Version classiqueVersion mobile

Studies in Semitic Vocalisation and Reading Traditions

 | 
Aaron D. Hornkohl
, 
Geoffrey Khan

The prosodic Models of Andalusi Hebrew Metrics

José Martínez Delgado

Note de l’éditeur

This study was carried out under the auspices of ERDF/Ministry of Science, Innovation and Universities–State Research Agency, Project: The Judeo-Arabic Legacy of al-Andalus: The Linguistic Heritage PGC2018-094407-B-I00. In this paper I will use the following abbreviations in a conventional way: C for consonant; V for vowel; S for sabab ḵafīf; L for sabab ṯaqīl; W for watid majmūʿ; V for watid mafrūq; LS for fāṣila ṣuġrā; LW for fāṣila kubrā; T for tenuʿa; Y for yated; – for long vowel; and ˘ for short vowel.

Texte intégral

1.0. Introduction

  • 1 This, for instance, was the approach used in the classic work by Schirmann (1995, 119–22, especiall (...)
  • 2 Although medieval sources had already attributed this adaptation to Dunash ben Labraṭ, the first mo (...)
  • 3 For a summary of the different theories on Andalusi metrics in general, see Corriente (1986; 1991; (...)

1Throughout the long history of Hebrew poetry of Andalusi origin, the methodological foci and analytical proposals found in manuals have changed and evolved, always adapting to the new realities of who was producing and consuming these types of compositions. Today, it is understood that Andalusi Hebrew metrics is based on a set succession and combination of long and short syllables, with the different sequences producing different metres.1 It is also widely accepted that metrical adaptation was carried out by Dunash ben Labraṭ in tenth-century Cordoba.2 However, arguments continue even today about whether this system is based on an opposition of long and short syllables (traditional quantitative pattern) or open and closed syllables with phonic accents giving the composition its rhythm (accentuated pattern).3

  • 4 Lippmann (1827, x–xi) edition, and Del Valle (1977, 145–58).
  • 5 See folios 45–50 of the Venice edition (1546).
  • 6 The theory of ten vowels, five large or long and five small or short, was first put forward in Sefe (...)
  • 7 On the metrical syllables, ḥarf or mora see Frolov (2000, 68–93). At the end of the chapter, Moshe (...)
  • 8 Edited by Allony (1966). The triple approach to vowels (Masoretic, grammatical, and metrical) can a (...)

2The traditional conception seems to have its origin in the introductions to Hebrew metrics written by Abraham ibn ʿEzra in his Sefer Ṣaḥot4 and Moshe Qimḥi in Mahalaḵ shevile ha-daʿat.5 As both authors, Ibn ʿEzra and Qimḥi, were ‘distributors’ of the Andalusi legacy in Europe, it is not surprising that they found a simplified formula to transmit and adapt the complex classical ʿarūḍ to an Arabic-speaking and Romance-speaking public who had either lost quantitative rhythm or never known it. According to this model, metres originated in the alternation of the medieval metrical units known in Hebrew as yated (a sequence correlated with CVCVC) and tenuʿa (a sequence correlated with CVC), producing what both men considered the Hebrew metres and, to some extent, this is still used today to scan any verse that employs this metrical system, whether the poet was Arabic-speaking or not. Both authors seem to have echoed the vowel theory of Moshe’s father, Yosef Qimḥi,6 converting the concept of ḥarf, understood in the metrical system used between the tenth and twelfth centuries as the smallest unit that can be scanned, into an alternation of short and long syllables.7 This form of scanning became established on the Iberian Peninsula as well, and was described in the early fourteenth century by David ben Yom Ṭob ben Bilya of Portugal.8 The model was very widespread and would become the version transmitted among the different Jewish communities in Europe during the sixteenth and seventeenth centuries.

  • 9 Yellin (1939) and Yellin (1940, 44–53).

3This alternation of short and long syllables is similar to what Orientalist William Jones did centuries later when he believed that the Arabic ʿarūḍ was simply a copy of Greek and Latin metrics, an idea that can be found even today in manuals on classical Arabic metrics written in Europe. Much more interesting is the mixed system to study Andalusi Hebrew metrics devised by David Yellin,9 who, after recognizing the Ḵalīlian metrical system in medieval Hebrew metrics, used the paʿal paradigm to scan and catalogue the metres.

4Thus, up to four different basic prosodic models of Andalusi Hebrew metrics can be identified:

  1. the original or indigenous model used between the tenth and twelfth centuries, faithfully conveyed in an anonymous manual,10 to a lesser extent in the first texts on metrics and even in later pieces such as Yaʿaqov ben Elʿazar ha-Bavli’s thirteenth-century work;11
  2. the Romance model, devised by Andalusi authors exiled in southern Europe after the arrival of the Almohads;
  3. the classical or European model, inspired by the methods of classical Greek poetry;
  4. the mixed or Israeli model codified by David Yellin, a hybrid of all the earlier models and the one used today.

2.0. The Indigenous Model

  • 12 For the description of the metrics in this work, I have followed, firstly, classic medieval treatis (...)

5This is the original model on which the Arabic metrical science (ʿilm al-ʿarūḍ) used by medieval poets in arabophone settings was based.12 The classical Arabic metrical system is composed of sixteen metres, fifteen of which, considered classical, are attributed to Al-Ḵalīl ibn Aḥmad al-Farāhīdī (718–791), with the last (mutadārak) attributed to his disciple ʾAbū al-Ḥasan Saʿīd ibn Masʿada al-Mujāšiʿī (d. 830), known as Aḵfash al-ʾAwsaṭ.

  • 13 For the concept of ḥarf as letter, as syllable, and as mora, and for how it is used to compose feet (...)
  • 14 The metaphor consists of understanding the verse as a tent (bayit) held in place with a sabab ‘rope (...)

6The units or metrical syllables are formed by joining two or more vocalised letters followed by a quiescent letter.13 The union of two letters is called sabab ‘rope’ (the type used to tie down a tent), that of three letters in two syllables (CVCVC or CVCV̄) is called watid ‘peg’ (the kind used to fix a tent rope), while the sequence of four or more letters (CVCVCVC or CVCVCV̄) is known as fāṣila ‘fastener’.14

  • 15 To catalogue the syllables and feet, I will use the Hebrew alphabet and reproduce the original Arab (...)

7Two types of sabab are recognised:15

1. Sabab ḵafīf ‘light rope’ (henceforth abbreviated as S): a succession of two letters, the first vocalised and the second quiescent, i.e., one closed syllable (CVC, represented as Image or its equivalent, the open syllable with a long vowel (CV̄, represented as Image, since lengthened letters are considered quiescent consonants. To vocalise the first letter, any of the vowels ImageImage and Image can be used. Therefore, for metrical purposes, Image and Image are identical.

  • 16 Jastrow (1897, 7).
  • 17 Derenbourg (1880, 277–90) and Alahmad Alkhakaf and Martínez Delgado (2018, 39–49 and 99–106).
  • 18 Bacher (1888, 17–18).
  • 19 Martínez Delgado (2017, 35).
  • 20 Martínez Delgado (2017, 51 and 84).

2. Sabab ṯaqīl ‘heavy rope’ (henceforth abbreviated as l): a succession of two vocalised letters, i.e., two short open syllables (CVCV, represented as Image. These occur rarely and, in fact, are always followed by a sabab ḵafīf, producing the sequence known as fāṣila ṣuġrā (see below). To create a sabab ṯaqīl, a compound shewa or a ḥaṭef vowel Imageand Image is used and this implies that the vowel that precedes it is also a vocalised letter, for example, the mem and the ʿayin in Imagewhich is equivalent toImage (henceforth abbreviated as ls). Licence to use this is reserved only and exclusively for feet that require the presence of this metrical syllable. The sequence seems to have been established on the basis of sequences involving a vowel and a following ḥaṭef as in Image and Image in which the vowel before the ḥaṭef was parsed as short. There is no consensus among grammarians, however, about the existence of the saba ṯaqīl, since it depends on whether the mobile shewa and ḥaṭefim were considered vowels. According to Ḥayyūj, the sequence is impossible.16 Ibn Janāḥ17 and Yosef Qimḥi,18 however, accept it. According to the anonymous manual,19 it is not possible in Hebrew in isolation, although in practice it is used.20

8Two types of watid are also recognised:

1. Watid majmūʿ ‘joined peg’ (henceforth abbreviated as W): a succession of three letters, where the first and second are vocalised and the third is quiescent, i.e., a CVCVC sequence (represented as Image or CVCV̄ (represented as Image Examples of this type are Image and Image As in the case of the sabab ḵafīf, an open syllable with a long vowel is considered to contain a final quiescent letter.

  • 21 Jastrow (1897, 7).
  • 22 Martínez Delgado (2017, 52 and 83).

2. Watid mafrūq ‘separated peg’ (henceforth abbreviated as v): a succession of three letters, where the first is vocalised, the second quiescent, and the third vocalised, i.e., CVCCV (represented asImage This only appears in two circumstances and there is no consensus among the grammarians concerning it. The first case is apocopated imperatives and imperfects of verbs whose first radical is yod, in either binyan qal, such as Image or hifʿil, such as Image However, acclaimed authors such as Ḥayyūj argue that this type of sequence is equivalent to CVCC,21 while the author of the anonymous manual claims to perceive a pataḥ /a/ sound after the last quiescent consonant22 and thus considers this a CVCCV sequence. The other case occurs in segolate nouns whose third radical is weak, such as Image in which, according to the phonological theory of the period, the accent on the first radical creates a weak letter and the final heh does not count for metrical purposes. It would scan, therefore, as Image i.e., Image or CVCCV.

9There are also two types of fāṣila:

1. Fāṣila ṣuġrā ‘small fastener’ (henceforth abbreviated as ls, i.e., sabab ṯaqīl + sabab ḵafīf): a succession of four letters, the first three of which are vocalised and the last quiescent (CVCVCVC, represented as Image. This commonly occurs where there is a vowel followed by a ḥaṭef, as in Image or in cases where in the scansion of the verse a shewa is read as vocalic after a short vowel, as inImage

  • 23 Derenbourg (1880, 277–90) and Alahmad Alkhakaf and Martínez Delgado (2018, 39–49 and 99–106).

2. Fāṣila kubrā ‘large fastener’ (henceforth abbreviated as lw, i.e. sabab ṯaqīl + watid majmū‘): a succession of five letters, the first four of which are vocalised and the last quiescent (CVCVCVCVC, represented as Image The only author who defends its existence is Ibn Janāḥ who argues that it occurs in the words Image and Image which for metrical purposes would be Image and Image respectively.23

The feet result from the succession of two or three of these prosodic units or syllables. The combination of metric syllables produces up to ten feet. Two of them are composed of five letters: Imagews and Imagesw; and the other eight are composed of seven:

Imagewss, Imagessw, Imagewls, Imagelsw, Imagessv, Imagevss, Imagesvs and Imagesws.

10Once inside the poem, these feet usually undergo a series of modifications that alter their original appearance, which are known as ziḥāfāt or ʿilal, depending on their position and constancy in the composition. The metres are formed by a succession of feet, sometimes eight (four in each hemistich) and other times six (three in each hemistich). In classical theory, these sequences serve to develop and organise the five metrical circles displayed by Al-Ḵalīl ibn Aḥmad al-Farāhīd in his now lost Treatise on Metrics. These five metrical circles are arranged as follows

  • 24 Theory would later include derived or modern mustaṭīl(...)

1. Muḵtalaf: two asymmetrical feet that are repeated twice per hemistich. It includes the classical metres ṭawīlImagews wss 2x in each hemistich), madīdImagesws sw 2x in each hemistich), and basīṭImagessw sw 2x in each hemistich).24

  • 25 The theory would later include the derived metre mutawaffir or mustawfir(...)

2. Muʾtalaf: two symmetrical hemistichs that repeat the same foot three times. It includes the classical metres wāfirImagewls 3x per hemistich) and kāmilImagelsw 3x per hemistich).25

3. Muštabah: two symmetrical hemistichs that repeat the same foot three times. It includes the classical metres hazajImagewss 3x per hemistich), rajazImagessw 3x per hemistich), and ramalImagesws 3x per hemistich).

4. Mujtalab: three feet (two of them always the same) are repeated in each hemistich. It includes the classical metres sarīʿImagessw sswssv 1x in each hemistich), munsariḥImagesswssv ssw 1x in each hemistich), ḵafīfImageswssws ssw 1x in each hemistich), muḍāriʿImagewss wss sws 1x in each hemistich), muqtaḍabImagessv ssw ssw 1x in each hemistich), and mujtaṯṯImagessw sws ssw 1x in each hemistich).

5. Muttafaq: the same foot repeated eight times. It only includes the classical metre mutaqāribImagews 4x in each hemistich). Some manuals add the mutadārak metre when it is included with the Ḵalīlian circles Imagesw 4x in each hemistich) along with its variant Imagess 4x in each hemistich).

  • 26 Unlike the Romance model, which understood that the verse is produced by the succession and alterna (...)
  • 27 In the medieval Hebrew tradition (Romance or later), there are no known names for these basic compo (...)

11The succession of these feet produces the verse or bayt.26 The verse is made up of two hemistichs; the first hemistich is known as the ṣadr (in Hebrew delet) and the second as ʿajz (in Hebrew soger). The term for the first two or three feet (depending of the length of the hemistich) is ḥašw ‘stuffing’, while the last feet in each hemistich have their own name, ʿarūḍ for the last foot of the first hemistich and ḍarb for the last foot of the second hemistich.27

  • 28 This form is used in the metres ṭawīl, basīṭ, wāfir, kāmil, rajaz, ramal, sarīʿ, munsariḥ, khafīf, (...)

12The verse can be complete (tāmm), if all its feet are used;28 in the case of ṭawīl its complete form is:

Image

13The verse can be partial (majzūʾ), if it has supressed a foot in each hemistich. In ṭawīl this would be:

Image

14It can be weak (manhūk) if two-thirds of the metre is supressed. In ṭawīl this would be:

Image

15It can be divided (mašṭūr) if a complete hemistich is eliminated In ṭawīl this would be:

Image

16Moreover, if the poet rhymes ʿarūḍ and ḍarb in both hemistichs at the beginning of the poem, i.e., both feet share the rhyme and foot, this rhythm is called taṣrīʿ.

17Finally, according to this model, Hebrew verse scans in the following way:

  • 29 Sáenz-Badillos and Targarona Borrás (1988, 42*).

(1) Image My heart burns in my bowels and my eyes spill tears because I am homesick for Hammot and Mefaʿat (Shemuel ha-Nagid)29

Image

  • 30 I took the following example from Kitāb al-ʿarūḍ by Al-Rabaʿī (Badrān 2000, 61). The same example i (...)

18The scansion would be the same as in an Arabic verse such as:30

Image

19As for the tribe of Tamīm, Tamīm ben Mur, the people found them sleeping soundly.

20This scans as follows:

Image

21Additionally, all of these feet can be modified. If the sabab in the stuffing (ḥašw) feet is affected, the modification is known as ziḥāfāt (these may be isolated), and if both the sabab and the watid of the ʿarūḍ and ḍarb are affected, it is called ʿilal (once applied, it must be maintained throughout the poem). The ziḥāf is always an elision or modification that affects the second letter of the sabab. Its use is not necessary and it never alters the metre. Although theoretically it should be avoided, it is reflected quite commonly in poetic lines. It affects only the sabab and is found especially in the ṭawīl, basīṭ, and hazaj metres.

22The modification or ziḥāf can be simple:

ṯalm: the first letter of the watid majmūʿ is eliminated. ImageImage
ʾiḍmār: the vowel of the second letter is eliminated. ImageImage
ḵabn: the second letter of the foot is eliminated when it is quiescent. Image
waqṣ: the second vocalised letter of the foot is eliminated. Image
ṭayy: the fourth letter is eliminated when it is quiescent. Image
ʿaṣb: the fifth letter is made quiescent when it is vocalised. Image
qabḍ: the fifth letter of the foot is eliminated when it is quiescent. Image
ʿaql: the fifth vocalised letter is eliminated. ImageImage
kaff: the seventh quiescent letter is eliminated. ImageImageor Image
ḥaḏf: the last sabab ḵafīf of the foot is supressed. ImageImage

  • 31 Not all the metres accept these double modifications; an incompatibility occurs and solutions like (...)

23However, it can also be compound, i.e., two modifications can apply in the same foot. This can only occur in the second, fourth, fifth, and seventh letter.31 There are four types:

ḵabl: ḵabn + ṭayy. Image
ḵazl: ʾiḍmār + ṭayy. Image
šakl: kaff + ḵabn. Image
naqṣ: kaff + ʿaṣb. Image

24In turn, ʿilal is an alteration that affects both the sabab and the watid of ʿarūḍ and ḍarb and once applied, it must be maintained throughout the poem. It can consist of an addition or suppression.

25The following are additions:

tarfīl: a sabab ḵafīf is added to the watid majmūʿ at the end of the foot. Image
taḏyīl: a quiescent letter is added to the watid majmūʿ at the end of the foot. Image
tasbīġ: a quiescent letter is added to the sabab ḵafīf at the end of the foot. Image or ImageImage

26Along with these three, another addition exists that can be applied to any foot, known as ḵazm. It consists of adding one or several letters to the beginning of the stich or hemistich and is used in the ṭawīl, madīd, basīṭ, kāmil, and ramal metres.

27The following are suppressions:

  • 32 Some theorists believe that qaṣr and qaṭʿ are the same; however, qaṣr applies to sabab ḵafīf and qa (...)
  • 33 Some theorists argue that this can only occur in the foot

ḥaḏf: the last sabab ḵafīf in the foot is supressed. ImageImage
qaṭf: a sabab ḵafīf in the foot is supressed and the preceding vowel disappears. Image
qaṣr: the second letter in the sabab ḵafīf is supressed and the vowel of the first letter is eliminated. Image
qaṭʿ: the last letter of the watid majmūʿ is supressed and the vowel of the second letter is eliminated. Image
Image.32 When this phenomenon occurs in mustaṭīl, it is called tašʿīṯ, i.e., to shorten one foot in a syllable.
ḥaḏḏ: the watid majmūʿ at the end of the foot is supressed.Image33
ṣalm: the watid mafrūq at the end of the foot is supressed. Image (only in the sarīʿ metre).
waqf: the vowel of the last letter of the watid mafrūq is eliminated. Image (specifically in the sarīʿ and munsariḥ metres).
kašf: the last letter of the watid mafrūq is supressed. ImageImage (specifically in the sarīʿ and munsariḥ metres).

28Just as there can be double additions, likewise double suppression can occur:

batr: ḥaḏf + qaṭʿ, the quiescent letter of the watid majmūʿ is suppressed and eliminated, leaving what precedes it quiescent. Image (madīd). Image (mutaqārib).

29Additionally, other suppressions exist that can be applied to any foot. These are:

kharm: the first letter of the watid majmūʿ of the first foot at the beginning of the verse is suppressed in the ṭawīl, hazaj, muḍāriʿ, muqtaḍab, and mutaqārib metres. Image When this occurs in the foot Image it is called ʿaḍb.
ṯarm: ṯalm + qabḍ. ImageWhen this occurs in the foot Imageit is called šatr.
ḵarb: ṯalm + kaff. Image
qaṣm: ṯalm + ʿaṣb. ImageThis should not be confused with ʿaḍb, which occurs only at the beginning.
jamm: ṯalm + kašf. Imageʿaqd: ṯalm + naqṣ. Image

  • 34 Traditionally catalogued as ṭawīl. Yaʿaqob ben Elʿazar catalogues it as a variant of hazaj (Yahalom (...)
  • 35 Sáenz-Badillos (1980, 1*; vocalisation mine).

30With this model, it is possible to affirm that the metre used in the first known compositions by Dunash ben Labraṭ (ca. 958) is a modified version of mustaṭīl (wssss).34 This was established as a classical formula and was reproduced and used exclusively in the musammaṭ genre (both muṯallaṯ and murabbaʿ) by the four great Hebrew poets of the Golden Age (mid-eleventh to mid-twelfth century), namely Samuel ben Nagrela, Shelomo ibn Gabirol, Moshe ibn ʿEzra, and Yehuda ha-Levi. An example of this scansion is as follows:35

(2)Image

31Know, my heart, wisdom, science and reflection. Follow the paths of intelligence, listen to the disciplined ones

Image

The second foot, Image has been modified by qaṭʿ (tašʿīṯ because it occurs in mustaṭīl), producingImage

  • 36 Brody and Schirmann (1974, 98).

32In the following composition, likewise, Ibn Gabirol has not created a hybrid metre,36 but rather has applied a ṯalm modification to the ṭawīl metre with a taṣrīʿ rhythm, as confirmed by the second verse:

(3)Image

Image ה

33Who is this who rises like the dawn and comes into sight, shines like a radiant sun, so beautiful

34Like a daughter of kings, noble and elegant; her aroma is like the aroma of burnt myrrh and a thurible

Image

  • 37 Millás-Cantera (1956, 77–78).

The same can be said about the funeral epitaph of Shemuel ben Shoshan of Toledo dated 1257 (lines 2 and 4):37 the metre is wāfir (with ʿarūḍ and ḍarb affected by qaṭf) with frequent modification of the original foot Image with ʿaṣb, producing Image in all its feet except the second, without any need to eliminate the ḥaṭef vowel beneath the guttural consonant during the scansion:

(4) Image

35Lord of might, of absolute supremacy, he and wisdom born at the same time, twins

Image

  • 38 I am following the edition by Sáenz-Badillos and Targarona (1994, 476–83).

36At times, editors argue that the compositions lack metre (¯ ¯ ¯ ¯ / ¯ ¯ ¯ ¯). However, this model makes it possible to identify the metre by taking into account the modifications. For example, the following composition by Yehuda ha-Levi38 is clearly a mutaqārib affected by ṯalm:

(5) Image

37Who decides and executes this in the high Heavens, and over the far-off seas, his justice shines

Image

3.0. The Romance Model

  • 39 See note 4.
  • 40 In fact, the first poems attributed to Dunash ben Labraṭ were scanned using this Eastern form that (...)

38The first allusions to this model are found in the writings of the Andalusi Jews who settled in Provence after the Almohad conquest of 1146. The oldest treatise that uses this model to explain Andalusi Hebrew metrics is Sefer Ṣaḥot by Abraham ibn ʿEzra, written in Mantua in the mid-twelfth century.39 This is an adaptation of the original Andalusi metrical system of feet appropriate for a non-arabophone context and is a much smaller, simplified version of the original model with a mixture of elements that were in vogue in Andalusi Hebrew poetry, such as the repeated use of musammaṭ.40

39Unlike the complexity of the original Andalusi model, the only rule specified by Abraham ben ʿEzra is ורמשי קר רפסמ תועונתה the number of‘ אושהו עעונתמה םע העונתה איהש וירחא ןירוק ותוא דתי vowels simply must be maintained; mobile shewa followed by a vowel is called yated’. After this brief clarification, he distinguishes seventeen forms that constitute eleven metres, of which only ten are classical. At no time does he speak of feet, only of an alternation of yated (henceforth abbreviated as Y) and tenuʿa (henceforth abbreviated as T). Ibn ʿEzra distinguishes the metres without naming them. The first three cases are Hebrew musammaṭ:

  1. YTTTT:41 Ibn ʿEzra shows that this sequence can repeat in a line three (muṯallaṯ or ternary) or four (merubbaʿ or quaternary) times.
  2. TTTT:42 the line must be a musammaṭ merubbaʿ (quaternary: bbba, ccca, etc.), it has no yated, according to Ibn ʿEzra, it is לק ‘light’ and may even dispense with the internal rhymes of the musammaṭ.
  3. YTYT: the line must be a musammaṭ merubbaʿ (quaternary: bbba, ccca, etc.) where a shewa has been added to each foot (YT). This is the classical mutaqārib metre in its complete form.
  4. TTYTT:43 this metre continues to have a merubbaʿ (quaternary) line, but it is ‘greater’ than the previous one. The internal rhyme alternates between the different feet that form the hemistichs of the verses (abab, abab, etc.). According to Ibn ʿEzra, a shewa can be added at the beginning of each sequence (YTYTT) and then the original, complete form of the ṭawīl reappears.
  5. YTTYTT:44 this is one of the most common metres in Hebrew and corresponds to the Arabic hazaj. Ibn ʿEzra asserts that some add YT to the end of this sequence, producing YTT YTT YT, which is in point of fact the wāfir metre.
  6. TTYTTYTT:45 this metre is, according to Ibn ʿEzra, the richest, because it has up to three types of variants. The first is obtained by adding a shewa to the beginning of the YTYTTYTT line;46 the second adds a vowel to the end, producing TTYTTYTTT;47 the third is the only one that preserves ḍarb and ʿarūḍ, producing a complete classical sequence TTYTTYTTY TTYTTYTTT.48
  7. TTYTYT: this is the classical mujtaṯṯ metre in partial form (majzūʾ) with complete ʿarūḍ פַאעִלַאחֻן and identical ḍarb.
  8. TYTTYT:49 this verse actually reproduces an incomplete TYTT sequence, but this is the ramal metre as it is most commonly used.50
  9. TTYTTY: in principle, it is not clear in the text if this is a variant of the previous case or a new metre. Given that it is clearly a rajaz in the majzūʾ, or partial, form with complete ʿarūḍ (מֻסְחַפְעִלֻן) and identical ḍarb, it should be identified as a different metre. According to Ibn ʿEzra, it is used with a merubbaʿ verse (quaternary musammaṭ).
  10. TTYTY: according to Ibn ʿEzra, in this sequence that repeats four times, the Y at the end must be replaced by T. The complete sequence would be TTYTY TTYTY TTYTY TTYTT.51 According to the author, this is a difficult metre to use (kaḇed) in Hebrew.
  11. TYTY:52 Ibn ʿEzra asserts that this is the most difficult of all. According to the example, it is used with a merubbaʿ line (quaternary musammaṭ).

40Moshe Qimḥi presents something similar in his work Mahalaḵ Shevile ha-Daʿat. As above, the prosodic units are Y and T. Qimḥi distinguishes the following metres:

  1. TTTTTTTT (mutadārak, 2 in Ibn ʿEzra’s classification);
  2. YTTTT (mustaṭīl, 1 in Ibn ʿEzra’s classification);
  3. YTTTTYTT (kāmil, 6 in Ibn ʿEzra’s classification);53
  4. YTTYTTYT (wāfir, 5 in Ibn ʿEzra’s classification);
  5. YTTYTT (hazaj, 5 in Ibn ʿEzra’s classification);
  6. YTYTTYTYTTT (ṭawīl, 4 in Ibn ʿEzra’s classification);
  7. YTYTYTYT (mutaqārib, 3 in Ibn ʿEzra’s classification);
  8. TYTTYT (ramal, 8 in Ibn ʿEzra’s classification);
  9. TTYTTTTYTT (basīṭ, 10 in Ibn ʿEzra’s classification);
  10. TTYTTTTYT (sarīʿ, not found in Ibn ʿEzra’s classification);
  11. TTYTTYTTT (rajaz, 6 in Ibn ʿEzra’s classification);
  12. TTYTYT (mujtaṯṯ, 7 in Ibn ʿEzra’s classification);
  13. TTYTTYTT (kāmil, 6 in Ibn ʿEzra’s classification);54
  14. TTYTTY (rajaz, 9 in Ibn ʿEzra’s classification);
  15. TTYTTTT (munsariḥ, not found in Ibn ʿEzra’s classification);
  16. TTYTY (basīṭ, 10 in Ibn ʿEzra’s classification);
  17. TTYTTYTTT (rajaz, 6 in Ibn ʿEzra’s classification);
  18. TYTY (mutadārak, 11 in Ibn ʿEzra’s classification).
  • 55 Allony (1966).
  • 56 See Cohen (2000, 66–76) for the Arabic version and ibid. (155–67) for the Hebrew version.
  • 57 A review of all these authors can be found in Del Valle (1988, 349– 459).

41This model was highly successful and is frequently found in other manuals on metrics written between the fourteenth and fifteenth centuries in the Iberian Peninsula. It appears, for example, in the work of David ben Yom Ṭob ben Bilya from Portugal, who divides the vowels into kings (qameṣ, pataḥ, ṣere, segol, ḥolem, and ḥireq) and servants (shewa and qibbuṣ śefatayim or shureq) and, as Qimḥi did, includes up to eighteen variants of what are today considered nine metres.55 In the last years of Nasrid Granada, Saʿadya ibn Danān wrote an introduction to his dictionary that included the tripartite conception of Hebrew vowels (Masoretic, grammatical, and metric). He dedicated an entire chapter to the art of writing poetry,56 following this model in broad terms and deviating in many respects from the indigenous model when he tried to merge them. Beginning in the fifteenth century, this was the model that was transmitted among the different Jewish communities dispersed around Europe in works by distinguished teachers such as Avshalom ben Moshe Mizraḥi, Abravanel, and David ben Yaḥya from Portugal, and in the sixteenth and seventeenth centuries in the works of the Italians Emmanuel ben Yequtiʾel, Azaría de Rossi, Samuel Archivolti, and Emmanuel Fransis, and the Dutchman Salomón de Oliveyra.57

4.0. The Classical or European Model

  • 58 On the history of the study of ʿarūḍ in Europe, see Frolov (2000, 1– 22).
  • 59 Cano (1987, 31–38).

42The inspiration for this model is classical Greek poetry and to some extent it is the heir to the thesis set forth by William Jones (1746–1794) for Arabic metrics in his Poeseos Asiaticae Commentariorum Libri Sex (Lipsiae, 17772) and continued, first by Heinrich Ewald (1803–1875) in his De Metris Carminun Arabicorum Libri Duo (Brunsvigae, 1825–1854), and then by William Wright (1830–1889) in his renowned A Grammar of the Arabic Language (Cambridge, 19693).58 With regard to Andalusi Hebrew poetry, the most complete and exact description of this model comes from María José Cano, who used this method to codify and scan the entire dīwān of the Andalusi poet Shelomo ibn Gabirol.59

43In this model, metre is understood to be quantitative and based on alternating long and short syllables. As with the Romance model, only two basic prosodic syntagms are recognised:

yated: the succession of short and long syllables (iambo in Greek and watid majmūʿ in Arabic), represented below by | – ˘ |.
tenuʿa: a long syllable (sabab ḵafīf in Arabic), represented below by | – |.

44The vowel in the long syllables can be any of the seven Tiberian vowels, while the vowels in the short syllables can only be the simple shewa, its three compounds and the waw conjunction vocalised with shureq (ּ ו ).

45The feet are composed of three or four syllables and, therefore, can, from a prosodic point of view, be binary or ternary.

Binary (reading from right to left)

Image

Ternary (reading from right to left)

Image

46According to this model, the metres can be symmetrical (the first two feet are identical), asymmetrical (the first two feet are different), or free (muwaššaḥ or šir ha-ʾezor).

47The line (bayit) is divided into two hemistichs. The first is known as delet and is responsible for determining the metre in this model. The second, known as soger, is usually a repetition of the first, carries the rhyme, and also usually takes most of the modifications (the addition or suppression of syllables, the suppression of a letter, or even the whole hemistich).

48The symmetrical or simple metres can be binary or ternary:

Binary feet (reading from right to left):

Image

Ternary feet (reading from right to left)

Image

The asymmetrical or compound metres alternate binary and ternary feet (reading from right to left):

Image

5.0. The Israeli Method

  • 60 Yellin (1939; 1940, 44–53).
  • 61 Brody (1895).

49This is a mixed system devised by David Yellin to study Andalusi Hebrew metrics60 based on the first codifications of Yehuda ha-Levi’s metrics by Heinrich Brody.61 It is, in short, a hybrid of the indigenous and classical forms, but with some confusion produced by the Romance model.

  • 62 See my list in Martínez Delgado (2017, 123–37).

50Brody’s initial conclusions about Andalusi Hebrew metrics were harshly criticised by Halper (1913). He said that Brody did not correctly identify many metres and manipulated the vocalisation because he paid more attention to theory than to practice, basing his analysis on Freytag and blindly following Ibn Danān. Brody concluded that some feet are impossible in Hebrew: two short vowels cannot follow each other (they are replaced by a long vowel in Arabic), and there can be no watid mafrūq (no Arabic verses end in a short vowel, since they are always scanned as long). According to Halper, viewed from the Arabic, all metres fit, but Arabic can use long and short syllables, while Hebrew prefers to use long ones. This, then, led Brody to assert that there are some impossible metres in Hebrew. Halper questioned the extent to which Arabic metres can be used to analyse Hebrew metres.62 He argued that the equation of Hebrew mobile shewa with an Arabic short vowel is an artificial equivalence used by Hebrew poets that should not affect the recitation of the poem. It would appear that Halper was describing Arabic metrics using Greek categories and understood that all syllables are long in Hebrew, including accentuated ones. He argued that the arabophone Jews were aware that shewa often corresponded to short Arabic vowels and made the change. Halper rightly suspected that a vowel followed by shewa could be two short vowels.

  • 63 For the Hebrew adaptation of these circles, see Yellin (1940, 47–52).

51In these circumstances, David Yellin saw no other choice but to catalogue all the Arabic metres used by Shemuel ha-Nagid following the indigenous model. This resulted in the new system that recognises the five Ḵalīlian circles in Hebrew metrics and uses the root paʿal to represent schematically the scansion and to catalogue the metres.63 However, the current use of the paʿal scheme to scan these compositions from the classical period (tenth–twelfth centuries) is unsatisfactory since it distorts the reality of the modifications applied. In Yellin’s system the metrical patterns are reduced to seven basic feet:

Image

The last two feet are actually forms that result from modifications (ziḥāfāt) in the indigenous model: Image (ss) and Image (sss), and in no case form part of the original metre. In fact, it is these modified forms that confirm that the Jewish poets were respecting the Arabic feet and did not Hebraise them as Yellin claims. If they had done so, the foot Image (never Image ) would have resulted in an impossible Image after a double batr-type modification.

  • 64 Yellin (1939, 189 n.1 and especially 195).

52Nonetheless, the exhaustive cataloguing of metres carried out by Yellin is one of the most important contributions to the study of Hebrew metrics in modern times. This system, however, creates some confusion when attempting to identify metres, such as mixed metres and bimetric compositions and, of course, when attempting to identify metres in the first Hebrew poetry written in the mid-tenth century. By way of example, Yellin correctly identified the šalem metre. He spoke, however, of another mixed metre devised by the Hebrew poets that he called ha-šalem we-ha-soʿer.64 The application of classical metre to these compositions shows that the poets were really using two different Arabic metres, kāmil and rajaz, on each occasion.

53Finally, the greatest contribution of this model has been in the area of nomenclature for the study of Andalusi Hebrew poetry. These are the Hebrew translations of the names of Arabic metres that are used today in any study in this field.

Arabic Hebrew (20th cent.) Ibn Danān (15th cent.)
Ṭawīl ʾAroḵ ʾAroḵ
Madīd Mitmoded Mašuḵ
Basīṭ Mitpaššeṭ Pašuṭ
Wāfir Merubbe ʿOdef
Kāmil Šalem/Šalem we-Soʿer Tamim
Hazaj Marnin Megil
Rajaz Šalem/Šalem we-Soʿer Ḥaruz
Ramal Qaluaʿ Ḥol
Sarīʿ Mahir Memaher
Munsariḥ Dohar Myuttar
Ḵafīf Qal Qal
Muḍāriʿ Dome Medamme
Mujtaṯṯ Qaṭuaʿ Pasuq
Muqtaḍab Meʾussaf
Mutaqārib Mitqareḇ Mitqareḇ
Mutadārak Mašlim/Mišqal ha-Tenuʿot Teʿuni
  • 65 For Ben Tehillim, see the edition by Sáenz-Badillos and Targarona Borrás (1997), and for the rest, (...)
  • 66 For example, his secular poetry, ed. by Brody and Schirmann (1974). See also Jarden (1971–1972); an (...)
  • 67 This is the case for his secular poetry edited by Brody (1934); Brody (1942); Pagis (1977); and his (...)
  • 68 See the classic edition by Brody (1894); and the anthology of Yehudah ha-Levi by Sáenz-Badillos and (...)

54All the important studies of Hebrew Andalusi metrics have been based on this system, including N. Allony’s work (1951) on the metrics of Dunash ben Labraṭ and other poets from the Golden Age, A. Mirsky (1961, 25–35) on the dīwān of the wandering poet Yiṣḥaq ben Ḵalfun, and Y. Yahalom (2001) in his study of Yaʿaqob ben Elʿazar. This is also the model used in the important anthology of Schirmann (1954–1959) and the main modern editions of the dīwāns composed by the four great Hebrew poets of the Golden Age, Samuel ibn Nagrela,65 Shelomo ibn Gabirol,66 Moshe ibn ‘Ezra,67 and Yehuda ha-Levi.68

Bibliographie

6.0. References

Alahmad Alkhalaf, Ahmad, and José Martínez Delgado. 2018. Risālat al-taqrīb wa-l-tashīl de Abū l-Walīd Marwān ibn Ǧanāḥ de Córdoba. Edición diplomática y traducción. Madrid: Diéresis.

Allony, Nehemya. 1951. The Scansion of Medieval Hebrew Poetry. Jerusalem: Mahbarot Lesifrut and Mosad Harav Kook (in Hebrew).

——. 1966. ‘Derex laʿasot ḥaruzim le-David ibn Bilya’. Qoveṣ al-yad 16: 225–46.

Álvarez Sanz y Tubau, Emilio. N.d. Tratado de la Poesía Árabe. Tetuan: La Papelera Africana.

Amīn, Aḥmad et al. 1948. Ibn ʿAbdrabbihi, Kitāb al-ʿIqd al-farīd. Cairo: Laǧnat at-ta·līf wat-tarǧama wannašr.

ʿAtīq, ʿAbd al-ʿAzīz. 1987. ʿIlm al-ʿarūḍ wa-l-qāfiya. Beirut: Dār al-Nahḍa al-ʿArabīya.

Bacher, Wilhelm. 1888. Yosef Qimḥi, Sefer ha-zikkaron. Berlin: Selbstverlag des Vereins M’kize Nirdamim (Dr.A. Berliner).

Badrān, Muḥammad. 2000. Al-Rabaʿī, Kitāb al-ʿarūḍ. Beirut and Berlin: Das Arabische Buch.

Basset, René. 1902. La Khazradjyah, Traité de Métrique Arabe par Ali el Khazradji. Algiers: Fontana.

Brody, Heinrich. 1894. Diwan des Abú-l-Hasan Jehuda ha-Levi. Dīwān. Kol shire Yehudah ha-Levi. Berlin: Bi-defus Tsevi Hirsh Itsḳoṿsḳi.

——. 1895. Studien zu den Dichtungen Jehuda ha-Levis. i. Über die Metra der Versgedichte. Berlin Druck: H. Itzkowski.

——. 1934. Moše Ibn ʿEzra. Shire ha-ḥol. Sefer rishon: Dīwān, shire ʾezor, miḵtavim, sefer ha-ʿanaq. Berlin: Shoḳen.

——. 1937. ‘‘Al mishqal ha-ʿaraḇi be-shirat ha-ʿivrit’. In Sefer ha-Yuval li-Shmuʾel Kraus, edited by S. Klein, 117–26. Jerusalem: Reʾuben Mas.

——. 1942. Moshe Ibn ʿEzra. Shire ha-ḥol. Sefer sheni: Beʾur la-dīwān, reshimot u-mafteḥot. Berlin: Shoḳen.

Brody, Heinrich and Jeffim Schirmann. 1974. Solomon ibn Gabirol, Secular Poems, with the participation of J. Ben-David. Jerusalem: Schocken Institute for Jewish Research.

Cano Pérez, María José. 1987. Selomoh ibn Gabirol, Poemas. Granada: Universidad de Granada.

Cohen, Moshe. 2000. The Grammatical Introductions to “The Book of sources” of Saʿadya ibn Danān. Jerusalem: Kfir Press, Meqor Baruch Publications (in Hebrew).

Corriente, Federico. 1986. ‘Métrica hebrea cuantitativa, métrica de la poesía estrófica andalusí y ‘arūḍ’. Sefarad 46: 123–32.

——. 1991. ‘Modified ‘Arūḍ: An Integrated Theory for the Origin and Nature of Both Andalusī Arabic Strophic Poetry and Sephardic Hebrew Verse’. In Poesía Estrófica, edited by F. Corriente and A. Sáenz-Badillos, 71–78. Madrid: Facultad de Filología. Universidad Complutense. Instituto de Cooperación con el Mundo Árabe.

——. 1998. Poesía dialectal árabe y romance en Alandalús. Madrid: Gredos.

Del Valle Rodríguez, Carlos. 1977. Sefer Ṣaḥot de Abraham Ibn ʿEzra. Salamanca: Universidad Pontificia.

——. 1988. El diván poético de Dunaš ben Labraṭ. Madrid: Instituto de Filología.

Derenbourg, Joseph and Hartwig Derenbourg. 1880. Opuscules et traités d’Abou’l-Walid Merwan ibn Djanah de Cordoue: Texte arabe, publié avec une traduction française. Paris: Imprimerie nationale.

Frolov, Dimitri. 2000. Classical Arabic Verse, History and Theory of ‘Arūḍ. Leiden-Boston-Cologne: Brill.

Halper, Benzion. 1913. ‘The Scansion of Medieval Hebrew Poetry’. The Jewish Quarterly Review (NS) 4 (2): 153–224.

Hāshimī, Aḥmad. N.d. Mīzān al-ḏahab fī sanāʿat shiʿr al-ʿarab. Beirut: Dār al-Fikr.

al-Hayb, Aḥmad Fawzī. 19892. Kitāb al-ʿarūḍ. Kuwait: Jamiʿat al-Kuwayt.

Jastrow, Morris. 1897. The weak and geminative verbs in Hebrew by Abû Zakariyyâ Yaḥyâ ibn Dâwud of Fez, known as Ḥayyûḡ; the Arabic text now published for the first time. Leiden: Brill.

Jarden, Dov. 1966–1992. Diwan Shemuʾel ha-nagid: ʿIm mavo, perush, meqorot, shinuye nusaḥ, reshimot, mafteḥot, milon u-bibliografya. Jerusalem: Hebrew Union College Press.

——. 1971–1972. Shire ha-qodesh le-rabi Shlomo ibn Gabirol. Jerusalem: Dov Jarden.

——. 1975. Shire ha-ḥol le-rabi Šelomoh ibn Gabirol: ʿIm perush meqorot u-maqbilot, mavo, mafteḥot, milon, u-vibliyografya. Jerusalem: Dov Jarden.

Levin, Israel and Tova Rosen. 2012. The Liturgical Poems of Moshe ibn Ezra. Tel-Aviv: The Haim Rubin Tel Aviv University Press.

Lippmann, Gabriel H. 1827. Abraham ben ‘Ezra Sefer Ṣaḥot. Fürt: Bi-defus D. Tsirndorfer.

Martínez Delgado, José. 2016. ‘Muestras del estrofismo andalusí y su métrica según la poesía hebrea’. Al-Qanṭara 37 (1): 39– 58.

——. 2017. Un manual judeo-árabe de métrica hebrea andalusí (Kitāb ʽarūḍ al-šiʽr al-ʽibrī) de la Genizah de El Cairo: Fragmentos de las colecciones Firkovich y Taylor-Schechter— Edición diplomática, traducción y estudio. Córdoba: UCO Press CNERU-CSIC.

Millás Vallicrosa, José María and Francisco Cantera Burgos. 1956. Las inscripciones hebraicas de España. Madrid: C.S.I.C.

Mirsky, Aharon. 1961. Itzhak ibn Khalfun Poems. Jerusalem: The Bialik Institute.

Pagis, Dan. 1977. Moshe Ibn ʿEzra. Shire ha-ḥol, edited by Haim Brody, Volume Three. Berlin: Hotsaʼat Makhon Shoken le-meḥkar ha-Yahadut le-yad Bet ha-midrash le-rabanim be-Amerikah.

Qimḥi, Moshe. 1546. Mahalaḵ shevile ha-daʿat. Venice: Daniel Bomberg.

Sáenz-Badillos, Ángel. 1980. Teshubot de Dunash ben Labrat. Edición crítica y traducción española, Granada: Universidad de Granada.

Sáenz-Badillos, Ángel and Judit Targarona Borrás. 1988. Šĕmu’el ha-Nagid, Poemas I: Desde el campo de batalla Granada 1038– 1056. Edición del texto hebreo, introducción, traducción y notas. Córdoba: El Almendro.

——. 1994. Yĕhudah ha-Levi, Poemas, Introducción, Traducción y Notas; Estudios literarios de A. Doron. Madrid: Alfaguara.

——. 1997. Šĕmu’el ha-Nagid, Poemas II: En la corte de Granada, Córdoba: El Almendro.

Shamseddīn, Ibrāhīm. 20082. Al-Tabrīzī, al-Kāfī fī al-ʿarūḍ wa-al-qawāfī. Beirut: Dār al-Kutub al-ʿIlmiyah.

Schirmann, Jeffim. 1954–1959. Ha-Shira ha-ʿivrit bi-Sfarad u-ve-Provans. Jerusalem: The Bialik Institute and Dvir.

——. 1995. The History of Hebrew Poetry in Muslim Spain. Jerusalem: The Magnes Press, the Hebrew University. Ben Zvi Institute (in Hebrew).

Sobh, Mahmud. 2011. Poética y métrica árabes. Madrid: Alderabán.

Yahalom, Joseph. 2001. Judaeo-Arabic Poetics, Fragments of a Lost Treatise by Elʿazar ben Yaʿaqob of Baghdad. Jerusalem: Ben-Zvi Institute, Yad Izhak Ben-Zvi and the Hebrew University (in Hebrew).

Yellin, David. 1939. ‘The metrical Forms in the Poetry of Shemuʾel Hannagid’. Studies of the Research Institute for Hebrew Poetry in Jerusalem 5: 181–208 (in Hebrew).

——. 1940. Introduction to the Hebrew Poetry of the Spanish Period. Jerusalem: Ḥevrah le-Hotsaʾat Sefarim ʿal yad ha-Universiṭah ha-ʿIvrit (in Hebrew).

Notes

1 This, for instance, was the approach used in the classic work by Schirmann (1995, 119–22, especially n. 105).

2 Although medieval sources had already attributed this adaptation to Dunash ben Labraṭ, the first modern academic to defend it was Brody (1937) in response to questions from Pinsker, Shamhuni, and Harkabi.

3 For a summary of the different theories on Andalusi metrics in general, see Corriente (1986; 1991; 1998, 90–121 and 31–37).

4 Lippmann (1827, x–xi) edition, and Del Valle (1977, 145–58).

5 See folios 45–50 of the Venice edition (1546).

6 The theory of ten vowels, five large or long and five small or short, was first put forward in Sefer ha-zikkaron (ed. Bacher 1888, 17–19).

7 On the metrical syllables, ḥarf or mora see Frolov (2000, 68–93). At the end of the chapter, Moshe Qimḥi himself confesses that he manipulated the Andalusi metrical system when he says: וראוי שתדע שזאת החלוקה למיני השיר איננה מה שעשו חכמי השיר בעצמה כי הם חלקו השיר לסיגים והכניסו תחת you should know that this‘ סוג מהם מינים׃ אבל ראיתי החלוק הזה יותר נקל division into types of verses (= metres) is not exactly what poetry experts used, since they divided the verse into sections and introduced the types (= feet) under each category; however, this division (yated-tenuʿa) seems easier to me’. [This passage is at the bottom of folio 50 just before the colophon in the Venice edition (1546)].

8 Edited by Allony (1966). The triple approach to vowels (Masoretic, grammatical, and metrical) can also be seen at the end of the fifteenth century in Saʿadyah ibn Danān (Cohen 2000, 66–76, for the Arabic version and 155–67 for the Hebrew version).

9 Yellin (1939) and Yellin (1940, 44–53).

10 Edition by Martínez Delgado (2017).

11 Yahalom (2001).

12 For the description of the metrics in this work, I have followed, firstly, classic medieval treatises like the work by Ibn ʿAbd Rabbihi, Kitāb al-ʿIqd al-Farīd (ed. Amīn et al. 1948), the annotated edition of La Khazradjyah (ed. Basset 1902), the Kitāb al-ʿArūḍ by Ibn Jinnī (ed. al-Hayb 19892), and that by al-Rabaʿī (ed. Badrān 2000) in addition to al-Kāfī by al-Tabrīzī (ed. Shamseddīn 20082) and, secondly, modern classical manuals like those by Álvarez Sanz y Tubau (n.d.); ʿAtīq (1987); Sobh (2011); and Hāšimī (n.d.).

13 For the concept of ḥarf as letter, as syllable, and as mora, and for how it is used to compose feet and metres, see Frolov (2000, 68–93). For its definition in Hebrew, see Jastrow (1897, 4).

14 The metaphor consists of understanding the verse as a tent (bayit) held in place with a sabab ‘rope’, which is, in turn, fixed with a watid ‘peg’ assisted by a fāṣila ‘fastener’.

15 To catalogue the syllables and feet, I will use the Hebrew alphabet and reproduce the original Arabic vocalisation with the Hebrew vowels qibbuṣ (ḍamma), pataḥ (fatḥa) and ḥireq (kasra). Bear in mind that in a case like Image shureq can never be used with the waw, since for metrical purposes in this foot, it is a quiescent letter like final nun, meaning that it cannot receive any vowels.

16 Jastrow (1897, 7).

17 Derenbourg (1880, 277–90) and Alahmad Alkhakaf and Martínez Delgado (2018, 39–49 and 99–106).

18 Bacher (1888, 17–18).

19 Martínez Delgado (2017, 35).

20 Martínez Delgado (2017, 51 and 84).

21 Jastrow (1897, 7).

22 Martínez Delgado (2017, 52 and 83).

23 Derenbourg (1880, 277–90) and Alahmad Alkhakaf and Martínez Delgado (2018, 39–49 and 99–106).

24 Theory would later include derived or modern mustaṭīlImagewss ws 2x in each hemistich) and mumtaddImage sw sws 2x in each hemistich) metres in this circle.

25 The theory would later include the derived metre mutawaffir or mustawfirImage swl 3x per hemistich) in this circle.

26 Unlike the Romance model, which understood that the verse is produced by the succession and alternation of yated and tenuʿa; this conception distorts the metrical nature of these compositions.

27 In the medieval Hebrew tradition (Romance or later), there are no known names for these basic components on which part of the rhythm is based, perhaps because they had already lost their original function in a non-arabophone context, as suggested by the words of Moshe Qimḥi. Nevertheless, it is possible that the terms delet and soger originally referred to these two feet and not to the hemistichs.

28 This form is used in the metres ṭawīl, basīṭ, wāfir, kāmil, rajaz, ramal, sarīʿ, munsariḥ, khafīf, and mutaqārib.

29 Sáenz-Badillos and Targarona Borrás (1988, 42*).

30 I took the following example from Kitāb al-ʿarūḍ by Al-Rabaʿī (Badrān 2000, 61). The same example is used by Elʿazar ben Yaʿaqob (Yahalom 2001, 111). This is not the only case where they coincide, which is why it is quite possible that one of the sources of Arabic verse that this Iraqi author had was the treatise by al-Rabaʿī (eleventh century).

31 Not all the metres accept these double modifications; an incompatibility occurs and solutions like muʿāqaba, murāqaba, and mukānafa are used, depending on the types of metres.

32 Some theorists believe that qaṣr and qaṭʿ are the same; however, qaṣr applies to sabab ḵafīf and qaṭʿ to watid majmūʿ.

33 Some theorists argue that this can only occur in the foot ImageImage

34 Traditionally catalogued as ṭawīl. Yaʿaqob ben Elʿazar catalogues it as a variant of hazaj (Yahalom 2001, 88), although metrical theory does not permit such a sequence in this metre.

35 Sáenz-Badillos (1980, 1*; vocalisation mine).

36 Brody and Schirmann (1974, 98).

37 Millás-Cantera (1956, 77–78).

38 I am following the edition by Sáenz-Badillos and Targarona (1994, 476–83).

39 See note 4.

40 In fact, the first poems attributed to Dunash ben Labraṭ were scanned using this Eastern form that resulted from the appearance of internal rhymes (sammaṭa) in the monorhyme lines of qaṣīdas (for the relationship between musammaṭ and muwaššaḥ, see Corriente [1998, 24–25], and for Hebrew poetry, see Martínez Delgado [2016, 39–58]). These compositions are formed by dividing the verse into sections with rhyme that is identical, but different from the end of the last foot, i.e., bbba, ccca, ddda, etc. These divisions of the verse can become murabbaʿ (quaternary) or muḵammas (quinary).

41 This is the derived or modern mustaṭīl form Image but applying tašʿīṯ, producingImage

42 Although known as mišqal ha-tenuʿot in Hebrew, this is the Arabic metre mutadārak modified according to Arabic norms, as Yellin (1939, 192) suspected. This foot does not have to repeat throughout the entire verse; it can alternate with Image , which is common in Arabic since its complete use is rare, even in that language. This metre accepts the qaṭʿ modification in all its feet. Because of this, once the modifications are applied, the metre changes its name.

43 This is a complete form of the Arabic ṭawīl metre modified with ḵarm at the beginning of each hemistichImage

44 In the Del Valle edition (1977, 150) yated + tenuʿa + yated + two tenuʿot, but the scansion of the poem itself confirms the error.

45 This is the kāmil metre with an ʾiḍmar modification in the stuffing (ḥašw) feet and ḥaḏḏ in the ḍarb, without any sign of its ʿarūḍ. In Hebrew, the version of this metre known as kāmil muḍmar is the most commonly used one. See Martínez Delgado (2012, 277–80).

46 This is really a recourse used by the Jewish poets to reconcile it with its original foot; in other words, in the first foot of the hemistich, the second vowel of the fāṣila ṣuġrā is replaced by a sabab ḵafīf, since there is no consensus about this sequence in Hebrew. According to Arabic theory, if the last two feet undergo an ʾiḍmār modification, the primitive form must appear in the poem so that it is not confused with rajaz; thus, the first foot is different and has a form that does not exist in Arabic.

47 This is a kāmil muḍmar where ḍarb and ʿarūḍ are first affected by qaṭʿImageand then by ʾiḍmārImage

48 This is really a rajaz. This confusion is very common in Hebrew metrics. On the šalem and ha-šalem we-ha-soʿer metres see below §5.0.

49 In the Del Valle edition (1977, 153), the final YT is omitted according to the scansion of the verse; as this sequence makes no sense; I am following Lippmann (1827, xi).

50 This is a majzūʾ, or partial, verse with complete ʿarūḍ ( Image ) and identical ḍarb.

51 This is the basīṭ metre with ʿarūḍ in complete form (tāmm) and ḍarb modified by qaṭʿ ( Image > Image = Image )

52 This is the mutadārak metre; its modified variant known as mišqal ha-tenuʿot is much more common in Hebrew, see 2. TTTT.

53 This is a modified version. The first stuffing (ḥašw) foot has been modified with qaṣr and the second with ʾiḍmār. The verse must be majzūʾ, or partial, with ʿarūḍ modified with ḥaḏḏ and ʾiḍmār, according to the classical norm.

54 This is a modified version. A syllable is added to kāmil muḍmar at the end (tarfīl) so that it is not confused with rajaz. TT TTY TT in the edition, the scansion of the verses used as an example is followed.

55 Allony (1966).

56 See Cohen (2000, 66–76) for the Arabic version and ibid. (155–67) for the Hebrew version.

57 A review of all these authors can be found in Del Valle (1988, 349– 459).

58 On the history of the study of ʿarūḍ in Europe, see Frolov (2000, 1– 22).

59 Cano (1987, 31–38).

60 Yellin (1939; 1940, 44–53).

61 Brody (1895).

62 See my list in Martínez Delgado (2017, 123–37).

63 For the Hebrew adaptation of these circles, see Yellin (1940, 47–52).

64 Yellin (1939, 189 n.1 and especially 195).

65 For Ben Tehillim, see the edition by Sáenz-Badillos and Targarona Borrás (1997), and for the rest, Jarden (1966–1992).

66 For example, his secular poetry, ed. by Brody and Schirmann (1974). See also Jarden (1971–1972); and Jarden (1975).

67 This is the case for his secular poetry edited by Brody (1934); Brody (1942); Pagis (1977); and his liturgical poetry, Levin and Rosen (2012).

68 See the classic edition by Brody (1894); and the anthology of Yehudah ha-Levi by Sáenz-Badillos and Targarona Borrás (1994).

Auteur

(PhD, Complutense University of Madrid, 2001) is Associate Professor at the University of Granada and works on the Sciences of the Biblical Hebrew Language in al-Andalus (tenth–twelfth centuries). He has edited and translated Kitāb al-Mustalḥaq by Ibn Janāḥ of Cordoba (Brill, 2020), Un manual judeo-árabe de métrica hebrea andalusí (UCO Press CNERU-CSIC, 2017), and Kitāb al-Taysīr by Shelomo b. Ṣaʿīr (Universidad de Granada, 2010).

Acheter

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search