^Yhile the same system was applied to all books
in Batjylonia, a special system was developed in
Palestine for Psalms, Proverbs, and Job. Many
differences still existed in the Kleventh Century
between the Babylonian school of Jacob ben
Xaplitali and the Palestinian school of Aaron ben
Asher. During the Seventh Century, the vowel-
signs seem to have been introiluced. The infra-
linear system, derived from the Xestorians, is
likely to" have been earlier than the supra-linear
system exhibited — e.g. by the Saint Petersburg
Codex of A.D. 916, wrongly supposed to have been
Babylonian. That the vowel-points are not
earlier than A.n. 500 is evident from the Talnuid,
which docs not know any such signs. In regard
to the vocalization, minor difTerences have been
lecorded between the Babylonian schools at
Xehardca. Sura, and Puniliedita, and between
these and the Palestinian school of Tiberias. It
was probably during the Sixth Century that the
!Massorah. in its narrower sense, was inserted
by the side of the text ( Massora parva ). or at
the bottom and the top of the page {Massora
magna), or at the end and beginning of the
book (Massora final is). This is a body of an-
notations, partly indicating how in doubtful
cases the text should be read (qerc' = 'it is read,'
sebhir = 'it is thought'), partly giving sum-
maries, eoncordantial matter, and the like. This
material was gathered during the Talmudic
period (c. 180.500 a.d.).
Quotations in Misbnah, the Palestinian Tal- mud, JUdrashim, and Babylonian Talmud, the Aranuiic Targuras, and Jerome, and less directly the later Greek versions, reveal the consonantal text that enjoyed canonical authority from the Second to the Sixth Century. It is probable that a standard text was gradually obtained, and that divergent types were naturally eliminated as a result of the critical process by which the canon was established, and of the demand that holy scriptures should be written exclusively in the Syrian ('Assyrian') characters to facilitate cor- rect reading and to distinguish approved codices. Certain ditiVrenees between the earliest Cireek version and the Massoretic text suggest that, in the case of .?oine of the Prophets and the Iliigiographa, the codices used by the Alexan- drian translators were written in the -Egv'pto- Aramaie alphabet, possibly already in the mid- dle of the Second Century B.C. As the Book of Esther (last tliird of the Second Century ii.c.) still associates the Hebrew language and its pe- culiar script (viii. !)), the change probably began somewhat later in Palestine. The lingering jnef- ercnce for the older letters yielded only to the practical considerations of the scribes in the course of the First Century A.D. All rolls were not transcribed from the old Semitic alphabet at the same time: but as new copies were made, the request to write them in the characters then in vogue in Syria (resembling, no doubt, the Palniyrene or Xabatipan) was more and more followed. While many copyists' errors undoubt- edly were committed, the text is not likely to have suflVred very greatly in this jjrocess of transcription. For the motive of the change was a deeper reverence for the text that was deemed canonical and a desire for greater accuracy. Tliis regard for the codices copied and the traditional reading is evidenced by the curious arrest of the tendency to indicate by vowel-letters the pro- nunciation, and by the introduction of spaces between words. Since, in the Talmudic period, only copies written in the later Syrian charac- ters would be used as models, the disappearance of older manuscripts is naturally accounted for. Concerning the state of the text from the Third Century u.c. to the Second Century a.d., we possess a certain amount of knowledge through the Samaritan Pentateuch and its Tar- gum, the earliest Greek version, and the trans- lations made from it, the Book of Jubilees, the earliest Syriae version, and some writers of the period. For the Pentateuch, the testimony is most direct, as the Samaritan copy is written in Hebrew and in the old cliaracters, slightly modi- tied. Our earliest manuscript, liowever, is of the Thirteenth Century, and the age of this recension is very uncertain. There is no evidence that the present text is identical with the one recognized by Manasseh at the time of Alexander, when the temple on Gerizim was built. But the fact that many of its 0000 deviations from the Jlassoretic text agree with the Greek version suggests that it gained a definite form at an earlier date than the Judican law. It is printed in the Paris and London polyglots. The difference between the ilassoretic text and that from which the oldest Cireek version was made is very great. The latter manifestly had no division of words, used vowel-letters very sparingly, was in some pas- sages longer, and in many more instances shorter, than the JIassoretic text, and fre(|uently pre- sented different consonants. It is evident, from the Book of Jubilees, that the text had not yet assumed its present form in Palestine at the beginning of our era. Whether the codices that were tirst rendered into Syriae presented a tj'pe akin to those used by the Alexandrian trans- lators, or corrections were made after the Second Century A.n. to bring the Syriae version into harmony with the Greek, cannot easily be de- termined.
The earlier condition of the Hebrew text can cnly be surmised, and the original itself ap- proximately restored, by a critical sifting of all this comparatively late material, by observation of parallelism, metre, and logical connection, and by scientific conjecture. In the attempt to re- construct the earliest form of the text, literary and historical criticism must come to the aid of the investigator, removing later glosses, inter- polations, annotations, editorial remarks, and mistaken inscrijitions, and thus revealing the growth and com])osition of the bililical books. Houbigant first undertook to publish a transla- tion based upon a critically restored text ( IV.iS). Since then the necessity <if such work has been in increasing measure felt by scholars. The greatest enterprise in this direction has been the publication of the Polychroiiie Bible, under the editorship of Paul Haupt, of which fourteen out of twentv parts have appeared (New York, 1898 sqq,).
The earliest division of the several books was occasioned by the practice of reading certain sections of the Law and the Prophets on the Sabbath. In Palestine the Law was read in about three years, in Babylonia in one year. Hence the 1.54 Palestinian srilarim of the Law and the corresponding hnphtaroth of the Projih- ets and the .54 Babylonian parashioth. There were also 379 shorter paragraphs called 'closed,' because the next section began on the same line, and 290 'open' paragraphs. These' Jewish sec-
