2.1 Levin's Argument

Levin introduces his study by setting forth the conundrum noted above: both Matthew and Luke portray Jesus as both virginally conceived and a son of David through Joseph. The linchpin of his argument comes in the second major section (‘Joseph's “Adoption” of Jesus: A non-Jewish Concept’). Here he notes that though most commentators assume that Jesus was a Davidide through adoption by Joseph, ‘when pressed for either precedence or proof of such adoption, the vast majority of commentators simply refer to “Jewish custom” or “Jewish Law”’.Footnote ⁵

‘However’, Levin contends, ‘Jewish law, both in antiquity and in the modern era, has no such legal institution.’ He explains:

To this he adds Jeffrey Tigay's assertion that ‘if adoption played any role at all in Israelite family institutions, it was an insignificant one … for the post-Exilic period in Palestine there is no reliable evidence for adoption at all’.Footnote ⁷ After quickly reviewing other potential evidence for Jewish adoption, Levin concludes that ‘there is nothing in Jewish law, in either the Hebrew Bible or in later Halakhah, which can be seen as the model by which Jesus, Son of God, could have been considered the legal, but not genetic, heir to the Davidic throne’.Footnote ⁸

Having eliminated the possibility of Jewish adoption, Levin advances his positive argument in three stages. He first demonstrates that Roman law ‘had an extremely well-defined concept of adoption’.Footnote ⁹ Second, Levin argues that Roman adoption did not exert any significant influence on Jewish law or practice. Although he recognises that Roman views of adoption impacted Paul, he sets the apostle aside as ‘an anomalous Diaspora Jew’ and asserts that Roman adoption had little impact on Jewish communities.Footnote ¹⁰ According to Levin, Jews in Judea largely resisted Romanisation so that ‘to a Judean of the first century ce, the very concept of legal adoption … would have been totally foreign’.Footnote ¹¹ Although he allows that Diaspora Jews may have been somewhat more likely to absorb Roman practices, he concludes against the possibility that they did so, both because they would have conformed to Jewish law and because ‘not a single one of the sources that we have from those communities mentions anything like the Roman concept of adoption’.Footnote ¹² Lastly, Levin argues that this Roman concept of adoption accords well with Jesus’ implied adoption in Matthew and Luke. Therefore, the evangelists must have been drawing on Roman, not Jewish, practices. This indicates that ‘the cultural and religious assumptions’ of Matthew and Luke, ‘at least in the matter of Jesus’ Davidic heritage, were far removed from those of contemporary Judaism’.Footnote ¹³

Levin's argument is straightforward and bold: (1) Matthew and Luke imply adoption, (2) Judaism had no such concept, (3) Roman law did, but (4) it did not impact Judaism as a whole, so (5) Matthew and Luke must be appealing to Roman law. If Levin is correct, then he has provided a much-needed background for understanding Jesus’ Davidic lineage in Matthew and Luke. However, as we will see below, this argument contains some significant flaws.

2.2 Critical Response

Let me begin by noting two points of agreement with Levin. First, he is correct that the vast majority of scholars who have treated Jesus’ Davidic lineage have surmised that Joseph adopted Jesus without substantiating that this was legitimate in early Judaism. Most interpreters provide no primary-source evidence for this claim; the most thorough cite m. B. Bat. 8.6 and perhaps one or two biblical passages.Footnote ¹⁴ To be fair, most of this discussion has taken place in commentaries, which are generally not the place for detailed treatments of Jewish adoption. Nonetheless, the dearth of evidence given is striking. Second, I fully agree with Levin that adoption was alive and well in Roman law and would add that it existed in Greek law and practice as well.Footnote ¹⁵

Beyond these points, however, I disagree quite sharply with Levin. The cornerstone of his argument is the assertion that adoption was a non-Jewish concept, and it is this point that is least convincing. To begin with, Levin's case against Jewish adoption is an argument from silence.Footnote ¹⁶ Levin presents no evidence that adoption did not exist in early Judaism; he simply argues that there is no evidence for it. This is key, because if there is no reason to think that early Jews opposed adoption, then Matthew, Luke and Paul, who in many ways seem to operate within a Jewish world of thought, may well bear witness to Jewish adoption. Since Levin's argument is from silence, the most he should conclude is, ‘At present we cannot substantiate that Jewish adoption existed from outside the New Testament; Matthew, Luke and Paul may bear witness to Jewish adoption.’ However, he goes far further; we should be wary of such leaps in logic.

Closely tied to this issue is the problem of scope: Levin draws bold conclusions from a small pool of data that he does not adequately discuss. Although Andrew Lincoln commends Levin for a ‘thorough review of the evidence’,Footnote ¹⁷ in reality Levin devotes only two pages to discussing adoption in Jewish literature and inscriptions.Footnote ¹⁸ Here is a list of the examples he mentions, in order of appearance:Footnote ¹⁹

A glance over this list reveals that several major bodies of Jewish literature are missing. Levin neglects the Scrolls and the targumim altogether. With regard to Philo, Josephus and the Old Testament Pseudepigrapha, he says only that Scott (who treats these corpora) ‘shows that Hellenistic Jewish sources such as Philo are aware of the concept [of adoption], but fails to show that they considered it to have become a part of normative Jewish practice’.Footnote ²³ In addition, several key passages from the Old Testament (e.g. 1 Chron 2.34–5; Ezra 2.61) and rabbinic literature (e.g. t. Sotah 11.20; b. Meg. 13a) are missing. So even to argue that there is no evidence for Jewish adoption outside the New Testament (which, as noted above, is the strongest conclusion that his argument can support), Levin would have to show that Jewish adoption does not exist in these writings.

Furthermore, Levin discusses the examples he does treat in a very cursory manner. He dismisses Gen 15.2, Gen 48.5–6, Exod 2.10, Ruth 4.16–17, Esth 2.7, 15 and Tob 14.13 with a single sentence. Similarly, although he admits that ‘one of the Jewish papyri from Elephantine mentions something like adoption’, he does not explain why this does not qualify as adoption or, if it does, how it squares with his thesis.Footnote ²⁴ Therefore, not only does Levin's argument from silence not support his strong conclusions; the paucity of his examples and the brevity with which he treats them cannot sustain any sweeping judgements about adoption in Judaism as a whole.

A fresh look at the evidence therefore seems necessary. First, however, we must clarify what sort of Jewish adoption Levin is arguing against, since he is unclear on this point. Levin seems to hint at a definition in the following statements:

By ‘adoption’ Levin seems to have in mind a mechanism by which a non-biological child inherits the legal status of the adoptive parent. Notice also that in the latter passage Levin describes this phenomenon as ‘legal adoption’. Since the two passages describe essentially the same phenomenon, it seems that in the former Levin considers Joseph's relationship with Jesus in Matthew and Luke (whatever the cultural background) to be legal adoption. Thus, in Levin's categories there is no place for saying (as some scholars who have cited Levin positively have done) that in Matthew and Luke Joseph adopts Jesus informally rather than legally.Footnote ²⁶ I suggest that Levin is prone to misinterpretation on this point because he is inconsistent in describing his thesis. In addition to the two passages above, consider the following excerpts (emphasis added):

Levin seems to be saying three distinct things: Jews did not (1) consider adoption to be a valid option for Jews (concept), (2) practise adoption (custom) or (3) have official legislation or a process for adoption (law/institution). However, the accent falls on the legal language, which sounds like (3) if one is not reading closely, and Levin never states all three straightforwardly in one place. Therefore, it is understandable why some scholars have cited Levin favourably while actually contradicting him on (1) and (2).

Levin's argument aside, I suggest that the question we should be asking is, ‘Did Jews recognise adoption (i.e. Jesus’ relationship to Joseph in Matthew and Luke) as legitimate for themselves and did they practise it?’ Even if adoption does not exist in Jewish law (i.e. biblical and halakic prescriptions), Jews could have recognised and utilised it. Further, a concept and custom of adoption may well imply the institution. (Indeed, I suspect that most scholars who have affirmed that Joseph legally adopted Jesus have meant this in the concept/custom sense: as far as Jewish society was concerned, Jesus was Joseph's rightful heir.)

3.1 Adoption in the Old Testament

Numerous texts in the Old Testament recount adoption. In Gen 15.2–3, Abram reminds YHWH that since he has given Abram no offspring, ‘the heir of my house is Eliezer of Damascus … a slave born in my house is to be my heir [יורשׁ]’.Footnote ³⁰ Although the text does not specifically mention lineal status, two factors suggest that an adoption, rather than merely the inheritance of possessions, is in view. First, in the four other instances in the Old Testament where ירשׁ has the sense of ‘inherit’ and Eliezer is not in view, it is used of sons (Gen 15.4; 21.10; 2 Sam 14.7; Jer 49.1).Footnote ³¹ Particularly relevant is Gen 21.10, where Sarah tells Abraham that Ishmael shall not inherit alongside Isaac. Here, being an heir is actually a more exclusive privilege than sonship. This suggests that for Eliezer to become Abram's heir would mean that he also becomes his son. A filial relationship also explains why Eliezer would be able to inherit ahead of Abram's relatives, particularly Lot, which would be the normal procedure.Footnote ³² Second, the situation in Genesis 15 resembles adoption customs from the Nuzi tablets. At Nuzi, a childless man could adopt a son to care for him during his life and inherit him at death. However, if the man were to beget a son, the adopted son would have to yield the place of primary heir to the biological son.Footnote ³³ This explains why despite being the heir apparent, Eliezer disappears from the narrative after Genesis 15. Therefore, in Gen 15.2–3, Abram has probably already adopted Eliezer or plans to do so if he continues childless.

Adoption also occurs in Gen 48.5, where Jacob adopts Joseph's sons Ephraim and Manasseh as his own, ‘just as Reuben and Simeon are’. With respect to lineage, Ephraim and Manasseh are now Jacob's children rather than his grandchildren. This has a far-reaching etiological effect: Ephraim and Manasseh become tribal heads with their uncles, a privilege not theirs biologically.Footnote ³⁴

One of the most obvious examples of adoption in the Old Testament occurs in Exod 2.10, where Pharaoh's daughter adopts a Hebrew boy (‘he became her son’) and names him Moses. As a result, Moses is apparently treated as a member of the royal family rather than a Hebrew (2.11). In this case the adoptive parent is obviously not an Israelite, but as we will see below, later Jewish traditions deemed Moses’ adoption as legitimate, amplified its privileges, and even cited it as precedent for Jewish adoption.

Levirate marriage is one of the most cited and contested examples of Old Testament adoption. This is perhaps because it is a legal institution (Deut 25.5–6) and therefore, if admitted, grounds adoption in Jewish law. The law states that if a man dies sonless, his brother shall marry his wife and the firstborn (son) ‘shall succeed to the name of the deceased brother’ (25.6). Levirate marriage thus assumes a posthumous adoption: the offspring of the deceased's wife and brother will be reckoned as the deceased's own child. Levin does not seem to see this as significant: ‘In the case of levirate marriage, the child remains with his birth parents, only assuming his dead childless uncle's name.’Footnote ³⁵ But the transfer of a name (i.e. lineal status) is precisely what constitutes an adoption! Regardless of social location, the levirate offspring is considered ‘son of [deceased uncle]’. Therefore, it seems reasonable to conclude that levirate marriage entails adoption, and thus adoption does exist (albeit implicitly) in Jewish law.

Adoption appears in several other passages as well. In 1 Chron 2.34–5, the sonless Sheshan gives his daughter in marriage to his Egyptian slave Jarha. The resulting son Attai is included in the Israelite genealogy (2.36), which suggests that Sheshan adopted either Jarha or Attai. Regardless, Attai is included in a lineage that should have been closed to him on biological grounds alone.Footnote ³⁶ Ezra 2.61 mentions one Barzillai, ‘who had married one of the daughters of Barzillai the Gileadite, and was called by their name’. The former Barzillai (not the Gileadite) was thus adopted in that he received the name (i.e. the lineal status) of his father-in-law.Footnote ³⁷

In sum, the Old Testament contains numerous cases of adoption.Footnote ³⁸ Although some of them occur within the family line, others do not, and all establish the basic point that non-biological offspring can be reckoned as one's child. These examples do not prove that early Jews recognised or practised adoption, but they form an important precedent that places the burden of proof upon those who assert that adoption was a non-Jewish phenomenon. Furthermore, since levirate marriage implies posthumous adoption, the claim that adoption does not exist in Jewish law must be discarded or at least qualified.

3.2 Adoption in Early Judaism

Two of the earliest non-biblical examples of Jewish adoption come from the Jewish colony at Elephantine in Egypt. The first is the story of Ahiqar (7th–6th cent. bce). Ahiqar is a wise counsellor, scribe and seal-keeper for Sennacherib, king of Assyria (Ahiqar 1–3).Footnote ³⁹ Although childless, Ahiqar declares, ‘I shall nevertheless have a son!’ (2; Lindenberger, OTP). When Sennacherib dies and is succeeded by his son Esarhaddon, the aging Ahiqar summons his nephew Nadin and adopts him so that Nadin can succeed him as scribe and seal-keeper for Esarhaddon (6–7). Ahiqar 8, where the adoption takes place, is unfortunately fragmentary. Yet it is clear that an adoption occurs since Esarhaddon commends Ahiqar, ‘who raised up his [neph]ew to be his son, since [he had] no son of his own’ (12; Lindenberger, OTP; brackets original), and Ahiqar refers to Nadin as his son (18; cf. 20, 22). Ahiqar later retires and Nadin is able to succeed him because of this filial relationship (14–22).

But what relevance does Ahiqar have for Jewish adoption? Clearly, the Jews who possessed the story knew of adoption, and the fact that they preserved it might suggest that they perceived its hero as one of their own. However, the Elephantine text (our earliest) is silent regarding Ahiqar's ethnicity, and the sayings in the latter part of the book presume Near Eastern polytheism, which weighs against Ahiqar's Jewishness.Footnote ⁴⁰ The question of how the sayings relate to the preceding narrative is vexed because the transition is missing. But since the story begins, ‘[These are the wor]ds of one Ahiqar’ (1; Lindenberger, OTP; brackets original), the story and the sayings are presumably connected, so it seems best to conclude that the Elephantine text does not identify Ahiqar as a Jew.

Subsequent authors, however, did view Ahiqar as Jewish. The book of Tobit, which depends on the story of Ahiqar, portrays him as an exiled Jew. The author introduces Ahiqar as Tobit's nephew and the ‘chief cupbearer, keeper of the signet, and in charge of administration of the accounts’ for Sennacherib and Esarhaddon (1.21–2; cf. Ahiqar 1–3). Ahiqar appears later in the story with his nephew, here called Nadab (11.18). Tob 14.10 mentions that Tobit ‘reared’ Nadab (cf. Ahiqar 8, 23, 25, 44) and alludes to Nadab's treachery that the story of Ahiqar recounts in more detail (Ahiqar 23–45). Though Tobit does not explicitly mention Nadab's adoption or call him Ahiqar's son, this is arguably assumed since it is an integral part of the Ahiqar story not only in the Elephantine text, but also the later Syriac, Arabic and Armenian versions.Footnote ⁴¹ The Syriac and Arabic versions also seem to portray Ahiqar as a Jew since at the beginning of the narrative he prays to the ‘Lord God’ and ‘the Most High God’, respectively,Footnote ⁴² and the polytheism of the Elephantine text is not present. Therefore, whatever Ahiqar's ethnicity may be in the Elephantine story, other texts including the staunchly Jewish Tobit apparently had no qualms about presenting Ahiqar (and, consequently, his adoption of his nephew) as Jewish.

The Elephantine papyri also contain legal documentation of a Jewish adoption. The papyrus in question (Brooklyn papyrus 8, ca. 416 bce) is an official declaration made before Widrang, commander of Syene, that records the transfer of the slave Yedoniah from one Jew (Zakkur b. Meshullam) to another (Uriah b. Maḥseiah) who simultaneously frees and adopts him:

The Jewishness of Zakkur and Uriah is certain, since both men appear in other Jewish papyri of the period,Footnote ⁴⁴ and there is widespread scholarly agreement that this papyrus witnesses a manumission to adoption.Footnote ⁴⁵ This document constitutes clear evidence that some early Jews did practise adoption.

A Jewish burial epitaph (2nd cent. bce to early 2nd cent. ce) from Leontopolis may indicate an adoption:

As the editors note, ‘The name Jesus confirms its Jewish character, and Dositheus was a popular Jewish name in Egypt.’Footnote ⁴⁷ The more difficult question is whether adoption or fosterage is implied. Étienne Bernand concludes that Dositheus ‘is probably an adopted child or a θρεπτός [i.e. slave or foster-child]’.Footnote ⁴⁸ Horbury and Noy follow Bernand but specify that Dositheus is ‘the heir (probably an adopted son or a slave brought up in the house. . …)’, citing Gen 15.2–3 LXX (where ἄτεκνος appears) as support.Footnote ⁴⁹ If Horbury and Noy are correct, then even if Dositheus was originally a slave raised by Jesus, adoption should probably be inferred for him to inherit his master. Levin, however, thinks adoption unlikely: ‘It is doubtful whether this can be taken as proof of legal adoption, and in any case no inheritance of status or property is involved.’Footnote ⁵⁰ In fact, inheritance of status or property may be involved; it is simply not mentioned, and this is hardly surprising given that the inscription is a burial epitaph. In my view, two factors tip the scales in favour of adoption. (1) The statement ‘You are my τέκνον, for I departed ἄτεκος’ in such a public and formal inscription suggests that Dositheus is related to Jesus as a natural child would have been. (2) Had the author(s) wanted to communicate that Dositheus was beloved by Joshua but was not necessarily his heir, the vocabulary of θρεπτός would have been far more apropos both lexically and poetically (creating a rhyming effect with ἄτεκος). Adoption therefore seems to be the more likely explanation.

A fascinating instance of adoption appears in the Melchizedek legend of 2 Enoch. The date and provenance of 2 Enoch are disputed. However, it is generally thought that the core of the work is Jewish, and a late first-century ce date is plausible.Footnote ⁵¹ While the Melchizedek legend was once thought to be a later addition to 2 Enoch, recent scholarship has recognised that some form of it constitutes an integral part the work.Footnote ⁵² In the story, Sopanim, wife of the priest Nir, conceives a child without her husband's involvement (2 En. 71.2). Sopanim dies before she gives birth (71.5–9) and her child emerges on his own, fully developed like a toddler and with a badge of priesthood on his chest (71.17–19). Nir and his brother Noe praise God for renewing the priesthood, dress the child in priestly garments and name him Melchizedek (71.20–1). Nir blesses the Lord, saying, ‘I had no child in this tribe who might become the great priest, but this is my son and your servant’ (71.31; Andersen, OTP),Footnote ⁵³ and goes on to include Melchizedek in his family's priestly genealogy (71.32). Nir thus effectively adopts Melchizedek (the son of his wife but not his own biological offspring) as his son to carry on his priestly vocation and line.Footnote ⁵⁴ This adoption is significant, for it has no precedent in the Old Testament and exhibits many similarities to Jesus’ situation in Matthew and Luke.

Jubilees (2nd cent. bce) presents Lot as adopted by Abraham. As Abraham is preparing to leave Haran, Terah instructs him, ‘Take Lot, the son of Haran your brother with you, (as) a son for yourself’ (Jub. 12.30; Wintermute, OTP; parentheses original). Abraham complies (13.1) and thereby initiates an adoptive relationship with Lot.Footnote ⁵⁵ Jubilees notes that after Lot left him, Abraham's ‘heart was sad because his brother's son had separated from him because he had no sons’ (13.18; Wintermute, OTP). This suggests that Lot's departure broke an adoptive relationship Abraham hoped would continue. Interestingly, this adoption appears in neither the MT nor the LXX.

The Old Testament Pseudepigrapha contain numerous references to Moses’ adoption by Pharaoh's daughter. Jubilees recounts how Moses became the son of Pharaoh's daughter (47.9). Likewise, Sibylline Oracles book 3 (2nd cent. bce) describes how the queen took Moses home, raised him and called him her son (3.253–4). According to LAB 9.16 (1st cent. ce), Pharaoh's daughter adopts and names Moses. In Artapanus (3rd–2nd cent. bce), Pharaoh's daughter is barren and adopts Moses, presumably because she has no children (fr. 3, apud Eusebius, Praep. ev. 27.3). Finally, in Ezek. Trag. 36–8 (2nd cent. bce), Moses recounts how Pharaoh's daughter provided ‘all things’ for him, ‘as though I were her own’ (38; Robertson, OTP).Footnote ⁵⁶ These examples show that early Jews were familiar with the concept of adoption and even developed Moses’ adoption beyond the biblical accounts.

Josephus recounts a number of adoptions. Like Jubilees, Josephus mentions that Abram adopted (εἰσεποιήσατο) Lot (Ant. 1.154). Ant. 2.232 narrates how Pharaoh's daughter Thermuthis makes Moses her son (παῖδα ποιεῖται).Footnote ⁵⁷ Interestingly, when Thermuthis tells her father about this, she says that she has adopted Moses not only as her son, but also as the heir (διάδοχος) of his kingdom, a significant addition to the Old Testament account (see also Ant. 2.237). In Josephus’ retelling of Genesis 48, Jacob commands his biological sons to ‘reckon among their number Joseph's sons, Ephraim and Manasseh, and to let them share in the division of Canaan’ (Ant. 2.195; Thackeray, LCL). Josephus returns to this point in Ant. 3.288, using it to explain why Moses includes Manasseh and Ephraim among the heads of the tribes. Thus, for Josephus the adoption of Manasseh and Ephraim had significant lineal implications. In Ant. 2.263, Jethro makes Moses his son (ποιεῖται δ᾿ αὐτὸν υἱόν), another addition to the biblical account. Finally, Josephus describes how the Essenes do not marry, but rather select and educate the children of others, deeming them to be their own relatives (συγγενεῖς ἡγοῦνται, J.W. 2.120). This could be either fosterage or adoption, but Josephus’ note that the Essenes consider such children to be their own relatives may weigh in favour of adoption.

Adoption also appears in Philo. Unsurprisingly, Philo recounts several adoptions from the Old Testament. In Philo's biography of Abraham, Sarah tells Abraham that if he has children with Hagar, ‘the offspring will be yours in full parenthood [γνήσια], but surely mine also by adoption [θέσει]’ (Abr. 250; Colson, LCL). Philo also retells the story of Moses’ adoption (Mos. 1.19; cf. 1.33) and, like Josephus, thinks that this adoption made Moses heir to the Egyptian throne (Mos. 1.32, 149).Footnote ⁵⁸

Many of the examples above from the Pseudepigrapha, Josephus and Philo retell or fabricate Old Testament adoptions. But what, if anything, do these accounts tell us about the early Jews who wrote them? At the very least, such stories are not consistent with Levin's thesis that early Jews uniformly refused to recognise or practise adoption. If early Jews eschewed adoption, we would expect them not to mention biblical adoptions at all or only to recount them when necessary (much as modern Jews and Christians do with biblical instances of polygamy, a practice that they eschew). But on Levin's thesis we would certainly not expect early Jews to stress biblical adoptions and introduce additional instances of adoption into the biblical narrative. That they do so in the ways catalogued above suggests a positive view of adoption.

In sum, early Jews recounted, accentuated and even fabricated biblical adoptions, transformed one gentile adoption into a Jewish adoption, and practised adoption themselves. Although we can hardly claim that all Jews everywhere recognised and practised adoption, some clearly did, and this falsifies the thesis that adoption is a non-Jewish phenomenon.

3.3 Adoption in the Targumim and Rabbinic Literature

Our study of the literature and inscriptions above has already provided substantial evidence that adoption existed in early Judaism. The following discussion of the later targumim and rabbinic literature is therefore not intended to provide primary support for adoption in early Judaism (as if rabbinic prescriptions could tell us precisely what first-century Jews thought and did) but rather plays a supporting role: if these later sources treat adoption in a way that is consonant with what we have already observed in the Old Testament and early Judaism, then it is highly unlikely that Jews in antiquity uniformly eschewed adoption as Levin asserts.

In several cases, the targumim underscore adoption in the Bible. Whereas in Gen 48.5 Jacob tells Joseph, ‘Ephraim and Manasseh shall be mine’ and subsequent children ‘shall be yours’, Targum Neofiti has Jacob declare, ‘Ephraim and Manasseh … will be (called) by my name’ and subsequent offspring ‘shall be (called) by your name’.Footnote ⁵⁹ Far from minimising the implications of Genesis 48, Targum Neofiti emphasises that Ephraim and Manasseh will now bear Jacob's name rather than Joseph's.

The Targum of the Prophets posits an adoption in an attempt harmonise 2 Sam 6.23 and 21.8. The former verse states that Saul's daughter Michal ‘had no child to the day of her death’. However, the latter lists five sons of Michal, ‘whom she bore to Adriel son of Barzillai the Meholathite’. Given the obvious tension between the two passages and the fact that Saul's daughter Merab, not Michal, married Adriel (1 Sam 18.19), most modern scholars assume that the text of 2 Sam 21.8 is corrupt and therefore read ‘Merab’ instead of ‘Michal’ with a few manuscripts. However, Tg. Neb. 2 Sam 21.8, apparently recognising the dissonance with 2 Sam 6.23, glosses ‘the five sons of Michal daughter of Saul’ with ‘the five sons of Merab, whom Michal the daughter of Saul raised’.Footnote ⁶⁰ Thus, on the Targum's interpretation, Merab bore five sons, Michal raised them, and Scripture therefore calls them by Michal's name.

The rabbinic saying most commonly cited to support Jewish adoption is m. B. Bat. 8.6: ‘If a man said, “This is my son,” he may be believed.’Footnote ⁶¹ This statement may be explained in two ways. (1) With respect to inheritance, the man is believed when he claims someone as his son (i.e. heir) because he has no reason to lie – if the person were not his son, he could simply give the inheritance to him as a gift. (2) With respect to Levirate marriage, the man is believed when he claims someone as his son and thereby exempts his wife from Levirate marriage because this does not benefit him, and he could have achieved the same result by divorcing her.Footnote ⁶² Of course, the saying is phrased so absolutely that in theory one might use it to provide a precedent or mechanism for adoption. However, it does not seem to have been intended in this way, so for the purposes of this argument I leave it to the side and focus on stronger examples.

The rabbis find numerous examples of adoption in Scripture. Confronted with the apparent contradiction of childless Michal's five sons (2 Sam 6.23; 21.8), the Tosefta concludes, ‘They were the sons of Merab. Merab gave birth to them, and Michal raised them, so they were called by her [Michal's] name’ (t. Sotah 11.20; brackets original).Footnote ⁶³ Two texts are cited to support this point: Ruth 4.17, where after Ruth gives birth the neighbourhood women declare, ‘A son has been born to Naomi’, and Num 3.1–2, which gives ‘the lineage of Aaron and Moses’ but only names sons of Aaron. For the Tosefta, the operative principle seems to be that one who raises children may rightly be called their parent (while not necessarily denying the legitimacy of the biological parents). Peppard argues that these examples constitute fosterage rather than adoption, and he is correct if one defines adoption in modern terms, as the transfer of the child from one parental unit to another.Footnote ⁶⁴ However, if one defines adoption as the inheritance of lineal status (as I have here), then for the Tosefta these are examples of adoption, not fosterage.

The Babylonian Talmud contains even stronger statements in favour of adoption. In b. Sanh. 19b, the Talmud reiterates the Tosefta's interpretation of 2 Sam 6.23 and 21.8 and concludes, ‘This teaches thee that whoever brings up an orphan in his home, Scripture ascribes it to him as though he had begotten him.’Footnote ⁶⁵ Here the implicit logic of the Tosefta is explicitly stated as a principle. To the Tosefta's proof-texts, the Talmud adds Exod 2.10 (Pharaoh's daughter reared Moses)Footnote ⁶⁶ and Ps 77.15 (Joseph sustained the sons of Jacob) and clarifies that in Num 3.1–2 the sons of Aaron are called by Moses’ name because he taught them. Levin discounts b. Sanh. 19b, saying, ‘The actual talmudic context … is a discussion of the assumed “adoption” of the children of David's wife Merab by her childless sister Michal, also a wife of David, and thus even this refers to adoption by women within an existing household.’Footnote ⁶⁷ However, the Talmud is clearly drawing a general principle from 2 Sam 21.8 that goes beyond the particular situation in that text, so one cannot say (as Levin wants to) that b. Sanh. 19b only pertains to adoption under the specific circumstances evinced in 2 Sam 21.8. Rather, the teaching is that whoever fosters an orphan may be named as its parent. Indeed, in b. Meg. 13a the Talmud derives a nearly identical principle from Moses’ adoption by Pharaoh's daughter, who is clearly not a relative of Moses: ‘This tells us that if anyone brings up an orphan boy or girl in his house, the Scripture accounts it as if he had begotten him.’Footnote ⁶⁸

The strong affirmations of adoption in b. Sanh. 19b and b. Meg. 13a seem to have created a number of problems for subsequent Judaism: which father should be named on a marriage contract? On a deed of divorce? Does the adopted child have the same obligations towards the adoptive family as towards a natural family (e.g. regarding incest or mourning)? Does adoption impact the adopted child's status as Kohen, Levi or Israel? And so on.Footnote ⁶⁹ Because of such difficulties, and perhaps also because b. Sanh. 19b and b. Meg. 13a are haggadic in nature,Footnote ⁷⁰ Judaism ultimately opted not to make adoption a legal institution.Footnote ⁷¹ However, it is important to note that this was a later decision that stood in tension with b. Sanh. 19b and b. Meg. 13a, not a reflection of what these texts actually say about adoption.Footnote ⁷²

The targumim and rabbinic literature therefore provide strong support for Jewish adoption. Numerous passages accentuate biblical adoptions, and b. Sanh. 19b and b. Meg. 13a clearly state that anyone who raises an orphan may be reckoned as his or her parent. The rabbis ground this claim in Scripture, even appealing to verses that modern exegetes might consider special pleading. It is difficult to see how such glowing assessments of adoption could have descended from an early Judaism that uniformly refused to recognise or practise adoption. It is, on the other hand, quite easy to see how they could have arisen from an early Judaism in which adoption was accepted and practised.