The Treatise Called Diqduqe ha-Teʿamim

At the end of a Tiberian Bible, or in the blank space left after a book, the Masoretes sometimes added a short work that does not read like a grammar for a beginner. One of those closing texts, transmitted under the name of Aaron ben Moses ben Asher of Tiberias and usually called Diqduqe ha-Teʿamim, has been printed, enlarged, reduced and reconstructed for more than a century, and it still does not exist in a single agreed form. The hub note on the treatise sets out what the work is and how it sits among the other Tiberian handbooks. This essay asks why the copies differ, and what the printed editions, from Venice in 1517 to Leipzig in 1879 and Jerusalem in 1967, have done with that difference.

The title is commonly rendered in two ways, and both already hint at the difficulty.

The fine details of the accents.

A usual English rendering of the title Diqduqe ha-Teʿamim

The other common rendering, “the grammatical rules of the accents,” is equally fair. A book of fine details is not a complete system. Geoffrey Khan, following Aron Dotan, has stressed that the treatise was never meant to describe the whole Tiberian pointing. It takes up selected points that were felt to be difficult, in the accents and also in the vowels, and it leaves the rest unsaid.¹ That is why later copyists could add rubrics, why editors could disagree about the original core, and why a reader who meets the work only in one printed form can be misled about its size.

A Book at the End of a Codex

The natural home of such a work was not a separate volume. It was the end of a Bible, or the space left after a book, where the Masoretes already collected lists that would not fit in the margins. The journal essay on the Masorah finalis describes that habit of gathering long notes when the page itself had no room. Diqduqe ha-Teʿamim belongs to the same family of closing texts, though its contents are rules rather than tallies. The Jewish Encyclopedia already treated the title as probably the author’s own choice, and it placed the work among the writings associated with Aaron ben Asher.²

That association matters, because it tells the reader what kind of authority the treatise claimed. Aaron ben Asher is the name attached to the model pointing of the Aleppo Codex, and the Tiberian system that later printed Bibles follow is the system his school taught. A short book of fine points, written in Hebrew and often in rhyme so that it could be remembered, is exactly the sort of thing such a school might leave behind. It is not a commentary on Scripture, and it is not a masoretic list of words. It is a set of remarks on the signs themselves, on shewa and gaʿya and on those accents whose use was easy to get wrong.

The treatise is therefore best read beside the later manuals that try to be more complete. The Hidāyat al-Qāriʾ, written in Judaeo-Arabic in eleventh-century Jerusalem, and the Hebrew Horayat ha-Qore that descends from the same tradition, set out the reading more systematically. William Wickes, in the nineteenth century, wrote two Oxford treatises that do what Diqduqe ha-Teʿamim does not attempt, namely a full account of the accents of the poetical books and of the twenty-one prose books. The older work is useful precisely because it is not those books. It shows which details a Tiberian teacher thought it necessary to fix in writing.

What the Treatise Will Not Do

Khan’s warning is the right place to begin any account of the contents. Because the work is selective, it cannot be used as a map of the whole Tiberian system. It has much to say about shewa and the ḥaṭef vowels, and about the gaʿya, and about particular accents, but it does not march through the twenty-one prose accents or the three poetical ones in order.³ A reader who wants that order must go to Wickes, or to the later Arabic and Hebrew manuals, or to a modern handbook such as Israel Yeivin’s introduction to the Tiberian Masorah.

The same selectivity explains the variety of the copies. A complete grammar has a shape that is hard to alter without being noticed. A collection of fine points can be extended. A copyist who had another list of difficult cases, or a rhyme that circulated independently, could add it at the end. An editor who found such additions in his manuscripts could print them as part of the treatise. None of this is unique to Diqduqe ha-Teʿamim. It is what happens to short technical works that are copied as appendices rather than as books with a fixed close. The Jewish Encyclopedia’s article on punctuation already pointed readers to this treatise as a medieval source for the signs, without treating any one printed form as the only text.⁴

The language of the work has a part in this history. Rhymed Hebrew is easier to memorise than prose, and it is also easier to detach. A stanza can travel on its own and come to rest in another copy. The result, visible once the recensions are set side by side, is a family of texts that share a core and disagree about the edges.

The First Printed Editions

Ginsburg’s starting point is the printed text, and it is already a warning. “As regards the separate Treatise called in some MSS. Dikduke Ha-Teamim which has come down to us in several Codices in the name of Ben-Asher,” he writes, “its text in the different MSS. and in the editio princeps is as hopelessly irreconcilable as that of the official Lists.”⁵ The official lists in question are the lists of differences between Ben Asher and Ben Naphtali, which precede this discussion in the same chapter. Ginsburg’s point is that the treatise, like those lists, does not present a single agreed wording.

The work “was first published in the editio princeps of the Rabbinic Bible by Felix Pratensis, Venice 1517, where it is described in the heading as the compilation of Ben-Asher.”⁶ That fact alone is worth lingering over. The treatise did not first appear as a slim independent handbook. It was printed as part of the great Rabbinic Bible of 1517, the edition in which a Christian editor at Daniel Bomberg’s press in Venice first put a wide range of masoretic and exegetical material around the biblical text. A short Tiberian work of fine points thus entered print in the company of Targum and commentary, under a heading that already named Ben Asher as its compiler.

A second edition was published by Leopold Dukes under the title Kontres Ha-Massoreth, at Tübingen in 1846, from a manuscript in the possession of Samuel David Luzzatto. In that manuscript, Ginsburg notes, no author’s name is given to the treatise. The two editions, moreover, “differ essentially in the text,” and “the recension published by Dukes barely contains one fourth of the text in the editio princeps.”⁷ The first two printed forms, in other words, already stand as far apart as some of the manuscripts. A reader who meets only the Venice heading will think of a compilation by Ben Asher. A reader who meets only Dukes will find a much shorter, anonymous work. Neither printed book can be treated as “the” treatise.

It is after these two editions that Ginsburg turns to the five recensions he himself printed in The Massorah. The printed history and the manuscript history tell the same story: the work was recognised, named, and copied, but it was not stabilised.

Five Recensions in Ginsburg

The fullest older account of that family is in chapter X of Christian D. Ginsburg’s Introduction of 1897. Ginsburg printed, in The Massorah, five recensions of the treatise, and he described them in enough detail for a later reader to see how far they diverge.⁸

The first recension, printed under the letter ṭet, §246, in the first volume of The Massorah, pages 654–660, was taken from British Library, Add. 15251, where the compilation forms an appendix, with other masoretic materials, on folios 444a–448a. The work is there ascribed to Ben Asher. Ginsburg notes that its arrangement and text stand closer to the editio princeps than the other recensions do, though the first printed edition contains about thirty-five more rubrics.⁹

The second recension, printed under the letter ṭet, §§44–75, in the third volume, pages 41–43, comes from the illuminated manuscript Orient. 2626–28, where it occupies the first and second lines of the ornamental square in volume I, folios 1b–22b. Not only does its text differ materially from the other treatises, Ginsburg writes, but the rubrics are fewer and are differently arranged. He could not exhibit it in a parallel column with the rest.¹⁰

The third is from Codex Tzufutkale no. 15, a transcript supplied to Ginsburg by Hermann Strack. An epigraph which, according to Strack, “proceeds from the clever hand of Firkowitsch,” ascribes the Masorah to Aaron ben Asher. The appendix consists of fifty-nine rubrics of sundry masoretic import, attached to an ancient fragment of the Pentateuch. Of these, only twenty-two correspond to Ginsburg’s first recension, and nine are to be found among the additions in the compilation of Baer and Strack.^{11 12}

The fourth recension, also in the third volume of The Massorah, is from Tzufutkale no. 17, again from a transcript by Strack. The codex is an imperfect Pentateuch of 213 folios, “one of the most important fragments of the Hebrew Scriptures.” An epigraph assigns to it the date 790, which would place it in the generation of Aaron’s grandfather. Ginsburg, following Strack, reports that the date has been tampered with: the letter shin (300) has been made out of an original tav (400). “No reliance whatever can be placed on the date,” he concludes, but the manuscript “is very important.” In this copy the rubrics of Diqduqe ha-Teʿamim are not grouped as a distinct whole. They are simply sundry parts of a somewhat extensive Masorah of ninety-six rubrics, of which only nineteen correspond to the first recension and eleven to the additions in Baer and Strack.^{13 14}

The fifth recension is the Masorah finalis of Tzufutkale no. 19, again from Strack’s transcript. Ginsburg’s point, taken as a whole, is not that one of these five is the treatise and the others are corruptions. It is that the work circulated as a set of rubrics that could be grouped, expanded, abridged and mixed with other masoretic matter, and that the ascriptions and dates attached to some copies cannot be taken at face value. The Firkovich epigraph dated 790 is the clearest case.¹⁵

The Leipzig Edition of 1879

Seligmann Baer and Hermann Strack published Die Dikduke ha-ṭeamim des Ahron ben Moscheh ben Ascher at Leipzig in 1879. Their edition presents the treatise in seventy-six sections.¹⁶ That number is not a count of what any one early manuscript contains. It is the size of a compiled text, in which material found in the copies has been arranged as a single work. Ginsburg, who used their edition and cited its introduction, was careful to distinguish the rubrics that correspond to his first recension from “the additions in the compilation of Drs. Baer and Strack.”¹⁷ The phrase is a judgement. It means that the Leipzig book is larger than the recension he took as his first point of comparison, and that some of its sections are additions.

The edition remains indispensable, because it made the treatise available as a printed book with a scholarly introduction, and because it gathered the evidence of manuscripts that few readers could see. It is also the source of a confusion that still appears in casual references. A writer who says that Diqduqe ha-Teʿamim has seventy-six sections is describing Baer and Strack, not the Leningrad copy, and not the core that Dotan later isolated. The same caution applies to the forty-two rubrics of the form Ginsburg printed from the tradition of the Leningrad Codex. Forty-two is the shape of one important copy, not a divine number.

Baer, whose printed Hebrew Bibles are discussed in the journal essay on his Leipzig text, was not a disinterested typographer. He had a view of what a correct Tiberian text should look like, and he was willing to adjust printed Bibles to it. The 1879 treatise should be read with that career in mind, not as an attack on the edition, but as a reminder that compilation and correction were part of the same nineteenth-century project of making the Masorah usable.

Dotan’s Core of Twenty-Six Sections

Aron Dotan’s critical edition, published at Jerusalem in 1967, is the modern attempt to work back from this variety to a core. The review of his later Thesaurus of the Tiberian Masora in the Revue des études juives cites that edition, and it is from that scholarly record that the figure of twenty-six original sections has entered the standard accounts.¹⁸ Dotan’s result is not a claim that every later rubric is worthless. It is a claim about what can be assigned, with some confidence, to the treatise as a short, coherent work, before copyists and compilers lengthened it.

Twenty-six, forty-two and seventy-six are therefore not rival estimates of the same object. They are descriptions of different stages. The core is the work as a critical editor can reconstruct it. The Leningrad form is the work as it stood in a great eleventh-century Bible. The Leipzig edition is the work as two nineteenth-century scholars compiled it from the manuscripts they had. A responsible account names the stage it is talking about. The hub note does that in a table; this essay has tried to show how the numbers were produced.

Dotan’s edition also belongs to a wider modern effort, in which the Tiberian reading is described from manuscripts and from the surviving manuals rather than from later printed convention. Khan’s volumes on the Tiberian pronunciation tradition are part of that effort, and they repeatedly send the reader back to the older treatises, including this one, for evidence of how particular signs were understood.¹⁹ The short book at the end of the codex has become, in that setting, a witness to be weighed, not a rule-book to be reprinted unaltered.

Lists, Rules and the Temptation of a System

Ginsburg’s remarks on the official lists of Ben Asher and Ben Naphtali, which immediately precede his account of the treatise, explain why he was so unwilling to treat Diqduqe ha-Teʿamim as a code. Those lists, he writes, often differ among themselves even in the Pentateuch, “where the greatest care has been taken to transmit the punctuation of Ben-Asher and Ben-Naphtali.” The attempt “to reduce these variants into a system, to formulate rules from these conflictingly recorded differences and to apply these rules to other passages” is, in his view, “a task far more in harmony with the superfine ingenuity of some mediaeval grammarians than with sober textual criticism.”²⁰

He illustrates the point from the oldest and most beautifully written Pentateuch in the British Library, Orient. 4445, a manuscript that is the subject of its own hub note. That codex “very rarely employs the Metheg or Gaya even before Chateph-pathach, and yet it is the presence or absence of the Metheg or Gaya which constitutes fully nine-tenths of the differences between these two redactors of the text.”²¹ If the oldest manuscript does not follow the lists uniformly, a short treatise of fine points cannot be used to manufacture rules that the manuscripts themselves do not observe.

The same caution should govern the modern use of Diqduqe ha-Teʿamim. The work is evidence of what a Tiberian teacher, or a later compiler writing under that name, thought it useful to record. It is not a key that will unlock every later copy of the Bible. Where it agrees with the great model codices, it confirms a practice. Where it diverges, or where the recensions diverge among themselves, it records a difficulty, not a licence to regularise the text. The journal essay on Mishael ben Uzziel is the better place to follow the differences between Ben Asher and Ben Naphtali as a set of recorded variants. The treatise belongs beside that evidence, not above it.

Authorship and Affiliation

The ascription to Aaron ben Asher is early and persistent, but it is not simple. Ginsburg’s first recension already carries the name. The Tzufutkale epigraphs repeat it, in one case in a hand that Strack did not trust. The Jewish Encyclopedia, in its article on Aaron, judged a Karaite identification of the Ben Asher family “untenable,” and it treated an ascription of the Aleppo Codex in this connection as unfounded.²² Later scholarship has reopened parts of that question, and this essay does not try to settle the affiliation of the family. What can be said from the sources used here is more limited. The treatise is transmitted under Aaron’s name; some of the most confident-looking dates and epigraphs have been challenged by the scholars who first published them; and the work itself, being a collection of fine points rather than a signed preface, does not carry a autobiographical statement that would close the matter.

The Jewish Encyclopedia’s longer article on the Masorah remains a convenient older survey of the literature in which the treatise sits, and of the printed Bibles and handbooks that used it.²³ For the signs that the treatise discusses, the same encyclopedia’s article on punctuation is still a useful starting point, though it must be read through the later work of Dotan, Khan and Yeivin.²⁴

The right conclusion is a modest one. Diqduqe ha-Teʿamim is a real Tiberian work, short, selective and often rhymed, that was copied at the ends of Bibles and then compiled by modern editors. It is not a complete grammar, and it is not a single text with a single number of sections. A reader who wants a first orientation will find it in the hub note. A reader who wants to see how the copies were counted and why the counts disagree will need Ginsburg’s chapter X, the Leipzig edition of 1879, and Dotan’s reconstruction of 1967. The fine details of the accents were never only a matter of the signs on the line. They were also a matter of how a short book was copied, enlarged and printed.

Notes

1. Geoffrey Khan, The Tiberian Pronunciation Tradition of Biblical Hebrew, vol. 1 (Cambridge: Open Book Publishers, 2020), introduction, §I.0.13. *
2. Aaron ben Moses ben Asher, in the Jewish Encyclopedia (New York: Funk and Wagnalls, 1901–1906), at JewishEncyclopedia.com. *
3. Khan, Tiberian Pronunciation Tradition, vol. 1, chapters on shewa and the ḥaṭef vowels. *
4. “Punctuation,” in the Jewish Encyclopedia (New York: Funk and Wagnalls, 1901–1906), at JewishEncyclopedia.com. *
5. Christian D. Ginsburg, Introduction to the Massoretico-Critical Edition of the Hebrew Bible (London: Trinitarian Bible Society, 1897), ch. X, pp. 278–279. *
6. Ginsburg, Introduction, ch. X, p. 278. *
7. Ginsburg, Introduction, ch. X, pp. 278–279. *
8. Ginsburg, Introduction, ch. X, pp. 278–280. *
9. Ginsburg, Introduction, ch. X, p. 278. *
10. Ginsburg, Introduction, ch. X, p. 278. *
11. Ginsburg, Introduction, ch. X, pp. 278–279. *
12. S. Baer and H. L. Strack, Die Dikduke ha-ṭeamim des Ahron ben Moscheh ben Ascher (Leipzig, 1879), Einleitung, p. XXXIII. *
13. Ginsburg, Introduction, ch. X, pp. 279–280. *
14. Baer and Strack, Die Dikduke ha-ṭeamim, Einleitung, p. XXXIV. *
15. Ginsburg, Introduction, ch. X, pp. 279–280. *
16. Baer and Strack, Die Dikduke ha-ṭeamim. *
17. Ginsburg, Introduction, ch. X, p. 279. *
18. Review of Aron Dotan, Thesaurus of the Tiberian Masora, Revue des études juives 147 (1988), at Persée, citing Dotan’s critical edition of the treatise (Jerusalem, 1967). *
19. Khan, Tiberian Pronunciation Tradition, vol. 1, introduction, §I.0.13. *
20. Ginsburg, Introduction, ch. X, p. 278. *
21. Ginsburg, Introduction, ch. X, p. 278. *
22. Jewish Encyclopedia, “Aaron ben Moses ben Asher”. *
23. “Masorah,” in the Jewish Encyclopedia (New York: Funk and Wagnalls, 1901–1906), at JewishEncyclopedia.com. *
24. Jewish Encyclopedia, “Punctuation”. *