This article is a survey of the history of Hebrew poetic forms from the Bible to the present time. The entry is arranged according to the following outline:introduction
The Variety of Formal Systems
The Specific Nature of Hebrew Literary History
The Major Periods of Hebrew Prosody
the rhymed piyyut
the spanish tradition
the area of ashkenazi jewry
the period of "revival"
"classical" verse in israel
some principles of biblical verse
the classical piyyut
The Formal Period
The Structure of One Cycle
Forms of Composition
an undivided poem
regular strophic structure
free strophic forms
the origins of rhyme in european poetry
the major norm: the discontinuous rhyme
language and rhyme
the history of kallirian rhyme
other kinds of rhyme
medieval hebrew poetry in spain
Kinds of Verse
The Hebrew Quantitative Meter
the basis of hebrew quantitative meters
the meters used in hebrew poetry
Other Metrical Principles Used in Strophic Poems
Rhyme in Medieval Poetry
the rule of maximum
The Dispersion of the Hebrew Terminal Rhyme
hebrew poetry in italy
The Syllabic System of Versification in Hebrew
the modern period
The Historical Setting
The Two Dialects of Modern Hebrew
Accentual Syllabic Meter in Hebrew
types of meters
definition of meter
Limited Free Verse: The "Ternary Net"
Rhyme in Modern Poetry
the basic norm of the exact rhyme
The Numeric Norm
The Historical Factor
The Morphological Norm
Three Criteria of the Rhyming Norm
the relativity of the morphological norm
minimum and maximum
a comparative perspective
rhyme in the ashkenazi pronunciation
the "inexact" rhyme
A Pan-Historic Synopsis of Hebrew Prosodic Systems
The Major Systems of Hebrew Rhyme
Hebrew poetry throughout the ages has used many forms of verse, rhyme, sound patterns, and strophic structure which changed from period to period, often from country to country, and from genre to genre. Since the close of the Bible, an enormous number of Hebrew poems have been written in Palestine and throughout the Diaspora, most of them following strict forms which were often quite complex and elaborate. To date no history of these forms has been published and while for the major periods some central concepts are known, they are usually framed in normative terms. The following survey should therefore be considered merely as a tentative outline.
The term "form," used here in a limited sense, refers to all poetic patterns which employ sound elements for the organization of the language material of a poem, such as rhyme, acrostic, meter, stanza, and other principles of composition. The term "poem" here refers to any text composed in such forms and does not necessarily imply aesthetic values in a modern sense. In the Middle Ages thousands of Hebrew texts, written as liturgy, chronicles, rhymed letters, dedications, etc., used the same formal norms employed in works which could be classified from a modern or aesthetic point of view as "poetic."
Of the few ancient literatures that have continued uninterruptedly throughout the ages, Hebrew poetry is the most variegated and versatile in its forms, due to its permanent creativity and to its interaction with different systems of language and poetry: Arabic, Italian, German, Russian, Yiddish, English, and others. The pronunciation of Hebrew as well as the norms of writing have undergone considerable changes during the wanderings of the centers of this literature. On the other hand, there were strong tendencies of continuity and conservatism in Hebrew forms and poetic genres, as well as in the language itself.
Unlike other languages, Hebrew, as a semi-"dead" tongue, has never changed the core of its vocabulary, or the written form of its words, its basic morphology, certain patterns of syntax and of idiomatic formulations, or the fundamental framework of its historical, semantic, and mythological allusions. Hebrew poems separated by a time span of a thousand years are from the point of view of their language comparable, and may be intelligible to the same reader. The major changes in the language (insofar as this survey) occurred in the field of pronunciation, but even these did not alter the basic form of the written word. However, due to its interaction with a variety of foreign prosodic and aesthetic norms, most known systems of verse have been created in Hebrew over the past 2,000 years.
The following prosodic systems are found in Hebrew poetry: (1) a purely accentual poetry with a free variation of the verse units (primarily in the Bible); (2) a meter based on a regular number of accents (in post-biblical poetry); (3) a meter based on the number of words (in the major tradition of liturgy); (4) a quantitative meter based on the opposition of short and long syllables (especially in medieval Spain and Italy); (5) a syllabic meter (in Italy since the Renaissance and in Central and Eastern Europe in the 19th century); (6) an accentual-syllabic system (in modern poetry in Eastern Europe, Israel, and the U.S.); (7) an accentual meter with restricted syllabic freedom, influenced by the verse of Russian modernism (in Israeli poetry since World War I); (8) a variety of free verse forms, based largely on a rhythm of phrase groups (which evolved in Europe in the 1920s during the vogue of Expressionism and in Israel since the 1950s under the impact of English imagism).
The earliest known systematic use of rhyme in poetry was invented in Hebrew sometime between the fourth and the sixth centuries c.e. It grew out of a cluster of principles of repetition, based on semantic, morphological, and sound elements. During its long history, Hebrew verse passed through a gamut of rhyme norms: terminal or accentual, continuous or discontinuous, grammatical or sound-autonomous, based on suffixes or on the lexical morpheme, using word repetition or excluding it. The same kind of variety runs through the rhyme patterns, strophic forms, and through the principles of composition of a poem.
A study of the changes in the forms of Hebrew verse should take into account the peculiar nature of its history. A Hebrew poet, regardless of his time, was at the crossroads of three lines of development. (1) There was the historical factor common to all literatures: the tension between synchrony and diachrony, i.e., trends of the poet's generation as juxtaposed to norms of the immediate past as well as classical works. The other two factors are specific to the geographic and sociological situation of the Hebrew writer: (2) the influence of Hebrew poetry written in other countries; (3) the impact of non-Hebrew poetry of his own time and place.
The tension between the three systems, of which the Hebrew poet was aware, was of primary importance to the history of Hebrew poetic forms. Quite often cardinal differences existed between the three. Thus, the 13th-century Hebrew poets in Rome wrote in the strophic forms of Byzantine Ereẓ Israel (canonized long before any Italian language existed) in which words rather than syllables were counted and in which each rhyme had multiple members but did not alternate with other rhymes. But they were confronted with two other poetic systems as well: Contemporary Italian poets used stanzas with alternating rhymes and syllabic meters; the Hebrew poets in Spain used a purely quantitative versification mostly without any strophic forms. It took time until Hebrew poets in Italy changed their poetic system and, typically enough, they adapted quantitative meters (developed in Spain in genres of Hebrew secular poetry) to write Hebrew strophic poetry in the Italian sonnet form. Similar dilemmas faced Jewish poets in other generations and countries.
A Jewish poet was closer to each of the three traditions than poets usually are when experiencing influences of a foreign literature. Thus, the impact of Hebrew poetry written in other countries was enhanced by the closeness of the language and the mobility of men of letters and of written and printed texts. The influence of aesthetic norms dominant in other languages was particularly strong, in spite of traditional Hebrew conservatism, because most Hebrew poets did not speak primarily Hebrew, but were intimately acquainted with other languages which they read and used in everyday life. In many cases they knew at least one more Jewish language and one or two foreign languages, e.g., Arabic and Spanish (in Christian Spain), or Yiddish and Italian (in 16th-century Venice), or Yiddish and Russian (in 19th-century Eastern Europe), or Yiddish, Russian, and English (in America since the late 19th century). Shifts in the forms of Hebrew poetry, whether gradual or drastic, were wrought by such factors as the influence of literary authorities, changes in the relationship to another culture, or changes in the system of genres of Hebrew writing itself. Such changes were usually accompanied by a sudden leap from one way of writing to another, brought about by a realization of a potential influence from one of the three above-mentioned directions.
The influences, however, were implemented neither automatically nor immediately. There was a strong awareness of the peculiar Hebrew tradition, and there usually was neither eclecticism nor chaos of forms. Forms created under the influence of one culture were transposed by Hebrew poetry into the domain of quite a different foreign culture; e.g., (1) quantitative meters, developed in Spain under Arabic influence, were used for centuries in Christian Europe where no such meters had been employed; (2) syllabic versification, developed in Italy, dominated 19th-century Hebrew poetry in Germany and Russia where such meters were no longer used. Even adaptations of poetic elements and themes from other literatures were not automatically introduced in their original forms.
Moreover, there was not necessarily an acceptance of a whole system of forms from the influencing source, but quite often a reconciliation, or a readjustment, of several traditions. Thus, (1) *Immanuel (b. Solomon) of Rome combined the form of the Italian sonnet with the Hebrew-Spanish quantitative meter which was of Arabic origin; (2) Italian strophic forms were used for several centuries with their original rhyme patterns but without the requirement of stress accord, which is compulsory in Italian rhyme.
Foreign influences on Hebrew poetry were not necessarily contemporaneous, e.g., while accentual-syllabic versification was introduced into Hebrew under Russian influence, it occurred only toward the end of the 19th century when this metrical system began to fall into disuse in Russian poetry. These influences should also not be considered as organic transplantations or imitations of a literary trend or poetic school. Belated as such an impact may have been, it was not necessarily accepted in all its aspects. Thus, Judah Leib *Gordon, though influenced by his Russian contemporary Nekrasov as to theme, genre, and even tone of language, did not accept the Russian verse system; the poetry of Abraham *Shlonsky of the 1920s and 1930s was strongly influenced by the imagery of Russian futurism, but in meter it was as classical as the verse of Pushkin. On the other hand, many Hebrew poets were very much aware of the relativity of prosodic systems. They knew how to use diverse, and sometimes even opposing, systems for different genres (such as religious and secular poetry) or for different languages (especially in the case of bilingual poets, such as Elijah Baḥur *Levita and J.L. *Gordon). Despite these complex circumstances and the great body of rhymed and versified Hebrew texts, the varying norms of Hebrew poetry can be described exactly, since in most ages these norms were conventional rather than individual and constituted a firm part of the language of Hebrew verse. The history of these forms epitomizes the worldwide scope of Hebrew poetry; the tensions between tradition and openness which were basic to its evolution; and the symbiotic, but autonomous, nature of Hebrew culture throughout the ages.
The peculiar nature of Jewish history does not permit the development of Hebrew poetic forms to be divided into pure historical "periods," but rather into "areas," determined by a combination of historical, geographic, and generic factors. Since the close of the Bible, the following major areas of Hebrew poetic traditions may be distinguished:
This is a rather amorphous area consisting of several distinct trends: *Wisdom poetry (*Ben Sira), the poetry of religious sects, the formulation of the basic prayers, and the beginnings of liturgy. A variety of rhythmic formulae, occasional rhyme, patterns of sound, and parallelism were widely used, but no established formal system of any kind can be discerned.
Created in Byzantine Ereẓ Israel sometime between the fourth and the sixth centuries c.e., the rhymed piyyut comprehends some clearly defined poetic genres which have specific functions in Jewish liturgy. Fundamental to it are large poetic cycles of a complex structure in which the poems use strophic patterns and obligatory rhyme. This kind of piyyut spread to the East (Babylonia and Egypt) and to Italy and Ashkenaz (the German Rhine area). A vigorous strain in this tradition, which used the difficult "Kallirian" rhyme, flourished in the 10th century and determined the formation of the Italian and Ashkenazi *maḥzor. The rhythm crystallized in this evolution was based on a strict number of words.
It is based on quantitative meters (under Arabic influence), which were used mainly in secular poetry but also in religious genres. Developed in Islamic Spain since the 10th century, it flourished in Christian Spain and Provence until the 15th century and dominated Hebrew poetry in Italy and throughout the Islamic East almost until the present time. Besides the long metrical poems which use one single rhyme, a peculiar strophic tradition evolved in Spain ("girdle" poems), as well as a major genre of rhymed prose (maqāmat).
Created from the 9th to the 20th century, Hebrew poetry in Italy passed through all possible stages of Hebrew poetic forms: several periods of forms stemming from the Palestinian piyyut; Italian strophic patterns; Spanish quantitative meters, which in time were transformed into Italian-like syllabic verse; and even onsets of accentual-syllabic iambs.
Hebrew poetry was written throughout the Middle Ages by Ashkenazi Jews, at first in Germany and France, then in the Slavic countries. The Palestinian-Italian tradition formed its early stages (10th to 12th centuries). A "weaker" line descending from the Palestinian piyyut, followed and continued until modern times, especially in several shorter genres (notably the *seliḥah and the *kinah). The forms of this tradition influenced other genres too, such as the Hebrew verse chronicle and some Yiddish poems.
Babylonia, North Africa, and other countries under Islamic rule passed easily from the old piyyut forms to the Spanish tradition (similar to Arabic forms). In the 16th century, however, the influence of Turkish song forms may be discerned in the writings of Hebrew poets in the Ottoman Empire (including Ereẓ Israel itself).
Toward the end of the 18th century in Central and Eastern Europe purely syllabic versification was introduced. It continued to be the medium of Hebrew verse until the end of the 19th century.
Since the early 1890s Hebrew poetry in Russia, using the Ashkenazi pronunciation, accepted accentual-syllabic meters and became receptive to all forms of modern European poetry. The system spread immediately from its Russian center to all countries of Hebrew creativity: Germany, Ereẓ Israel, the U.S., etc.
Accentual-syllabic meters, transferred to the Israeli (basically "Sephardi") pronunciation, appeared at the beginning of the 20th century, but started to dominate Hebrew poetry only since the late 1920s. The system spread from its center in Israel to other countries where Hebrew literature was being written (Poland, U.S.S.R., U.S.).
There are two varieties: (1) the Russian influenced strophic and rhymed free verse which is close to regular meters (the so-called Russian Dolnik); (2) free irregular verse beginning with some poems by H.N. Bialik (written in period 8), developed in Europe in the 1920s and in Israel and the U.S. especially since the 1950s.
These forms followed both foreign examples (English, German) and Hebrew antecedents (notably some of the so-called biblical verse of the Period of "Revival").
Approaching the present day, the periods become shorter; different forms, traditions, and influences become more intermingled, frequently coexisting in time, in place, often in one literary journal, and even in the writings of one poet.
The forms of biblical poetry constitute a world of their own, at the same time however, a discussion of post-biblical verse must consider the Bible which had an overpowering influence on Hebrew poetry of all periods. The language of the Bible has dominated the language of Hebrew poetry, more often than it did prose, in a variety of poetic conceptions, at least since *Saadiah Gaon (10th century) and almost to the present. Despite this fact, however, post-biblical Hebrew poetry has not relied on biblical rhythm and verse forms. With few exceptions, post-biblical Hebrew prosody at every stage of its development was based on highly formal conventions, and it could not have been satisfied with the fluid, though rich orchestration of biblical verse. Nevertheless the patterns of biblical poetry, its syntactic-rhythmical tendencies, its typical word groups, its alliterations, loomed large behind the language of the Hebrew poets in subsequent generations. These patterns did not prevail or mold the new forms, but embellished and imbued Hebrew poems with the power of internal rhythm. The strength of the biblical example was not merely in its sanctified status, but in the very "weakness" of "impurity" which its rhythm had from any normative or classicistic point of view: the intimate, almost inseparable relationship between the semantic, syntactic, and accentual aspects of its rhymic patterns of language.
Though including writings which range nearly over a millennium, the Bible has been viewed by later ages as primarily a unified work with basically a common language. Whatever may have been the developments in phonetics and prosody during the time of its creation, the Bible for post-biblical readers was the canonized text with its system of stresses, intonation marks, and vocalization. In this survey of post-biblical poetry the major principles of biblical verse, as seen from the point of view of a reader of later times, shall merely be mentioned and illustrated.
The foremost principle dominating biblical poetry is parallelism. Usually two versets (sometimes three or even four) are parallel to each other in one or several aspects. The parallelism may be either complete or partial; either of the verset as a whole or of each word in it; of words in the same order or reversed. It may be a parallelism of semantic, syntactic, prosodic, morphological, or sound elements, or of a combination of such elements. In most cases there is an overlapping of several such heterogeneous parallelisms with a mutual reinforcement so that no single element – meaning, syntax, or stress – may be considered as completely dominant or as purely concomitant. The parts of the parallelism may be equal or unequal in their size or form; they may be related to each other in a variety of ways: synonymous, antithetic, hierarchic, belonging to a category of some kind, etc. The principles of the parallelism used may change from verse to verse. The basis of this type of rhythm may be described as semantic-syntactic-accentual. It is basically a free rhythm, i.e., a rhythm based on a cluster of changing principles. Its freedom, however, is clearly confined within the limits of its poetics. The following is an example of a rather ordered type:
.וְתִשְׁמַע הָאָרֶץ אִמְרֵי־פִי
,תִּזַּל כַּטַּל אִמְרָתִי
הַאֲזִינוּ הַשָּׁמַיִם וַאֲדַבֵּרָה
יַעֲרֹף כַּמָּטָר לִקְחִי
Give ear, o, ye heavens, and I will speak;
And hear, O earth, the words of my mouth.
My doctrine shall drop as the rain,
My speech shall distil as the dew,
As the small rain upon the tender herb,
And as the showers upon the grass
(Deut. 32: 1–4).
There are 3:3 stresses in the first two pairs of versets, and 2:2 stresses in the last pair. But syntactically the last two pairs are linked. The words הַאֲזִינוּ ("give ear"), and וְתִשְׁמַע ("hear") are synonymous in meaning though not in morphology; "I will speak" and "the words of my mouth" are not synonyms, but their meanings are parallel. "Heavens" and "earth" are parallel by opposition. "Rain" and "dew" both express fruition by water, but one is strong and the other is subtle, these are two poles of one scale. There is also a concatenation of the three parts: versets 3 and 4 unfold the theme of the first pair ("the words of my mouth"); versets 5 and 6 develop the images of 3 and 4. But the versets of the last pair are parallel only to one member of the previous pair ("the rain" or "the dew").
The parallelism of meaning in the last four versets is chiastic: the water is strong (3) – weak (4) – weak (5) – strong (6). In the last pair דֶשֶׁא and עֵשֶׂב are on one level, but שְׂעִירִים and רְבִיבִים, though morphologically alike, are quite different in degree. Some additional devices of rhythm and sound reinforce the effect of this passage.
If the equivalent meaning or syntactic pattern of parallel versets draws the reader's attention to the parallelism and its reinforcing quality, it is the rhythmical structure proper which embodies it. The major rhythmic element is stress. The rhythm is accentual, but the number of stresses in each verset is not necessarily fixed or permanent. There may be an exact repetition: 3:3 stresses, or a freer relationship: 3:4, as well as changing numbers throughout the poem. The specific numerical relationship is however important. The numbers are quite often equal or similar. Moreover, whenever there is freedom it is confined within fixed boundaries. Each verset is usually a phrase, a basic syntactic and logical unit, consisting of 2, 3, or 4 stressed words. The smallness and compactness of the verset lends each stress conspicuous force. The condensed, laconic nature of biblical Hebrew also contributes to the prominence of each word within the line, the more so when it is reinforced by the parallel verset. The versets are static, independent units, well balanced against each other. This is supported by the nature of biblical syntax which favors parataxis to the subordination of clauses and phrases.
Is stress the only sound element determining biblical rhythm? For many generations scholars have argued over the "secrets" of biblical prosody; there have been attempts to correct or rewrite the text so that it might conform with pseudo-classic ideas of rhythm which require strict numbers of some kind: regularized "feet," equalized hemistichs, or stanzas of recurring numbers of lines. Such attempts seem pointless today since no exact regularity of any kind has been found and since rhythm need not be based on strict numerical regularity. Considering the rhythm to be based on free variation, it is clear, however, that stress is not enough to describe the effects of biblical rhythm. The number of unstressed syllables between two stressed ones, though not fixed in the sense of modern accentual-syllabic versification, is certainly limited: By rule no two stresses are permitted to follow each other, on the other hand long words have secondary stresses. Thus, each stress dominates a group of 2, 3, or 4 syllables; there are 2, 3, or 4 such groups in a verset, and 2, 3, or 4 parallel versets in a sentence. It is a three-stage hierarchy of simple, indivisible, though flexible groups. Within this free framework there are clearly functional specific patterns, such as the so-called rhythm of elegy based on an opposition of 3:2 stresses. The rhythm of major stresses is so strong that sometimes it may be the only supporter of the parallelism of two versets, without any actual repetition of meaning or syntax.
Within this framework of rhythmical parallelism there is a whole gamut of sound repetition and sound patterns, freely distributed, but clearly embellishing the text. Whatever the origins of Hebrew rhyme and puns or sound patterns in later poetry, the later poets were able to draw on a variety of such devices in the Bible. There is (1) simple alliteration: (2) ;הוֹד וְהָדָר, חֵן וָחֶסֶד a chain of one repeated sound: צַדִּיק מִצָּרָה נֶחֱלָץ; (Prov. 11:8); (3) a repetition of the same root which is syntactically justified: אָחוּדָה־נָּא לָכֶם חִידָה ("I will riddle you a riddle," Jud. 14:12), חוּדָה חִידָתְךָ וְנִשְׁמָעֶנָּה ("riddle your riddle and we will hear," ibid. 14:13); (4) puns on similar sounding roots: אַל תַּחֲרוֹשׁ עַל רֵעֲךָ רָעָה (Prov. 3:29); (5) פַּחַד – פַּחַת פָּח ;אִישׁ – אֵשׁ – אֵשֶׁת; root rhyming בָּבֶל – בָּלַל (cf. Gen. 11:9), צְדָקָה־צְעָקָה (cf. Isa. 5:7); (6) occasional rhymes in modern sense יֵינָהּ – שֻׁלְחָנָהּ (cf. Prov. 9:4), צֶמַח־קֶמַח (cf. Hos. 8:9), etc.
Rhyme is sometimes obviously linked to the parallel structure, e.g.,
/פֶּן־תִּתֵּן לַאֲחֵרִים הוֹדֶךָ
/פֶּן־יִשְׂבְּעוּ זָרִים כֹּחֶךָ
Lest thou give thy vigor unto others,
And thy years unto the cruel;
Lest strangers be filled with thy strength,
And thy labors be in the house of an alien.
The two sentences are similar in rhythm (3:2 stresses) and are linked by an anaphora, as well as by parallel syntax, meaning, morphology, and rhyme. Though the symmetry is pervasive and multiple, it is however neither regular nor permanent: the first versets of each line are parallel in meaning as a whole but not in each word; אחרים ("others") and זרים ("strangers") are parallel in morphology and rhyme but not in their syntactical function; וּשְׁנוֹתֶךָ ("thy years") and וַעֲצָבֶיךָ ("thy labors") are not parallel in the same sense as אחרים ("others") and זרים("strangers"); אכזרי ("the cruel") and בית־נכרי ("the house of an alien") are not synonymous in the language but become so when enforced by this context. In the same way all parallel words rhyme with each other, except for the second word.
This is an extreme example of order; usually the patterns are less symmetrical and the sentence that follows may not have any of the above devices. Rhyme, as it is known at present, i.e., as a regular organizing principle of a poem which is not an internal ornament of a line but links lines together, was created as concomitant to an unequivocal strophic structure and a formalization of poetic patterns. This occurred centuries later in the Palestinian piyyut of *Yannai and Eleazar *Kallir.
Piyyut (from paytan (poet) from the Greek Ποιητής) is the common term applied to a variety of genres of Hebrew liturgical poetry which originated in Ereẓ Israel under Byzantine rule. Some scholars distinguish between piyyut and seliḥah (a penitential prayer), including under the former all kinds of hymns and under the latter several types of elegies, supplicating or exhortative religious poems. For the purposes of this survey it is convenient to include the entire range of Hebrew religious poetry of the Middle Ages under the general term piyyut. The chronological division of the earlier periods of the piyyut from a formal point of view is as follows: (1) the so-called beginnings of the piyyut, primarily in Ereẓ Israel and in Babylonia from the close of the Bible until the creation of the formal rhymed piyyut; (2) the formal period, employing formalized, strophic, and rhymed poems patterned in highly complex piyyut cycles, apparently originating in their complete form in Byzantine Ereẓ Israel somewhere between the fourth and sixth centuries c.e. Only the latter will be discussed here.
Various forms of rhythm, sound patterns, sporadic rhyme, acrostic, and strophic patterns have been developed in biblical literature and during the first centuries of the piyyut. But only by an act of formalization were the new complex structures created. Even if it were possible to trace every single device of the formal piyyut to earlier examples, there is no precedent to any of the complex structures as a whole. Rhyme, refrain, stanza, etc., whatever had been sporadic, was now formalized and organized in complex cycles of poems, governed by strict rules which set the formal conventions of all poems belonging to a given genre.
In the same way as the period is characterized by the introduction of unequivocal rules of formal structure, differences in genres are marked by differences in form. Moreover, some diversities between poets, or successive generations, or local traditions are marked by minor or major changes both in the complex structures, as well as in the use of particular devices or genres of the piyyut.
The large variety of genres, formal differences, and historical changes in these structures does not permit a complete, even schematic, description here. Since the complex structures are determined by the genre, it is preferable to describe the formal structure together with the thematic aspect of each genre, its liturgical function and the particular way of its inclusion in the basic text of the prayer book. The difficulty of such a detailed description is underlined by the present limited knowledge of the history of the piyyut. While there are many scattered studies and insights, there is no detailed up-to-date historical description of the whole field. The objective circumstances were a contributing factor to this state of scholarship: tens of thousands of poems and fragments, found in the Cairo Genizah, are in the process of being deciphered. These poems, written over many centuries, are by and large undated, often fragmentized or written in a cryptic language, and are either anonymous or only have the first name of the poet who in most cases is unknown from other historical sources. On the other hand a considerable number of piyyutim were known for centuries because they were included in the prayer book. More and more of them have been published in recent years. Below only the principles of some basic patterns employed in the complex structures will be outlined and only schematic examples of major formal principles will be given.
The widespread forms of the older formal piyyut, especially the *kerovah with its varieties and the *yoẓer, are cycles of a complex nature, e.g., a kerovah by Yannai is a superstructure of nine parts with a permanent set of rules for each. Yannai wrote hundreds of kerovot – a different cycle for each week of the triennial cycle of the Torah reading. The structure of each of these poems is governed to the smallest detail by one set of rules. Other poets wrote cycles of poems for the Sabbath, the festivals, and often several different sets of poems for the same purpose, apparently written for the services of different years or different synagogues.
Yannai's kerovot are mostly of the kedushta type. The kedushta is a poetic cycle incorporated in the prayer in which the Kedushah is recited. The kedushta has a fixed theme for every week based on the weekly biblical portion. The theme and its language are integrated into the poems of the cycle. It consists of the following parts:
(1) A poem to the first benediction of the Amidah com-posed of 3 stanzas of 4 versets each. Every stanza has a separate fourfold end rhyme, linking all its versets. Every verset begins with a separate letter following the order of the Hebrew alphabet; the poem is thus linked by an unfinished acrostic from the letters א to ל. The concluding verset alludes to the first sentence of the weekly portion which follows and introduces a series of biblical sentences in their original form, having neither rhyme nor meter. The biblical passage gives, as it were, authoritative support to the content of the poem. This chain of biblical sentences is linked to a closing stanza of 3 or 4 versets, with the last word of the chain of sentences repeated at the beginning of the closing stanza. The last verset of this stanza is again linked to what follows, alluding to the Magen Avot benediction (the second) recited after this poem.
(2) A poem to the second benediction which is similar in its strophic structure to the first poem. It continues the interrupted acrostic (from the letter מ to ת) and uses the last two letters twice in order to fit the 22-letter Hebrew alphabet into a framework of a series of four-verset stanzas (altogether 24 versets). This poem too is linked to the weekly portion with the final verset alluding to its second sentence. A chain of explanatory sentences also lead toward a closing stanza in which the final verset anticipates the following benediction, Meḥayyeh ha-Metim.
(3) A short poem of 4 stanzas, each starts with a letter which is part of an acrostic of the poet's name יניי. Every stanza consists of 4 short cola of 2 or 3 words each, rhymed either with a fourfold rhyme or with a twofold rhyme. The poem ends with an allusion to the first word of the haftarah which follows, together with an explanatory passage.
(4) A poem of a rather free structure, having no fixed rules for its rhyming though usually consisting of 3 fourfold stanzas. Concluding the first part of the piyyut, the poem is marked by the obligatory use of the final word – Kadosh.
(5) A poem traditionally called asiriyyah because it is composed of 10 stanzas which are linked by an acrostic of the first 10 letters of the Hebrew alphabet. The stanzas are rhymed couplets having quite often a large variety of internal rhymes.
(6) A poem consisting of 11 stanzas, each using a separate fourfold rhyme. Every couple of versets is linked to a complete alphabetical acrostic. The poem introduces a group of poems and is preceded by the biblical statement which it discusses. Frequently, the biblical statement or parts of it are interwoven into the poem either as beginnings of the first lines of the poem or of its stanzas.
(7) This part consists of 1 to 3 pattern poems (rehitim), each of which has an individual structure, usually of a complex form, which permeates the text in every detail. Variegated and individual in their composition, they follow a fixed set of rules (described below). Only in this category is the poet allowed to use strictly organized poems without rhyme.
(8) The silluk, a kind of free verse poem that introduces the kedushah. It has a free structure which varies from kerovah to kerovah and is richly rhymed in an unrestricted manner. In the poetry of Yannai's follower (or disciple), Eleazar Kallir, the silluk developed into a very long, exuberant, richly orchestrated, yet unrestricted poem. Yannai's silluk, however, is rather short.
(9) The Kedushah; it has neither rhyme, nor strophic structure. In the period of Yannai there was no fixed version as yet, and the poet was free to formulate his Kedushah in every cycle anew. It was based on an exegesis and elaboration of the formula Kadosh, Kadosh, Kadosh (Holy, Holy, Holy).
The above is a simplified account of a poetic cycle, as described by M. *Zulay. It is impossible to delineate here in detail the forms of other cycles, their liturgical functions, and their development throughout the ages. Each cycle is determined by a combination of certain thematic, verbal, and formal elements. Though the intricate rules for each cycle changed from genre to genre, there was no free combination in each new creation: The basic forms used in these compositions were quite restricted. Some of these basic forms are discussed below:
Within a given cycle the form and length of each poem was restricted, depending on its place in the cycle, its use of acrostic, and its strophic form. The following strophic forms existed in the piyyut:
One single end-rhyme runs throughout the poem (ḥaruz mavri'aḥ). At the beginnings of the lines there is a compulsory acrostic which covers the whole Hebrew alphabet, each letter is repeated one or several times; the number of versets being either 22, 44, or 88.
Each poem is composed of a number of stanzas of a permanent form and length. Every stanza has its own independent rhyme both differing from and not interfering with members of other rhymes: aaaa bbbb cccc, etc. The length of stanzas may vary from 2 to 10 versets. A stanza of 4 versets became the major form of the seliḥah, especially in the variety created throughout the ages in Ashkenazi Europe. The multiple (fourfold or eightfold) repetition of each rhyme and the lack of rhyme alternation in the rhyming piyyut create an effect quite distinct from standard European strophic poetry which uses its rhymes in alternation.
Stanzas may have an additional internal structure and may be molded by means additional to rhyme, primarily meter, acrostic, and the refrain. Thus, in Kallir's kerovah to the *Musaf prayer of the Day of Atonement (in the Ashkenazi prayer book), one of the poem Essa De'i le-Meraḥok is based on a stanza of 9 versets. The following is one stanza of the poem:
חֲדָשִׁים וְגַם יְשָׁנִים
בְּפִלּוּל אֲשֶׁר מְשַׁנְּנִים
וְעַל מִבְטָחֵימוֹ שְׁעוּנִים
/שֶׁל כָּל־יְמוֹת הָשָּׁנִים
/וְיוּשְׁבוּ לְתַעֲרָם שְׁנוּנִים
/לְאִוֶּלֶת מִהְיוֹת שׁוֹנִים
/יֻשְלְגוּ אָדְמֵי שָׁנִים
/יֻלְבְּנוּ כִּתְמֵי שׁוֹשַׁנִּים
/רַחֲצוּ וְהִזַּכּוּ מֵעִשּׁוּנִים
Each verset has 3 words, 3 versets form a line, each line begins with a letter from the acrostic (in this case it is the end of the poet's name קליר, the י repeated twice), 3 lines form a stanza by means of the particular rhyme which is repeated 9 times.
This particular poem belongs to the kiklar (from the Greek κύςκλο, cycle) genre, in which there is a refrain-like shorter stanza of 3 versets after each regular stanza. There are 3 different refrains in the above poem, alternating between the 7 regular stanzas. In the following general scheme of the poem each verset of the regular stanzas shall be represented by letters from a to g, according to the rhyme patterns; the refrains by letters p to r; capital letters represent versets linked by an acrostic:
The poem is organized, as it were, both vertically and horizontally. The triadic principle is dominant throughout: 3 words make a verset, 3 versets a line, 3 lines a stanza, 3 stanzas complete a refrain cycle. The third refrain cycle is, however, not completed since there are 7 stanzas.
An unusual kind of formal poem – the pattern poem – was developed in the piyyut, especially by Yannai. A pattern of a line elaborated in all its details – syntactic, semantic, morphological, and sound devices – was established in the poem and was then repeated throughout its 22 lines (the number being determined by the acrostic). A great variety of such patterns appear in Yannai's poetry, all in the seventh part of the kerovah cycle. The following is an example of a very simple kind:
If you loved who would hate? אִם אָהַבְתָּה מִי יַשְׂנִיא
If you blessed who would curse? אִם בֵּירַכְתָּה מִי יַאֲרֹר
If you fenced who would break out? אִם גִּדַּרְתָּה מִי יִפְרוֹץ
If you joined who would separate? אִם דִּבַּקְתָּה מִי יַפְרִיד
Every line is a rhetorical pattern with two fixed and two free words. The initial letter of the first free word depends on the place of the word in the alphabetically arranged poem; this word is a verb in the second person past, and its meaning has a positive connotation. The second free word is opposite in meaning, it has a strongly negative connotation, and is a verbin the third person future.
There are more complex patterns, such as this:
אָהוּב (ל…) חס וְשָׂנאוּי לְמַאַס
בַּחוּר לְסֶגֶל וּבָזוּי לְסֶגֶר
גְּדִי לְרִצּוּי וְנָמֵר לְנִיצּוּי
תּוֹלְדוֹת אָהוּב וְשָׂנאוּי
תּוֹלְדוֹת בָּחוּר וּבָזוּי
תּוֹלְדוֹת גְּדִי וְנָמֵר
The story of a loved one and a hated one/
loved for respect and hated for neglect
The story of a chosen one and despised one/
chosen for virtue and despised for rejection
The story of a lamb and a tiger/
a lamb for pleasing and a tiger for strife
Every line consists of two parts, of 3 and 4 words, respectively. The first part refers to a story of two personae: one positive, the other negative. The second part elaborates on the first, repeating the two personae and modifying the description of each. The first word is permanent, creating an anaphoric chain ("the story of…"); the second word is positive in meaning and is strung on an acrostic; the third word is either a direct or indirect opposite of the second. The other hemistich repeats words (2) and (3) and qualifies them, explaining the reason for the opposition: In what perspective are the personae to be cast. The modifiers do not provide a full explanation, but allude to a biblical text. Both modifiers – words (5) and (7) – are introduced by a preposition of purpose (-ל) and are linked to each other by some kind of rhyme, though the rhyming principle changes throughout the poem. It may be:
a terminal sound rhyme רִצוּי– נִיצוּי
or an initial rhyme עזוז – לעזאזל, סגל – סגר
or a semantic rhyme זכות – חובה, חיים מות
Since the morphology of most words of the pattern is fixed, in all cases where suffixes are used the rhyme is inevitable, e.g.,
וירא אהובים ויאמל ואבה לאררם
וירא באים ויבעת וביקש לבלעם
וירא גרים ויגר וגמר לגרשם
The meaning of the pattern is "He saw ("positive personae") and he was frightened (or shocked, or worried, etc.) and he wished (or planned, or hurried, etc.) to curse them (or swallow them, or uproot them, etc.)." All four changing words are linked to the acrostic, the second and the fifth, using plural suffixes, create each a chain of rhymes. In the rhyming chain of the second words there are, however, two exceptions: המון, רבבה – words which designate plurality but do not have the grammatical plural form. Indeed, rhyme is concomitant to grammatical parallelism but is not an absolute necessity. It is possibly the only structured poem in a Yannai cycle which may have no rhyme, if rhyme does not appear with the grammatical form. In most cases there is rhyme, but of a peculiar variety; a morphological rhyme based on a suffix.
The form of the pattern poem is derived from biblical parallelism, but two new principles were applied: (a) the symmetry of two versets was turned into a chain of synonymous sentences; (b) there was a rigorous formalization of the pattern, and all deviations are excluded.
These are of two kinds: (a) the unrhymed piyyut, an exceptional form, fulfilling strict liturgical functions and employing phrases of a formulaic nature; (b) the rhymed free poem, especially the silluk, developed by Kallir into a long chain of rhymed versets, with changing rhymes and shifting rhyme principles and without any strophic structure or measure of length. Each rhyme usually has many members (e.g., 25 lakh+20 mu+18 šev+13 mem, etc. – in Kallir's silluk to Parashat Zakhor). Besides sound rhyme there may be in the rhyming position semantic rhymes (names of rivers or of time periods), word repetition, words of one root, etc.
Rhyme, the great innovation of the piyyut, had impact on the history of world poetry. Since not many piyyutim were known before the recent studies of the Genizah (where over 50,000 liturgical poems were discovered), and also since the external circumstances of the piyyut were obscure and its language almost puzzling, it was not until recently that scholars have become aware of this original contribution of Hebrew poetry. All its aspects, however, have not yet been fully explored.
It is clear by now that rhyme grew out of the internal development of Hebrew poetry and became in Hebrew a permanent, even obligatory, feature of poetry earlier than in any other language. It is assumed that the principle of rhyme was then transferred to the poetry of the Syriac Church, written in Aramaic (a language closely related to Hebrew, spoken inter alia by Jews and written in the same area; i.e., in the Middle East) and through this mediation introduced into Latin poetry and then into all other languages of Europe.
Not one, but several kinds of rhyme existed in the piyyut, each associated with different strophic forms. The most important was the rhyme of the strophic poems. The basic norm of this rhyme is unknown in the poetry of other languages. Each rhyme of a strophic piyyut had to meet two requirements: (1) parallelism of all the sounds of the last syllable, beginning with the consonant preceding the last vowel; (2) parallelism of two consonants belonging to the root of each rhyming word, e.g., in Eleazar Kallir's stanza quoted above (from the piyyut Essa De'i le-Meraḥok), the rhyming words are: ŠaNIM – ha-ŠaNIM – ye-ŠaNIM – Šo-ŠaNIM – ŠenuNIM – meŠaneNIM – meiŠuNIM – So-ŠaNIM – ŠenuNIM – meŠaneNIM – meiŠuNIM – Šo-NIM – ŠeuNIM. For the sake of identification capital letters represent the rhyme (all the sounds repeated in all members of one rhyme).
The rhymeme in this system is both terminal and discontinuous. The principle of terminality implies that the rhymeme covers the final syllable of each rhyming member whereas in European or in modem Hebrew poetry its basis is the stressed syllable rather than the final one. In most cases in Hebrew, though, the two overlap, but in instances of discrepancy, stress in the rhyme of the piyyut is disregarded. The principle of discontinuity of the rhymeme is unique in rhyming systems and is based on the nature of the Hebrew lexical morpheme, which is discontinuous, consisting merely of consonants. Thus, the changes of vowels in such Hebrew words as Ša-Var – Še-Ve-R – Šo-Ve-R cause morphological differences only (Š+V+R is a root meaning "break," the vowels in the example creating: past, noun, present), whereas in English the differences between, e.g., "lever – liver – lover" are lexical.
Though rhyme may have had, as one of its sources the puns on words of one root, rhyme became an autonomous pattern, independent of grammar or word repetition. The discontinuous rhyme is merely similar in structure to the Hebrew root, but is not necessarily based on words of one root. In the above case the rhyme is S+NIM. Between the discontinuous sounds of the rhymeme there appeared changing vowels and even consonants, though usually consonants of the kinds found in the rhymeme (as in our case).
Thus, rhyme was based on sound parallelism of the roots of words as well as of their endings. Since a Hebrew root can have no more than three sounding consonants, only one (at most) is given to variation. In stanzas with many rhyme members, it is extremely difficult to find enough words which may meet such requirements, especially when the rhyming words are at short distances from each other. Such rhyming was possible in the piyyut due to the difficult "Kallirian" style which allowed, on the one hand, for an almost unlimited number of neologisms and, on the other hand, was abundant in allusions and ellipses which permitted the bringing together of words from quite distant semantic areas.
There are in this system five major forms of rhyme, dependent upon the morphological structure of the rhyming words:
(1) if the final syllable is open (e.g., LA), an additional preceding consonant was necessary, e.g., in Kallir's rhyme: GoLA – Geula – beGiLA – niGLA – GiLA – veeGLA – aGuLA – meGiLA – veGoLA (the rhymeme is G + La); (2) if the final syllable is closed (e.g., NIM) and one consonant belongs to a suffix, a root consonant has to be preceded as in the above case with plural endings: ŠaNIM – haŠaNIM – ŠenuNIM, etc. The rhymeme is Š+NIM; (3 and 4) if the final syllable has no root consonants, two discontinuous root consonants are added, e.g., צוּרֵינוּ – נַעֲצְרֶנּוּ – יוֹצְרֵינוּ – מִצָּרָתֵינוּ – ẒuReNU – naaẒReNU – yoẒReNU – ẒaRateNU (the rhymeme is Ẓ+R+NU). The same holds for a suffix in a closed syllable (e.g., M+R+HEM); (5) if the final syllable is closed and includes no suffix, then it meets in itself both requirements. There is no discontinuity, but the difficulty in finding or inventing rhyme words remains, e.g., the famous stanza which served as a symbol of Kallir's unintelligible (or even cacophonous) style:
/אץ קוצץ בן־קוצץ / קצוצי לקצץ / בדבור מפוצץ
/רצוצי לרצץ / לץ בבוא ללוצץ
.פלץ ונתלוצץ / כעץ מחצצים לחצץ / כנץ על צפור לנצץ
Due to the neologisms, allusions, and the elliptic syntax, this passage is almost unintelligible without a commentary. On the other hand, the richness of rhyme and sound effects is obvious.
Using the symbols N – the norm; R – a root consonant; C – a morphological consonant; V – a vowel; + – a possible discontinuity in the sound string, the above five forms of the Kallirian rhyme may be summarized as follows:
A typical case of discontinuous rhyme can be found in the Hoshanot read on the first and second days of Sukkot where the poet rhymes 22 times עִי + שׁ (in the poem אֶעֱרוֹך שַׁוְעִי) or 22 times עוֹת + שׁ (in the poem (מוֹשָׁעוֹת – שְׁבוּעוֹת – בְּשַׁוְעוֹת שׁוּעוֹת – שַׁעְשֻׁעוֹת אֵל לְמוֹשָׁעוֹת, etc.
In Yannai's poetry, repetitions of four equivalent or similar words are often found in rhyme (such as לְפָנִים – לִיפְנִים פָּנִים – פָּנִים). A repeated word obviously meets both requirements of the rhyme norm. But Yannai's pupil Kallir excluded word repetition as a substitute for a strophic rhyme, thus enforcing his difficult norm. Word repetitions remained a device of rhyming, but in a distinct kind of rhyme chain.
The rhyme norm described above was primarily based on sound. Sound was not identical with letter or with the later canonized vocalization. For the sake of rhyme the qameš (ָ) and the pattah (ַ) were equivalent (a); also the ṣere (ֵ) and segol (ֶ) (e). The letters א ,ה ,ח ,ע lost their consonantal qualities and in the rhyme of the piyyut they are interchangeable and may be either disregarded or counted as consonants. Kallir rhymes אֶלוֹהַּ – לִשְׁלוֹחַ, טְמֵאָה – קִמְעָה, etc., ב and ו seem to be equivalent מַחֲשָׁבָה – שִׁוָּה, אֵבָה – גַאֲוָה, etc. On the other hand, however, consonants with or without a dageš rhyme freely with each other, thus פּ פ, e.g., נֶפֶשׁ – טִפֵּשׁ and בּ ב, e.g., מַרְבֵּץ – רוֹבֵץ – מַשְׁבֵּץ – קוֹבֵץ. The equivalence of בּ ב, according to the graphic principle, and that of ב ו, according to the sound principle, established a new equivalence for the sake of rhyme: ב וִ, e.g., Kallir rhymes דְּבָרִים שְׁוָרִים – גִּבּוֹרִים or רְבִיד – מַעֲבִיד – לְהַאֲבִיד – הִרְבִּיד – דָּוִד.
This tradition of equivalents for the sake of rhyme was carried with the piyyut into Italy and Franco-Germany. Thus, *Meshullam b. Kalonymus of the 10th century (born in Italy, lived in Mainz) rhymes freely הִרְבָּה – רָוָה – מְרִיבָה ;בבִּוְ, etc.; Rabbi *Meir b. Baruch of Rothenberg in the 13th century rhymes: טְבוּעִים – מִצְבִּיעִים – מְשַׁוְעִים , etc.
The forms of "Kallirian" piyyut spread throughout the Diaspora to the East and to the West. In the East they were superseded in the 10th and 11th centuries by the forms of Arabic versification, especially as adapted by the Hebrew poetry of Spain. In Italy and Franco-Germany they dominated the basic form of the maḥzor and do so to the present day. With time, the difficult rhyme norm was simplified: poets dropped the requirement to include two root consonants; rhyme was based on a repetition of the final syllables and became terminal, i.e., the standard Hebrew rhyme of the Middle Ages. The process of simplification apparently originated in Ereẓ Israel. (Thus, in the ninth century the Palestinian-influenced piyyut of southern Italy was based on final syllables only.) But the "strong" norm prevailed again in the 10th century in Babylonia, Italy, and Franco-Germany to be dropped finally toward the end of the 11th century.
In Yannai's poetry the discontinuous rhyme of the strophic poem is not the only rhyme form. All aspects of the Hebrew word were employed in one form or another for the sake of rhyme: the root, the suffix, the meaning, the sound. In pattern poems it is obvious: not only the final sounds of the parallel words are repeated, but also their meaning and morphological structure. In this genre, however, rhymes, as other kinds of repetition, are tied: they serve the composition of the poem not independently but as a whole cluster. For the later development the untied free rhyme is of primary interest.
The following kinds of rhyme may be discerned: (1) Sound Rhyme, a rhyme based on parallelism of sounds especially in the discontinuous terminal form described above. (2) Morphological Rhyme. This rhyme is based on a suffix. It appears sporadically in the Bible and was used several times at considerable length in the Dead Sea Scrolls and in Ben Sira. It became a legitimate variety of rhyme in non-strophic piyyutim, especially in the pattern poems. (3) Semantic Rhyme. The relations of the rhyming members are in parallelism of meaning rather than in sound: זמר – רנן אכילה – שתיה, etc. Even in a strophic poem, in a chain of sound rhymes, Yannai writes suddenly: סוס – סוס – חמור – חמור rhyming horse with donkey! (4) Root Rhyme. Found in the rhyming of words of one root, they do not necessarily have similar sound endings, e.g., תּוּצְדַּק – יִצְדַּק – צַדִּיק – צֶדֶק or גֵּאָה – גָּאָה – גַּאַוָה – גֵּאִים. Semantic rhyme and root rhyme are used only occasionally, especially in free strophic forms, such as the silluk, or as an additional device within the line. They are of particular interest for the understanding of the origins of sound rhyme which grew in an environment of repetition of any possible aspect of the language. (5) Word Rhyme. This rhyme is based on the repetition of one word, usually a key word (life, death, night, war, etc.), throughout a poem. It is older than systematic rhyme and is often employed in piyyutim – either in poems of 22 lines or in free strophic patterns.
A distinct kind of piyyut uses a word rhyme together with sound rhymes in one single rhyme chain of a long poem, such as Kallir's rhyme of the word טל ("dew") (in his "prayer for dew" "תפלת טל"), repeated endlessly and interwoven with words ending with same sounds: טלטל, נטל, etc.
Biblical rhythm was accentual but with free variation of the numbers of stresses in parallel versets. It seems that later developments led in two directions:
(1) The rhythm as found in the poetry of Ben Sira where there are usually 2 versets to a line, with 4 metrical stresses being the optimal limit of a verset. A 4-stress pattern is achieved if long words are seen as having two metrical stresses (in a way quite similar to our reading of iambs in modern Hebrew poetry), e.g.,
אח́די תאו́תיך ́אל ת́לך ו́מחמדו́תיך ́המ́נע
́אם תע́שה ד́צון נפ́שך ́תשיג́ך שמ́חת שונ́א
This was apparently a tradition of a poetry which sensed an inherent semi-regular meter, approaching syllabic regularity. (The syllabic principle is said to be underlying Syriac poetry of the early Christian centuries.)
(2) The tradition of early liturgical poetry which was based on the number of major stresses. Here, too, 4 stresses were a common optimal frame, but these were major stresses, each dominating a word or a group of words. Thus, *Yose b. Yose, in his famous Avodah le-Yom Kippur has regular stanzas of 4 lines each (determined by a fourfold repetition of each letter in the alphabetical acrostic). Each line consists of two versets (or hemistichs), each verset having 4 major stresses:
,א́תה כו́ננת עו́לם ברוב-́חסד / ו́בו יתנ́הג עד ́קץ הי́מים
אשר ́לא ימ́וט מע́ון יצו́רים / ולא ימ́עד מ́כבד פשע́ים וחט́אים
In Ben Sira there are often two stresses on one long word; in Yose b. Yose two smaller words are linked by one stress. The number of stresses is similar, but the interpretation of the rhythm in the language is quite different.
Yannai usually has no regular rhythm, except for pattern poems in which the number of words is fixed by the pattern. Neither is there any syllabic regularity, similar to the one which is supposed to govern Syriac meters, discernible. But in Kallir's poetry there are already piyyutim (beyond the pattern poems) which have a fixed meter, based on the number of words, such as the meter of the kiklar analyzed above. In Italy, this meter became obligatory from the ninth century. While this meter may have grown out of earlier stress regularity, it was now strictly based on the graphic division of words, requiring a permanent number of words in each verset. It became the dominant form of the Ashkenazi piyyut in the Middle Ages. Some genres had norms peculiar to them, e.g., the seliḥah was usually written in Italy and in Franco-Germany in 5-word lines and 4-line rhymed stanzas.
Hebrew poetry entered a new era with its emergence in Islamic Spain, in the 10th century. The Arabic rules of versification were adopted by the Hebrew verse; quantitative meter became the dominant system in Hebrew poetry in Spain from its beginnings, through the "Golden Age," until the destruction of Jewish life in Spain at the end of the 15th century. Due to the authority and the achievements of Hebrew culture and poetry in Spain, its poetic language and metrical system spread to other countries. It dominated Hebrew poetry throughout the Islamic world – Egypt, Babylonia, Yemen, North Africa, the Ottoman Empire – until recent times. It ruled Hebrew poetry in Provence, spread throughout Europe, and reigned in Italy until the 19th century.
Hebrew literature in the East in the first centuries of Arabic rule, though flourishing in the very heart of Islamic culture and strongly influenced by Arabic science and literature, shows no trace of having come under the sway of the forms of Arabic poetry. *Saadiah Gaon of Babylonia (10th century), a distinguished philosopher and linguist in Arabic, followed the norms of the pre-Islamic piyyut in his Hebrew poetry. He used the strophic structure of the piyyut in fourfold or manifold rhymes which change from stanza to stanza and there was no trace of any syllable-counting meter. However, Saadiah Gaon's pupil, *Dunash b. Labrat, a native of Fez, who was educated in Baghdad and went to Cordoba, Spain, introduced there in the middle of the 10th century the Arabic quantitative metrical system into Hebrew poetry. Ben Labrat's innovation, which became the subject of a fierce polemic, was seen as violating the nature and grammar of the Hebrew language. But even Ben Labrat's opponents used quantitative meters in their caustic polemical poems against this very same system. Arguments against quantitative metrics, raised time and again, especially emphasized the biblical tradition and the accentual nature of the Hebrew language. The opponents themselves, however, notably *Judah Halevi, seldom strayed from this metrical system in their secular poetry.
Hebrew literature in Spain was written in a variety of genres, secular as well as liturgical. Generic properties included theme, forms of composition, attitude of the speaker, use of language. But hardly any thematic genre had its own peculiar meter or rhyme scheme. On the other hand, there was a strong distinction between several kinds of literature, based on principles of meter, rhyme, and strophic structure. The following major types may be discerned: (1) Non-strophic poems using one single rhyme throughout, linking all the lines of the poem. In this type quantitative meter in one of the classical regular forms was obligatory (used in most of Spanish Hebrew poetry, especially secular poetry). (2) Strophic poetry of the type of the "girdle" poem, employing a quantitative metrical pattern which may be irregular in itself but permanent throughout the poem. (3) Poems with a plain syllabic meter, primarily in strophic forms. (4) Strophic poetry in "free" verse, i.e., without syllable counting, used primarily in liturgical genres. (5) Rhymed prose, used primarily in genres of Oriental storytelling.
A quantitative meter is based on a regular pattern of short and long (rather than stressed and unstressed) syllables. Hebrew quantitative meters though derived from Arabic versification were quite different from their prototypes as well as from Greek quantitative patterns. This was basically due to the different properties of the Hebrew language.
Traditional descriptions of Hebrew verse in Spain did not distinguish between problems of diachrony and synchrony. The derivation of a particular meter from this or that Arabic prototype seemed to be more relevant than the assessment of its place in the synchronic system of Hebrew verse. The existing classification of medieval Hebrew meters, basically unchanged for the past 800 years, relies on medieval Arabic cataloguing. One finds usually long taxonomic lists of patterns rather than structural rules to explain the nature of the Hebrew quantitative meters.
Twelve of the 16 basic meters codified in the theories of classical Arabic versification were adopted by Hebrew poets (the other being impossible to imitate in Hebrew). With their many derivations, the number of particular regular meters runs into several dozens (Yellin's list has 67), whereas the irregular patterns of the "girdle" poems may account for several hundred forms. No explanation is usually given as to why no other meters existed.
The traditionally identified meters will not be enumerated here but rather an attempt will be made to explain the basic rules and tendencies. One reason for the large number of metrical types is that each is used as a label for a pattern of a whole line which has not been analyzed into its distinctive features. Three such features should be considered: (a) the basic metrical units, or recurring groups of syllables ("feet"); (b) the number and order of such groups (i.e., the length of the line); (c) the form of the final group – whether complete, short, or changed (cf., in accentual-syllabic poetry an analytical term such as "iambic pentameter" – one word signifies the basic foot, the other the length of the line, whereas the nature of the end of the line is described in terms of rhyme gender: "feminine" or "masculine"). Thus, the difference between the two traditionally distinct meters ha-merubbeh (/ ∪ – – – / ∪ – – – / ∪ – – /) and ha-marnin (∪ – – – / ∪ – – –)(∪ stands for a short syllable, – for a long, the direction of the symbols here is from left to right) is one of length of line only; whereas for a parallel difference between (– ∪ – – / – ∪ – – / – ∪ – –) and (– ∪ – – / – ∪ – –) only one term (ha-kalu'a a' and b') has been used. While a difference in the length of a line may be an important rhythmic factor, it should not justify the use of unanalyzed terms.
The Hebrew poets in Spain did not resort to the distinction between short and long vowels of the biblical vocalization. Only the mobile šewa, the hataf, and the conjunction וּ (when pronounced u) were considered short vowels. All full vowels were considered long, e.g., in Ibn *Gabirol's poem (from right to left; short syllables are unmarked):
̆מְלִֿיצֿתֿי ̆בְדאֿגֿתֿי ̆הֲדוֿפֿה, / ̆וְשֿמֿחתֿי ̆בְאֿנחֿתֿי ̆דְחוֿפֿה
In comparison with Arabic or Greek quantitative meters, the number of short syllables in Hebrew is conspicuously small. Moreover, whereas in Greek verse a long (i.e., a strong) syllable constitutes the distinguishing element of a foot, in Hebrew it is a short (i.e., a weaker) element. It is hard to conceive how such a weaker element could provide a rhythmic basis for a foot. Indeed, there was a different way of describing this kind of meter, through use of another kind of contrast, namely that of cord (C = –) (tenu'ah, i.e., vowel) and peg (P = / ∪ – /) (yated). A short followed by a long was called a peg. (In traditional Hebrew grammar a peg is considered one syllable.) All other syllables are cords. The above quoted line from Ibn *Gabirol can be rewritten (from left to right): PCC PCC PC/PCC PCC PC, the basic foot consisting of one peg and two cords (rather than one short and 3 longs). Besides the rhythmic factor considered above, this system of description is justified because in Hebrew there is practically no short "syllable" which is not followed by a long one, i.e., a short is not an independent unit (even though modern pronunciation may create such an illusion). Except for experimental poems, there are no meters of pegs only. Thus, a quantitative meter may be described as based on a regulated opposition of pegs and cords.
The basic group of syllables recurring several times in a line is called a foot. A foot may consist of either 2 or 3 syllables. The basic feet are:
|Number of Syllables||Place of the Peg|
No foot has more than one peg. (In some meters there may be a substitute of 2 pegs: PP, but not as a regular recurring unit.) In Hebrew, as opposed to Arabic, there was a meter without pegs, the so-called mishkal ha-tenu'ot (meter of cords) of 8 long syllables, but it retained the quantitative opposition, since the text avoids all pegs in the language (i.e., all the mobile šewa'im and the like). No ternary foot without pegs may constitute a regular metrical scheme (though CCC may occur as a substitute within a meter with pegs).
The verse form of classical poetry is a distich (bayit) consisting of 2 lines identical or differing slightly at the end: the first line is called délet, the second soger. At the end of the soger there is a rhyme member linking it to the whole poem, which has one rhyme with a long chain of members, connecting the ends of the distichs as a string. A poem often consists of several dozens of distichs repeating again and again the same rhyme, with the typical effect of emphasis and monotony. In the first distich usually both lines are rhymed.
Any meter in this system is based on the principle of rhythmic impulses of recurring groups of pegs and cords, combined with a tendency to repeat each group at least twice, but an exact repetition of the same group more than twice in each line is avoided in most meters. The variations are codified within the metrical pattern of a line. This pattern, however, is in all its details permanent throughout the poem.
In each line the basic foot is repeated several times, but the last foot may be incomplete. The feet of a line may be of one kind only (as in the accentual-syllabic system of modern poetry), or of two kinds, unknown in modern poetry. If the first two feet are identical, the meter is regular; if they are different, the meter is variegated.
The length of a line is not as varied as in Hebrew or European poetry of modern times. There is a strong interrelation between the nature of the feet and their number. The rules governing the length of a line are the following: (a) if one of the first two feet is binary, there are 4 feet in a line; (b) if both first feet are ternary, there are 2 or 3 feet in a line, and the third may be either complete, shortened, or changed.
The last foot of a line can be described separately; if it is the third foot, it is either complete, shortened, changed, or avoided. But for this element, the length of a line can be seen as automatic. Indeed, there are 5 basic regular meters. These can be illustrated by using the list of all meters and variants which appear in J. Schirmann's famous anthology of Hebrew poetry in Spain and Provence (see following table).
|CC binary||CC·CC·CC·CC||משקל התנועות||mishkal ha-tenu'ot|
|CPC ternary||CPC·CPC·CP||הקלוע א׳||haa-kalu'a (a)|
|medial||CPC·CPC||הקלוע ב׳||ha-kalu'a (b)|
|CCP ternary||CCP·CCP·CCP||השלם א׳||ha-shalem (a)|
|final||CCP·CCP·CC·P/C||השלם ב׳||ha-shalem (b)|
|CCP·CCP·CCC||השלם ג׳||ha-shalem (c)|
|CCP·CCP·CPC||השלם ז׳||ha-shalem (g)|
|CCP·CCP·CPCC||השלם ח׳||ha-shalem (h)|
|CCP·CCP·CP||המהיר א׳||ha-mahir (a)|
|CCP·CCP·CC||המהיר ב׳||ha-mahir (b)|
|CCP·CCP -/C||השלם ה׳||ha-shalem|
(The symbol P/C shows two differing endings of the two lines in a distich.) In spite of the different labels, all the variations are in the third foot and can be described separately. Binary meters have practically no variations (due to rule a) and there is practically no binary final meter. Only the rule of length (a) can explain why the meter of cords (binary neutral) uses 8 syllables. There is no regular ternary meter of cords. Moreover, the ternary medial meter is rare in Hebrew. No other regular meters are possible. In the ternary meters the third foot may be complete only if the peg is at the very end. But this weak position of the last peg calls for a great variety of substitutes.
From the rules of length follows also the structure of the variegated meters of which there are two kinds:
(1) Alternating meters: if one of the first two feet is binary and one is ternary there are 4 feet in a line (according to rule a); the whole pattern is repeated twice (e.g., PC PCC PC PCC in the meter ha-arokh).
(2) Changing meters: if both first feet are ternary (and different) there cannot be 4 feet (according to rule b), i.e., there can be no repetition of the whole group within each line. In this case, if there is a third foot it either repeats the first or is changed (as the last foot of a line).
In alternating meters the following rules hold for the basic patterns: (1) there are 4 feet in a line; (2) the meter is based on a regular alternation between ternary and binary feet; (3) both kinds of feet are either initial or final; (4) in each hemistich, if there are 2 pegs they are removed from each other by only one cord. From rules (3) and (4) follows that there may be only 2 basic meters:
alternating initial PC PCC PC PCC הארוך ha-arokh
alternating final CCP CP CCP CP המתפשט ha-mitpashet
There can be no medial feet since there must be a common rhythmical denominator; if it is not the foot, it is its direction. Since each line has two symmetrical hemistichs, variations of the scheme may be accepted at the end of the line as well as in the second foot, e.g., the alternating falling meter has a variant PC CCC PC CCC (the meter of Dunash b. Labrat). The variants will not be listed here.
If the first two feet are different but ternary, the whole group cannot be repeated twice. Only two ternary feet combine: the medial and the final. Hence the two basic metrical schemes are as follows:
medial and changing CPC CCP CPC הקל א׳ ha-kal (a)
final and changing CCP CPC עוטקה ha-katu'a
Variations occur in the third foot (the end of the line) and in the second (the end of the basic group). These meters are however rare and will not be enumerated here.
All the basic meters practically used in Hebrew poetry may be summed up as in the following table.
The structural symmetry is obvious. There are practically no meters beginning with a binary final foot; no medial foot in the alternating patterns, and no initial foot in the changing meters (rare exceptions may be found).
In actual poetry the situation is even simpler. Indeed, some poets liked experimenting. As Yellin has shown, *Samuel ha-Nagid used 57 different metrical schemes. The bulk of Hebrew poetry in Spain, however, employed only a small number of basic meters, with some variations, of which the most widespread are (in this order): initial and final ternary meters and the meter of cords. More precisely:
|(1)||ternary initial, esp.||PCC PCC PC||הַמְּרֻבֶּה||ha-merubbeh|
|but also||PCC PCC||הַמַּרְנִין||ha-marnin|
|(2)||ternary final, esp.||CCP CCP CC P/C||השלם ב׳||ha-shalem (b)|
|also||CCP CCP CC||המהיר ב׳||ha-mahir (b)|
|(3)||cords (binary neutral)||CC CC CC CC||התנועות||ha-tenu'ot|
These three groups, with a few variations, account for 94% of Moses *Ibn Ezra's meters in his secular poetry. The major meter, ha-merubbeh, found in about half of the Hebrew poems in Spain, later gave way to ha-shalem (b).
Following these three groups, though far behind, are the alternating meters, initial (PC PCC PC PCC – ha-arokh), and final (CCP CP CCP CP – ha-mitpashet). The preference of initial over final meters is due to the structure of the Hebrew word; the majority of vocal šewa'im are at the beginnings of words and a šewa may easily be added before a word with a preposition or conjunction (ב, כ, ל, ו). Medial or changing meters are quite rare.
The bulk of the poetry uses ternary meters, with 3 (in-complete) feet in a line. Since the length of the line is regulated, it varies only within narrow limits. There are only lines of 6 to 10 syllables. If the short syllables are also counted (as they were later, in Hebrew poetry in Italy), the limits are 8–14. Since in contemporary Israeli poetry about half of the šewa'im are considered syllables, those limits are comparable with 7 to 12 syllables today. If the special effect of the ha-marnin, which has the typical rhythm of a short line, is excluded, all other meters compare well with the variations given in modern poetry between 4 iambics and 4 anapests. Thus, the length of a line in Spanish Hebrew poetry as well as its rhythmic-syntactical form are similar to the length of typical lines in modern poetry. The optimal line has 8 to 9 long syllables (or 11, counting the short ones), which is similar to a line of 4 or 5 iambs.
There are many variations of verse endings in the last foot of the basic metrical schemes. Any such variant creates a permanent pattern, repeated in all the lines of a poem. As opposed to modern poetry or to Greek and Arabic quantitative meters, Hebrew poets allowed very rarely for changes from line to line (feet-substitutes) or deviations from a given metrical scheme (i.e., changes occurring only in some lines, e.g., in Hebrew in the changing meter CPC CCP CPC the second foot may be substituted by PP). The variations in Hebrew in the third foot are felt not against the pattern of the poem but against the rhythmic impulse of the first 2 feet of the same line.
Variations of feet in verse endings are of several kinds: (1) the last foot is short (catalectic), PCC→PC, CPC→CP; CCP→CC; this change occurs almost only in ternary meters where it is the usual case (unless it is the last peg which is shortened, as in the final ha-shalem);
(2) hypercatalectic: CCP→C which is very rare; (3) a peg substituted by a cord: CCP→CCC; CP→CC, which occurs quite often since short syllables are scarce in Hebrew; (4) two cords are substituted by a peg, CCP→PP; PPC→PP; a rare variation, occurring in changing meters, especially in the second foot; (5) a peg is advanced, CCP→CPC (or: CP), e.g., in ha-shalem (g) CCP CCP CPC. The most widespread changes are a catalectic foot (1) or a substitute by a cord (3) (cf., instances in the list of regular ternary meters).
The quantitative opposition provides the Hebrew poet with a metrical framework rather than with a pervasive rhythmic movement. The role of short syllables in the Hebrew language is much less than in Arabic. Thus, the two major meters in Arabic, tawil and basīt (equivalent to our alternating final and initial meters) are far from being major meters in Hebrew. Moreover, every possible substitution of longs for shorts is resorted to in Hebrew. Thus, the scheme of the Arabic basīt is (from left to right):
its Hebrew derivation:
Instead of 8 shorts to 6 longs, the proportion became 4:9. A common variation of this meter has even less shorts:
/ C̄C̄ P̆ ¯ / C̄C̄ / C̄C̄ P̆ ¯ / C̄C̄ /
No two consecutive short syllables are possible in Hebrew, therefore some Arabic meters could not be reproduced. There is also the favorite Hebrew innovation: the meter of cords in which all short syllables are avoided.
On the other hand, in many Hebrew poems can be distinguished a strong tendency of regulating stress order and word boundaries. Although no permanent laws hold in this area, the tendencies are clearly felt, e.g., in the poem by Solomon ibn *Gabirol (short syllables are unmarked):
נֿחֿד / ̆בקֿר אֿי ̆גרֿונֿי / דֿבֿק ̆לחֿ / כֿי ̆לשֿונֿי
היה לב בי סחרחר מרב כא בי ואוני
The formal division of the quantitative meter in this poem (katua'), though consistent, seems artificial. The language of the poem follows quite clearly a different pattern:נֿ / ̆בקֿר / ̆גרֿו / דֿ / ̆לחֿי ̆לשֿו
The accents are clearly regulated, and so are the word boundaries. Though it is not an absolute rule in this poem, in 85% of the cases there are word boundaries in marked places (whereas only 36% observe the formal foot boundaries of the meter). This kind of regularity in stress order and word boundaries is partly due to the correlation between the following factors: (1) short syllables cannot be stressed in Hebrew; (2) short syllables are most common at the beginning of a word, therefore, a boundary usually precedes them. It seems, however, that the major force behind this tendency is the subconscious rhythmical sense for stress order felt especially in the works of the great poets, Ibn Gabirol and Judah Halevi.
Relative regularity in stress order may be felt as a rhythmic substitute in meters without the peg/cord alternation. Thus, in the meter of cords there are lines which are clearly "iambic" in the modern sense:כֿתת פֿם לֿ הֿ/ וֿכת רֿק מֿ דֿש
Though the bulk of Hebrew poetry in Spain used regular meters and one rhyme running throughout the poem (with as many as 60 or 80 rhyming members – distichs), several kinds of strophic forms also flourished.
The muwaššaḫ, or "girdle" poem (שִׁיר אֵזוֹר, shir ezor), an original development of Arabic Andalusian poetry, was represented in Hebrew poetry almost from its beginnings (11th century) and was the form of some of the best Hebrew lyrical poems in the 12th century. Though originally used in love poetry, it was employed widely for religious poems. The girdle poem combines in its composition both the strophic principle of changing rhymes and the principle of the "running" rhyme, which runs through all parts of the poem in a refrain-like manner. There are two kinds of stanzas: (1) the changing stanzas with changing rhymes. Every stanza has one or several distinct rhymes, different from the rhymes of other stanzas;
(2) the girdle stanza, a strophic pattern recurring after every changing stanza, with the same rhyme or rhymes repeated in all girdle stanzas throughout the poem. A girdle stanza often appears at the beginning of the poem; it is the "guiding" stanza. In many poems the final girdle, the so-called ḫargˇa, is written not in Hebrew but in popular Arabic or in the old Romance language of Spain. Usually it is a quotation of a love conversation. The ḫargˇa thus determines the meter and rhyme of the girdle, as well as the melody (most girdle poems were apparently created as songs).
The meter of the changing stanzas and of the girdle may be identical, but often is not. Each line may consist of 1, 2, or 3 parts, rhymed or unrhymed. The metrical pattern may either be regular or highly irregular: Within the line there is a free combination of all kinds of feet, which seems often to be a kind of "free verse." But the same irregular pattern is repeated throughout the poem, in the stanzas of each kind separately. The two metrical schemes are often related to each other, in a variety of ways, e.g., one may include a partial repetition of the other. A simple example, in Judah Halevi's song "בִּי הַצְּבִי בִּי אֲדוֹנִי" ("Bi ha-Ẓevi, Bi Adoni")
|the meter of the changing stanza is||CCP CPC|
|and that of the girdle:||CC CPC/CCP CPC|
The stanza has a simple but irregular scheme. The metrical pattern of the girdle repeats the meter of the stanza in its second part, but the first part of the girdle is different (in this case a slight variation).
An example of a complex rhyme scheme can be found in a poem by Joseph ibn Jacob ibn *Ẓaddik (1075–1149), which begins with a guiding girdle,
|מֵאָהֳלִי/||בָּרַח, אֲהָהּ, גּוֹזָל/||– נוּמִי, אֲהָהּ, נִגְזַל|
|?מִי גוֹאֲלִי/||– עָפְרִי אֲהָהּ, אָזַל/||– דִּמְעִי, אֲהָהּ, יִזַּל|
|!טוּב מַעֲנֶה/||לָעוּת חֲלִילֶךָ/||נוֹגֵן, שְׁלַח אֶצְבַּע|
|.כֵּן יַעֲנֶה/||צַחוֹת, כְּקוֹלֶךָ/||אִלֵּם – אֲבָל יַבַּע|
|.בִּשְׂמׂאל מְנֵה/||עַל פִּי נְבָלֶיךָ/||שָׁלשׁ וְגַם אַרְבַּע|
|מִפִּי כְלִי־||שָׂפָה, וְאַל יֶחְדַּל/||שִׁירִים נְצׂר עַל דַּל|
|! לֹא מֵחֳלִי||– יִגְדַּל עִתִּים, וְעֵת יִדַּל||שִׁיר – קוֹל, אֲשֶׁר|
The rhyme pattern is seen in the following table (capital letters represent the girdle rhymes):
In the genres of religious poetry the metrical principles varied: (1) the quantitative principle, using regular or irregular patterns of pegs and cords; (2) patterns using free numbers of cords where no pegs appeared; (3) syllabic meter where the opposition P/C was disregarded (i.e., the short syllables šewa and ḥataf appeared irregularly, but were not counted); (4) a free verse in the vein of older Hebrew liturgy, though usually tending toward a syllabic semi-regularity.
The maqāma is a genre of rhymed prose, usually written as a chain of stories in the Oriental manner, and interwoven with anecdotes, fables, and metrical poems. Many books in this genre were written during the Middle Ages or translated and adapted from Arabic (notably by Judah b. Solomon *Al-Harizi and *Immanuel of Rome). Usually the prose text of the maqāma rhymes throughout, though it has no meter. The number of members of each rhyme is not fixed; the distance between the rhyming members constantly changes and the sound patterns of such rhymes also vary, from a mere minimum to near-homonyms. On the other hand, the poems, which are frequently introduced into this rhymed prose, are clearly marked by their strict adherence to classical meters and rhyming.
A typical case of a different kind of rhymed prose is the religious philosophical poem "Keter Malkhut" by Ibn *Gabirol. Though rhyme and rhythm play an important role in this work, their use is neither permanent nor regular; it may be considered a kind of richly adorned free verse, changing its rhythmical tone from a densely rhymed sound-orchestration to mere prose employing parallelism.
Rhyme in Hebrew poetry in Spain, and throughout the Diaspora in the Middle Ages, was terminal. It disregarded stress or morphology. The rhymeme included all sounds from the consonant preceding the last vowel to the end of the line:
|DO||נוּדוֹ||– לְהַגִּידוֹ||– הוֹדוֹ|
|DOT||חֲרָדוֹת||– חִידוֹת||– חֲמוּדוֹת|
|DOD||וּנְדוֹד||– כִּידוֹד||– מִדּוֹד|
The norm is N=CV (C). The number of sounds (2 or 3) depends on the language: whether the final syllable is open or closed. The principle was obviously derived from the rhyme of the Palestinian piyyut, after it dropped the requirement of including two root consonants in the rhymeme.
In order to make rhyming easier, the poets made wide use of rhymemes with open syllables or with suffixes (as in the first two of the above examples), thus having to change only one root consonant. This tendency was motivated by other principles of medieval poetics. Since there was no requirement for individuality in imagery or theme, the poets could widely use the Hebrew plural suffixes throughout their long poems. The same holds for possessive particles, such as ־ֶיךָ (yours, when addressed to God) אַיךְ (to Zion), etc.
The obligatory requirement that a consonant precede the final vowel, similar to the French consonne d'appui, was peculiar to medieval Hebrew poetry and was not required in other languages. It was, as it were, a compensation for the missing stress principle required in modern Hebrew poetry. A typical example is found in the following table:
|Medieval Hebrew||Medieval Hebrew|
|סוֹרֵג||מְדַלֵּג||סוֹרֵג – מְדַלֵּג||ÉG|
|דֶּרֶג||שֶׁלֶג||שֶׁלֶג – פֶּלֶג||ÉLEG|
In modern Hebrew ÉG is a perfectly sufficient rhymeme, in medieval poetry an additional preceding consonant had to be included in the rhyme. On the other hand, in penultimately stressed (feminine) rhymes in modern poetry the inclusion of one syllable is not enough. The principles changed, but the overall proportion between the vocabulary of the language and the rhyming patterns remained similar.
This relationship between modern and medieval Hebrew rhyme may be compared to the difference between English and French rhyme. Whereas in French rime riche (using consonne d'appui) was highly welcome, in English it was often excluded from rhyme. The situation is similar: French rhymemes are based practically on the last syllable, the words are longer, and an addition to the minimal rhymeme is welcome in order to avoid trite rhyming. Only in Hebrew, however, was the use of the consonne d'appui obligatory; hence it may be called the Hebrew terminal rhyme. Its peculiar impact was felt especially against the background of Italian, German, Yiddish, or Russian rhyme, where such enrichments were discouraged.
If the final syllable was based on a suffix, the poets often strove to enrich the rhymeme, adding to it at least some part of the root. Though this was not a necessary rule (there appeared rhymemes of pure suffixes too), it was a strong tendency.
But rich rhyming was limited by unwritten rules:
(1) if the normally required final syllable (N) had two root consonants, no sound could be added to N; rhymemes such as מִיד, דּוֹד, בָּל are both minimal and maximal;
(2) if the final syllable (N) included one root consonant, a preceding vowel could be added; thus, there are rhymemes, such as: לִי, כִי, לִים, רים; but also לִי, ָ–רִים–ָ, etc.;
(3) if the final syllable (N) included no root consonant, one root consonant could also be added; thus, besides rhymemes such as הֵם, נוּ, –יִךְ there are: יהֶם, –ֵינוּ–ֵ and ,לֵיהֶם, רַיִךְ נֵינוּ.
Though stress was disregarded, in meter as well as in rhyme, a secondary tradition developed a "feminine" rhyme, which is based on penultimately stressed endings (which are a small minority in the language, but are represented in several suffixes and in word endings with ע, ח (עַ, וֹעַ, ֵ–חַ, וּחַ–ֵ).
Feminine rhyme became obligatory on one kind of meter composed of unequal hemistichs in which the final foot had a cord instead of a peg appearing in the first hemistich, e.g., the ha-shalem (b): CCP CCP CCP / CCP CCP CCC. The end presents a change in the regularity of the meter – where a P was expected a C appeared instead. As a compensation for this frustrated expectation, the poet used in this case feminine rhyme.
Feminine rhyme appeared occasionally in other meters too. But with the meter ha-shalem (b) it became prominent in Romance-speaking countries, especially in Italy, where feminine rhymes were the dominant rhyming form.
Elsewhere, i.e., in the majority of Hebrew medieval poems, stress was disregarded, words ultimately and penultimately stressed rhymed freely with each other.
The Hebrew terminal rhyme originated in the Palestinian liturgy as an alleviated form of its "difficult" rhyme. It may be found both in Ereẓ Isael after Kallir and in ninth-century Byzantine in Southern Italy. It developed again, as a simplification of the "Kallirian" rhyme in 10th-century Babylonia (Saadiah Gaon) and in 11th-century Germany. It was strengthened by the comparison with the Arabic rhyming norm (basically requiring a consonant and the vowel following it) and later with European terminal rhyme which knew no discontinuous rhymeme.
This norm persisted in Hebrew throughout the world until the end of the 18th century, except for Italy, where stress was accepted in rhyme since the 17th century (but in Italy, too, no violation of the Hebrew norm could be found). The norm also remained obligatory throughout the Ashkenazi domain (Germany, Poland), though Hebrew had become a penultimately stressed language there. The penultimate stress caused a neutralization of all final vowels. Nevertheless rhyme remained exclusively in the final syllable. Thus, Meir b. Samuel of Sczebrzeszyn in his historical chronicle rhymes in 8-line stanzas words such as: פָּקִיד – נִפְקָד – עוֹקֵד, apparently pronounced: pokәd-nifkәd-oykәd. Though the original i, o, ey (or i, a, e) were blurred in an unstressed position, rhyme remained terminal: a repetition of final sounds. This Hebrew conservatism is even more astounding in bilingual poems, such as the Megillat Vinẓ (1616), with regularly alternating Hebrew and Yiddish stanzas. In the Yiddish stanzas all rhymes are stress-bound (feminine and masculine), according to the standard European norm; even Hebrew words follow this rule. But in the Hebrew stanzas the same Hebrew words disregard stress: terminal rhyme is preserved.
Only in some cases under the influence of foreign poetry did Hebrew rhyme relinquish the requirement of the consonne d'appui (in closed syllables only), rhyming N=VC. Such was the case in some of the girdle poems (patterned on rhymes in a foreign language), e.g., Judah Halevi rhymes: צַח – פַּח – נֶאֱנַח, (AḤ). The same holds for the bilingual Hebrew-Arabic strophic poems of the Yemenite classical poet Shalem *Shabazi and for the strophic songs the 16th-century kabbalist poet of Safed, Israel b. Moses *Najara, who was apparently influenced by Turkish songs.
The Jewish community in Italy was probably the oldest in Europe; though small in number, it was an important center throughout the Middle Ages. Located in a central position, between Israel, Yemen, and Babylonia in the East and Spain in the West, between North Africa in the South and Germany and France in the North, Italy was on the crossroads of the major cultural trends in Jewish history. Hebrew poetry in Italy, the first examples of which are from the ninth century, continued to flourish uninterrupted until the 20th century. The changes of poetic systems in Italy may be representative of the shifts in Hebrew prosody throughout the centuries. The major formal periods in Italy will be briefly listed below:
(1) The poetry of Byzantine Southern Italy in the ninth century consisted of strophic piyyutim, from 2 to 10 lines in a stanza, each stanza having one separate rhyme. The rhymes were simple (terminal norm). Usually an acrostic was required and sometimes a permanent refrain was used to close all stanzas of a poem. Contrary to the "Kallirian" piyyut, the early Italian piyyut required a compulsory meter, based on a constant number of words in a line.
Though strophic poems were known in Latin and in Greek-Byzantine poetry of the period, in these languages rhyme was not yet a required, regular, or permanent device. Only in Hebrew did rhyme serve as a criterion for strophic structure and was obligatory.
(2) In the 10th and beginning of the 11th centuries Italy accepted again the "difficult" "Kallirian" rhyme. It was, as it were, a "reversed evolution." But strict meters were required too. *Solomon b. Judah ha-Bavli and other poets of this period composed in this vein. Their followers who moved to the Rhine area introduced this norm into the Hebrew piyyut of Franco-Germany. These circles edited the mahẓor and apparently included in it only such rhymed strophic poems which were written by Kallir or followed his rhyming norm. A number of yoẓerot and a large number of seliḥot were created in this style.
(3) In the 11th and 12th centuries the norms of the ninth century were again revived: The piyyut used strophic poems with changing but separate rhymes, written in exact meters, based on the number of words. The simplification of rhyme was apparently due to a variety of factors, the foremost being (a) The influence of the Hebrew rhyme of Spain and Provence (though neither the pattern of one running rhyme nor the quantitative meter was accepted). The Spanish scholar and poet Abraham *Ibn Ezra propagated the simpler rhyme in Rome in the 11th century. (b) The decline of the difficult enigmatic style of Ha-Bavli, which occurred in Franco-Germany too. Without this style "Kallirian" rhyming was almost impossible.
In the 13th century Italian poetry in the vernacular emerged and flourished. Hebrew poets living in Rome could not have been unaware of the differences in the respective prosodic systems: (a) Hebrew strophic poems used changing but separate rhymes (aaaa bbbb cccc, etc.), whereas Italian rhymes were usually alternating (abba; aba bcb, etc.); (b) Hebrew meter was based on the number of words, Italian – on the number of syllables; (c) Hebrew rhyme was terminal but required a consonne d'appui: N=CV(C), whereas Italian rhyme was stressed, usually feminine: N=V´CV. Thus, Hebrew rhyme was based on one syllable, Italian on two. Hebrew leaned primarily on consonants, Italian on vowels. In these three respects Hebrew poetry in Italy adopted the Italian norms, but it was done over a period of centuries, primarily through the transformation of forms existent in some Hebrew tradition.
(4) Alternating rhyme was introduced into Roman Hebrew poetry in the 13th century. The major poet who initiated this change was Benjamin b. Abraham *Anav. But Benjamin Anav did not directly imitate Italian forms; he switched to alternating rhymes, meeting thus an Italian aesthetic norm by adopting the patterns of the girdle poem which had been developed in Hebrew poetry in Spain and Provence. However the poets of this generation did not transfer the system of quantitative meters from Spain; only a semi-regular syllabic meter, as in many a strophic piyyut of Judah Halevi or Isaac *Ibn Ghayyat, was employed. Many such poems, both of Italian and Spanish origin, were by that time absorbed into the Italian maḥzor.
(5) *Immanuel of Rome (end of 13th–beginning of the 14th century) was the major Hebrew poet who shifted to the use of both quantitative syllabic meters as well as of Italian strophic forms, primarily the sonnet. In both techniques he had predecessors, but the major achievement was his. With him Hebrew poetry in Italy switched from liturgy to secular poetry. It seems that in order to find an equivalent for Italian poetic forms, Immanuel had to seek a language for secular poetry in Hebrew; this he found in the Spanish tradition, which he accepted with its rhymed prose (maqāma), quantitative meters, Oriental storytelling, and imagery. Suddenly Hebrew poetry discovered exact syllabic meters, required by the Italian aesthetic taste, in its own language and tradition. Though the distinction between long and short syllables, a vestige of Arabic influence, was apparently disregarded in the Italian Hebrew pronunciation, its patterns persisted until the 20th century: Isaac Ḥayyim (Vittorio) *Castiglioni wrote his poem on the death of Theodor Herzl in 1904 in a quantitative meter.
Moreover, Hebrew poets in Italy found in the Spanish tradition meters which fitted the lengths of line favored in Italian poetry. The major meter, especially in the sonnets, was the endecasyllabic line for which a Hebrew poet was able to use either ha-merubbeh or ha-shalem (b), counting both "long" and "short" as whole syllables. Ha-shalem (b): / – – ∪ –/ – – ∪ – / – – – / became the major meter of Hebrew poetry in Italy, due to its compulsory feminine rhyme which fitted both Hebrew-Spanish and Italian taste. Immanuel accepted it for his sonnets, breaking each distich into two lines, with a rhyme for each; lines with an even number of syllables had a masculine rhyme, and those with uneven numbers – a feminine rhyme.
But Immanuel introduced stressed rhyme and alternating rhyming only for his Italian strophic forms. In other parts of his book he completely accepted the Spanish Hebrew tradition, employed widely the running rhyme, rhymed prose, disregarded stress, etc. His was a combination of two systems with a common denominator: the quantitative metrical system.
(6) After Immanuel of Rome, Hebrew poetry adopted a variety of other Italian strophic forms (besides the sonnet which became the most popular), the ottava rima, Dante's terza rima, the sestina, the canzonetta, and some others. Nevertheless, strophic forms of the piyyut on the one hand and the Spanish running rhyme on the other lived on for centuries:
(7) Despite the domination of the quantitative meter, a new syllabic meter evolved. Its first major poet was Moses b. Isaac *Rieti (beginning of the 15th century) called "Il Dante Ebreo" for his book Mikdash Me'at written in the form of Dante's terza rima. Rieti understood that Hebrew had to rhyme primarily in the ultimately stressed form (masculine rhymes) and accordingly reduced his line to 10 syllables.
Whereas previous Italian attempts at syllabic meters (13th century) disregarded short syllables altogether, according to their Spanish prototypes, Rieti counted short syllables as completely equivalent to long ones. Hence he abolished the limitation of short syllables to particular spots in the metrical scheme. But Rieti's innovation, i.e., syllabic meters without quantitative distinctions, did not become prominent in Italian Hebrew verse until the 18th century.
(8) For several centuries Hebrew poets proceeded to retain in their rhyme forms the distinction between Italian strophic patterns and the Spanish or the liturgical tradition. On the whole, rhyme was terminal, stressed rhyme being reserved for the sonnet and other Italian patterns. Thus, *Joseph ha-Zarefati (12th century) writes his octaves in masculine rhymes but still follows the Hebrew rule of a required consonne d'appui (thus, פֶּר and בֶּר are for him two different rhymemes). Only in the 17th century was the change completed; stressed rhyme according to the European norm became compulsory. Despite the nature of the Hebrew language, which favored masculine rhymes, under Italian impact feminine rhyme became dominant. Since the 18th century feminine rhyme was almost exclusive. It was employed primarily in the ha-shalem meter, or in derivations of it: either dropping one or both short syllables, or shortening the line, e,g., in Moses Ḥayyim *Luzzatto's La-Yesharim Tehillah there are two kinds of verse line: (a) the 11-syllabic: – – ∪ – / – – – – / – – – –, and (b) the 7-syllabic: – – ∪ – / – – – – each retaining merely one short syllable. Some of his followers in Italy and in Amsterdam dropped this last vestige of quantitative metrics, thus paving the way for the forms of the new era, the Haskalah.
(9) Another Italian development of great interest should be noted: the earliest invention of accentual iambs in Europe was accomplished in Yiddish rhymed romances by the Venetian poet Elijah Baḥur Levita, about 1508/09. Northern Italy was at that time a center of Yiddish literature. Elijah Levita, a grammarian, a versatile scholar, and a poet was fluent in several languages. He wrote Hebrew verse both in the Sephardi pronunciation, using quantitative meters, and in the Ashkenazi vein, using free accentual verse. When adapting long Italian strophic romances, such as Buovo d'Antona (the Bove Bukh) and Paris un Viene, and creating stanzas in pure ottava rima in a quite modern Yiddish, he merged the Italian syllabic principle with the Germanic accentual principle (which ruled Yiddish poetry until his time) and developed his iambic tetrameter. The process of this invention is of major interest to comparative prosody, but with the decay of the Italian center, it did not last in Yiddish poetry. Accentual-syllabic meters reappeared in Yiddish and in Hebrew under Russian influence only as late as around 1890 (with the one exception discussed below).
(10) In the later centuries of Italian Hebrew poetry iambic pentameters began to appear.
The combination of principles from three metric systems in one verse line – the Hebrew quantitative, the Italian syllabic, and the biblical accentual – did not hamper but encouraged the creation of a fourth system in the same verse: the accentual-syllabic meter, in its iambic form. How did it come about? The major quantitative meter used in Italy was: – – ∪ – / – – ∪ – / – – –. Since short syllables cannot be stressed in Hebrew, the third and seventh syllables were unstressed. Italian poetry opposed a stress on the fifth syllable. On the other hand, the 10th syllable was stressed by the rule of rhyme. Since the biblical rule precluded two adjacent syllables from being stressed (if it happens, the first stress would move backward) and the ninth and 11th were also excluded from stress, the following pattern – with the third, fifth, seventh, ninth, and 11th positions unstressed – emerged: – – ∪ – – –∪ – – – – –. Thus, only even syllables were allowed to receive a stress and a perfect iambic pentameter evolved. (The first foot only was free for variation, but this is the case in English or German iambics too.) Only when the Italian requirement for not stressing the fifth syllable was disregarded did these iambs not materialize.
Hebrew stress was apparently strong and this tendency was felt and spread to other meters too (primarily meters without fixed short syllables). Despite this obvious iambic tendency, it was never formulated as such, being rather an automatic, unintentional result of rules of quite a different nature. The 19th-century Haskalah poets were strongly influenced by late Italian Hebrew poetry, but having a different pronunciation (Ashkenazi as opposed to the Italian "Sephardi"), they could not feel this underlying iambic meter. Though they dropped entirely all distinctions of a quantitative nature, they interpreted this verse as purely syllabic. Only poets in Ereẓ Israel of the 1930–40s, such as J. *Fichmann writing again in a "Sephardi" dialect, rediscovered the iambs of their Italian predecessors.
The modern age of Hebrew literature began with the revival of Hebrew poetry in Germany in the second half of the 18th century. It is regarded as a "secular" period (though many of the poets were religious and some of their themes were of a religious nature) since there was a conscious creation of poetry and prose written in the genres of contemporary European literature which were conspicuously different from the genres of liturgy. Haskalah poetry was a direct descendant of Hebrew poetry in Italy and Holland. However, since this poetry emerged with a new social and cultural trend, the Enlightenment movement, and flourished closer to the center of the Jewish population in Eastern Europe, it expressed a re-orientation of Hebrew literature and may rightly be considered a new period.
Haskalah literature was written and published by small groups of writers and their followers in Germany, Austria, Hungary, and Russia (including Poland) throughout the 19th century. Though their ideas were, to some extent, typical of the European Enlightenment of the 18th century, Haskalah poetry cannot be considered of a monolithic nature, but rather as an eclectic body of verse. This new poetry was indeed, from its beginnings, influenced primarily by 18th-century German literature, especially in the typical genres of epic and fable. It embraced, however, also genres developed previously in Hebrew, in Italy, such as allegorical drama, and absorbed themes and motifs from 19th-century European lyrical and social poetry.
It seems that Hebrew literature lagged considerably behind the evolution of European poetry, going into the stages of the development of neighboring literatures only after those had been established as "classical." Thus, one of the major Hebrew writers of the Haskalah in Russia, Judah Leib *Gordon, who knew Russian well and lived for many years in the capital of Russia, wrote poetry in the vein of the Haskalah in the 1860s and 1870s, i.e., in the time of Tolstoy and Dostoevski and after Pushkin, Lermontov, and Tyutchev. But though Haskalah verse seemed to be a fossilized remnant of the 18th century, untouched by the poetics of Russian classical poetry, Gordon also absorbed some influences from the social and "civic" poetry of his Russian contemporary Nekrasov. On the other hand, he continued to use forms which antedated the Haskalah.
The meter of Hebrew poetry throughout the Haskalah was syllabic. Thus, the poets continued the basic form of Hebrew versification in Italy in spite of the fact that their prototypes in German and Russian were written in accentual-syllabic meters. Even translations from German poetry were transposed in Hebrew into syllabic meters, regardless of the German prototype, e.g., Schiller's Glocke (written in accentual-syllabic meters) was translated into Hebrew in syllabic meters, without stress regularity, and into Yiddish in accentual meters, without syllable counting. Stress, strangely enough, played no role in the Hebrew meters of this period, despite the fact that it was prominent in the speech of these writers and even dominant in the meter of folk song in their spoken language, Yiddish, as well as in the small amount of Yiddish poetry which they wrote, and despite the fact that it ruled the versification of German and Russian poetry which they strove to imitate. There was no traditional Hebrew poetic authority to back up this choice of syllabic versification – the venerated poetry of the Bible was accentual and relatively free in its verse forms. The only explanation could be the sense of continuity and the typical conservatism of Hebrew verse.
A few attempts were made in the second half of the 19th century to introduce accentual-syllabic meters (notably by A.B. *Gottlober). But only S. *Frug, well-known as a Russian poet, transferred the Russian system of versification into Yiddish when he started writing in this language. Ḥ.N. *Bialik in the 1890s, strongly influenced by Frug, was among the first to use predominantly accentual-syllabic versification in his Hebrew poetry.
The new meter, influenced by the Russian prototype, swept Hebrew and Yiddish poetry in the 1890s, paradoxically enough at the same time when the symbolist movement, which tried to break away from meter altogether, emerged in Russia.
Haskalah poetry used the pure syllabic system, which had developed in Hebrew literature in Italy, after the last vestiges of quantitative versification have been dropped. Every poem had its own meter, i.e., a permanent number of syllables in each line, with a stress on the penultimate syllable. Otherwise, stress was not regulated. A permanent caesura was rarely implemented.
Apart from marginal uses of quantitative meters, inherited from medieval Spanish Hebrew poetry, the poetry of the Haskalah did not apply the distinction between short and long syllables. The traditionally short syllables, šewa and ḥataf, confused the poets and were considered sometimes as syllables and sometimes as non-syllables. Naphtali Hirz *Wessely, who introduced this system, used the ḥataf at the beginning of words as a syllable and in the middle of words as a non-syllable; the mobile šewa as a syllable in the middle of a word and as a non-syllable at the beginning. It seems that Wessely tried to avoid the mobile šewa at the beginning of words in order to eschew the problem. With time, however, it became impossible to refrain from using a whole group of words (beginning with the mobile šewa) in Hebrew poetry.
Shirei Tiferet, Wessely's epic, was the classical prototype of all the poetry of the Haskalah. The prologues to each of the 18 parts of his epic were written in 11-syllabic rhymed stanzas, but the epic itself was composed in 13-syllabic un-rhymed verse. Wessely used feminine endings exclusively both in his rhymed and unrhymed poetry. Unrhymed feminine endings in his poem conformed with the Hebrew tradition inherited from Italy (where the penultimate syllable was stressed) and with Wessely's German prototype, Friedrich Gottlieb Klopstock's Der Messias. Wessely, however, in accordance with his Italian prototypes, resorted only to words which were penultimately stressed in the Sephardi pronunciation; these constitute a small and very specific part of the Hebrew vocabulary.
The firm grip of tradition led to the exclusive use of penultimately stressed (mille'eil) endings in the Sephardi pronunciation in much of Haskalah poetry; it was an absolute rule in the higher genres of poetry, especially poetic drama and epic verse, and in the "higher" circles of the Haskalah, such as the centers of Germany and Vilna, though not in Galicia or Hungary. The paradox of this use of mille'eil in the Sephardi pronunciation is underscored by the fact that in other respects Hebrew was obviously pronounced according to the Ashkenazi dialect, even in specific Ashkenazi subdialects. Thus in the poetry of Lithuania (a major center of Haskalah literature in the second half of the 19th century), the rhymes often betray the poet's pronunciation: חֹפֶש – נֶפֶש (both ḥolam and segol being pronounced as ei), רַעַשׁ – כַּעַס (שׁ and ס being equal to s). Though Lithuanian Ashkenazi pronunciation is evident in their poetry and the penultimate stress was the general rule of the Hebrew words in all varieties of the Ashkenazi dialect, the Haskalah poets did not dare use Hebrew in their own pronunciation as a natural resource for the compulsory use of feminine rhymes. Consequently most of the vocabulary of their language was excluded from final verse positions.
When the young poet Mikhal (Micah Joseph *Lebensohn) was negligent in this respect and rhymed words which were perfectly equivalent in his own dialect, but appeared as a mixture of millera and mille'eil (masculine and feminine) to a distant Sephardi ear, he was scolded by the Italian scholar S.D. *Luzzatto. Both Mikhal and J.L. *Gordon, in his later period, broke the rule and on and off used words which are feminine only according to the Ashkenazi pronunciation. But even then the majority of rhymes were still based on words which are considered mille'eil in the Sephardi pronunciation: What used to be a compulsory rule became a habit, or a matter of merit in poetic style.
This phenomenon had a strong effect on the style of Haskalah poetry. Words of the mille'eil form exist only in several specific groups: (1) a small group of nouns penultimately stressed (אָרֶץ – פָּרֶץ – קָרֶץ) which recurred endlessly in the rhymes of the Haskalah and became trite symptoms of this poetry; (2) a variety of archaic forms (אֶחֱלוֹמָה, בָּמוֹ, לָמוֹ, מֶנְהוּ); (3) several forms of the verb (הפריעה, אמרתי), notably in the biblical end-stop pronunciation (יִנְהָרוּ, יִבְעָרוּ); (4) a group of feminine endings (אוֹמֶרֶת, תָּבִינִי). Since other sources were limited, the penultimately stressed forms of the verb became prominent in the rhymes of Haskalah poetry. As a further result, the rules of rhyme caused sentence inversion, since the verb was closing a verse line, and enjambement was excluded; all complements of the verb, similes, etc., preceded the verb rather than followed it (as the usual word-order would require). The following is a typical stanza of this kind:
וּבֵין כֹּה וָכֹה הַדְּמָמָה הִפְרִיעוּ
;תּוֹפְשֵׂי הַמָּשׁוֹט בַּמַּיִם יַחְתּוֹרוּ
,גַּם קוֹל עַל הַמַּיִם עַם רָב הִשְׁמִיעוּ
.גַּם דָּוִד גַּם רֵעוֹ מִשְׁנָת נֵעוֹרוּ
Meantime the silence they interrupted,
The crew who in the water rowed;
A voice on the water a multitude emitted,
And David and his friend from their sleep awoke
Thus, in spite of the accepted Ashkenazi pronunciation in the later poetry of J.L. Gordon, most of the words in his poetry were excluded from rhyme position: a group of words constituting 8% of the normal language continuum was used in 90% of his rhymes. With the abolition of the restriction to the Sephardi stress, no revolutionary change occurred in the rhymes of the Haskalah since the typical rhymes became part of poetic style as such. Only a fundamental change in poetic style and in the very conception of poetic language, introduced by Ḥ.N. Bialik and his generation, was to alter radically the resources of Hebrew rhyme, making available for rhyme practically the whole range of the Hebrew language. During the Haskalah, this freedom was enjoyed only sometimes in minor genres and by poets on the geographical "periphery" (poets from Galicia or Hungary).
Though the tradition of epic poetry in the Haskalah began with Wessely's blank verse, in time rhyme became dominant in this domain too, especially in the Lithuanian center. The poets of the Haskalah used a variety of strophic forms very often of more than 4 lines, notably the stanza developed by Wessely consisting of 6 lines and rhyming aabccb, the ottava rima inherited from Italy, and other strophic patterns, especially of 6 or 8 lines. The 4-line stanza was also widespread, primarily in the form of abab, but it was not as predominant as in the poetry of later generations.
The strophic forms in all their variety usually used alternating rhymes, the members of one rhyme alternating with the members of another rhyme. Rarely did a Haskalah poet systematically use one rhyme more than twice without alternating. In this respect Hebrew poetry conformed to the prevalent European sense of rhyme variation. On the other hand, there was no alternation whatsoever or rhymes insofar as their rhythmical properties. Though Russian poetry alternated, as a rule, not only the rhymes, but also their rhythmic patterns, i.e., combining throughout a poem feminine and masculine, or masculine and dactylic rhymes, Hebrew poetry of this period did not accept this norm. Feminine rhymes were the absolute rule, except for sporadic, non-systematic uses of masculine-rhyming words.
In Italy such a restriction of the language could be understood as influenced by a taste formed through the reading of Italian poetry in which predominantly feminine rhymes are used; this is a concomitant of the structure of the Italian language. In Russia, however, the restriction made no sense and can only be explained through the compulsion exerted by the internal Hebrew tradition. Paradoxically enough, this requirement continued to obtain even at a time when feminine rhymes were drawn merely from the words regarded as feminine according to the Sephardi pronunciation, i.e., when the bulk of the language could have been used for the purpose of masculine rhymes. It was only the generation of Bialik that, again paradoxically enough, attempted to alternate between feminine and masculine rhymes, in spite of the scarcity of the latter in the Ashkenazi dialect, which became in this generation the accepted language of Hebrew verse.
Nevertheless, though stress was disregarded by the syllabic system, it may subconsciously have played a role in forming the rhythmic nature of Hebrew verse in the Haskalah period. The most widespread meters were 13 and 11 syllables. Such a length of lines conformed very well with the structure of 4 major accents, or 4 words, grouped mostly in 2 pairs. This condition may have played a role in the acceptance of Haskalah poetry, for there was, to the ear, as it were, an underlying quasi-biblical meter, felt even more because of the biblical language used in this poetry. A line of 11 or 13 syllables, using 4 major stresses, in a language in which the average number of syllables to each stress is about 3, can easily be brought to approximate an amphibrachic tetrameter, e.g., the first stanza of A.D. *Lebensohn's poem לַבֹקר רנה in the Ashkenazi pronunciation can be read (unstressed syllables are unmarked):קוֹֿל ̆חִּשְׁמַ̆ע נַֿפְשִׁ̆י// ̆הֲמוֹֿן ̆צִבְא̆וֹת חַֿ̆יִל
;רִֿ̆נַּת ̆כּוֹכָֿבִ̆ים// וּ̆בְרֿאשָׁ̆ם ̆יָרֵֿ̆חַ;̆עַל מִ̆שְׁמֶֿרֶ̆ת אֶֿרֶ̆ץ// עָֿמְד̆וּ בַּֿ̆לָּיִלֿ;עַֿתָּ̆ה כִּ̆י בָ̆א שַֿׁחַ̆ר// לִבָּ̆ם ̆שָׂמֵֿ̆חַ,גַּֿם ̆נָ̆גַע זֶֿה// ̆עַל עֵֿינַ̆י ̆וַיֵּ̆עוֹֿר̆וּ̆וּמְלֹֿא ̆כָל ̆הָאָֿרֶ̆ץ// אֶֿראֶ̆ה ̆כִי אוֹֿר̆וּ
This is a typical Haskalah stanza (aba bcc), rhyming Sephardi mille'eil. But an Ashkenazi reading reveals the underlying dactylic-amphibrachic meter, sidestepped only in the first hemistichs of lines 3, 4, and in the second hemistich of line 5.
Toward the end of this period the unregulated stresses within the line became more and more often "ordered," many lines approximating 4 amphibrachs (such were, for example, the poems of S.L. Gordon in the early 1890s).
Thus, for the second time in its history, Hebrew syllabic verse developed again toward an accentual-syllabic meter, dominated however by the amphibrach rather than the iamb, prevalent in Italy, and it followed the Ashkenazi rather than the Sephardi pronunciation.
Such is the story of the transformations of a major Hebrew metrical form: The quantitative meter of Spanish Hebrew poetry, originating in Arabic versification, was reinterpreted in Italy as syllabic, under the influence of Italian versification. Hebrew syllabic poetry stretched over a period of centuries (from the 12th to the 19th), adopting Italian strophic forms without relinquishing the quantitative patterns of the Arabic heritage, but shifting time and again into accentual-syllabic iambs. The Haskalah took up the same syllabic verse forms, continued to use them in spite of a literary environment which accepted exclusively accentual-syllabic meters and imbued them with an underlying semi-biblical (that is accentual) rhythm. It finally brought Hebrew poetry again to the verge of accentual-syllabic meters.
The development of Hebrew poetry should be considered as a chain in transformation rather than a series of totally opposed and separate periods. In regard to sentence structure and syntactical rhythm, there was no fundamental change: Similar groups of words could constitute a verse line, since a line of 11 syllables was the most frequent length in all these periods.
Though echoes of European literature and of European poetics of the modern age had reverberated in Hebrew poetry since the days of Dante in Italy, and throughout the literature of the Haskalah, Hebrew poetry had not accepted fully, until the very end of the 19th century, the consequences of the lyrical revolution accomplished by *Goethe, Pushkin, or the English romantics.
From the limited point of view of this survey, one can observe the striking fact that Hebrew meter was based, until the end of the 19th century, on syllable-counting rather than on the subtle and complex instrument developed in other languages with "free" stress in the form of accentualsyllabic (or tonic-syllabic) versification. English poetry since the 16th century, German poetry since Opitz (17th century), Russian poetry since the middle of the 18th century – the whole modern period of these poetries – cannot be imagined without the metrical system. It is an instrument whose exact structures made possible the clear-cut distinction of a large variety of forms and also provided the background for clearly pronounced effects of particular rhythmical variations. These two assets were of primary importance for a poetry characterized by the individuality of the writer, the individuality of the poem, the reliance on a living language, and the immediate appeal in concrete sensuous images to the imagination of the reader. Even free verse was rich and effective when playing on this background. Such a poetics was accepted and absorbed by the Hebrew poets who, through an externist's secondary education, came to know the classical Russian heritage of Pushkin, Lermontov, and their followers. The first poet to write consistently in accentual-syllabic meters was Ḥ.N. Bialik whose first poem "El ha-Ẓippor" ("To the Bird") was published in 1894, in the very year when Russian symbolism emerged, i.e., a movement which strove to break away from the regularities of this very same metrical system.
In Odessa before World War I where Hebrew poets wrote some of the best of Hebrew poetry in the poetic mode of Russian literature of the 1830s, young Jewish poets, writing in Russian, launched modernistic journals such as The Flying Omnibus. The beginning and end of the major cycle of modern Russian poetry seemed to meet at one time and in one place. Obviously, Hebrew poetry could not for long be excluded from the general developments, the more so because Yiddish poetry was sometimes written by the same poets in a language more alive and actual and therefore absorbed the waves of modernism more rapidly. Thus, in one generation Hebrew poetry not only caught up with the European classical heritage as conceived by the early 19th century, but at the same time landed, in one grand leap, into the European 20th century. The struggle and interaction between a variety of poetic trends – evolution turned into contemporaneity – make it one of the most panoramic and interesting periods of Hebrew poetry. But, as regards this survey, it is a handicap: It is difficult to keep apart the "generations" of poetry in which unequivocal norms persist. Free verse was developed almost contemporaneously with exact meters or Greek verse forms. Modernistic rhymes were intermingled with exact "classical" rhyming and with blank verse. Hence it is advisable to discuss forms and formal systems – "regular" or "modernistic" – rather than periods of poetry.
A second objective difficulty in discussing the rhythms of Hebrew poetry in the 20th century is due to the revolution in the pronunciation of the Hebrew language which undermined its whole prosodic foundation. The rhythm and sound orchestration, so essential to the concept of poetry of the classics of the last generation – Ḥ.N. Bialik, S. *Tchernichowsky, J. *Fichmann, J. *Steinberg, Z. *Shneour – is lost to the ears of Israeli readers. The poetry of the Hebrew revival in Russia at the end of the 19th century, which is concrete and sensuous and employed the Russian sensibility for subtleties of rhythm and sound, unfolded the sound values of the language in the Ashkenazi dialects. But almost at the same time there was in Israel a revival of Hebrew as a spoken language, which used the "Sephardi" or "Israeli" pronunciation. The clash between the two dialects was sometimes fierce. The changing laws of language will be discussed below. Here one example may suffice. The two words עֹז־רְתֵת were a perfect rhyme in the Lithuanian Ashkenazi Hebrew of U.N. *Gnessin E YS-rs-E YS (both ḥolam and ẓere being pronounced ey; ת sounding like s; and z = s in rhyme, according to the Russian convention of neutralizing voiced consonants in an end position); but the same rhyming pair lost all sound identities to the ears of an Israeli who reads it: oz-retét. The Sephardi pronunciation was also employed in Hebrew meters in this period, as early as 1900 (not to count some experiments during the Haskalah period as well as the unintentional iambs of Hebrew poetry in Italy). The Ashkenazi pronunciation was still dominant in the 1920s and was alive with some poets until the 1960s. This coexistence again complicates our discussion.
The shift of dialects was a revolution in the sound system of a language, which did not occur elsewhere in such an abrupt manner. In this process most of the poetry of the period of Revival was lost from the point of view of its musicality and rhythm. But the poets who moved from one Hebrew tongue to quite a different one, despite the pangs of readjustment, remained the same, and so did the poetic ideals and norms. These had no time to change. It was simply a matter of readjusting to the new sound system, of regaining a modus vivendi with the spoken language. Therefore it is possible, despite the crucial shift, to discuss the prosodic norms, using at first illustrations from Hebrew poems in the contemporary Israeli pronunciation.
When Hebrew became a spoken language in Ereẓ Israel, it adopted the principles of the "Sephardi" pronunciation in which the location of stress is based on the accentuation marks of the Bible. The majority of words in this Israeli pronunciation have a stress on their final syllable. Only a small group of words are penultimately stressed; these are of two varieties; the so-called segoliyyim having two e vowels, patterned like dégel, and words with the furtive pattaḥ ending in an originally guttural consonant, like תַּפּוּחַ (tappú'aḥ), רֵיחַ (rei'aḥ) לָנוּעַ (lanu'a). A larger group of penultimately stressed words is provided by several suffixes, such as the feminine forms אוֹמֶרֶת (omeret – versus the masculine Omer), verbs in some perfect forms: אָמַרְנוּ, אָמַרְתָּ, אָמַרְתִּי (amarti, amarta, amarnu), or nouns in the plural with some possessive pronouns (דְּבָרַיִךְ, דְּבָרֶיךָ – devarekha, devarayikh). A new group of penultimately stressed words consists of foreign borrowings: akadémya, gemnázya, etc. On the other hand, several distinctions in the quality of sounds marked in the biblical vocalization system, are blurred in the Sephardi dialect. Thus, both pattaḥ and qameṣ are pronounced a; ṣere and segol are pronounced e; תּ and ת are both t.
The Ashkenazi pronunciation of Hebrew, developed in Europe since the 14th century, is based primarily on the penultimate stress. With a few exceptions, a penultimate stress is absent only when it is impossible to implement it: (a) in monosyllabic words (a small group in Hebrew); (b) in bisyllabic words, if the first syllable is a short one (ḥataf or šewa), e.g., בְּנִי, אֲנִי; (c) in longer words where the penultimate syllable is a short one, the stress moves to the third-to-last syllable, e.g., הַמְּחֿוֹ̆נְ̆נִים (ha-meḥónenim), נַֿ̆עֲ̆רָה (ná'arah). In (b) and (c) there are exceptions, based on the fact that historically short syllables became normal and may be stressed, like in Hebrew words in Yiddish (e.g., חֲלוֹם should be khalóym, but it is often pronounced as in Yiddish: khólem). On the other hand, there is in Ashkenazi a wider range of vowel qualities: qameṣ is distinguished from pattaḥ; ḥolam and ṣere are diphthongs. The weak ת is pronounced s (rather than the Israeli t). Within the Ashkenazi domain there were several dialects, on the whole resembling the dialects of Yiddish. In poetry these sub-dialects are felt not in the meter but in the rhyme.
Ashkenazi Hebrew with its diphthongs and penultimate stress was felt by the poets to be "softer" and more "musical" than the "harsh" Israeli Hebrew which is ultimately stressed and in which a makes up 50 percent of its vowels. Until the late 1930s some tried to keep poetry in the traditional Ashkenazi dialect, but finally they had to give in to the spoken language of Israel. Several poets attempted "translating" their poetry into the new dialect. On the whole they succeeded in making a mechanical meter, but in most cases the poem was severely harmed in the process. A variety of interesting transitional forms developed. Thus, U.Z. *Greenberg, though still writing in Ashkenazi Hebrew, let the Israeli workers in his poems speak in their authentic Israeli pronunciation.
The dominant system of Hebrew prosody since the 1890s was accentual-syllabic, though throughout the period other forms also existed. Accentual-syllabic versification came to Hebrew poetry under the influence and in the forms of the Russian tradition of the 19th century. However, some rhythmical characteristics of these meters are due to the structural properties of the Hebrew language. Accentual-syllabic meters are based on the ordering both of the number of syllables and of the location of stresses in a verse line. But it is rare, especially in Hebrew poetry, that the actual stresses in the language of a line constitute a neatly ordered pattern, copying exactly the metrical scheme. There is a discrepancy between the units of the language and the units of meter: stress and word boundaries on the one hand and metrical accents and feet on the other. A meter exists in a poem if its actual stresses and word boundaries meet certain rules of correlation with the underlying metrical scheme.
A meter is a permanent order of accented and unaccented syllables, underlying all lines of a poem (or part of it). The sign (–) represents a metrically accented syllable, the sign (∪) an unaccented one. The elementary recurrent group of syllables is called a "foot." Thus, in a line of the type: ∪ ∪ – ∪ ∪ ∪ – ∪ – there are three feet / ∪ ∪ – /. A foot is not a rhythmical unit; its boundaries do not mark any stop in reading; it is a mere abstraction of the basic principle underlying the pattern of a line. Each foot has one accented and one or two unaccented syllables.
There are two binary feet: iamb / ∪ – /
trochee / – ∪ /
and three ternary feet: anapest / ∪ ∪ – /
amphibrach / ∪ – ∪ /
dactyl / – ∪ ∪ /
A meter of a line is determined by the kind of feet and their number, e.g., an iambic pentameter is a line of five iambs. The number of feet is determined by the number of accents; the last foot may be incomplete and may vary throughout a poem. Thus, in an iambic pentameter there may be either 10 or 11 syllables: ∪ – ∪ – ∪ – ∪ – ∪ – (∪) (depending on the gender of rhyme or on the line ending). Usually in Hebrew poetry there is only one kind of foot in a poem, i.e., one form of alternating accented and unaccented syllables (binary or ternary).
A poem has a certain meter, when it can be read according to a metrical pattern without contradicting its language. The general rule of correlation is: If a word receives metrical accents, at least one of them must fall on the stressed syllable. This rule implies that a word (1) may be unaccented; or (2) may have an accent on its stressed syllable; or (3) may have several accents, one of them falling on the stressed syllable, e.g., Nŏsh, e̅rĕv ăf v̆e-ŏr(e)vim ăl t(e)r̆an. This is a line of four anapests (by *Alterman) in which the stress (marked) of the first word is disregarded by the meter, the stresses of other words are employed by the meter as regular accents. The following is an example of a Hebrew stanza meter of 4 iambs (by Alterman):
Ăz ḥivvăr gădl h͞e'i
Ět ha-r(e)hŏvt vě-hԾa-sh(e)v̆akm
Ămd năti ăl p(e)ni
Năḥshl shămyl̆m y͞erŭkm
The first word of the first line is unaccented by the meter, the second word has 2 accents (one of them on the stressed syllable). The second line provides only two stresses in its language.
Obviously, an expressive reading of a poem will consider language stress and word boundaries rather than the mechanical pattern of the metrical accents. Thus, rhythmical variation is created primarily by the fact that not all accents of the meter are realized in the language, the division of actual stresses and word boundaries may vary from line to line, e.g., a trochaic stanza (Alterman):
Dumiyyáh la-merḥavím shoreket
Bóhak ha-sakkín be-éin ha-ḥatulím
Láylah, kámmah láylah! Ba-shamáyim shéket
It has 5–6–6–4 trochees, but the number of stresses is 3–4–5–2, irregularly dispersed. When read according to its language, every line seems to be rhythmically different (every box represents a word, X denotes a syllable, Y an accented syllable).
In modern Hebrew there is, on the average, one stress to each three syllables. In binary meters, which constitute the bulk of modern Hebrew metrical verse, rhythmical variation is based primarily on avoiding stresses in accented positions. This tendency usually follows the Russian symmetrical pattern of variation. Thus in a meter of four iambs or trochees, the fourth and the second accents are almost always stressed, the third and the first are quite often unstressed. (This is obviously different from English binary meters where variation is largely based on the opposite possibility: stressing unaccented syllables.) In short, Hebrew iambs and trochees are not based on a regular number of stresses to each line, but on changing deviations from a regular abstract scheme.
Whereas in binary meters variation is built into the system (a 3:2 relationship between accent and stress), in ternary meters almost all accents coincide with stresses. In the Israeli pronunciation where in most words stress coincides also with word boundary, the effect becomes tedious, especially in the anapestic meters where almost every foot is a word and every accent a stress. Poets did their best to create variation here too, but the solution came in the form of a kind of free verse adapted from Russian modernist poetry (Blok, Akhmatova, Yessenin). In this system, the number of accents in a line remains regular, but the number of syllables is free to a certain extent. Usually an impulse of a ternary meter is created, to be disturbed on and off; instead of two unaccented there are occasionally one or none (and in some poets also three). The abstract pattern looks like a "net": a ternary scheme with "holes" in it which appear without any regularity, but rarely enough, so as not to destroy the underlying ternary pattern. Beginnings of lines are usually free too, thus abolishing any distinction between anapest, amphibrach, or dactyl, e.g., two stanzas by the poetess Raḥel:
|Hen damáh be-dami zorém||∪ ∪ – ∪ ∪ – ο ∪ –|
|Hen koláh bi rán||∪ ∪ – ο ∪ –|
|Rahél, ha-ro'áh zón laván||ο ∪ – ∪ ∪ – ∪ ∪ –|
|Rahél – em ha-ém||ο ∪ – ∪ ∪ –|
|Ve-al kén ha-bayit li ẓar||∪ ∪ – ο ∪ – ∪ ∪ –|
|Ve-ha-ir zarah||∪ ∪ – ο ∪ –|
|Ki hayah mitnoféf, sudaráh||∪ ∪ – ∪ ∪ – ∪ ∪ –|
|Le-ruḥot ha-mid-bár||∪ ∪ – ∪ ∪ –|
The conversational tone is achieved here by breaking the anapestic flow. But the same principle may be used for a variety of rhythmic tendencies and poetic themes and tones. It became a major form of Hebrew poetry since the 1920s, developed by *Raḥel, Alterman, Zusman, Lea *Goldberg, *Bat-Miriam, and other poets of the Russian tradition.
Since the beginnings of Hebrew accentual-syllabic meters, varieties of freedom from their strictures were sought. Thus, Tchernichowsky used widely the dactylic hexameter, varying often two or one unaccented syllables. The effect was similar to the ternary net, but the "excuse" was an interpretation of the Greek meter followed by German poets, which varied the dactyls by using trochees (instead of the Greek spondee). Bialik developed his so-called biblical rhythms; but, unlike in the Bible, the number of accents was fixed and the number of unaccented syllables varied in a limited way: 1 or 2 (and occasionally 3) syllables in each interstress interval. About 10 years after the initiation of the accentual-syllabic meters, a Hebrew poet appeared who wrote purely free verse: Avraham *Ben Yiẓḥak. This trend, based on the balancing of small word groups and phrases, was enhanced by the influence of German expressionism (exerted on such poets as David *Vogel). It was renewed in some of the young poets of the Palmaḥ generation (1948) and in the 1950s, under the influence of English modernism. The forms of free verse are too varied to be discussed here. Basically they lean on syntactic patterns, strengthened by parallelism and sound orchestration. At present the whole scale from strict meters to prose-like free verse is productive in Hebrew poetry.
Though rhyme in Hebrew was older than in any of the surrounding languages, and though its forms changed throughout the centuries, it was not before the 1890s that Hebrew rhyme accepted fully the European rhyme system. (As has been seen, the principle of stressed rhyme was adopted already in Hebrew poetry in Italy in the 17th century, but it actually applied to feminine rhymes only and did not involve the whole language until the end of the 19th century.)
In modern Hebrew poetry it is convenient to distinguish "exact" from "inexact" rhymes. In "exact" rhymes the rhymeme always extends to the very end of the rhyming members (x=ḥ or kh; c=ẓ): sma MA-ey-MA, novÉYAX- KerÉYAX; in "inexact" rhymes some of the final sounds are not identical, i.e., the rhymeme does not always reach the end of the verse line:meso-RÉGEt- baRÉGEv; la ḥaDSÍ- kiDuŠIn; ba-xÓFEn- ha-OFEk, etc. The inexact rhyme, a symptom of modernism, should be discussed after the basic "exact" norm from which it deviated.
The rhymeme is the basic norm in modern Hebrew poetry, as it is in most European languages; it includes all sounds from the last stressed vowel to the end of the line: N=V´ ( ). (The parentheses represent all sounds, which may come after the stressed vowel.) As opposed to the terminal Hebrew rhyme of the Middle Ages, this is an accentual-terminal rhyme norm, e.g., šovÁX- heÁX; šenavÓXA- kamÓXA; dÁY LA- LÁYLA; lirKOŠET- xarOSET. The rhyme in these cases includes 2, 3, or 4 sounds (V´C, VCV, V´CCV, VCVC). All of these are minimal rhymemes: they may be enlarged, but deducting one sound destroys the rhyme. Thus, the basic norm is not determined by the number of sounds but by their position. The number of sounds following the stressed vowel depends on the structure of the words; it is a matter of language rather than of rhyming norm. In this system, a rime riche is based not on the number of sounds in the rhymeme, but on the employment of sounds additional to the required norm. Thus, tic-Nax – aNAX (תִּצְנַח – אֲנָךְ) is a rich rhyme, though its rhymeme Nax has only three sounds, since AX would be good enough; whereas ESET in IESET – nogÉSET (לֶסֶת – נוֹגֶסֶת) is not rich, though it has four sounds, since it is the minimal sound group in such feminine rhymes.
In addition to the basic norm, several secondary norms are at work, some are more general, others less obligatory or more restricted to certain poets or trends.
Hebrew poetry in the Israeli pronunciation requires a minimum of two sounds in the rhymeme. In English, German, or Yiddish poetry, one stressed vowel is enough, if it comes at the end of the word: free – tranquility – sea, be – we, go – snow are perfect rhymes in English. But in Israeli Hebrew, as in Russian, in such cases, a consonant has to precede the final vowel: bitfi-LÁ-leo-LÁ is a minimal rhyme (N2 = CV). Two sounds are enough, even when there is no consonant: ligvÓA- elÓA.
Hebrew poetry in the Ashkenazi pronunciation did not require this numeric norm. Bialik rhymed lÍ – bnÍ (לִי – בְּנִי), hazE – hapE (הַזֶּה – הַפֶּה), etc. The reason is obvious: there are in the Ashkenazi pronunciation very few ultimately stressed words (primarily monosyllables, a rather small group in Hebrew) and even fewer such words with open syllables. With the additional rule it would be almost impossible to rhyme these words. On the other hand, in the Israeli pronunciation most of the words are ultimately stressed, the number of vowels has been reduced to 5, and there are an enormous number of words which terminate in (resulting both from the historical pattaḥ and qameṣ). The use of such an á as a rhyme would be too easy and trite to be effective.
There was a historical factor to this development, too. Bialik and many of his contemporary "Ashkenazi" poets at the end of the 19th and the beginning of the 20th century (Z. Shneour, I. *Katzenelson, Jacob Steinberg, J. Fichmann) wrote Yiddish as well as Hebrew poetry. Yiddish, as other Germanic languages, does not require the numeric norm. But the Israeli poets of the next generation (Raḥel, A. *Shlonsky, Lea Goldberg) were overwhelmingly influenced by Russian poetry where this norm is required.
The historical factor is felt again in the "young" Israeli poetry of the 1950s. Hebrew poetry now moved from the Russian to the English sphere of influence and away from rich "colorful" rhyming to a rather "prose-like" poetics. Here again rhymes appeared, based on a single stressed vowel: kÍ-civonÍ; lezokhrÓ be-motÓ, etc.
In the Israeli pronunciation sIR – kabIR, niM – alIM (נִים – אַלִּים) are perfect rhymes, but mexusIM – alIM (מְכוּסִים – עָלִים) is not, since IM in this case is a morphological ending: the non-feminine plural suffix of nouns, verbs, and adjectives. This secondary norm requires the participation of at least one stem consonant in a rhyme: mesuSIM – ma'aSIM or aLIM – keLIM are minimal rhymes in this case.
The basic norm with the secondary norms may now be combined: "exact" rhyme in modern Hebrew poetry in the Israeli pronunciation requires in its rhymeme all the sounds from the last stressed syllable to the end of the verse line, provided that there are at least two sounds and at least one is part of the root of the rhyming word. This complex rule makes three kinds of demands: (a) a norm for the place of the rhymeme; (b) a norm for its minimal size; (c) a norm concerning its morphological structure.
The three heterogeneous norms have different degrees of validity in different poets or generations. The morphological subnorm seems to be the most flexible. Some suffixes are less susceptible to this norm. Thus, Raḥel, a poetess of the 1920s who preceded the "young" generation in the use of "prosaic" language in her lyrical poetry (influenced not by English Imagism but by the Russian Acmeists, especially the poetry of Akhmatova), strictly applies the morphological norm to the non-feminine plural suffix Im (e.g., raVIM – asaVIM), but disregards this requirement for the feminine plural OT (kallOT – netivOT) and other suffixes. Some "young" poets of the 1950s use even obvious grammatical rhymes, with the plural suffix IM as a rhymeme.
On the whole, the more widespread the use of a suffix in a language, the stronger is the tendency not to rely upon that suffix alone. The opposition to grammatical rhyme, inherited from Russian poetry, was strongest in the plural IM. On the other hand, the requirement to add a root consonant is entirely weak in two-syllabic suffixes, which are penultimately stressed, such as the dual: áyim.
The "mistrust" of grammatical rhyme is often expressed in poetry far beyond the required norm. Thus, Shlonsky uses rich rhymes especially when a suffix occurs, as if to compensate for the very use of a suffix in rhyme: DIRIM – aDIRIM, MeDuROT – MiDoROT (where RIM or ROT would be sufficient), etc.
The norm described above concerns the minimal group of sounds required in a rhymeme. In Hebrew poetry in the Israeli pronunciation there is a wide discrepancy between the minimum sounds required and the maximum actually used. Some poets, such as Shlonsky or Alterman, influenced by the poetics of Russian modernism, employ rhymes as rich as possible: tiKTEFÉNU – Ktefenu, AKuMÓT – hAKOMÓT, etc. The words in modern Hebrew, having no secondary stress, are usually long (3 syllables and more). Most of the sounds in the rhyming words of the poets under discussion are employed in the rhymeme. This tendency is doubly connected with the poetics of Hebrew "imagistic" poetry: (a) It is part of the general "colorful" aesthetics which abounds in striking imagery, rich sound patterns, "strong" themes, etc. There is a strongly expressed "set toward the message," a high "density" of the poetic language. (b) Since many sounds are involved in each rhyme, it is quite difficult to find words rhyming with each other; only a poetic language with a high degree of flexibility in imagery and elliptic combination could enable such freedom in connecting rhyming words drawn from distant spheres of meaning.
The maximal limit for a rhymeme consists in leaving a minimal difference between the rhyming members. In such rhymes as Yehudah *Karni's K'ILU MAT – KIL'UMAT (כְּאִלּוּ־מַת – כִּלְעוּמַת) or YIF'AM – YIF AM (יִפְעָם – יִיףעָם), the difference may only be a junction between words. In Alterman's rhyme בְּגוּמַת לְחָיַיִךְ – אֶת יִיִנָן הַלּוֹהֵט לְחַיַּיִךְ (LEXAYAYIX), the difference lies merely in the different morphological structure: in the first case ("your cheeks") the l is part of the root לחי, in the second case ("for your life") the l is a separate morpheme ("for") connected only graphically with the word חיים.
Rich rhyme in Israeli Hebrew can be explained not merely by the influence of one kind of modernist poetics. The properties of the language also encourage this trend. Most of the words in this pronunciation are stressed ultimately and most of the words are multisyllabic. Rhyming merely one syllable time and again would be tedious. Moreover, since most of the words are stressed ultimately, there is a multitude of words available for each rhyme ending. It may be compared to other languages. In Russian, where many words are also multisyllabic, stresses may occur on any syllable of a word, there-fore the number of words rhyming ultimately is relatively smaller, and multisyllabic rhyme is usual. In Yiddish, too, the number of feminine and dactylic rhymes is incomparably higher.
In English the number of monosyllables is so high that masculine rhymes are usual, as in Hebrew. But the "neutral" sounds of each number are not felt strongly, since they are few. In a usual Hebrew word one or three syllables do not participate in a minimal rhymeme: cf., the English pARTS – mARTS (though the rhymeme is monosyllabic most sounds of each member are covered by it) with the Hebrew mešuxrÁR – veaxzÁR where the nonparticipating, "neutral," sounds of each member are conspicuous. Moreover, English has some 13 different rhyming vowels (as compared to the mere five of Israeli Hebrew) and many consonant clusters, preceding and following the vowel, which make the number of possible rhyme endings incomparably higher and the number of words available for each relatively much smaller. Since there are very few possible rhyme endings in Israeli Hebrew, it is much easier to meet the minimal rhyme requirements and also easier, and more necessary, to add sounds and "enrich" the rhymeme. French with its ultimate stress, though it is more abundant in rhyme endings than Hebrew, also tends to prefer rime riche.
The basic norm of the accentual-terminal rhyme is identical in the poetry of the "Sephardi" pronunciation (in Italy since the Renaissance) or in the Israeli, as well as in the Ashkenazi, pronunciation, which was accepted in European Hebrew from the 14th century, but entered rhyme only in the 19th century. But the realization of the norm differed strongly due to the difference in the rhythmic structure of the Hebrew word.
Since the Ashkenazi stress falls on the second or third syllable from the end, most rhymes were automatically polysyllabic and most sounds of a word were included in the rhymeme. Thus, mÍDBOR – nÍDBOR is a very common rhymein Bialik's poetry, but the same pair makes a very rich rhyme in the Israeli pronunciation: mIDBÁR – nIDBÁR (since AR would be enough). Therefore, rich rhymes are few in Ashkenazi but may abound in Israeli Hebrew (at least in the practice of some poets). Moreover, the necessity to include in most cases at least two syllables in the rhymeme leads the poets to search for alleviating devices. Thus, Bialik in his early poetry tends to use grammatic rhymes which have already one syllable given in the morphological suffix, and the poet has to find words which differ in one syllable only. This necessity also leads to the use of archaic endings, feminine forms, etc. In short: any manqué form of a word, e.g., the Israeli rhyme צוֹהֵל – נוֹזֵל is not a rhyme in Ashkenazi (cohel – nozel), but the feminine form, with an added syllable, is צוֹהֶלֶת – נוֹזֶלֶת (coheles – nozeles). Therefore Bialik uses not אוֹר נוֹזֶל ("running light") but a more archaic form, which is feminine אוֹרָה נוֹזֶלֶת. Feminine verbs or adjectives in rhyme position obviously bring about feminine nouns in the middle of the line. The same holds for plurals and archaic forms.
Modernistic Hebrew poetry uses a large number of inexact rhymes, such as Alterman's ŠKuFÁ hI-miŠKaFAyIM: KoS Ha-MÁyIM – KSuMÁ hI, etc. In such rhymes at least one member ends with a "neutral" sound not participating in the rhymeme. But the effect is strong, since such rhymes usually have many sounds. There is a great variety of concrete forms, but in all cases the stressed vowel is constant, i.e., it is an accentual rhyme. In most cases the rhymeme is discontinued and the system may be called: accentual-discontinuous. The rhymeme, in addition to the fixed-stressed vowel, is based primarily on consonants. This phenomenon of the discontinuous rhymeme, particular to Hebrew poetry, is based on the nature of the Hebrew lexical morpheme, which is discontinuous and purely consonantal.
This system, representing a strong break from the standard European norm where rhyme is accentual-terminal and is usually continuous, had its forerunner in the earliest system of rhyme, in old Hebrew liturgy (the piyyut). But the concrete immediate influence which created this norm in Hebrew modernism came from the poetry of Mayakovski and *Pasternak, where rhymemes were also inexact in their endings and moved deeper into the middle of the line. While discontinuity in Russian rhyme was an occasional form of deviation rather than the rule, for Alterman and his contemporaries it became the norm, based on the characteristics of the Hebrew language. Alterman uses both exact and inexact rhymes in the same poems. The minimal requirement for a rhymeme now is: two sounds, at least one of which is the last stressed vowel. But only rarely was the minimal rhymeme employed (e.g., an exact rhyme: zahÁV – yadÁV; an inexact rhyme: koXÓ harXÓv). Most of the rhymes are very rich. The sound contrast between the rhyming members was strongly emphasized through the introduction of neutral sounds in between the sounds of the (discontinuous!) rhymeme, e.g., SiPuNÉXA – kaSE PaNÉXA, LEORÉR – LEOR nER; or by changing the order of the parallel sounds, e.g., miTPARECET – TRAPECIo-T (TPAR Trap).
In short, modernist rhyme cannot be described merely in negative terms, as a deviation from a "classical norm." The norm of the accentual-discontinuous rhyme creates a system as consistent and as effective as the accentual-terminal one, though the range of variation given to particular poets may be considerably wider now.
The preceding historical survey, though simplified as much as possible, presents a long chain of changes. When pan-historical comparisons are made, one finds logical relationships; similarities and contrasts between systems which are distant in time and place, but created in the forms of one language and culture. The table below, The Major Systems of Hebrew verse, may present the basis of such a comparison. The major systems of Hebrew verse are arranged in this diagram clockwise, in the order of their emergence in the history of Hebrew poetry. Except for a meter based on pitch, all known verse systems were productive in Hebrew. As can be seen from the diagram, there is a logical pattern, a kind of cyclic movement in this history. The major basis of the meter moved from phrase to word to syllable and vice versa.
The earliest and the latest verse systems were based on a free rhythm of phrase groups, though in the Bible there was a strong symmetricity of parallelism, whereas in modernist free verse there is a typical flow of continuity and lengths of lines may be highly varied. On the other hand, in modernist free verse poets often employ changing segments of accentual-syllabic meters as well as effects of irregular rhyme. It is not the freedom of "primitive" poetry, preceding any system, but the freedom of a "late" post-classical period, which is also free to employ any device developed in the "classical" rules of previous periods.
From biblical rhythm, based on semantic-syntactic-accentual free parallelism of phrases, the development of Hebrew verse moved toward basing its meters on more and more exact measures, i.e., ordering smaller elements of the language from phrases to the number of stresses, through the exact number of words, to the number of syllables, to a distinction of syllables according to their prosodic features.
Meters based on syllable counting ruled Hebrew poetry from about 950 almost to 1950. These were the most exact and variegated systems of Hebrew versification. Within this tradition, the change in the internal organization of the verse line from a quantitative principle to an accentual principle represented the general development of European poetry, but marked also the shift from the artificial "high" style of reading poetry to the intrusion of the cadences of the spoken language. (In religious poetry of Franco-Germany throughout the Middle Ages, a system based on the number of words persisted, i.e., a rhythm which, though numerically rigorous, was closer to representing some phrase patterns and clearly resembled the rhythm of medieval Yiddish and German poetry.)
In modernist poetry the movement of the early centuries of the Christian era was reversed: from strict syllable counting through a semi-regular meter, relying almost exactly on the number of major word stresses (though with a still limited freedom of syllable numbers 7), to a free verse system, based primarily on a rhythm of phrase groups, relying on the tension between the verse line and syntactic units. But in this period, even within the domain of free verse (8), the previous regular norms (6 accentual-syllabic and 7, accentual net) were still widely employed. On the other hand, the essential difference between the major systems of Hebrew verse should not lead to the overlooking of some basic consistent trends which cut across several systems. Within each system not all possibilities were equally employed. In any system, a rather small number of all possible forms were prevalent in poetry. Observing the syntactic possibilities of Hebrew verse in different periods, one finds a predilection for a certain optimal length of line, persistent throughout the ages: three or four major stresses in the Bible, four or five graphic words, 11 or 13 syllables (including short ones), three or four amphibrachs, and five iambs which are very similar in length of line and conveniently accommodate similar groups of words and phrases.
|Length of a line||Free||Fixed|
|Major Basis of the Meter||phrase||Word||syllable|
|Antiquity and Middle Ages (From phrase to syllable)||I. Bible: free accentual (Phrase – parallelism group of stresses)||II. Early Piyyut: accentual Number of major stresses||III. Rhymed Piyyut: word meter Number of words||IV. Spain: quantitative Number of syllables of long/short syllable|
|V. (Italy) Haskalah: syllabic Number of syllables|
|Modern Age (From syllable to phrase)||VIII. Modernist: Free Verse Changing balance of phrase groups||VII. Modernist ("Russian") accentual net Number of major stresses (+ limited freedom of syllables)||VI. Modern: accentual= syllabic Number of syllables order of stressed/ unstressed|
|No Regular Rhyme||Form of Rhymeme||Discontinuous||Discontinuous|
Free sound orchestration
|final||II. Kallirian Piyyut:|
R+RV מ + רִי
R+RVC מ + רִיס
R+R+CV(C) (מ + ר + הם (תִּי
|III. Medieval: terminal|
CV(C) (תֵם (תִּי
|stressed||V. Modernist ("Russian"):|
CV́+ דְּבָרִים רִי + מוֹרִי מַבְרִיק
V́C(V)+ (וֹר + (אֹרַח עוֹרֶק
CV́C(V) (דֶר + (דֶּרֶךְ − קוֹדֵר − אַץ
C+C+CV́+(V) ק + ס + תָ + יִ
(Ko-S ha Malm-KSu-MAhl)
|IV. (Italy, Haskalah) Modern:|
A similar pattern can be discerned in the history of Hebrew rhyme norms. Here, again, Hebrew poetry completed a whole cycle in its development. But rhyme was not as obligatory as meter. The earliest and the latest periods have no regular rhyme, i.e., no rhyme in the strict sense of a sound device used regularly for the strophic composition of a whole poem. The Regular Rhyme section of the table Major Systems of Hebrew verse represents typical rhymemes, using the following symbols: V – vowel, C – consonant, R – root consonant (only where relevant), V́ – stressed vowel, ł – discontinuity in the rhymeme. When read clockwise, the diagram represents the history of Hebrew rhyme.
Disregarding some secondary developments, there were four major rhyme systems. The similarities and the differences between these systems are related to the form and location of the rhymeme. The upper part of the diagram is opposed to the lower part from the point of view of the decisive vowel: in the Middle Ages the rhymeme relied on the final vowel, in the modern age on the stressed vowel. On the left hand (the extremes of this history) the rhymeme could be discontinuous, whereas on the right hand (in the "classical" periods) the rhymeme had to be a continuous and a terminal chain of sounds.
There is also a correlation (though not overlapping) between the corresponding major systems of meter and rhyme, as may be seen from a comparison of both diagrams. At the extreme ends of this cycle, when rhythm was based primarily on phrases, i.e., was dominated by a balancing of syntactic and semantic patterns, no regular rhyme was necessary. In the "classical" periods, when meter was based on the number of syllables, rhyme, too, was syllabic: the medieval rhymeme was based on one (terminal) syllable; modern rhyme based its major distinction of rhyme gender (masculine-feminine) on the number of syllables. Typically enough, in verse systems in which the prominence of the word was basic, discontinuous rhyme developed, i.e., rhyme based on the nature of the Hebrew word. However, this parallelism, essential as it was, was by no means automatic, e.g., word meter continued a long time after the suppression of the early discontinuous rhyme.
general: (There is no general history or survey of Hebrew versification). On rhyme: B. Hrushovski, "Ha-Shitot ha-Rashiyyot shel he-Ḥaruz ha-Ivri min ha-Piyyut ad Yameinu," in: Hasifrut, 4 (1971), 721–49. add. bibliography: E. Cogan, "Migilgulei Mishkal ha-Hat'amot be-Shiratenu," Heker ve-'Iyyun be-Madda'ei ha-Yahdut – Sifrut, Mikra, Lashon (1976), 107–117; T. Carmi (ed.), The Penguin Book of Hebrew Verse (1981). bible: W.H. Cabb, A Criticism of Hebrew Metre (1905); J. Begrich, Zur hebraeischen Metrik, in: Theologische Rundschau, 4 (1932), 67–89; I. Gabor, Der hebraeische Urrythmus (1929); B. Levin, Zivvug ha-Millin ba-Tanakh (1926). add. bibliography: F.M.Cross, "Toward a history of Hebrew prosody," Fortunate the Eyes That See (1995), 298–309 (=From Epic to Canon (1998)). piyyut: J. Schirmann, Hebrew Liturgical Poetry and Christian Hymnology, in: jqr, 44 (1953/54), Zunz, Lit Poesie; A. Mirsky, "Maḥẓavtan shel Ẓurot ha-Piyyut," in: ymḤsi, 7 (1958), 1–129; M. Zulay, ibid., 2 (1936), 213; idem, Piyyutei Yannai (1938); E. Fleischer, "Le-Ḥeker Tavniyyot ha-Keva be-Fiyyutei ha-Kedushta," in: Sinai, 65 (1969), 21–47; idem, "Iyyunim bi-Ve'ayot Tafkidam ha-Liturgi shel Sugei ha-Piyyut ha-Kadum," in: Tarbiz, 40 (1971), 41–63; idem, "Mivnim Strofiyyim Me'ein-Ezoriyyim ba-Piyyut ha-Kadum,",in: Hasifrut, 2 (1970), 194–240; B. Hrushovski, "Ẓurot ha-Piyyut ha-Kadum ve-Reshit ha-Harizah ha-Ivrit," in: Hasifrut, 3 (1971). add. bibliography: E. Fleischer, The Yoẓer – Its Emergence and Development (Heb., 1984); J. Yahalom, Poetic Language in the Early Piyyut (Heb., 1985); S. Elizur, Poet at a Turning Point – Rabbi Yehoshua Bar Khalfa and His Poetry (Heb., 1984); A. Mirsky, Ha-Piyyut – The Development of Post Biblical Poetry in Ereẓ Israel and the Diaspora (Heb., 1990); M. Zulay, in: E. Hazan (ed.), Ereẓ Israel and its Poetry – Studies in Piyyutim from the Cairo Genizah (Heb., 1995); N. Katusmata, The Liturgical Poetry of Nehemiah ben Shelomoh ben Heiman ha-Nasi (Heb., 2002); idem, Hebrew Style in the Liturgical Poetry of Shemuel ha-Shelishi (Heb., 2003). spain: Schirmann, Sefarad; idem, "La métrique quantitative dans la poesie hebraique du Moyen-Age," in: Se-farad, 8 (1948), 323–32; B. Halper, "The Scansion of Mediaeval Hebrew Poetry," in: jqr, 4 (1913/14), 153–224; D. Yellin, Torat ha-Shirah ha-Sefaradit (1940); idem, "Ha-Mishkalim be-Shirat Shemu'el ha-Nagid," in: ymḤsi, 5 (1939); I. Davidson, in: jqr, 30 (1939/40), 299–398; N. Aloni, Torat ha-Mishkalim (1951); S. Almoli, Shekel ha-Kodesh (1965); K. Heger, Die… Ḫargˇas… (1950); A. Mirsky, "Mashmaut he-Ḥaruz be-Shirat Sefarad," in: Leshonenu, 33 (1969). add. bibliography: E. Hazan, The Poetics of the Sephardi Piyut, according to the Litrugical Poetry of Yehuda Halevi (1986); J. Schirmann, in: E. Fleischer (ed.), The History of Hebrew Poetry in Muslim Spain (Heb., 1996); idem, in. E. Fleischer (ed.), The History of Hebrew Poetry in Christian Spain and Southern France (Heb., 1997); Y. Yahalom (ed.), Judaeo-Arabic Poetics – Fragments of a Lost Treatise by Eleazar ben Jacob of Baghdad (Heb., 2001). italy: A Mirsky, Mishkal ha-Tenu'ot ha-Italki, in: Sefer Ḥanokh Yalon (1963), 221–7; B. Hrushovski, "The Creation of Accented Iambs in European Poetry and their First Employment in a Yiddish Romance in Italy (1508–09)," in: For Max Weinreich, on his Seventieth Birthday (1964), 108–46. yiddish poetry. U. Weinreich, On the Cultural History of Yiddish Rime, in: Essays on Jewish Life and Thought (1959), 423–42; B. Hrushovski, "On Free Rhythms in Modern Yiddish Poetry," in: U. Weinreich (ed.), The Field of Yiddish (1954), 219–66. the modern age: B. Benshalom, Mishkalav shel Ḥ.N. Bialik (1945), idem, "Keri'at Deror la-Ḥaruz ha-Monosilabi," in: Hasifrut (1968–69), 161–75; S. Span, Massot u-Meḥkarim (1964); B. Hrushovski, "'Ritmus ha-Raḥvut' Halakhah u-Ma'aseh be-Shirato ha-Ekspresyonistit shel U.Z. Greenberg," in: ibid., 176–205. add. bibliography: U.Shavit, "Bein Ḥaruz le-Mashma'ut – le-Ofyah ve-li-Mkomah shel ha-Siyomet ha-Daktilit be-Shirat Bialik," in: Z. Malachi (ed.), Al Shira ve-Sipporet – Meḥkarim be-Sifrut ha-Ivrit (1974), 229–262; R. Zur, "Ḥaruz Anti-Dikduki u-Khshirut Leshonit," in: Ha-Sifrut 22 (1976), 59–62; M. Barukh, "Iyyun Nosaf al ha-Kesher bein Ḥaruz u-Mashma'ut [besifrut yeladim ivrit]," in: Sifrut Yeladim va-No'ar, 7 (1980), 10–14; D. Bregman, The Golden Way – The Hebrew Sonnet during the Renaissance and the Baroque (Heb. 1995); idem, A Bundle of Gold – Hebrew Sonnets from the Renaissance and the Baroque (Heb., 1997).