here:

Several biblical texts clearly speak – directly or indirectly – of infant killing in Tophet, and the picture resulting from the all the information (on the basis of a rigorous textual critique) is coherent and totally indisputable. The reality of child sacrifices in the “Canaanite” and Israelite context is now accepted by most scholars; as far as the “fundamentalists” are concerned, the question to be argued is whether or not this kind of rite was related to the Yahwistic cult.

Nevertheless, some scholars (above all, in the field of Punic studies) continue to claim that biblical texts do not speak explicitly of ritual infant killing: this statement is simply wrong. (265)

(The fact that Xella characterizes the question of whether child sacrifice “was related to the Yahwistic cult”—that is, done in the name of YHWH and/or with an understanding that this was in fact his will—as one with which the “fundamentalists” are concerned to argue, I think speaks to the fact that he understands dispute of this to be marginal.)

[7] A particularly instructive use of this (עָבַר) is found in Jeremiah 32:35; Leviticus 18:21, where it’s not explicitly associated with “fire” as it is elsewhere. Ezekiel 16:21 is another interesting text, where it’s used in conjunction with שָׁחַט, “slaugher,” and נָתַן, “give (over).” The latter is clearly used to refer to ritual sacrifice: see also Ugaritic ytn, “to offer (as a sacrifice)”; Akkadian nadānu as “to make a sacrifice”; Micah 6:7; 1 Chronicles 21:23; TDOT, X, 102-3. All of that being said, it’s noteworthy that LXX Ezekiel 20:31 doesn’t include the “by fire” clause. In his disseration, Heath Dewrell suggests 

The mention of “passing sons by fire” in the MT of [Ezekiel] 20:31 appears to be the work of a late glossator. The person responsible for the gloss appears not to have understood the distinction between the Law of the Firstborn and the lmlk ritual. (“Child Sacrifice in Ancient Israel and Its Opponents,” 194)

[7b] Francesca Stavrakopoulou, in her King Manasseh and Child Sacrifice: Biblical Distortions of Historical Realities, notes that

Van Seters (“Law on Child Sacrifice”, 368) prefers to relate the language of Ezek. 20:25-26 with Exod. 13:12 alone, whereas De Vaux (Studies, 72) and Zimmerli (Ezekiel, vol. 1,411) associate the practice to which Ezekiel refers with Exod. 22:28-29; see also Greenberg, Ezekiel 1-20, 368-370. W. Eichrodt (Ezekiel [trans. C. Quin; OTL; Philadelphia: Westminster, 1970; German original, 1966], 270) and Mein (Ezekiel and the Ethics of Exile, 117-118) identify Exod. 34:19 as the law underlying Ezek. 20:25-26. 

Of course, much of the material in my post that follows this discusses alternative proposals. To add to this, Stavrakopoulou notes that “Weinfeld even claims that Ezek. 20:25-26, 31 refers to dedication, not sacrifice” (191, emphasis mine). Stavrakopoulou herself admits that “[i]t is notoriously difficult to ascertain the precise relationship between sections of Ezekiel and Pentateuchal texts and traditions” (184 n. 173); however, she does write that Ezek 20:25-26 ’employs the terminology shared by the firstborn laws of Exod. 13:2, 12 and 34:19, for it describes the sacrificed child as the פטר רחם, “the first-birth of the womb”.’ See also Patton, “‘I Myself Gave Them Laws That Were Not Good’: Ezekiel 20 and the Exodus Traditions.”

[8] Levenson’s response to Greenberg applies just as equally to Milgrom here:

But it is the latter opinion that better fits the biblical data: YHWH once commanded the sacrifice of the first-born but now opposes it. Without recourse to modern historical reasoning, the only explanation for this that preserves the continuity of YHWH’s will is the one that Ezekiel, in fact, offers: YHWH’s command and Israel’s obedience to it were in the way of punishment, a means to bring about the death of those who had turned away from the means to abundant life.

(Of course, if perhaps the author of Ezekiel wants his contemporaries to discontinue following these laws, this would be different.)

[9] We might also note that the analogy with God hardening Pharaoh’s heart is particularly apt here. In Exodus 14:4, God says “I will harden Pharaoh’s heart, and he will pursue [the Israelites]”—the purpose of which being that, through Pharaoh’s inevitable defeat which is to follow this, “the Egyptians shall know that I am the LORD.” The final clause here bears an unmistakable resemblance to Ezekiel 20:26’s “…in order that I might horrify/desolate them, so that they might know that I am the LORD.”

[10] Dewrell writes that “those who argue that the plain sense of Exod 22:28-29 (along with Exod 13:1-2, if one views these verses as originally separate from 13:11-16) implies that children were to be offered to Yahweh are undoubtedly right”; and yet also that “despite the plain sense of the text, however, scholars are right to question how the obvious reading can be the correct one, given the overwhelming evidence that firstborn children were not regularly sacrificed” (121). This again, however, is a fallacy, as was demonstrated in Note 2.

[11] See Numbers 3:12; 8:16; and 18:15.

[12a] I’ll explore this more in a later post.

[12b] In fact, if the structure of Exodus 13:12-13 (in the form that we currently find it) were intended to be chiastic, this would be a strong argument that the “opener of the womb” here is synonymous with human children:

A you shall set apart to the LORD all that “open the womb”
   B and all the firstborn of your livestock that are males shall be the LORD’s
   B’ But every firstborn donkey you shall redeem with a sheep . . .
A’ and every firstborn male among your children you shall redeem.

Again, the idea here is that if 13:12a simply denoted (solely) firstborn animals—as is argued by those who don’t prefer to see a command of child sacrifice here—then 13:12b (“all the firstborn of your livestock”) is redundant. With the addition of 13:13, however—even though it’s precisely the addition of this verse that introduces the possibility of redemption/substitution into the original law—we would see here an elaboration on the kinds of things that 13:12a and 13:12b were referring to: the donkey apparently being a special example of firstborn livestock, and “every firstborn male among your children (בניך)” either being a (special) subset of “all that ‘open the womb'” in 13:12a, or even synonymous with it.

Of course, appealing to the chiastic structure itself as an argument here is somewhat speculative, although it probably wouldn’t be that difficult to “test,” by seeing if there are other laws with a similar structure.

For similar deep literary/redactional analyses of Exodus 13, see Finsterbusch, “The First-Born between Sacrifice and Redemption in the Hebrew Bible.”

[13] Stavrakopoulou lists those verses in which it’s used to refer to human sacrifice as Lev. 18:21; Deut. 18:10; 2 Kgs 16:3; 17:17; 21:6; 23:10; Jer. 32:35; Ezek. 16:21; 20:26, 31; 23:37; 2 Chr. 28:3; 33:6 (King Manasseh and Child Sacrifice, 142). Further, when we properly convey the syntax of Ex 13:12 in translation—”וְהַעֲבַרְתָּ to the LORD all that ‘open the womb’, and all the firstborn of your livestock that are males shall be the LORD’s”—the first part (with which the verb עָבַר belongs) can easily be understood as referring only to human/child sacrifice.

[14] In rabbinic tradition, children being “passed through the fire” was understood quite literally, not as their immolation in sacrifice, but rather as their being passed in between a fire (or two fires: b. Sanh. 64b), to then be conscripted into temple service. This may be indebted to a common feature of ancient Near Eastern treaties where the parties passed in between an animal that has been cut in half. See Jeremiah 34:18 on this; also Snaith 1966, and Leviticus 18:21.

[15De Abrahamo 196, translation by Colson (Philo, Volume VI, 197). Niehoff, while acknowledging the broader context in which this statement was made, still suggests that “[t]his indicates his mental proximity to the original notions of the Akedah. Philo was clearly more sympathetic to the idea of child sacrifice than both Josephus and the rabbis” (Philo on Jewish Identity and Culture, 174).

[16] …καὶ εἰς ὄψιν ἄγεται τοῦ Πατρὸς, ὡς ἄνθρωπος καθ’ ἡμᾶς, ὁ συνεδριάζων αὐτῷ Θεὸς Λόγος, ὁ καὶ διὰ τῆς τοῦ νόμου σκιᾶς ἐν τοῖς πρωτοτόκοις γραφόμενος. Ἅγια γὰρ καὶ ἱερὰ τῷ Θεῷ καὶ τὰ πρὸ τῆς ἐνανθρωπήσεως πρωτότοκα, ἅπερ ἔθυον αὐτῷ κατὰ νόμον: “…and brought into the Father’s presence in human nature like unto us, and by the shadow of the Law is numbered among the firstborn. For even before the Incarnation the firstborn were holy, and consecrated to God, being sacrificed [ἔθυον] to Him according to the Law.” (Even in patristic usage, θύω still seems to refer exclusively to material sacrifice: animal, human, etc.) See also the larger context of Cyril here.

Of course, this seems like an outgrowth of a much more common early Christian interpretation, perhaps going all the way back to (an interpretation of) things like Matthew 5:17.

Also, interestingly, some patristic figures—like Ambrose—connected Jesus’ circumcision and his crucifixion: “[Christ] was circumcised first according to the Law, in order not to dissolve the Law [ne legem solveret]; afterward [he was circumcised] through the cross, so that he might fulfill the Law…” Origen of Alexandria suggests

Poposcit ergo pretium nostrum sanguinem Christi. Verum donec Jesu sanguis daretur, qui tam pretiosus fuit, ut solos pro omnium redemptione sufficeret, necessarium fuit eos qui instituebantur in lege, unumquemque pro se, velut ad imitationem quamdam futurae redemptionis, sanguine suum dare

Therefore [Satan] demanded the blood of Christ as the price for us. So then, until the blood of Jesus was given, which was so precious that it alone would suffice for the redemption of all, it was necessary for those who were being trained up in the Law to offer their own blood for themselves as a kind of foreshadowing of the future redemption (Commentary on Romans, 2.13)

Although someone’s “own blood” can signify their offspring themselves (cf. again Acts 20:28), in this context Origen goes on to specify circumcisiom:

And therefore for us as those for whom the price of Christ’s blood has been furnished, we do not have need to offer a price for ourselves anymore, that is to say, to offer the blood of circumcision [id est, sanguinem circumcisionis offerre]. But if it seems criminal to you that the God of the law should command that wounds be inflicted upon infants and that their blood be shed

(Interesting also is Origen’s variant citation of 1 Peter 1:18-19 before this, Redempti sumus non corruptibili pretio argenti et auro, sed pretioso sanguine Unigeniti, “We have been redeemed not at a corruptible price of silver and gold but with the precious blood of the only begotten.” The original text of 1 Peter here doesn’t have “of the only-begotten” but “of Christ, like that of a lamb without defect or blemish.” While we shouldn’t read too deeply into Origen’s citation itself, both “precious blood” and “only-begotten” have integral connections with child sacrifice.)

In relation to all this, again see my post here, on a connection between circumcision and (substitutions for) child sacrifice.

Finally, Stroumsa, in his essay “The End of Sacrifice: Religious Mutations of Late Antiquity,” makes some compelling observations:

In strong opposition to post-Yavneh Rabbinic Judaism, early Christianity unabashedly presented itself as a sacrificial religion, although one of a new kind, in which the central ritual was called anamnēsis, a re-actualization, or even re-activation—rather than our weaker term ‘memory’—of Jesus’ sacrifice. It was a religion without temples, in which the same sacrifice was offered perpetually, on a daily basis. It was offered by priests, organized in a hierarchy (in contradistinction to the basic equality of rank between the Rabbis). The very metaphorisation of biblical traditions by Christian thinkers permitted the preservation of the terms of Israelite religion. In Christian literature of the first centuries, one can follow the clear development of sacrificial vocabulary. The language of martyrdom, strikingly, is replete with allusions to sacrifice. The clearest testimony of this perception of martyrdom as a sacrifice is probably the Martyrdom of Perpetua and Felicity. Imitatio Christi, when it goes up to the willingness to give one’s life for one’s faith, transforms the martyr, like Jesus, into a sacrifice, more precisely, a human sacrifice. (41)

. . .

one should perhaps point out once more the deep ambiguity of sacrifice. Transformed, reinterpreted, metaphorized, memorized, it seems never to have died out completely. Late antiquity experienced the end of public sacrifice as the core of religious praxis, but that did not mean the end of the very idea of sacrifice. (42)