The Hidāyat al-Qāriʾ as a Guide for the Reader

In the first half of the eleventh century, in Jerusalem, a Karaite grammarian named ʾAbū al-Faraj Hārūn ibn Faraj sat down to describe what a pointed Bible does not, by itself, teach: the length of the vowels, the precise realisation of the consonants, the way a shewa was to be sounded, and the force of an accent. Those things were taught, and he wrote a manual so that a reader could learn them. He called the work, in one of its titles, a guide for the reader. The hub note on the Hidāyat al-Qāriʾ sets out what the treatise is and how it is organised. This essay follows the book as it was rewritten, abridged and renamed, from Jerusalem to the Genizah, to Yemen and to Europe.

Hidāyat al-Qāriʾ, a guide for the reader.

One of the titles by which ʾAbū al-Faraj’s manual is known

The work is not a single text. Its author wrote a long version first and then a short one. Later readers made further recensions in Arabic and in Hebrew. The copies that survive carry several titles, and one European Hebrew version was even given to the wrong author. A modern edition has had to reconstruct the Arabic text from many manuscripts, most of them without colophons. The history of the guide is, in that sense, a history of its forms.

A Book Written after the Last Masoretes

ʾAbū al-Faraj wrote several decades after Aaron ben Asher and Ben Naphtali, the last Tiberian authorities. He had no direct contact with them. Geoffrey Khan, whose work on the treatise is presented in two open-access volumes and on his Cambridge page, is careful on the point that matters next:¹ teachers of the Tiberian reading could still be found in Palestine in the eleventh century, and the earlier Masoretic treatises were still available. The guide is therefore a description made at one remove from the Masoretes of Tiberias, yet close enough to their oral tradition to record what the signs alone cannot show.² His authorship is not a later guess. It is stated explicitly in a scribal note on a Genizah fragment of the short version now in the John Rylands Library, Manchester, the fragment JRL A 694, as reported by Ilan Eldar and cited by Khan.³

That setting explains the book’s confidence. In the introduction to the long version, ʾAbū al-Faraj identifies the reading of the Land of Israel with the Tiberian reading and calls it the original one. The people of the communities of the exile, he writes, would press any teacher who travelled from Tiberias to those distant lands to teach their children that reading, and would make him sit down so that they could learn it from him. A traveller who came from the exile to the Land of Israel, he adds, was equally eager to abandon his own tradition and to learn the reading of the land.⁴ The same introduction claims that the accents have been fixed since the time of the prophets, and offers three arguments for the claim, one of them being that a reader who gave the first word of Genesis the wrong accent, or who confused one accent with another, would be declared in error by the congregation.⁵ Modern scholarship does not share the dating. Khan, like the journal essay on the vowel-point controversy, places the invention of the signs in the early Islamic period; the publisher of the Stuttgart edition has put it around 700, and elsewhere around 600.⁶⁷ The author’s claim is still evidence, not for the age of the signs, but for the authority that the Tiberian reading still carried in his day.

The three arguments themselves are worth following, because they show how a Karaite grammarian in Jerusalem made a historical case for a living reading. The first is that the nation has not been cut off from the holy land, except in the period between the two temples. After the returns under Ezra and Nehemiah, Abu al-Faraj writes, the people were not cut off from the land “until this day,” and the reading that Ezra and his congregation used is the reading of the Land of Israel in his own time, transmitted from generation to generation. An objection is anticipated: Edom deported the people and cut them off from Jerusalem. The reply is that Edom cut them off from Jerusalem only. There was still pilgrimage to Tiberias from the Levant and the East, and to Gaza from the West, during the period of Edomite control, “about five hundred years, in accordance with what is mentioned in their chronicles.” Those who dwelt in the land, he says, taught their children the Torah and the rest of the twenty-four books, including “the correct division of the accents and their musical rendition.”⁸

The second point turns on the communities that did not return. Those who remained in exile, he says, lost knowledge of the accents and began to read with different cantillations. The Jews of Byzantium read in a way different from the people of Iraq; the Jews of Persia differ from both; the people of the West differ from all of these; and there may be still other communities, driven to remote countries, with a way of their own.⁹ The diversity of diaspora readings is, for him, the exception that proves the rule: the reading of the land is one, and it is the reading a travelling teacher from Tiberias was pressed to teach. The third point, already mentioned, is the congregation’s refusal to accept a wrong accent on the first word of Genesis, or a substitution of one accent for another.¹⁰ Taken together, the arguments are not a modern history of the signs. They are a defence of a standard that, in the author’s hearing, could still be taught, travelled with, and enforced.

Before the three arguments, the introduction also offers a brief speculation on the origin of the melodies themselves. Asked what the origin of the accent melodies was before the later arrangement was fixed, Abu al-Faraj answers that the people of the language may have paused on one letter and made it slightly longer than another, enough to make their intentions known.¹¹ That is a grammarian’s hypothesis, not a Masoretic list, and it already shows the hybrid character of the book: a practical guide that is willing, when asked, to think about how a reading could have begun.

A Long Version of Three Discourses

The long version was written first. In its introduction ʾAbū al-Faraj says that the work consists of three discourses (maqālāt), one on the consonants, one on the vowels and one on the accents. The introduction to the third discourse refers to a slightly different division into two parts (ʾajzāʾ), with the first two discourses in part one and the third in part two. The first part, it explains, contains a discussion concerning the letters and a discussion concerning the vowels; the second part contains one discourse on the accents of the twenty-one books and the accents of the three books.¹² The “three books” are Job, Proverbs and Psalms, whose poetic accent system is treated separately from the twenty-one prose books, a distinction also followed in the journal essay on the poetic accents.

The literary form of the book is mixed, and Khan’s comparison is the right place to begin. The Arabic genre that might have been expected to match a guide for the reader most closely is the literature of correct Quranic recitation, known as tajwid. Manuals from roughly the same period, such as those of al-Dānī (d. 1052–3) and al-Qaysī (d. 1045), open with an introduction on the requirements of correct recitation and then treat the pronunciation of the consonants by place of articulation. That opening is clearly analogous to the first discourse of the Hidaya. The resemblance does not go much further. The tajwid manuals have no real counterpart to the philosophical sections on letters, to the morphological discussions of interchange, augmentation and elision, or to the treatment of verbal conjugations that occupy parts of the first discourse. Their remaining chapters are concerned mainly with contextual variations in the realisation of vowels, especially length. That does not correspond closely to the second discourse of the Hidaya, which treats philosophical aspects of the vowels, their role in morphological formations, the shewa, and the production and places of articulation of the vowels. The tajwid manuals have no section on cantillation accents at all, because pitch variation in the recitation of the Quran is not prescribed but improvised.¹³

The parts that have no counterpart in the tajwid manuals were brought in from elsewhere: from the Masoretic tradition, from morphological material in ʾAbū al-Faraj’s other grammatical works, from the Arabic grammatical tradition (the theory of the production of vowels can be traced, Khan notes, following Eldar, to the physician Ibn Sīnā), and from philosophical treatments of language in Muslim speculative theology (kalam). The habit of folding morphological material into a treatment of pronunciation is older than this book. It is already visible in earlier Masoretic treatises such as Kitāb al-Musawwitāt and Seder ha-Simanim. Khan’s conclusion is therefore measured. The Hidaya is hybrid in the nature and origin of its contents. It does not replicate the format of the Quranic manuals, and it is not a simple compendium of Masoretic lore, even though a sentence in the introduction might be read that way.¹⁴

The grammatical works of ʾAbū al-Faraj show, more generally, a considerable influence from contemporary Arabic linguistic thought. Khan points to convergences in literary form as well, including the authorial preface. The guide belongs to that wider corpus, not only to the narrow history of Masoretic lists.¹⁵

Why the Short Version Was Copied More Often

ʾAbū al-Faraj later abbreviated the long version. The short version has survived in many more manuscripts, a fact Khan reads as a sign of its greater popularity. The same pattern holds for his grammar: the shortened al-Kitāb al-Kāfī fī al-Lugha al-ʿIbrāniyya, “The Sufficient Book concerning the Hebrew Language,” is extant in many more copies than the original comprehensive grammar, al-Kitāb al-Mushtamil.¹⁶ Abridgement, in this author’s practice, was not an afterthought. It was a way of giving the same material a second, more usable life.

The short version is not a mechanical cut. In its introduction ʾAbū al-Faraj indicates that he may take the opportunity to include, where necessary, “what was not mentioned in the (original) Hidāyat al-Qāriʾ.” Khan compares this with the author’s practice in al-Kitāb al-Kāfī, where he sometimes expands on the earlier grammar, rearranges it, or even supersedes a view he had expressed before.¹⁷ The first two discourses of the long version were made more concise by removing the philosophical discussions of letters and vowels and the sections on morphological patterns and processes. What remains, in those first two discourses, is the practical teaching on the pronunciation of the consonants, the vowels and the shewa. The division into two discourses was itself eliminated.¹⁸ A reader who wants the theory must go back to the long version. A reader who wants to recite will find the short one enough, and the copies suggest that most readers did.

A Book of Many Names

The first part of the introduction to the short version records a small cluster of titles, and they should be kept distinct. The long version, it says, was called Kitāb al-Shurūṭ, “The Book of Conditions” or “The Book of Rules,” but was generally known (mulaqqab) by the name Hidāyat al-Qāriʾ. The term sharṭ, “rule,” is used frequently in the work. The short version (mukhtaṣar Hidāyat al-Qāriʾ) is said to have been known as Kitāb al-Muhja, “The Book of the Essential Requirements,” literally “the lifeblood.” That title corresponds closely to Muhjat al-Qāriʾ, “The Essential Requirements of the Reader,” which appears in a booklist datable to the thirteenth century, published by Allony. A Genizah fragment of the short version, the same JRL A 694 that preserves the authorship note, has a scribal note referring to the work as Kitāb al-Hidāya fī ʿIlm Shurūṭ al-Miqrā, “The Book of Guidance regarding the Knowledge of the Rules of Scriptural Reading.” A Genizah fragment of an anonymous Masoretic treatise, Cambridge T-S NS 311.113, refers to what appears to be the Hidaya by yet another title, Kitāb al-ʾAlḥān, “The Book of the Accents.”¹⁹

None of this is unusual for ʾAbū al-Faraj. Khan notes that some of his other works are referred to under different forms of title, including his glossary of biblical words, which appears in the sources under more than one name.²⁰ A title, in this tradition, is often a description rather than a fixed heading. The guide is the same work under several descriptions, and the descriptions tell the reader what each community thought the book was for: rules, essentials, guidance, or accents.

The Manuscripts without Colophons

Khan’s edition of the long version is based on all the manuscripts of the work then known to him. They are extant predominantly in the Second Firkovitch collection of the National Library of Russia in St Petersburg. He lists six St Petersburg copies (L1–L6), one Cambridge fragment (T-S NS 301.23 = L7), and a New York manuscript at the Jewish Theological Seminary (JTS 8110, EMC 829, 832 = L8). Some of these had already been identified by Eldar; Khan identified the remainder in the Firkovitch collection. The majority of sections covered by the edition are extant in their entirety, with only a few, apparently relatively small, gaps. The result is a nearly complete reconstruction of the long version, and a complete reconstruction of the short one.²¹

The short version is far more abundant. Khan used a selection, again predominantly from the Second Firkovitch collection, which preserves the longest manuscripts: twenty-eight St Petersburg copies (S1–S28) and a Cambridge Mosseri fragment (Mosseri I, 57.2 = S29). There are also several Genizah fragments of the short version, which Eldar had used in his edition of selected passages. For his own edition Khan preferred the longer Firkovitch manuscripts now available.²² The Cambridge Genizah collections remain, even so, part of the story of how the short version survived.²³

The manuscripts used for the edition of both versions are written in Judaeo-Arabic, that is, Arabic in Hebrew script. They do not include any colophons. Their hands and their codicology are similar to the manuscripts from the same collection that formed the basis of the edition of ʾAbū al-Faraj’s al-Kitāb al-Kāfī. On that comparison they can be dated to a period ranging from the eleventh to the thirteenth centuries. Their ultimate origin, Khan writes, would in all likelihood have been the Karaite synagogue in Cairo, which appears to have been the source of most of the Second Firkovitch collection.²⁴ The book is therefore recovered from a corpus that is rich, late enough to be copied and recopied, and silent about the names of the copyists.

Those copies are not uniform in their pointing of Hebrew words. Where vocalisation appears in a base manuscript, Khan reproduces it as the scribe wrote it. The signs are always Tiberian, but in many cases they deviate from the standard Tiberian Masoretic vocalisation, and so record the idiosyncrasies of individual scribes. The most common deviations include the interchange of segol and pathaḥ, and a non-standard distribution of shewa and ḥaṭef signs. The standard Tiberian forms are given in the facing translation.²⁵ The manuscripts of the guide are, in that limited way, witnesses to pronunciation twice over: once in what the treatise teaches, and once in how its scribes pointed the Hebrew they quoted.

From Jerusalem to Yemen

The Hidaya was conveyed, Khan writes, beyond the confines of Palestine to Yemen and to Europe. The two routes took different versions of the book, and they produced different later forms. The long version was transmitted to Yemen, probably in the thirteenth century. Two abridgements were made of it there, one in Arabic, edited by Neubauer in 1891, and one in Hebrew, edited by Derenbourg in 1871. Each was known as Maḥberet ha-Tījān, “The Composition of the Crowns,” because they were copied at the beginning of Bible codices known as “crowns” (Arabic tījān).²⁶ Khan adds, in a note, that another derivative Arabic version was published in Ginsburg’s 1885 corpus of Masoretic material.²⁷

A treatise copied at the front of a Yemenite “crown” has changed its function. It is no longer only a book for a grammarian. It has become a preface to Scripture, a short account of the reading placed where a reader would find it before the text itself. That is one reason the Yemenite abridgements matter for the history of the Tiberian tradition: they show the long version still being used, in a new climate, as a practical introduction rather than as a theoretical encyclopaedia. They also show that the long version, which is rarer in the Firkovitch corpus than the short one, had a second life in a community that kept both the name “crown” and a conservative reading tradition.

Hebrew Translations in Europe

The short version took the European road. It found its way to central Europe, and two full Hebrew translations were made of it. One was made in Mainz and was given the title Horayat ha-Qore, “Guide for the Reader,” in the manuscripts, the earliest being datable to the thirteenth century. The other translation was given the title Tokhen ʿEzra, “The Ruling of Ezra,” in a manuscript dated 1145, and the title Ṭaʿame ha-Miqra, “The Accents of the Bible,” in a manuscript dated 1285–1287. Both of those copies were made in Italy. In the version entitled Ṭaʿame ha-Miqra the work is erroneously attributed to the Spanish grammarian Yehudah ibn Balʿam.²⁸

The misattribution had a long afterlife. Horayat ha-Qore and Tokhen ʿEzra were for a time treated as if they were independent Hebrew works, and the version under Ibn Balam’s name entered the history of the poetic accents under a Spanish authorship the book does not possess. The journal essay on the poetic accents still records that older attribution, which Eldar and Khan have since corrected. The correction does not make the European Hebrew copies less important. It makes them what they are: translations and rearrangements of the short Arabic guide, through which a Jerusalem treatise taught a European readership how the Tiberian signs were to be read.

The dates of the Italian copies are early enough to matter. A manuscript of 1145 is not far from the world in which the Arabic short version was still being copied in the East. The book did not wait centuries to be translated. It moved while the Karaite copies in Egypt were still being written, and it changed language as it moved.

What a Modern Edition Can Recover

The sections on the consonants and vowels, Khan argues, are of great importance for the reconstruction of the Tiberian pronunciation tradition, and this applies in particular to the original Arabic long and short versions. When he wrote the introduction to the first volume, no full edition of those sections was available. He therefore included a critical edition, with a facing English translation, as the second volume of the same work. Eldar had already published selected passages from both versions, with a Hebrew translation, and in 2018 he published the section on the musical accents. Khan’s edition complements that work by making available the entire extant text of the introduction and of the sections on consonants and vowels, based on many newly identified Firkovitch manuscripts.²⁹³⁰

The edition is accompanied by a commentary that clarifies difficult passages and sets ʾAbū al-Faraj’s ideas in the linguistic and philosophical thought of his day.³¹ That is the right use of a hybrid book. A sentence about the shewa, or about the length of a vowel, may be Masoretic in substance, Arabic in its grammatical framing, and philosophical in its categories. To read it only as a list of rules is to miss the kind of book it is.

The same volumes are also a reminder of why the guide was written at all. The early model Tiberian Bible codices are, Khan writes, an important starting point for the reconstruction of the pronunciation, but they do not disclose everything. Additional sources are needed for those aspects of the reading that the signs do not show. Among those sources are the treatises written by Tiberian Masoretes, or by scholars close to their circle who still had access to the tradition. The Hidaya is one of the latest and most systematic of that group. It stands beside the earlier rhymed Hebrew tracts, including Diqduqe ha-Teʿamim, and beside the tenth-century Arabic treatises on the vowels and the shewa.³² It does not replace the Babylonian or the Palestinian pointing, which are different systems; it describes the Tiberian reading as a living practice, at the moment when that practice could still be heard in Palestine and was already being written down for readers who would not hear it.

A guide for the reader is, in the end, a book that expects to be used. The long version keeps the theory. The short version keeps the teaching. The Yemenite abridgements place that teaching at the front of a Bible. The European translations give it a Hebrew voice, and, in one case, a borrowed name. The modern edition returns the Arabic text to a state in which those later forms can be compared with their source. The encyclopedia note remains the place to begin for the shape of the work. The copies themselves, and the names they travelled under, are the reason the work still has a history.

Notes

1. Geoffrey Khan, faculty page, University of Cambridge, at cam.ac.uk. *
2. Geoffrey Khan, The Tiberian Pronunciation Tradition of Biblical Hebrew, vol. 1 (Cambridge: Open Book Publishers, 2020), introduction, §I.0.13.1. *
3. Geoffrey Khan, The Tiberian Pronunciation Tradition of Biblical Hebrew, vol. 2 (Cambridge: Open Book Publishers, 2020), introduction, §II.Int.0.1, citing Eldar 1994, 41. *
4. Khan, Tiberian Pronunciation Tradition, vol. 2, long version, §II.L.0.4. *
5. Khan, Tiberian Pronunciation Tradition, vol. 2, long version, §§II.L.0.2–II.L.0.5. *
6. Khan, Tiberian Pronunciation Tradition, vol. 1, introduction, §I.0.13. *
7. Biblia Hebraica Stuttgartensia, Deutsche Bibelgesellschaft, at die-bibel.de. *
8. Khan, Tiberian Pronunciation Tradition, vol. 2, long version, §II.L.0.3. *
9. Khan, Tiberian Pronunciation Tradition, vol. 2, long version, §II.L.0.4. *
10. Khan, Tiberian Pronunciation Tradition, vol. 2, long version, §II.L.0.5. *
11. Khan, Tiberian Pronunciation Tradition, vol. 2, long version, §II.L.0.2. *
12. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.2. *
13. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.2. *
14. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.2. *
15. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.2. *
16. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.3. *
17. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.3. *
18. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.3. *
19. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.3. *
20. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.3. *
21. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §§II.Int.0.1–II.Int.0.4. *
22. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.4. *
23. Cairo Genizah collections, Cambridge University Library, at Cambridge Digital Library. *
24. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.4. *
25. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.5. *
26. Khan, Tiberian Pronunciation Tradition, vol. 1, introduction, §I.0.13.1, citing Eldar 1994, 15–16. *
27. Khan, Tiberian Pronunciation Tradition, vol. 1, introduction, §I.0.13.1, n. 141. *
28. Khan, Tiberian Pronunciation Tradition, vol. 1, introduction, §I.0.13.1, citing Busi 1984 and Eldar 1994, 16–18. *
29. Khan, Tiberian Pronunciation Tradition, vol. 1, introduction, §I.0.13.1. *
30. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.1. *
31. Khan, Tiberian Pronunciation Tradition, vol. 2, introduction, §II.Int.0.1. *
32. Khan, Tiberian Pronunciation Tradition, vol. 1, introduction, §I.0.13.1. *