Abstract
It is commonly asserted that Targum Song of Songs was composed in Palestine in the seventh or eighth century CE. This article surveys the most significant criteria used to posit that assertion (such as language, Jewish education, and messianism), and it argues that these criteria are either inconclusive or point to a different Sitz im Leben for the Targum. The article then adds one element to the discussion, the use of the late Latin term olibanum, ‘frankincense’, in 4.11. Ultimately, this article argues that the traditional dating and provenance of Targum Song of Songs' composition should be adjusted. A tenth- or eleventh-century time period and a location in either southern Italy or Byzantium better fit the evidence.
Keywords
1. Introduction
Targum Song of Songs (TgSong) is one of several targumim that, in their final form, were composed some time after the Arab conquests. These late targumim include individual targumim to the books of the Writings and Targum Pseudo-Jonathan (TgPsJ) to the Pentateuch. There are certain linguistic and thematic features that unite these individual targumim, and in fact, at the end of this study I will highlight recent work on Targum Chronicles (TgChron) to bolster claims made here. 1 It is important to note, however, that these are individual texts, and one should assess the internal evidence for the Sitz im Leben of each targum before appealing to the other late targumim. Therefore, while the discussion in the present study may ultimately help us better place the late targumim as a group, in the short term, this article is concerned specifically with the Sitz im Leben of TgSong.
The text of TgSong survives in two recensions: a Yemenite recension and a Western recension. The Yemenite recension is comprised of Yemenite manuscripts, and Melamed has produced a diplomatic edition of six manuscripts, using MS London Or 1302 as his base manuscript. 2 He also included Lagard's presentation of the textus receptus (which is closer to the Western recension) for the sake of comparison. The other recension, the Western recension, is comprised of European and North African manuscripts. Using MS Hέb Paris 110 (henceforth, Paris 110) as his base manuscript, Fontela produced a diplomatic edition that includes ten Western manuscripts, as well as two Yemenite manuscripts for the sake of comparison. 3 While the two recensions are largely similar in content and ultimately derive from the same Urtext, the Yemenite recension has suffered from several scribal mistakes and alterations. 4 As such, the Western recension is the superior recension, and in the examples that follow, I use Paris 110 as the base manuscript for translations.
As for the Sitz im Leben of TgSong, it is commonly asserted that it was composed in Palestine during the seventh or eight centuries. 5 There are, however, reasons why this common stance should be questioned, and in what follows, I will briefly survey the most pertinent criteria that have been used to date the Targum. I will show that some of the criteria merely point to an earlier date for the text, rather than the actual date. Other criteria point to a later date than has been proposed, as well as a location outside of Palestine. Ultimately, I argue that we should date the Targum to the tenth century or later and outside of Palestine, perhaps in southern Italy or Byzantium.
2. Language
Many scholars have adopted the theory that the language of TgSong and the other late targumim is Late Jewish Literary Aramaic (LJLA). 6 This is a dialect that, as Alexander succinctly states in the introduction to his translation of TgSong, is ‘a purely literary dialect…created by someone who knew Jewish literature in both eastern and western dialects as well as in the Targumic Aramaic of Onq[elos] and Jon[athan], and was quite happy to combine for literary purposes elements drawn from all three of these dialect groups’. 7 This is indeed the case, but there is one particular way in which the old models of describing the so-called mixed language are often still adhered to, and that is the prioritization of Jewish Palestinian Aramaic (JPA).
Some used to explain the artificial nature of the language of the late targumim (including TgSong) by asserting that it was originally written in JPA but was later corrupted. 8 This prioritizes JPA over the other dialects and assumes that (1) the text must have been written in Palestine and that (2) Jews outside of Palestine would never have utilized Palestinian features in their own compositions. This position is still occasionally maintained, albeit in a more nuanced way. In a recent survey of Jewish Aramaic dialects, Steven Fassberg uses the term ‘Jewish Palestinian Aramaic—Third Stratum’ to describe the language of these late targumim, and in his work on TgSong specifically, Alexander also highlights JPA and objects to any linguistic treatments which would place the Targum outside the land of Israel. 9
This is misleading as it relates, at the very least, to TgSong. In fact, the fundamental, or we might say ‘unmarked’, dialect of TgSong is that of Jewish Literary Aramaic (JLA), the language of Targum Onqelos to the Pentateuch and Targum Jonathan to the Prophets. Every aspect of the grammar and lexicon begins with JLA until it diverges. 10 There are, in fact, a number of consistent and frequent elements from JPA, the second-most influential dialect on TgSong's language, but as Kaufman notes in his initial article on LJLA, ‘these common “Yerushalmite” [JPA] characteristics are merely part of a larger group of features that characterize the literary dialect’. 11 The fact that the unmarked dialect of the Targum is JLA (and not a specific late vernacular dialect), that it is JLA-first and not JPA-first has two relevant ramifications. First, it makes TgSong more universal. Onqelos was the most widely studied Jewish text (see the discussion below), and this may have contributed to the fact that TgSong was an incredibly popular text throughout all of Jewry. Second, it opens up the distinct possibility that the Targum was composed outside of Palestine.
Another linguistic matter is the presence of Arabisms, which indicates a date following the Arab conquests. Alexander suggests that these Arabisms point to a linguistic window between the eighth and ninth centuries when most Palestinian Jews had adopted Arabic as their own language but were not yet utilizing the language for religious purposes. 12 While the presence of Arabisms is significant, one might argue that there is more to take into consideration. Every one of the uncontested Arabisms occurs in a list of gemstones in 5.14 in the Western recension (the Yemenite recension uses Hebrew terms for the gemstones), and due to the odd syntax surrounding the list, the list is likely a later addition. 13 Furthermore, their presence only provides us with an earliest date for the text. In fact, the lack of more Arabisms may be significant as well (assuming that Arabic had become the native language of the targumist, wherever he was). Perhaps it points to a time after this ‘window’ had closed and Arabic was beginning to be used for religious purposes. In such a setting, it is possible that the targumist felt the need to provide a true targum, that is, a targum in Aramaic, not Arabic (or Greek?). If we were to approach the matter from this perspective, perhaps this Targum was composed as a response to such linguistic encroachment into Jewish religious life that the targumist may not have approved of. 14 This lack of Arabisms elsewhere in the book is striking, and it may be a clue as to why TgSong was written in the first place.
3. Themes
3.1. Passover
Even the most cursory reading of TgSong shows that Exodus themes are pervasive, and there is a strong possibility that the Targum was utilized in some way, whether in the synagogue or in private devotion, during the festival of Passover. 15 Elbogen points to the creation of the midrashim on the Megilloth as suggesting a relatively early date for the use of these books in the liturgy, but even if this were the case and a targum were created to accompany the Hebrew, it does not follow that we are dealing with the same TgSong, which is a text that is a unified composition and clearly postdates the Arab conquests. 16 Furthermore, while the biblical text was closely associated with the Exodus quite early, the stipulations that incorporate Song of Songs into the liturgy are only laid out in the minor tractate Massekhet Soferim 14. The setting of this tractate has been a matter of some debate, but its dating is likely no earlier than the eighth century. 17 In fact, it is significant for the purposes of the present study to note Blank's proposal regarding the composition of this tractate. She argues that the tractate was composed at different times and in different places, and that the relevant section of chs. 10–21 was written in Europe only some time prior to the eleventh century. She suggests that Italy or Byzantium are logical locations, and these are precisely the locations that I am proposing for TgSong. 18 It is likely that TgSong was composed after this formal link was established between Song of Songs and Passover, and furthermore, this formal link may have provided the impetus for the Targum. Based on these points, an earliest possible date would be somewhere between the eighth and tenth centuries, but this unfortunately does not provide us with an actual date for the Targum.
3.2. Messianism
Another theme that is quite strong in the Targum is messianism. The text opens with a long midrash on the ‘Ten Songs’, the ninth song being the biblical book Song of Songs, and the culminating tenth song ushering in the eschaton. While the Messiah is not mentioned in this opening verse, the entire midrash is suffused with an eschatological fervor that sets the stage for an overarching exposition on messianic hope that is laid out in the remainder of the Targum. 19
Following the cataclysmic failure of the Bar Kokhba revolt in 135 CE, the rabbinic conception of the Messiah is muted in the Tannaitic literature. In contrast to this, it seems that messianic expectations were always present with varying emphases in popular religious conceptions, as well as in the daily Jewish prayers. 20 At various times over the centuries, particularly during times of persecution, waves of messianic and eschatological texts appear, and later rabbinic literature (such as the Babylonian Talmud) incorporates more and more references to the Messiah. It is not until the tenth century, however, that we find a more definitive exposition of messianic expectations in Sa‘adia's The Book of Beliefs and Opinions. 21 This work quickly became an influential and guiding interpretation throughout Jewry.
After reviewing the messianic emphases in TgSong, Alexander points to one of these waves of messianic literature, a time period that begins with the Persian–Roman wars and flows into the wake of the Arab conquests. It is during this time that a number of messianic texts were composed, such as Sefer Zerubbabel, Pirqei deRabbi Eliezer, and 3 Enoch, and according to him, TgSong. 22 Alexander then argues that this points to a date for TgSong during the seventh or eighth centuries. At the same time, however, he highlights that the eschatology of the Targum is most like the formulation laid out by Sa‘adia, who, as just stated, was from the tenth century. 23 A more reasonable explanation is that our targumist drew from this important rabbinic thinker rather than that their interpretations coincided by happenstance. Accordingly, this places TgSong after Sa‘adia's work in the tenth century.
3.3. Schools
One final theme that deserves mention is the pervasive emphasis on education and proper study. Following the midrash on the ‘Ten Songs’ in 1.1, the text indicates in 1.2 that the Lord gave the Torah to Moses, as well as the Six Orders of the Mishnah and the Talmud. 24 The emphases in these two verses set the stage for much of what follows in TgSong (7.3, 5, etc.). In fact, focused study is the Israelites' requisite task as they await the Messiah (7.14; 8.13, etc.), and when he comes, he arrives as a scholar who will help them ‘suck out’ the reasonings of the Torah (8.1).
Alexander argues that this emphasis on schools and some of the terminology used to describe them points to a Palestinian setting. 25 He goes on to say, ‘the most telling argument in favor of a Palestinian [setting]…is the simple fact that our text mentions only one Yeshibah. There were two great ecumenical schools in Babylonia in our period… but only one in Palestine.’ 26 This may be a bit too literal of a reading, though. This is a Targum concerned with all of Israel, and there are no clear indications of the rivalry between Palestinian and Babylonian schools. Furthermore, such a position not only makes the false assumption that a text like TgSong could only have been created in one of these three schools, but it is also an understatement and misrepresentation of Jewish education, which was not limited to the three main Yeshivot.
Already in the first century of the Common Era, Josephus boasts that all Jews had been educated in the Law and could recall all of the laws if asked (C. Ap. 2.18, 178), and the Babylonian Talmud praises a certain Joshua b. Gamla of the first century CE, for setting an ordinance whereby each district and town was to employ a teacher for a secondary school that all boys should attend (b. B. Bat. 21a). 27 While these literary accounts must be taken as exaggerations and idealizations, there is likely a degree of truth as it pertains to the emphasis on education in Jewish communities. In fact, this emphasis only increases over the centuries, and education becomes a central component in Jewish identity in Late Antiquity and the Medieval Period. 28 In the twelfth century, a Jewish traveler by the name of Benjamin of Tudela traveled from his home in Spain, through Europe, Asia, and Africa, and he kept a detailed account of his travels. Included in the itinerary is a list of the populations of the many communities that he visited and the names of the scholars who lived there. 29 While Benjamin of Tudela's travels postdate the standard dating of TgSong by several centuries (though closer to where I believe it should be dated), many of the Jewish communities that he visited were quite old. Some of the communities in Italy, for instance, existed well before the Common Era and had a longstanding reputation for good scholarship. Clearly, scholarly activity was not restricted to the three main academies in Babylonia and Palestine. 30
The curriculum of these schools is not entirely clear, and it likely varied among different communities. 31 However, there was a place for targumic studies in all of the curricular descriptions that survive. In places where Aramaic was still a spoken language of the people, the targumim were present in academic studies, as well as in private preparations for the upcoming synagogue services. They were also used as a preliminary introduction to the Oral Torah. 32 In descriptions of education in Medieval Europe, where Aramaic was not spoken as a native language, the targumim still had an important role to play. 33 Targum study was grouped together with the study of the Bible, and the study of Targum Onqelos, in particular, was a prerequisite to study of the Talmud, the ultimate destination of Jewish education. 34 Onqelos was therefore the most widely studied Aramaic text, and it would be unsurprising, then, that an Aramaic text composed by someone who was not a native Aramaic speaker would be in a dialect that largely mirrors that of Onqelos. This is particularly significant given the fact that, as stated earlier, the dialect of Onqelos is the unmarked dialect of TgSong.
As to the specific features that possibly indicate an affinity with the Palestinian Yeshivah, Alexander posits two features: the setting of the liturgical calendar and the use of the term ḥaberim (8.13), which seems to refer specifically to scholars affiliated with the Palestinian academy. 35 On these features, one should keep two points in mind. First, the ḥaberim were men who had earned particularly high academic credentials but were not required to live in Palestine; in fact, they were commissioned to be leaders of their local communities. 36 The ḥaberim would, however, gather at various points of the liturgical year. 37 Second, Jewish communities among the diaspora (and their schools) were often able to achieve a degree of independence while maintaining affinities to either the Babylonian academies or the Palestinian one. The case of Byzantium and southern Italy is particularly intriguing in this regard. 38
By the Middle Ages, the schools of Byzantium and southern Italy had become well-respected institutions. 39 While they were independent entities in their own right, they maintained their historic ties to Palestine for centuries. 40 Beginning in the tenth century, as Palestinian influence began to dissipate and Babylonian influence began to increase throughout Jewry, Byzantine communities capitalized on the strengths of both centers. 41 As Holo states, ‘Byzantine Jews felt no inconsistency in participating in the scholarly realms of both of the great Talmudic traditions, and contributing to their respective institutions’. 42 Furthermore, even after the Babylonian schools became more influential throughout the diaspora, Byzantine Jews maintained their relationship with Palestine. 43 Such a situation would actually fit well with TgSong, which draws from both Babylonia and Palestine but seems to have an overall affinity towards Palestinian institutions and practices.
With all of these considerations taken into account, the assumption that TgSong must be written in one of the three great Yeshivot is dubious at best. In fact, I would argue that the emphasis on education in the Targum simply suggests that our targumist is looking at the world around him and seeing that the primary way to maintain Jewish identity—and perhaps a particular rabbinic one at that—is to stress the importance of a proper Jewish education, wherever it is that Jews reside.
4. Other Criteria
4.1. Esau and Ishmael
One of the obvious indicators that TgSong is composed after the Arab conquests is that the targumist refers to Christians and Muslims using the nomenclature of Esau and Ishmael, respectively. One verse that has caused a deal of confusion is 1.7, which reads as follows:
When the prophet Moses's time had come to depart from the world, he said before the Lord, ‘It is revealed before me that this people is going to sin and go into exile. Now tell me, how will they be supported, and how will they reside among the nations, whose laws are harsh as heat and as the scorching noon-day sun in the summer solstice of Tammuz? Why should they wander among the flocks of the sons of Esau and Ishmael, who ally their idols to you for kinship (אירבחל ןוהוועט דל ןיפשמיד)?’
Moses is speaking to the Lord about a future time when the Israelites will once again be in exile, but this time it will be ‘among the flocks of the sons of Esau and Ishmael, who ally their idols to you for kinship’. The targumist unfortunately does not indicate what he means by ‘idols’ or what it would mean for them to ally or associate themselves with the Lord, but it should be noted that the term ועט does not necessarily refer to actual idolatry. It can simply refer to erroneous worship. Regardless, the charge that Muslims are idolaters is particularly startling given Islam's vigorously monotheistic, anti-idolatrous stance. Furthermore, the charge that Muslims are idolaters in any sense is unprecedented in other Jewish literature. 44
There are a few ways to deal with this conundrum. The targumist could simply be unfamiliar with Islam. Loewe proposes that the targumist's ‘formulation implies a Sitz im Leben in which Jewry had not yet come to appreciate the monotheistic insistence of Islam’. 45 The implication is that this could only have been shortly after the Arab conquests and, once again, in Palestine. 46 An alternate way of positing a basic unfamiliarity with Islam would be to set the Targum in a place that is geographically separate from Muslims, though the targumist is aware that Muslims held control over a broad swath of land where Jews were living.
Besides a basic unfamiliarity with Islam, another possible explanation is that the targumist is making a unique, literary extension of an older trope in rabbinic literature whereby Ishmael is regarded as an idolater. The interpretations of Gen. 21.9 are a good exemplification of this. The Hebrew text states as follows:
:קחצמ םהרבאל הדלי רשא תירצמה רגה ןב תא הרש ארתו
Sarah saw the son whom Hagar the Egyptian had borne to Abraham playing. (JPS)
Commenting on קחעמ, Jerome states already in the fourth century that some Jews interpreted this as a reference to idolatry. 47 This is confirmed by a statement in Genesis Rabbah, for instance, where R. Ishmael reportedly says, ‘This term SPORT [קחעמ] refers to idolatry, as in the verse, And rose up to make sport (Ex. xxxii, 6)’. 48 In varying degrees, this is picked up in the Palestinian targum tradition, as well as the late TgPsJ. It is possible that the TgSong targumist is simply taking this tradition and uniquely extending it to include not merely the ancestors of the Muslims, but contemporary Muslims as well. While other late Jewish texts (like TgPsJ) may continue this trope with an oblique insult against Muslim ancestors, the TgSong targumist is going much further. 49 Once again, this could be due to ignorance, but it does seem unlikely that the targumist would know that Muslims claim Ishmael as their ancestor but was unaware of their monotheistic, anti-idolatrous stance. I would argue that we have a targumist who is neither ignorant nor careless. He is polemical. The question is whether the targumist is using a wholly unique approach or whether he was living in a milieu in which such an accusation was acceptable.
When one looks outside of Jewish literature, one occasionally encounters Christian statements alleging that Muslims are idolaters. 50 The most rigorous polemical approach with such a charge only emerged in about the ninth century in Byzantium. 51 This polemic became popular in Greek-speaking Christian communities and indeed charged Muslims with idolatry. 52 Using the Qur'an's ardent monotheistic claims in Surah 112 (al-Tawḥid, ‘The Unity’) against itself, the polemicists argued that the Arabic word ⋅amad does not point to monotheism, as Muslims claim, but to idolatry. 53 The Qur'anic use of ⋅amad, which is only attested here in the Qur'an, is not entirely clear, but according to Hanson, the term likely ‘means “solid, massive” or more figuratively, “permanent, everlasting, eternal”’. 54 The eighth–ninth century Arabic-speaking writer Theodore Abu-Qurrah translates it into Greek with the term σϕυρόπηχτος, a word meaning ‘indivisible’ or ‘hammered together’. Later, Nicetas of Byzantium, who did not speak Arabic and probably lived in ninth century Byzantium, takes the figurative translation further by using the term ὁλόσϕαιρος, ‘all-spherical’, and then ὁλοσϕυρος, ‘made of solid, hammer-beaten metal (i.e. entire, a compact whole)’. 55 This later term is frequently used in Greek polemics to charge that the Qur'an teaches the worship of a material god—idolatry. In fact, it even becomes part of the Byzantine conversion ritual in the statement, ‘And before all, I anathematize the God of Muhammad about whom he [Muhammad] says, “He is God alone, God ὁλοσϕυρος (made of solid, hammer-beaten metal); He begets not and is not begotten, nor is there like unto Him any one”’. 56
In the end, I am tentatively inclined to favor an interpretation whereby the targumist is uniquely and polemically extending the earlier rabbinic interpretation of Hagar and Ishmael's idolatry. However, the possibility that our targumist was living in a place where such a charge was acceptable is enticing and plausible, given the popularity and effectiveness of this polemic. This would have allowed the targumist to utilize Christian polemics to his advantage and then say that both Christians and Muslims were idolaters. If this is the case, then the implications are not only a later dating of TgSong, but also a familiarity with Greek intellectual discussions and a location outside of Israel. 57
4.2. ‘This Unclean Land’
This brings us to another phrase that has caused a deal of controversy, the final verse of TgSong, 8.14:
At that moment, the elders of the congregation of Israel will say, ‘Flee from this unclean land, my beloved, Lord of the world. Settle your Presence in the heavens above, but when we pray before you in the time of our distress, may you resemble a gazelle which keeps one eye closed and one eye opened when it sleeps, or like a young gazelle of a stag which looks backward when it flees. May you so gaze on us and observe our distress and misery from the heavens above until the time that you are pleased with us. Then may you redeem us and bring us to the mountain of Jerusalem. The priests will send up sweet incense to you there.’
The targumist, in the mouth of the elders, says to the Lord, ‘Flee from this unclean/polluted land’. Melamed argued that the targumist would not have referred to Israel as ‘impure’, and the logical conclusion is that the targumist must have been living in and speaking about Babylonia. 58 Others have countered that the targumist was speaking like one of the prophets and making just such a case: the land of Israel has become corrupted and polluted. 59 All of this, however, may be a case of being too atomistic in the reading. In fact, the word אערא can have a more global sense, and furthermore, it is notable that the entire Targum has an exile-to-Palestine motion. As Alexander posits in his fine outline of the book, there are three cycles that occur, all going from exile to Palestine: the exile in Egypt to Solomon's temple; the exile in Babylon to the Hasmonean temple; the exile among the sons of Esau and Ishmael to the Messiah's temple. 60 Why is it then required that this text is precluded from being written in the diaspora by someone who is longing to go to Palestine, a longing that is tempered by a targumist who denounces presumptuous entrance into Palestine (2.7; 3.5)? In fact, the last statement of this very verse maintains this motion: until the time that you are pleased with us and you redeem us and bring us to the mountain of Jerusalem’. Are we to think that the author is in Palestine, but not quite in Jerusalem? It is more likely that this verse exemplifies the tenor of the entire Targum which is going from exile to Palestine.
4.3. Olibanum and a Proposed Sitz im Leben
I would like to add to this discussion a heretofore unexplained term in the Targum which may shed some additional light on the text's Sitz im Leben. TgSong 4.11 ends with the following statement in the Paris 110 manuscript and several other manuscripts of the Western recension:
ןונבילוא םשוב חירכ דינהכ ישובל חירו
And the scent of your priests' robes is like the scent of the spice of frankincense.
The Aramaic translation is clear enough, except for the final term, ןונבילוא, which is not attested in other Aramaic or Hebrew texts. 61 Before looking at the term's meaning, we should note the textual variants that are attested among the manuscripts. Note the following:
Paris 110: ןונבילוא
Western variant: ןינבלוא
Yemenite: ןנבל (ןימסוב/תרוטק) חירכ: as the scent (of the incense/spices) of Lebanon
Yemenite MS London Or 2375: ןנׇבלִוׄא
Sperber: ןנבל וא ןימסוב: spices or incense
Some of the other Western manuscripts, as well as Largard's presentation of the textus receptus, have a form similar to that which is attested in Paris 110, which Levy vocalizes as ןינִבׇּלְוּא in his Chaldäisches Wörterbuch. 62 The Yemenite recension, which tends to eliminate or change unusual words and forms in TgSong, largely avoids the difficult word. Most of the manuscripts have the phrase ןנבל חירכ, ‘like the scent of Lebanon’, as in the underlying Hebrew. A few manuscripts insert תרוטק, ‘incense’, or ןימסוב, ‘spices’, resulting in ‘like the scent of the incense/spices of Lebanon’. 63 One Yemenite manuscript (MS London Or 2375), however, has the word in question in the phrase ןנׇבלִוׄא חירכ, to which a second hand has added ןימסב. This particular manuscript is Sperber's base text, and being unfamiliar with the word himself, Sperber unhelpfully adds a space between the וא and the ןנבל so that his text reads ןנבל וא ןימסוב, ‘spices or incense’. The biggest problem with this reading is that הנובל, ‘incense’, is a mass noun that is never attested in a plural form. With all of this said, it is safe to assume that this unusual form is original to TgSong—given its attestation in both recensions and its pervasive presence throughout the more reliable Western recension.
While it is largely agreed that this word is related to frankincense in some way, neither the particular form of the word nor why it means ‘frankincense’ have been adequately described. 64 Levy states that the word means ‘Weihrauchbaum’ or ‘incense tree’, and he lists the Greek term λίβανος in parenthesis. 65 Jastrow posits that the term is some kind of ‘unguent made of incense’ and states that the term derives from אתׇּנְוּבלְ. 66 Alexander simply gives the gloss ‘frankincense’ and indicates that the form ‘is probably simply a variant of lbwnt' = “frankincense”’. 67 None of these explanations is sufficient.
In actuality, it is likely that this is the Latin term olibanum, which does indeed mean ‘frankincense’. 68 Given that the final /-um/ of Latin terms becomes a final /-n/ as they enter Greek, this is the most likely explanation for the term. 69 Note that in this word, the change to final /-n/ would result in olibanun, which matches the form as attested in the Paris 110 manuscript (ןונבילוא). The forms with a final ןי- attested in several manuscripts evidence an unfamiliarity with the word and an attempt to understand it as a masculine absolute plural.
What is significant for the purposes here is that this term is a post-classical Latin term that is first attested in tenth- and eleventh-century manuscripts in Italy. 70 The term eventually spreads to a number of romance languages, but the startling issue is the incredibly late date (and location) of the term. The standard classical Latin term for frankincense is tus/thus. 71 Though rare, Latin occasionally attests the use of libanos, which entered the Latin lexicon via Greek and is ultimately of Semitic origin. 72 The form with the initial /o/, however, is unique, and it is this form, the form in TgSong 4.11, that is incredibly late. Scholars have proffered several explanations for this initial /o/, but that is of no real consequence for our purposes. 73 What is more important is that the mere presence of this term has a number of potential implications for our understanding of the Targum's original setting. Though one word can only carry so much weight, I would suggest that the interpretation here may better account for all of the evidence thus far presented regarding the Targum's Sitz im Leben.
I therefore suggest that we push the date of the Targum forward by at least two centuries to the tenth century. We should also take into consideration the possibility, and even the likelihood, that the Targum was composed outside of Palestine. As I have noted in the earlier survey, there are significant reasons to doubt the assignment of the Targum to eighth-century Palestine. Some of the topics more seamlessly fit a later date, and furthermore, the overarching thrust of the Targum, which repeatedly goes from diaspora or exile to Palestine, suggests a setting outside of Palestine. Furthermore, the particular form of olibanum, with a Greek morpheme, suggests more specifically that TgSong may have been composed in a location where Greek was actively spoken, and southern Italy or Byzantium are viable options for the proposed time period.
If we allow ourselves to look at other LJLA texts, Leeor Gottlieb's recent work on TgChron is instructive, where he highlights the use of the ethnic appellation ייארגנוה, ‘Hungarians’, in both TgChron and TgPs. This is another datable term, like olibanum, and it does not come into existence until the ninth century. 74 He asks, ‘Is there any reason that a learned scholar, well versed in rabbinic lore and intimately familiar with earlier Targums and their translation techniques, could not produce such a work in Byzantium, or even in Italy, for that matter?’ 75 It is precisely these two locations that seem particularly appealing to me. Southern Italy was a learned, multi-cultural, multilingual area, and we know of important and popular Jewish texts that were created there at about this time, such as Sefer Josippon—another text quite interested in the history of the Israelite people. 76 As for Byzantium, this would account for the possibility that the targumist may have been familiar with Greek polemics which charged Muslims with idolatry. Furthermore, this would be geographically closer to communities who continued to speak Aramaic, and this may account for the Syriac words that are attested in LJLA texts, though not frequently in TgSong. 77 Significantly, both Italy and Byzantium maintained strong relationships with Palestine, while also incorporating Babylonian influences. This could account for the broad and diverse influences that are discernible in TgSong.
5. Conclusion
In conclusion, I would suggest that both the common dating of TgSong to the seventh or eighth century and the insistence that the text must have been composed in Palestine are assertions grounded on weak claims. In fact, a very different setting for TgSong comes together when all of the criteria are assessed. It is a setting in which the targumist, who possibly speaks Greek, is living in a diaspora community in the tenth or eleventh century, and he is longing for the Messiah's arrival and a return of all Jews to Jerusalem.
Footnotes
1.
Concerning shared linguistic features, see Dalman 1960 [1905]: 27–39; Cook 1986: Chapter 5. Concerning shared thematic features, see Levine 1973: 3–5; Brady 2009, 2014.
2.
R.H. Melamed 1921.
3.
Fontela 1987.
4.
R.H. Melamed 1921: 15; Fontela 1987: 139; Alexander 2003: 6–7.
5.
Loewe 1966: 163–69; Alexander 2003: 55, 58; Lieber 2014: 37; Fishbane 2015: 267.
6.
Kaufman 2013a [1993].
7.
Alexander 2003: 10. In the introduction to his translation of TgSong, Alexander oddly does not use the term ‘Late Jewish Literary Aramaic’ to describe the language, and in fact, he seems to disagree with Kaufman's approach (without stating Kaufman's position or citing his work) which would place the Targum outside of Israel. He does, however, use the term to designate TgSong's dialect in a later profile of TgSong (2011: 120). Aside from this quote, there are many issues with Alexander's presentation of TgSong's language. See Litke 2016: 16–19.
8.
Fontela 1987: 116 and Heinemann 1971: 129. Similarly, see Knobel's assessment of TgQoh (1976: 89–96) and Levine's of TgRuth (1982: 377).
9.
Fassberg 2015: 85; Alexander 2003: 11–12.
10.
Litke 2016: 423–39.
11.
Kaufman 2013a [1993]: 5.
12.
Alexander 2003: 55.
13.
Alexander 2003: 210–13. Alexander also argues that ןמ רב, ‘except’, is being used in a way more like Arabic bada in 4.3 and 6.7, but an Aramaic understanding of this particle is acceptable in the context.
14.
On the use of Greek in diaspora Medieval synagogues, see de Lange 2012. On Jewish translations of the Bible into Arabic, see Griffith 2013: 155–74.
15.
The ties to Passover are made clear in the introductory verse of the Targum, which opens with a list of ten songs. The culminating tenth song unites Passover with eschatological hope in the following statement: ‘And the exiles are going to utter the tenth song when they go out from their places of exile, as it is written and delineated by the hands of the prophet Isaiah: “This song shall become a joyful thing for you, as on the night when the festival of Passover was sanctified… (Isa. 30.29).”’
16.
Elbogen 1993: 150; Alexander 2003: 13.
17.
Strack and Stemberger 1996 [1991]: 227–28.
18.
Blank 1998: 59–72, 1999: 4–5. See also Lerner 1987: especially 397–400.
19.
Brady 2009. On the midrash in general, see Alexander 2003: Appendix A; Kugel 1982.
20.
On the eschatological nature of the Amidah in particular, see Alexander 2007: 228–29.
21.
See the translation by Rosenblatt (1948: 290–322); see also Sarahek 1968: 27–50.
22.
See Reeves 2005; Himmelfarb 2017; Alexander 2009 [1983].
23.
See Alexander 1998: 466; 2003: 56.
24.
Arguments that this must be a reference to the Palestinian Talmud are unfounded given the presence of grammatical and lexical features from Jewish Babylonian Aramaic (and more specifically, the Babylonian Talmud); see Litke 2017: 92–94.
25.
Note the mention of the ḥaberim in 8.13 and references to the setting of the calendar (see the discussion below).
26.
Alexander 2003: 60 (emphasis his); Loewe 1966: 163. Both Alexander and Loewe point to a fragmentary midrash from the British Museum (MS Or 5554a) in their arguments. This text elevates the Babylonian schools to the exclusion of the Palestinian school when commenting on Song 8.14. It does not necessarily follow, however, that the targumist was consciously writing a response to this interpretation or even aware of it. In fact, the Targum does not address schools in this verse at all, which would then ultimately be an argument from silence.
27.
The Palestinian Talmud has a similar account about a different individual (y. Ketub. 8.11, 32c).
28.
Generally, see Moriel 2007; Bortniker 2007. See also Hirshman 2009: 109–20. For a thorough analysis of Medieval education that is drawn from the Cairo Genizah materials, see Goitein 1971: II, 171–272. While they likely overstate the case and oversimplify what is a complex historical situation, the stress on education among Jewish communities allows Botticini and Eckstein (2012) to argue that widespread education was the primary explanation for the unique features of Jewish economic history.
29.
See Adler 1907.
30.
Another example of high literary skill and scholarly acumen is the creation of piyyutim, which occurred throughout the diaspora. Generally, see Rand 2014; Münz-Manor 2010; Lieber 2010.
31.
Note, for instance, that Sephardic and Ashkenazi Jews had different educational emphases. See Talmage 1987; Zimmels 1958: 136–63.
32.
Houtman 2014: 81–86.
33.
See Goitein 1971: II, 205–11. For a treatment of the bilevel educational system presented in ‘The Book of the Statutes of the Torah’ (Sefer Ḥuqqei ha-Torah), see Kanarfogel 2001; for an educational description specific to twelfth-century Byzantium, see de Lange 1994.
34.
Houtman 2014: 86–93.
35.
Others have pointed to the mentioning of the Av bet din and the Sanhedrin as other examples indicating a Palestinian affinity (Loewe 1966: 163), and the comments below apply to them as well.
36.
Goitein 1971: II, 201; Gil 1997 [1983]: 505–506, 656.
37.
Goitein 1971: II, 201.
38.
For the following section, I largely draw from Holo 2009: 78–96.
39.
For Byzantium, see de Lange 1994: 116–17.
40.
Gil 1997 [1983]: 547.
41.
This contrasts with communities in the Iberian Peninsula, which heavily favored the Babylonian school. On the decline of the Palestinian Yeshiva in the late eleventh and twelfth centuries, see Gil 1997 [1983]: 774–76.
42.
Holo 2009: 87.
43.
Gil 1997 [1983]: 545–49.
44.
There is one other late targum which shares the same terminology, TgPs 69.10, דרקיאל ןוהתוועט ןוהפתשמב דל ןידסחמד איעישר, ‘the wicked dishonor you by allying/associating their idols to your glory’. This text is also concerned with exile, but it is unclear which exile the targumist is referring to.
45.
Loewe 1966: 165.
46.
Alexander 2003: 57.
47.
Hayward 2010 [1989]: 115.
48.
See Gen. Rab. 53.11, translated by Freedman in Freedman and Simon 1961 [1939] (italics his).
49.
Compare Hayward (2010 [1989]) and Ohana (1975) on whether TgPsJ is polemical towards Muslims.
50.
One possible early example is a text known as De Haeresibus, which is attributed to John of Damascus. In it the author defends the veneration of the cross against the Muslim charge of idolatry by countering that Muslims similarly rub and kiss the Black Stone, the Ka'ba. See Sahas 1972: 136–37; Hawting 1999: 83–87.
51.
One should also note that there are later Christian texts in Medieval Europe which refer to Muslims as idolaters. See Southern 1962; Daniel 2000 [1960]; Akbari 2009: especially Chapter 5.
52.
Many thanks to Sidney H. Griffith for bringing this polemic to my attention (personal communication). See Meyendorff 1964; Khoury 1969, 1972; Rigo 2009.
53.
For reference, note the following two English translations of the Surah: ‘Say, “He is Allah, [who is] One, Allah, the Eternal Refuge [⋅amad] He neither begets nor is born, Nor is there to Him any equivalent”’ (Sahih International), and ‘Say: He, Allah is One Allah is He on Whom all depend [⋅amad]. He begets not, nor was he begotten; And none is like him’ (Shakir). I have added the ⋅amad in square brackets for clarity.
54.
Hanson 1996: 62. Concerning John of Damascus' earlier understanding of the oneness of God as it relates to Christianity and Islam, see Sahas 1972: 75–77. Wehr (1994: 613) gives the translation ‘lord, eternal, everlasting’.
55.
Hanson 1996: 63–64. For the latter term, see Lampe 1961: 950; Sophocles 1900: 803.
56.
Hanson 1996: 55. This terminology was used in the formal abjurations into the twelfth century. For the text of the abjuration, see Montet 1906.
57.
On the decline of Greek in the Near East, see Wasserstein 2003. Though rare, there are instances in which an oddly spelled word in TgSong may reflect a Greek phonological influence, as may be the case with רימ (TgSong 4.14; 5.13) for the more common רומ, ‘myrrh’; see Litke 2017: 99–100.
58.
E.Z. Melamed 1971: 215.
59.
See Heinemann 1971; Loewe 1966; 163–64; Alexander 2003: 58.
60.
Alexander 2003: 14–19.
61.
This final phrase is translating the underlying Hebrew ןונבל חירכ דיתמלש חירו (‘And the scent of your robes Is like the scent of Lebanon’, JPS).
62.
Levy 1866: I, 29. Some of the Western manuscripts do not have םשוב (alternatively spelled םסוב).
63.
Note that the latter option with ןימסוב is grammatically incorrect since it is in the absolute state rather than the construct state.
64.
Though not evidenced in Song of Songs Rabbah, it should be noted that the interpretation of ‘Lebanon’ as ‘frankincense’ at 4.11 is attested as early as the Vulgate, where it is translated as tus/thus (on which, see below).
65.
Levy 1866: I, 29.
66.
Jastrow 2005 [1903]: 25.
67.
Alexander 2003: 140 note jj.
68.
Many thanks to Edward M. Cook for making this initial recommendation (personal communication).
69.
Butts 2016: 6.
70.
du Cange 1983–87: 6, 42; Blatt 1978: 440.
71.
Lewis & Short 1987 [1879]: 1, 919.
72.
This is seen as early as the Vulgate of Sirach 24.21; 39.18.
73.
The options, none of which are entirely convincing, include the possibility that it derives from the Greek article ὁ, an abbreviation for oleum libani, or the Arabic article al. See Müller 1974: 59.
74.
Gottleib 2012: 348–50. Many thanks to Leeor Gottlieb for making his dissertation available to me.
75.
Gottlieb 2014: 7.
76.
See Dönitz 2012, 2016; Bowman 1993.
77.
See Kaufman 2013b; Bhayro 2014; Litke 2017: 95–96.
