_Text._--The form in which the Hebrew text of the Old Testament is
presented to us in all MSS. and printed editions is that of the
Massoretic text, the date of which is usually placed somewhere between
the 6th and 8th centuries of the Christian era. It is probable that the
present text became fixed as early as the 2nd century A.D., but even
this earlier date leaves a long interval between the original autographs
of the Old Testament writers and our present text. Since the fixing of
the Massoretic text the task of preserving and transmitting the sacred
books has been _carried_ out with the greatest care and fidelity, with
the result that the text has undergone practically no change of any real
importance; but before that date, owing to various causes, it is beyond
dispute that a large number of corruptions were introduced into the
Hebrew text. In dealing, therefore, with the textual criticism of the
Old Testament it is necessary to determine the period at which the text
assumed its present fixed form before considering the means at our
disposal for controlling the text when it was, so to speak, in a less
settled condition.
Massoretic text.
An examination of the extant MSS. of the Hebrew Old Testament reveals
two facts which at first sight are somewhat remarkable. The first is
that the oldest dated MS., the _Codex Babylonicus Petropolitanus_, only
goes back to the year A.D. 916, though it is probable that one or two
MSS. belong to the 9th century. The second fact is that all our Hebrew
MSS. represent one and the same text, viz. the Massoretic. This text was
the work of a special gild of trained scholars called Massoretes
([Hebrew: baalei hamasoreth]) or "masters of tradition" ([Hebrew:
masorah] or less correctly [Hebrew: masoreth]),[2] whose aim was not
only to preserve and transmit the consonantal text which had been handed
down to them, but also to ensure its proper pronunciation. To this end
they provided the text with a complete system of vowel points and
accents.[3] Their labours further included the compilation of a number
of notes, to which the term Massorah is now usually applied. These notes
for the most part constitute a sort of index of the peculiarities of the
text, and possess but little general interest. More important are those
passages in which the Massoretes have definitely adopted a variation
from the consonantal text. In these cases the vowel points attached to
the written word (_Kethibh_) belong to the word which is to be
substituted for it, the latter being placed in the margin with the
initial letter of _Qere_ (= to be read) prefixed to it. Many even of
these readings merely relate to variations of spelling, pronunciation or
grammatical forms; others substitute a more decent expression for the
coarser phrase of the text, but in some instances the suggested reading
really affects the sense of the passage. These last are to be regarded
either as old textual variants, or, more probably, as emendations
corresponding to the _errata_ or _corrigenda_ of a modern printed book.
They do not point to any critical editing of the text; for the aim of
the Massoretes was essentially conservative. Their object was not to
create a new text, but rather to ensure the accurate transmission of the
traditional text which they themselves had received. Their work may be
said to culminate in the vocalized text which resulted from the labours
of Rabbi Aaron ben Asher in the 10th century.[4] But the writings of
Jerome in the 4th, and of Origen in the 3rd century both testify to a
Hebrew text practically identical with that of the Massoretes. Similar
evidence is furnished by the Mishna and the Gemara, the Targums, and
lastly by the Greek version of Aquila,[5] which dates from the first
half of the 2nd century A.D. Hence it is hardly doubtful that the form
in which we now possess the Hebrew text was already fixed by the
beginning of the 2nd century. On the other hand, evidence such as that
of the _Book of Jubilees_ shows that the form of the text still
fluctuated considerably as late as the 1st century A.D., so that we are
forced to place the fixing of the text some time between the fall of
Jerusalem and the production of Aquila's version. Nor is the occasion
far to seek. After the fall of Jerusalem the new system of biblical
exegesis founded by Rabbi Hillel reached its climax at Jamnia under the
famous Rabbi Aqiba (d. c. 132). The latter's system of interpretation
was based upon an extremely literal treatment of the text, according to
which the smallest words or particles, and sometimes even the letters of
scripture, were invested with divine authority. The inevitable result of
such a system must have been the fixing of an officially recognized
text, which could scarcely have differed materially from that which was
finally adopted by the Massoretes. That the standard edition was not the
result of the critical investigation of existing materials may be
assumed with some certainty.[6] Indeed, it is probable, as has been
suggested,[7] that the manuscript which was adopted as the standard text
was an old and well-written copy, possibly one of those which were
preserved in the Court of the Temple.
But if the evidence available points to the time of Hadrian as the
period at which the Hebrew text assumed its present form, it is even
more certain that prior to that date the various MSS. of the Old
Testament differed very materially from one another. Sufficient proof of
this statement is furnished by the Samaritan Pentateuch and the
versions, more especially the Septuagint. Indications also are not
wanting in the Hebrew text itself to show that in earlier times the text
was treated with considerable freedom. Thus, according to Jewish
tradition, there are eighteen[8] passages in which the older scribes
deliberately altered the text on the ground that the language employed
was either irreverent or liable to misconception. Of a similar nature
are the changes introduced into proper names, e.g. the substitution of
_bosheth_ (= shame) for _ba'al_ in Ishbosheth (2 Sam. ii. 8) and
Mephibosheth (2 Sam. ix. 6; cf. the older forms Eshbaal and Meribaal, 1
Chron. viii. 34, 35); the use of the verb "to bless" ([Hebrew: bareh])
in the sense of cursing (1 Kings xxi. 10, 13; Job i. 5, 11, ii. 5, 9;
Ps. x. 3); and the insertion of "the enemies of" in 1 Sam. xxv. 22, 2
Sam. xii. 14. These intentional alterations, however, only affect a very
limited portion of the text, and, though it is possible that other
changes were introduced at different times, it is very unlikely that
they were either more extensive in range or more important in character.
At the same time it is clear both from internal and external evidence
that the archetype from which our MSS. are descended was far from being
a perfect representative of the original text. For a comparison of the
different parallel passages which occur in the Old Testament (e.g. 1 and
2 Samuel, 1 and 2 Kings, and 1 and 2 Chronicles; 2 Kings xviii. 13-xx.
19 and Isaiah xxxvi.-xxxix; 2 Sam. xxii. and Ps. xviii.; Ps. xiv. and
liii., &c.) reveals many variations which are obviously due to textual
corruption, while there are many passages which in their present form
are either ungrammatical, or inconsistent with the context or with other
passages. Externally also the ancient versions, especially the
Septuagint, frequently exhibit variations from the Hebrew which are not
only intrinsically more probable, but often explain the difficulties
presented by the Massoretic text. Our estimate of the value of these
variant readings, moreover, is considerably heightened when we consider
that the MSS. on which the versions are based are older by several
centuries than those from which the Massoretic text was derived; hence
the text which they presuppose has no slight claim to be regarded as an
important witness for the original Hebrew. "But the use of the ancient
versions" (to quote Prof. Driver[9]) "is not always such a simple matter
as might be inferred.... In the use of the ancient versions for the
purposes of textual criticism there are _three_ precautions which must
always be observed; we must reasonably assure ourselves that we possess
the version itself in its original integrity; we must eliminate such
variants as have the appearance of originating merely with the
translator; the remainder, which will be those that are due to a
difference of text in the MS. (or MSS.) used by the translator, we must
then compare carefully, in the light of the considerations just stated,
with the existing Hebrew text, in order to determine on which side the
superiority lies."
Samaritan.
_Versions._--In point of age the Samaritan Pentateuch furnishes the
earliest external witness to the Hebrew text. It is not a version, but
merely that text of the Pentateuch which has been preserved by the
Samaritan community since the time of Nehemiah (Neh. xiii. 23-31),
i.e. about 432 B.C.[10] It is written in the Samaritan script, which
is closely allied to the old Hebrew as opposed to the later "square"
character. We further possess a Samaritan Targum of the Pentateuch
written in the Samaritan dialect, a variety of western Aramaic, and
also an Arabic translation of the five books of the law; the latter
dating perhaps from the 11th century A.D. or earlier. The Samaritan
Pentateuch agrees with the Septuagint version in many passages, but
its chief importance lies in the proof which it affords as to the
substantial agreement of our present text of the Pentateuch, apart
from certain intentional changes,[11] with that which was promulgated
by Ezra. Its value for critical purposes is considerably discounted by
the late date of the MSS., upon which the printed text is based.
Aramaic.
The Targums, or Aramaic paraphrases of the books of the Old Testament
(see TARGUM), date from the time when Hebrew had become superseded by
Aramaic as the language spoken by the Jews, i.e. during the period
immediately preceding the Christian era. In their written form,
however, the earlier Targums, viz. those on the Pentateuch and the
prophetical books, cannot be earlier than the 4th or 5th century A.D.
Since they were designed to meet the needs of the people and had a
directly edificatory aim, they are naturally characterized by
expansion and paraphrase, and thus afford invaluable illustrations of
the methods of Jewish interpretation and of the development of Jewish
thought. The text which they exhibit is virtually identical with the
Massoretic text.
Septuagint.
The earliest among the versions as well as the most important for the
textual criticism of the Old Testament is the Septuagint. This version
probably arose out of the needs of the Greek-speaking Jews of
Alexandria in the 3rd century B.C. According to tradition the law was
translated into Greek during the reign of Ptolemy Philadelphus
(284-247 B.C.), and, though the form (viz. the _Letter of Aristeas_)
in which this tradition has come down to us cannot be regarded as
historical, yet it seems to have preserved correctly both the date and
the locality of the version. The name Septuagint, strictly speaking,
only applies to the translation of the Pentateuch, but it was
afterwards extended to include the other books of the Old Testament as
they were translated. That the interval which elapsed before the
Prophets and the Hagiographa were also translated was no great one is
shown by the prologue to Sirach which speaks of "the Law, the Prophets
and the rest of the books," as already current in a translation by 132
B.C. The date at which the various books were combined into a single
work is not known, but the existence of the Septuagint as a whole may
be assumed for the 1st century A.D., at which period the Greek version
was universally accepted by the Jews of the Dispersion as Scripture,
and from them passed on to the Christian Church.
Versions of Aquila, Symmachus, Theodotion.
The position of the Septuagint, however, as the official Greek
representative of the Old Testament did not long remain unchallenged.
The opposition, as might be expected, came from the side of the Jews,
and was due partly to the controversial use which was made of the
version by the Christians, but chiefly to the fact that it was not
sufficiently in agreement with the standard Hebrew text established by
Rabbi Aqiba and his school. Hence arose in the 2nd century A.D. the
three new versions of Aquila, Symmachus and Theodotion. Aquila was a
Jewish proselyte of Pontus, and since he was a disciple of Rabbi Aqiba
(d. A.D. 135), and (according to another Talmudic account) also of
Rabbi Eliezer and Rabbi Joshua, the immediate predecessors of Aqiba,
his version may be assigned to the first half of the 2nd century. It
is characterized by extreme literalness, and clearly reflects the
peculiar system of exegesis which was then in vogue among the Jewish
rabbis. Its slavish adherence to the original caused the new
translation to be received with favour by the Hellenistic Jews, among
whom it quickly superseded the older Septuagint. For what remains of
this version, which owing to its character is of the greatest value to
the textual critic, we have until recently been indebted to Origen's
_Hexapla_ (see below); for, though Jerome mentions a _secunda editio_,
no MS. of Aquila's translation has survived. Fragments[12], however,
of two codices were discovered (1897) in the genizah at Cairo, which
illustrate more fully the peculiar features of this version.
The accounts given of Theodotion are somewhat conflicting. Both
Irenaeus and Epiphanius describe him as a Jewish proselyte, but while
the former calls him an Ephesian and mentions his translation before
that of Aquila, the latter states that he was a native of Pontus and a
follower of Marcion, and further assigns his work to the reign of
Commodus (A.D. 180-192); others, according to Jerome, describe him as
an Ebionite. On the whole it is probable that Irenaeus has preserved
the most trustworthy account.[13] Theodotion's version differs from
those of Aquila and Symmachus in that it was not an independent
translation, but rather a revision of the Septuagint on the basis of
the current Hebrew text. He retained, however, those passages of which
there was no Hebrew equivalent, and added translations of the Hebrew
where the latter was not represented in the Septuagint. A peculiar
feature of his translation is his excessive use of transliteration,
but, apart from this, his work has many points of contact with the
Septuagint, which it closely resembles in style; hence it is not
surprising to find that later MSS. of the Septuagint have been largely
influenced by Theodotion's translation. In the case of the book of
Daniel, as we learn from Jerome (_praefatio in Dan._), the translation
of Theodotion was definitely adopted by the Church, and is accordingly
found in the place of the original Septuagint in all MSS. and
editions.[14] It is interesting to note in this connexion that
renderings which agree in the most remarkable manner with Theodotion's
version of Daniel are found not only in writers of the 2nd century but
also in the New Testament. The most probable explanation of this
phenomenon is that these renderings are derived from an early Greek
translation, differing from the Septuagint proper, but closely allied
to that which Theodotion used as the basis of his revision.
Symmachus, according to Eusebius and Jerome, was an Ebionite;
Epiphanius represents him (very improbably) as a Samaritan who became
a Jewish proselyte. He is not mentioned by Irenaeus and his date is
uncertain, but probably his work is to be assigned to the end of the
2nd century. His version was commended by Jerome as giving the sense
of the original, and in that respect it forms a direct contrast with
that of Aquila. Indeed Dr Swete[15] thinks it probable that "he wrote
with Aquila's version before him, (and that) in his efforts to recast
it he made free use both of the Septuagint and of Theodotion."