The Targums are the Aramaic translations - or rather paraphrases - of the books of the Old Testament, and, in their earliest form, date from the time when Aramaic superseded Hebrew as the spoken language of the Jews (see Hebrew Language). In their origin they were designed to meet the needs of the unlearned among the people who had ceased to understand the Hebrew of the Old Testament. In the absence of any precise evidence on the point it is impossible to give more than a rough estimate as to the period at which Hebrew, as a spoken language, was finally displaced by Aramaic. It is, however, certain that the latter language was firmly established in Palestine in the 1st century A.D. By that time, as we know from many sources, Aramaic was not only the language in common use, but had also received official recognition,' despite the fact that Hebrew still remained the learned and sacred tongue. Hence we may reasonably infer that the mass of the people had adopted Aramaic at a considerably, earlier period, probably, as early as the 2nd century B.C., and that the need of Aramaic translations of the sacred text made itself felt but little later. By the Jews 2 the introduction of Targums is ascribed to Ezra; but this tradition, which probably owes its origin to the Talmudic explanation of Neh. viii. 8, 3 is inconsistent with the linguistic evidence furnished by the postexilic literature of the Old Testament, and must be rejected as unhistorical, if only because the process by which Aramaic took the place of Hebrew was admittedly a very gradual one. The Talmudic tradition, however, is, doubtless, correct in connecting the origin of Targums with the custom of reading sections from the Law at the weekly services in the synagogues, since the need for a translation into the vernacular must first have arisen on such occasions. As we know from the New Testament, the custom of reading in the synagogues both from the Law 4 and from the Prophets 5 was well established in the 1 st century A.D.: its introduction, therefore, will date from a much earlier period. The practice of accompanying these readings with a translation into Aramaic is, further, so generally recognized by the 2nd century A.D. that the Mishna 6 takes it for granted, and merely inculcates certain regulations to be observed by the Meturgeman (translator), who had by this time acquired a definite status. From it we learn that the Meturgeman, who was distinct from the reader, translated each verse of the Law into Aramaic as soon as it had been read in Hebrew: in the readings from " the Prophets " three verses might be read at a time. Later regulations are also laid down in the Talmuds in order to prevent any appearance of authority attaching to the translation, and also to ensure reverential 1 Cf. Dalman, Die Worte Jesu, p. 2 f.; Grammatik des jud.- palcist. Aramdisch, 2nd ed., p. 9 f.
2 Sanhedrin, Jer. Meg., i.
3 Nedarim, 37b; Jer. Meg., iv. - " and they read in the book, in the law of God, this is the Scripture, wntrz (R.V. distinctly), this is the Targum." Acts xv.
Luke iv. 16 f.; Acts xiii. 14, Meg. iv. 4-6, 10. treatment on the part of the translator.' Elsewhere, 2 we only find references to certain passages of Scripture, viz., the stories of Reuben and Tamar (Gen. xxxv. 22 and xxxviii.), the two accounts of the golden calf (Exod. xxxii.), the blessing of the priests (Num. V. 22 f.), the stories of David and Amnon (2 Sam. xi., xii. and xiii.), which might be either read and translated, or only read and not translated, or (according to a different tradition) neither read nor translated. It is noticeable that none of the passages cited conveys any rules or information as to the character of the translation to be employed. Judging by the contents of our existing Targums, and the Targumic renderings given in Jewish literature, it is improbable that any definite system of interpretation was ever formally adopted, the rendering into the vernacular being left to the discretion of the individual Meturgeman. At first, no doubt, the translator endeavoured to reproduce the original as closely as possible, but, inasmuch as his object was to give an intelligible rendering, a merely literal rendering would soon be found to be insufficient, and he would be forced, especially in the more difficult passages, to take a more elastic view of his obligations. To prevent misconception he must expand and explain what was obscure, adjust the incidents of the past to the ideas of later times, emphasize the moral lessons to be learned from the national history, and, finally, adapt the rules and regulations of the Old Covenant to the conditions and requirements of his own age. As time went on the practice of introducing additional matter of an edifying character grew in popular favour, and was gradually extended. Thus, by degrees, the reproduction of the original text became of secondary importance, and merely served as a pretext for the discussion of topics that had little or no bearing on the context. The method, by which the text was thus utilized as a vehicle for conveying homiletic discourses, traditional sayings, legends and allegories, is abundantly illustrated by the Palestinian and later Targums, as opposed to the more sober translations of Onkelos and the Targum to the Prophets.
It would, however, be incorrect to suppose that the translation of the text was left entirely to the individual taste of the translator. The latter is rather to be regarded as the representative of the age in which he lived, and his interpretation is to be taken as reflecting the exegesis of that period. That there were certain limits beyond which the translator might not venture, without incurring the censure of the authorities, may be inferred from the few instances of translation which are mentioned with disapproval in the Mishna and elsewhere. Thus the rendering of Lev. xviii. 21 a by " Thou shalt not give any of thy seed to an Aramean woman to make her conceive " is censured, presumably because the prohibition of Molech worship is thereby ignored. 3 In the same Mishnic passage it is forbidden to render Lev. xviii. 7 as if the text had " his father " and " his mother." 4 Yet another translation (that of Lev. xxii. 28) is mentioned with disapproval in the Jerusalem Talmud, 5 though it has been preserved in the Targum PseudoJonathan ad loc.° A definite rule for guidance in translating is apparently preserved in the Tosefta, 7 where it is stated that " he who translates quite literally is a liar, while he who adds anything is a blasphemer," Exod. xxiv. 10, " and they saw the God of Israel " is cited as an example. It is argued that the literal rendering of this passage is inadmissible, because no man has ever seen God; on the other hand, the insertion of the word " angel " before God would be blasphemous. The correct rendering is stated to be " and they saw the glory of God." But it is doubtful if the rule here given was ever intended to Tos. Meg., 3; Jer. Meg., iv. 3--3; Sota, 39h; Sopherim, xi. 1, xii. 7, xiv. 2.
Meg., 25, 25b; cf. Ginsburger, M.G.W.J., xliv. I f.
3 Meg., iv. 9; cf. Jer. Meg., iv. 9; Sanhed., ix. 1, where the meaning is given as - " He who marries an Aramean woman and raiseth up children by her raiseth up enemies to God "; for another explanation, see Ginsburger, M.G.W.J., xliv. 5 f.
4 Cf. Berliner, Targum Onkelos, ii. p. 85 f.
5 Meg., iv. to. Cf. Ginsburger, l.c.
7 Tos. Meg., end.
apply to more than the particular type of passage exemplified: if it had been applied generally, it would have clashed with the whole trend of Midrashic and Targumic paraphrase.
There can be little doubt that the Targums existed for a long time in oral form. They belonged to the class of traditional literature which it was forbidden to write down, and, so long at least as the Targum tradition remained active, there would be little temptation to commit it to writing. But it is highly probable that this prohibition, in the case of the Targums, was mainly enforced with respect to those parts of the Old Testament which were read in the synagogal services, e.g. the Law and the Prophets, and that it was less rigidly observed in regard to the other portions of Scripture: a written translation of the lafter would be of special value for the purpose of private study. Hence there is no need to reject the tradition as to the existence of a written Targum on Job in the time of Gamaliel I. 8 (1st century A.D.), especially as references to Targum MSS. occur in the Mishna and elsewhere. 9 But, as Dalman has pointed out,' 0 it was not these manuscripts, but the living tradition of the learned which was recognized as authoritative throughout the period which closes with the compilation of the Talmud.. .. The official recognition of a written Targum, and therefore the final fixing of its text belongs to the post-Talmudic period, and is not to be placed earlier than the 5th century.
I. Targums On The Pentateuch (t) The so-called Targum of Onkelos admittedly owes its name to a mistaken reference in the Babylonian Talmud." In its original context, that of the Jerusalem Talmud, 12 the passage refers to the Greek translation of Aquila. With the exception of this one reference, the Targum is always introduced in the Babylonian Talmud by the phrase " as we translate " (irn:inr_-r7), or " our Targum " (p' 1 ?urn): it is probable, therefore, that the name of the author, or authors, was unknown to the Babylonian Jews. It is first quoted under the title of the Targum of Onkelos by Gaon Sar Shalom (d. A.D. 859). According to Dalman, 13 its language differs in many material particulars from the Aramaic dialects of the Palestinian and Babylonian Talmuds, and is more closely allied to the biblical Aramaic. On the linguistic side, therefore, we may regard Onkelos " as a faithful representative of a Targum which had its rise in Judaea, the old seat of Palestinian literary activity." It is not, however, to be regarded as a reproduction in written form of a Palestinian translation, but rather as an official translation of the Law, in the Judaean dialect, which was carried out in Babylon, probably about the 4th century A.D.: in its final form, according to Dalman (l.c.) it cannot be earlier than the 5th century. The translation, as a whole, is good, and adheres very closely to the Hebrew text, which has not been without its influence on the Aramaic idiom; at times, especially in the poetical passages, a freer and more paraphrastic method is employed, and the version shows evident traces of Halakhic and Haggadic expansion. The Hebrew text used by the translators appears to have been practically identical with the Massoretic. The version was held in high esteem in Babylon, and, later, in Palestine, and a special Massora was made for it. The latest edition is Berliner's reprint (1884) of the Editio Sabbioneta (1557) Of all the extant Targums that of Onkelos affords perhaps the most characteristic and consistent example of the exeget i cal methods employed in these works. Two principles may be said to have guided the translators. On the one hand, they had, as their primary object, to produce a faithful rendering of the original which at the same time would be intelligible to the people: for this purpose a purely literal translation would be insufficient. On the other hand, they regarded it as necessary to present the sacred text in such a manner as best to convey the particular form of interpretation then current. But later Jewish exegesis was especially concerned to eliminate everything in the sacred writings that might give rise to misconception with respect to God on the part of the unlearned. Hence we find various expedients adopted in the Targums for avoiding any reference to the Deity, which might be misunderstood by the people, or which involved apparent irreverence. Examples of this peculiarly Targumic method are: (I) the insertion of " word " (x1n^n), " glory " (siip'), " presence " (x7':w) before the divine name, when God is referred to in his 8 Tos. Shabb.; cf. Jer. Shabb., xvi.; Bab. Shabb., I I 5a; Sopherim, v. xv.
9 Jad. iv. 5, and see the preceding references.
"° Grammatik des jiidisch-paldstinischen Aramdisch, p. 12 f.
" Meg. 3a. n Meg. i. 9. u Gramm. p. 12 f.
dealings with men; (2) the insertion of the preposition " before " (c1p) when God is the object of any action; (3) the use of the passive for the active voice, e.g. nnp for yii' or mtn; "p y'aw for ynr'; ' uni for -Inv, ras', in', tin; n'nn for i:i; (4) the use of periphrasis for the more pronounced anthropomorphisms, such as " to smell," " to taste," or when the use of the status constructus might seem to bring God into too close connexion with men or things; (5) the use of different expressions, or the insertion of a preposition before the divine name, when God is compared to man, or the same action is predicated of God and man; (6) the use of " for non' and n'n5rr, and the rendering R i ni or r;Iya when a'n (21. e denotes heathen gods. Instances of this endeavour to maintain, as it were, a respectful distance in speaking of God occur on every page of the Targums, but cases also occur, by no means infrequently, where human actions and passions are ascribed to God. The explanation of this phenomenon is to be found in the fact that anthropomorphisms, as such, were not necessarily avoided, but only in those cases where they might be misunderstood by the people.
(2) In addition to the Targum of Onkelos two other Targums to the Pentateuch are cited by Jewish authorities, under the titles of the Targum Jerushalmi and the Targum of Jonathan ben Uzziel. Of these the former contains only portions of the Pentateuch,' and is therefore usually designated the Fragmentary (Jerusalem) Targum. In a large number of cases this Targum gives merely a variant rendering of single words: where longer passages are given it presents a very paraphrastic translation, and bears all the marks of a late Haggadic composition. Its fragmentary character arises from the fact that it is simply a collection of variae lectiones and additions to the version of Onkelos, intended possibly for use at public services.' That this Targum was redly intended to supplement that of Onkelos is shown by comparing the two texts. For the former is frequently unintelligible without the latter, since it offers no translation of those words, or clauses, for which it gave the same rendering as Onkelos. On the other hand, the version of Onkelos affords just the supplementary material that is required to restore sense to the shorter text. Moreover, in not a few cases the Fragmentary Targum itself attaches to its variant rendering the succeeding word from Onkelos, thus indicating that from this point onwards the latter version is to be followed. More conclusive still is the fact that in a number of old Mahzor MSS. we find Targums to the Song of Moses and to the Decalogue, in which this process has been fully carried out, the text of Onkelos being given as well as the variants of the Fragmentary Targum.
The second Jerusalem Targum, or the so-called pseudo-Jonathan, admittedly owes its ascription to Jonathan ben Uzziel to the incorrect solution of the abbreviated form by which it was fre quently cited, viz. '"n, or Targum Jerushalmi ('t?5wn' ?unn), This Targum represents a later and more successful attempt to correct and supplement the Targum of Onkelos by the aid of variants derived from another source. It is not, however, a revision of the Fragmentary Targum - for it is clearly independent of that version - but is rather a parallel, if somewhat later, production, in which the text of Onkelos is already combined with a number of variants and additions. It is noticeable that this Targum has been considerably influenced by the Targum of Onkelos, and in this respect, as in others, is far less trustworthy than the Fragmentary Targum, as a witness to the linguistic and other peculiarities of the source from which they were both derived. It exhibits, to a marked degree, that tendency to expand the text by additions of every kind, which has been already noted as characteristic of the later stages of Targumic composition. Homilies, legends, traditional sayings and explanations, in fact every form of Haggadic expansion are utilized by the Targumist, so that at times his works convey the impression more of a late Midrash than of a translation. This impression is fully confirmed by (a) a comparison of the Talmud and later Midrashic works with which it has obvious points of contact, and (b) the historical allusions, such as the mention of Constantinople (Num. xxiv. 19), of a wife and daughter of Mahomet (Gen. xxi. 21), and the references to Esau and Ishmael as representative world-powers (Gen. xlix. 26; Deut. xxxiii. 2; cf. Fragm. Tg. to Gen. xlix. 2; Deut. xxxiii. 2).3 In its translation of the Hebrew pseudo-Jonathan is careful to avoid anthropomorphisms and to give the sense of all but the most simple metaphors, though his method is not so thorough as that of Onkelos. Every endeavour is made to gloss over, or modify, expressions which seemed derogatory to the ancestors of ' According to Zunz, Gottesdienstliche Vortrdge, 2nd ed., p. 80, its contents bear the following proportions to Genesis, z o o to Exodus, about 1 1 4 to Leviticus, s to Numbers, and 4 to Deuteronomy.
2 Seligsohn, De duabus Hier. Pent. paraphrasibus (1858): for a fuller discussion see Bassfreund, " Das Fragmenten Targum " in M.G.W.J. xl.
3 The view that Deut. xxxiii. 11 could only have been written by a contemporary of John Hyrcanus cannot be maintained; cf. Dalman, Gramm. p. 30 f., and, more fully, Bassfreund, M.G.W.J. xliv. (1900), pp. 481 f.
Israel, and to amplify everything which redounded to their credit. On the other hand, pseudo-Jonathan shows a tendency to condense those additions which it has in common with the Fragmentary Targum: in particular he omits all quotations from Scripture.
In regard to the source of the two Palestinian Targums to the Pentateuch, we must accept the conclusion of Bassfreund 4 that they both derived their variants from a complete Targum Jerushalmi. This conclusion is based on the following grounds: (i) Various Jewish works dating from the iith to the 14th century contain a large number of quotations under the heading '"n, i.e. Targum Jerushalmi. Of these rather less than a quarter are found in the Fragmentary Targum, the remainder being mostly taken from passages for which no translation of that Targum exists. This completer work, however, cannot be identified with the pseudo-Jonathan, for more than half of these quotations are missing from the latter; and further, in passages for which we possess both the Targums, the text of the Fragmentary Targum agrees much more closely with the quotations: the linguistic evidence also shows that the Fragmentary Targum is a more faithful representative of the original source; (2) the pseudo-Jonathan displays a curious inconsistency in its rendering of particular words and phrases, at one time following Onkelos, at Another a different source. That this latter source is the Targum Jerushalmi is proved, in the majority of cases, by a comparison with the Fragmentary Targum; (3) quotations from Scripture preserved in the Fragmentary Targum point to a completer version than our present Fragmentary Targum. But though the existence of an older Targum Jerushalmi cannot be denied, it is clear that the form in which it was utilized by the two Palestinian Targums cannot be of an early date, for many of the latest elements in the Fragmentary and pseudo-Jonathan Targums were undoubtedly derived from their common source. Moreover, the existence of a written Palestinian Targum at an early date is expressly excluded by the evidence at our disposal. In the middle of the 2nd century A.D. R. Simon ben Gamaliel forbade the translation of the Pentateuch in any language but Greek; and this command was upheld by R. Johanan in the 3rd century. Even in the time of the later Amoraim there is no mention of a written Palestinian Targum, though the official Babylonian Targum is repeatedly referred to in the Babylonian Talmud, in the Midrashim, and at times also by Palestinian Amoraim. These considerations are sufficient to disprove the theory of Geiger, 6 which has for so long been accepted in one form or another, that the Targum of Onkelos was merely a reproduction of the old Targum Jerushalmi revised in accordance with the " new I-Ialakha " introduced by R. Aqiba. Yet it is impossible to hold that the Targum of Onkelos was the only representative of Targum tradition that existed among the Jews down to the 7th century A.D., the period to which the internal evidence compels us to assign the Targum Jerushalmi as used by the Fragmentary Targum and the pseudo-Jonathan. We must rather assume that a tolerably fixed Targum tradition existed in Palestine from quite early times. The language employed in the Targum of Onkelos is, admittedly, Palestinian or Judaean, and since language and thought are ever closely allied, we may conjecture that the current Judaean exegesis, which, in part at least, must go back to the 2nd century A.D., was not without its influence on the Babylonian translation. This old Targum tradition, however, never received official recognition in Palestine, and was unable, therefore, to hold its own when the new Babylonian version was introduced. We may infer that, as time went on, a reaction in favour of the older renderings made itself felt, with the result that these were collected in the form of variants and appended to Onkelos. But the authority enjoyed by the latter rendered it secure against any encroachments; hence any later expansions, especially those of a popular Haggadic character, naturally found their way into the less stereotyped Targum Jerushalmi. Unfortunately, we possess but little material for controlling the texts either of the Fragmentary Targum or of the pseudo-Jonathan. Of the latter only one manuscript (Brit. Museum Add. 27031) is known to exist, and this has been utilized by Ginsburger in his Pseudo-Jonathan (Berlin, 1903). The same scholar has also edited the Paris manuscript (110) of the Fragmentary Targum (Das Fragmententhargum, Berlin, 1899), to which he has added the variants from Cod. Vat. 440 and the manuscripts at Nuremberg and Leipzig. In the same edition are collected the various fragments of the Targum Jerushalmi, which are to be found in the early editions of the Pentateuch and in part also in various manuscripts.