Note to the casual reader: this is a technical post intended for Old Testament scholars.

In the eighteenth century, Benjamin Kennicott led a project to catalog all the variants in the consonantal text of Masoretic Hebrew manuscripts that were accessible throughout Europe, including only those manuscripts copied before the invention of the printing press. In the end, more than 600 Hebrew manuscripts were collated. The resulting two-volume work published in 1776–1780, Vetus Testamentum Hebraicum, included a printed Hebrew text (the van der Hooght edition) with a critical apparatus below. In the Pentateuch, Kennicott also prints the Masoretic text (MT) and Samaritan Pentateuch (SamP) in parallel columns, with differences noted.  Only textual variants within the Masoretic or Samaritan manuscripts were recorded; Kennicott’s edition does not note variants from the LXX, the Vulgate, or other ancient translations. While many more Hebrew manuscripts have been discovered or become accessible since the time of Kennicott, no other complete catalog of Hebrew textual variants has ever been produced. Thus, someone who wants to see a complete apparatus for (consonantal) textual variants in the Masoretic Text will have to consult Kennicott. Even the editors of BHQ and BHS still rely on Kennicott when noting textual variants within the Masoretic tradition; they do not have their own catalog of textual variants in the MT.

I first learned of Kennicott after I was already accustomed to using BHS. BHS and BHQ print only a small fraction of the total number of variants within the MT; when they do note a variant, they give comparatively little information about the manuscript evidence and usually present only a negative apparatus. The thing that immediately struck me about Kennicott’s work was that there are as many or more textual variants among Hebrew manuscripts in the Old Testament as the Nestle-Aland editions show for Greek manuscripts in the New Testament. Scholars who claim that Kennicott’s work is of little value because the Masoretic manuscripts are “practically uniform” are simply wrong. It is true that most of the variants are very small (one word or one letter), but changing one letter of a Hebrew word often changes it to a completely different word, making a significant difference for the translation and meaning of the Hebrew text. Even spelling (orthographical) variations within the manuscript tradition can offer insights into the original inspired text. I soon realized that if I wanted to do serious study of a passage in the Old Testament, I would have to check Kennicott for textual variants that might affect my translation or exegesis.

A scholar who uses Kennicott will quickly develop a sense of patterns of textual variation within the Masoretic tradition, accruing invaluable insights. He will gain an understanding of which letters are commonly confused, which spellings are commonly changed, and which words or letters are commonly omitted or duplicated by mistake. In short, he will gain a sense of the nature of textual variants within the MT, which will help him greatly in deciding which readings are original. One will find that commentators often speculate about textual variants and propose emendations without having a solid evidential basis for their hypotheses.

When I first discovered Kennicott, I bought a reprint edition. Now Kennicott’s work is conveniently available online: vol. 1, vol. 2, and both volumes. These volumes can be saved as PDF files on your computer as a backup. Unfortunately, Kennicott has not been incorporated into any Bible software program, and probably will not be due to the lack of interest in it by contemporary OT scholars.

Kennicott’s work is entirely in Latin and Hebrew/Aramaic, as Latin was the standard language of scholarly writing in the eighteenth century. The good thing is that one does not have to translate any Latin sentences to use Kennicott; one only needs a key to Latin terms used in the apparatus, as with BHS. Here is my understanding of the sigla in Kennicott’s apparatus:

  • ‸ = “omits”
  • * = “omitted” (in the parallel MT/SamP text)
  • 1º = “first occurrence”
  • 2º = “second occurrence”
  • bis = “twice”
  • forte = “accidentally,” referring to a copyist’s mistake
  • marg. habet = “margin has”
  • nunc = “now,” referring to a corrector
  • primo = “at first,” referring to the original reading of a manuscript
  • spat. post = spatium post = “a space after the . . .”
  • sup. ras. = supra rasura = “erasure above”; there is a sign above a word or letter indicating that it should be deleted
  • videtur = “it appears”

Kennicott’s edition has been criticized for not including variant readings of the Masoretic pointing. In my view, this criticism is unfair; there are so many consonantal variants, that attempting to collate all the variants in vowel points would have made the apparatus too large and the project too time-consuming. More importantly, the vowel points were not part of the original inspired text; thus, if one only wants to get back to the original text, the vowel points are of secondary importance. To see some variants in the vowel points, one can consult Giovanni de Rossi’s 1784–1788 work Variae lectiones Veteris Testamenti (vol. 1, vol. 2, vol. 3, vol. 4). Compared with Kennicott, de Rossi is more comprehensive for the variants he describes, but he only treats a fraction of the variants that are in Kennicott, and his work is much more difficult to understand for scholars who do not read Latin because it is not arranged as a formal apparatus. De Rossi includes a more expansive catalog of manuscripts, variants in vowel points, variants in ancient versions, and a written analysis or explanation of the variants.  De Rossi is also helpful for his list of Kennicott’s manuscripts and his descriptions of them (in vol. 1). Don’t forget to check de Rossi’s supplement at the end of vol. 4, as well as his Appendix.

Benjamin Kennicott was a strong Christian, and it was his Christian faith that motivated him to collate Hebrew manuscripts in order to determine what was the original inspired text of the Old Testament. Today, Kennicott’s work is basically ignored by Old Testament scholars. The reason for this is directly related to higher criticism, which seeks to destroy biblical faith by treating the Bible as a purely human product. The critical view of the formation of Old Testament books is that each book was composed by many different authors/editors, in many different editions, over a long period of time. There is no direct evidence for the critical hypotheses, but they are considered dogma in contemporary OT scholarship, given that the alternative is accepting biblical faith. A component of redaction criticism and other forms of higher criticism is the hypothesis that the Masoretic text is corrupt, and in order to get past the hypothetical layers of editing to a hypothetical “initial” text, scholars must make substantial emendations to the received Hebrew text. Since the ancient versions, particularly the LXX, are often substantially different from the MT, they are a major source for the large editing changes contemporary critics wish to make to the Old Testament text. But some scholars want to essentially rewrite the whole OT to make it fit with their view of what the text should mean and how it was formed, so they add many of their own conjectural emendations even to what is in the LXX, often going so far as to propose entirely new Hebrew words on the basis of words in cognate languages. The textual variants in Kennicott are ignored by critical scholars because they are the sort of small variants one would expect to arise by mistake when copying manuscripts, rather than the large variants that would be created by someone who is editing the biblical text.

The discovery of the Dead Sea Scrolls tempered some of the most extreme assertions of critical scholars about the MT, at least for a while. Most of the Dead Sea Scrolls contain a text that is virtually identical to the medieval Masoretic manuscripts, with only small textual variants. This proved that Hebrew manuscripts were faithfully copied for at least 1,500 years. Thus, the fact that many of Kennicott’s manuscripts date to the late medieval period does not mean that they do not preserve readings of manuscripts that were copied in antiquity. The main reason why older Hebrew manuscripts are not abundant is because the Jews “retire” old manuscripts when they begin to wear out. However, the readings of medieval manuscripts are valuable because they are part of a manuscript tradition that was very carefully copied. By way of analogy, in NT textual criticism medieval manuscripts which have a text that is very similar to the earliest extant manuscripts are given significant weight. I would personally argue that the Masoretic text is a faithful representation of the original inspired text, and I would only propose emendations of the MT or follow readings of the LXX on rare occasions.

For scholars who hold a high view of the fidelity of the MT, Kennicott is a much more valuable tool for textual criticism than the Greek Septuagint, the Latin Vulgate, the Syriac Peshitta, the Aramaic Targums, BHS, or even BHQ. I still consult other critical editions, ancient versions, and commentaries, but Kennicott contains most of the variants that I would consider important for recovering the original inspired text. The 600+ manuscripts collated by Kennicott are a large enough sample so that the inclusion of more manuscripts likely would not change the balance of evidence significantly. Yet I hope that someday all extant Hebrew manuscripts will be collated in a digital critical apparatus that will make the variant readings they contain more accessible to scholars. As it is, Kennicott’s work remains the most comprehensive and authoritative textual apparatus of Hebrew manuscripts ever produced. I hope that along with a new collation of Hebrew manuscripts, there will also be a return in evangelical seminaries to doing textual criticism primarily from an apparatus of Hebrew textual variants, without telling students that they can only analyze OT textual variants if they do extensive work in Latin, Aramaic, Syriac, Ugaritic, and so forth.

I will close this post with an extended quotation from Benjamin Kennicott, The State of the Printed Hebrew Text of the Old Testament Considered (Oxford: Oxford, 1759), 2:295, 298-300. Kennicott’s rational faith is clearly evident here, and is refreshing in comparison with the unreasonable skepticism of modern scholars:

The Original of the Pentateuch, in the hand-writing of Moses, was preserv’d with great care, being deposited in the side of the ark [Deut 31:26]; and with the ark was probably introduc’d into the temple at Jerusalem. . . . That this MS, wrote by the hand of Moses, was not stolen by the Philistines, but safely deposited in the temple; and that (after being conceal’d in the dangerous days of the idolatrous kings of Judah) it was found in the days of Josiah—-this seems clearly pointed out in the account given in 2 Chron, 34, 14. For there the copy of the law thus found by Hilkiah the priest is call’d ספר תורת יהוה ביד משה liber legis Jehovæ in manu (or per manum) Mosis. ’Tis scarce possible for words more naturally to describe a book written by Moses himself; or to vouch more fully, that the MS of the law then found was in the hand-writing of Moses. And perhaps all doubt will be remov’d, when ’tis consider’d farther—that, tho’ there are 15 places in the old Testament, which mention the words law of Moses and book of Moses, yet this one place only mentions the book of the law in the hand (or by the hand) of Moses: the reason of which seems to be that the other places speak of that law in general; but this place speaks of one particular MS, namely the original. . . . As to the point of age, this MS certainly might be the original; distance of time leaving it very possible. For the most extended chronology does not make the interval from the death of Moses to the death of Josiah 950 years; an age exceeded by that of several MSS preserv’d at this day.