introduced in Isa. xxxviii, and thrust into an improper place after v. 20, when they should have been written after v. 6. Thus also in iy

in Ps. xxviii. 8., is in some manuscripts changed into iny iy from

-:

Ps. xxix. 11. Alterations of this kind however are not to be unnecessarily multiplied, since the very numerous discrepancies between the books of Samuel and Kings and those of Chronicles, show that transcribers have not always indulged their inclinations in harmonizing parallel places.

§ 134. Abbreviations incorrectly understood.

Although abbreviations are of infrequent occurrence in Hebrew manuscripts, yet there are some, which occasionally appear to have been incorrectly understood. Thus the abbreviation which is used in the Targums for the name Jehovah, ', seems to have been found by the Alexandrine translator of Jonah, in i. 9, where instead of y, a Hebrew, it would appear that he read "y, and incorrectly translated it dλs us, as if it were intended for "y and not in. So

also in Jer. vi. 11, '', meant as an abbreviation for лin лën, að

pears to have been read 'non, Supov μs. The manuscript of Kenni

cott, marked 76, often omits the last letters, as 'ia for 'is, 'wid

:

,3 .has arisen in Ps. cvii מים In this way .במות for במו' מושיע for .מימין for מימ' from the abbreviation

.T.

§ 135. Marginal notes introduced into the text.

The explanatory and traditionary notes which are so common in the margin of manuscripts, have been sometimes considered by transcribers as a part of the discourse, and introduced into the text. Thus in Isa. xl. 7. pyn van ¡ɔ8, truly the people is grass, seems to be an interpolation of this kind, as it is wanting in the Septuagint; and in vii. 17. л, the king of Assyria. Such is also the case

with I Sam. vi. 19. x qhx own, 50,000 men, which is not accord

ing to the order of numbering, and is wanting in JOSEPHUS, Ant. VI. ii. 4.[a] Larger portions also have been introduced in the first book of Samuel, (xvii. 12-31, 50, 55-58, xviii. 1--5, 9-11, 17-19.) which did not exist in the manuscript used by the Alexandrine translator, but were introduced into his version, as is evident from many Paris manuscripts, from versions of the second century, by Origen. They seem to have been taken from the Chaldee Paraphrase, as the Targum of Jonathan on these passages contains many other things of the same kind. However this may be, they do not belong to the text; for they are at variance with the context both preceding and subsequent, as has been remarked by Houbigant on I Sam. xvii. 11.[b] So also much has been interpolated in Jeremiah which was not read by the Alexandrine translator.

[a) See LE CLERC and DATHE, in loc. Tr.]

[6) See Germ. Introd. P. I. § 135. S. 480. f. EICHHORN II. Th. S. 532. ff. BERTHOLDT III. Th. s. 897. Tr.]

§ 136. Improper division of words.

Although the oriental writers left spaces between the words, or made use of points in order to indicate the end of each word, it fre quently happened, that through the haste or carelessness of the transcribers they were neglected, and words were connected. These, being separated again by other transcribers, have been sometimes inaccurately divided, so as to connect with a preceding word a letter which belonged to the following, or the reverse. Sometimes, also, improperly, a word has been separated into two, or two have been combined in one. Thus by unto death has arisen in Ps. xlviii. 15, from nihy, as it is read in 143 manuscripts, and expressed in the Alexandrine version and in the Chaldee Paraphrase. On the other hand, in Ps. lv. 16. nin devastations, ought to be divided according to the k'ri into or let death seize,* which is given also

* [Or, according to MICHAELIS, let him constitute death their rigid creditor; exactorem constituat mortem super illos; from the Arabic form Suppl. ad Heb. Lex. No. 1655. Tr.]

نسا

in many editions, and by the Alexandrine translator, by Aquila, Symmachus, Jerome, and the author of the Peshito Syriac. So also in Hos. vi. 5, the words Yi T, and thy judgments, a light

goeth forth are more accurately divided by the Septuagint, the Syriac and the Chaldee thus, xxx and my judgment (my righ

teous declaration), goeth forth as the light (the Sun).

§ 137. Erroneous corrections.

It has been already remarked (§ 108 and 111), that here and there the text has been improperly corrected to make it conformable to grammatical analogy, or to the Masora. A reading therefore which departs from grammatical principles and the Masora is less to be suspected of having been altered, and consequently is more probable than one which agrees with them. Another source of corruption has been the disposition of transcribers and readers of manuscripts to alter any thing difficult, or of rare occurrence, or which was unknown or unintelligible to themselves, or seemed likely to give offence. On this account a reading which is difficult, of rare occurrence, and which might give offence, is in itself preferable to one which is common, easy, and liable to no exception, the very character of which is sufficient reason for suspecting it of having suffered alteration by some modern hand. This rule is as it were the touchstone of the able and judicious critic.

§ 138. Whether various Readings have arisen from Corruption.

Although the Hebrew text has not been corrupted, yet corruptions may have been attempted, and this may have given rise at least to various readings. But such an attempt cannot be proved by examples of various readings produced from this source, unless it be thought proper to consider as such the change of □”, Gerizim into ha'y Ebal,

in Deut. xxvii. 4., and of n, Moses, inton, Manasseh in Jud.

xviii. 30.; in the latter of which places the transcribers themselves

have candidly intimated by the suspended Nun that n is the genu

ine reading, and the Masorets also expressly state, that the Nun was inserted, in order that the first idolatrous priest should not be said to be a grandson of Moses. But although we have no other instances of readings originating in corruption, yet there is no want of examples of readings less favourable to Christians having been preferred by transcribers and widely propagated in manuscripts, as on in Ps. xvi. 10., in Ps. xxii. 17., and in Gen. xlix. 10.

Wherefore,

the reading which is less orthodox in the Jewish sense is more probable than the orthodox, so as to admit of an exception against even a majority of witnesses in its favour.

§ 139. Critical Argument from the Series of Discourse.

Whatever may be the result of the depositions of witnesses or of investigation of the question which reading could most readily arise from another, the series of discourse must be compared. For although the truth of a reading cannot be inferred solely from its agreement with the context, because it is possible that a false reading may have very plausible pretensions in this respect, yet a reading which is in opposition to the context is certainly spurious, for no author of a sound mind would write words without meaning or at variance with the connexion of his discourse. Yet great circumspection is required that a reading, which at first view seems to be at variance with the context or not to give a suitable sense, be not immediately rejected; for it may be the more difficult reading, which, when correctly understood, exactly coincides with the context, and is preferable by the rule just stated (§ 137) on account of its very difficulty. The context, it is to be remembered, comprises not only the connexion of those parts of the discourse which immediately precede and follow the passage in question, but also the syntactical construction of the sentence. The comparison of this requires by far the greatest degree of caution. On the one hand no rule is without its exception, and what is supposed to be an anomaly may be a legitimate exception to the rule. On the other hand, admitting the existence of a solecism, it may have

originated even with the author himself, provided it be not too great an aberration from the genius of the language, as that in I Sam. vi.

.שבעים איש חמשים אלף איש .19

§ 140. Critical Arguments derived from the Poetic Parallelism.

In poems, the poetic parallelism affords no contemptible aid in a comparison of the context. For as in Hebrew poetry the members are either synonymous or antithetic, or at least conformed to each other in the order of construction, they serve for mutual illustration, and a reading which is at variance with the parallel member is spurious. It is not however on the other hand to be concluded, that a reading which coincides with the parallel member is necessarily genuine, since a spurious reading, especially if it have arisen from an erroneous judgment of transcribers, may exhibit a perfect coincidence. In alphabetical poems omissions of verses are easily observed, (See Ps. xxv. 4, 5, 17, 18. xxxvii. 28, 29. cxlv. 13, 14.) and even additions and transpositions of words. So in Ps. xxxvii. 39. the addition of the prefix Vau in nyn is evident, as the verse ought to begin with Thau; and

in Ps. xxv. 2. the transposition of the words 72 h for 2 is

discoverable from the necessity of beginning the verse with Beth.

§ 141. Argument from Parallel Places.

As transcribers have often altered parallel places with the view of making them correspond, this has given rise to the rule of criticism, that a reading which varies from the parallel place is, cæteris paribus, more probable than one which entirely agrees with it. Still the frequent disagreement of parallel places shows that the eagerness of transcri bers to render them conformable to each other has not been without some limitation, even manifest errors having been occasionally left untouched. Undoubtedly, therefore, faulty readings may be corrected from the parallel places. Thus, in II Chron. xxii. 2., where the son is said to be forty-two years old, and therefore according to II Chron. xxi. 20., two years older than his father, the text is without doubt to be restored from II Kings viii. 26., so that the age of the